(L5 , \ nmRs csz£L(i~
NOAA TR NMFS CIRC 371
A UNITED STATES
DEPARTMENT OF
COMMERCE
PUBLICATION
NOAA Technical Report NMFS CIRC-371
U.S. DEPARTMENT OF COMMERCE
National Oceanic and Atmospheric Administration
National Marine Fisheries Service
Ocean Fishery Management:
Discussions and Research
ADAM A. S0K0L0SKI (Editor)
SEATTLE, WA
April 1973
NOAA TECHNICAL REPORTS
National Marine Fisheries Service, Circulars
The major responsibilities of the National Marine Fisheries Service (NMFS) are to monitor and assess the
abundance and geographic distribution of fishery resources, to understand and predict fluctuations in the quan-
tity and distribution of these resources, and to establish levels for optimum use of the resources. NMFS is also
charged with the development and implementation of policies for managing national fishing grounds, develop-
ment and enforcement of domestic fisheries regulations, surveillance of foreign fishing off United States coastal
waters, and the development and enforcement of international fishery agreements and policies. NMFS also
assists the fishing industry through marketing service and economic analysis programs, and mortgage insurance
and vessel construction subsidies. It collects, analyses, and publishes statistics on various phases of the industry.
The NOAA Technical Report NMFS CIRC series continues a series that has been in existence since 1941. The
Circulars are technical publications of general interest intended to aid conservation and management. Publica-
tions that review in considerable detail and at a high technical level certain broad areas of research appear in
this series. Technical papers originating in economics studies and from management investigations appear in
the Circular series.
XOAA Technical Reports NMFS CIRC are available free in limited numbers to governmental agencies, both
Federal and State. They are also available in exchange for other scientific and technical publications in the ma-
rine sciences. Individual copies may be obtained (unless otherwise noted) from NOAA Publications Section, Rock-
ville, Md. 20852. Recent Circulars are:
315. Synopsis of biological data on the chum salmon,
Oncorhynchus keta (Walbaum) 1792. By Rich-
ard G. Bakkala. March 1970, iii + 89 pp., 15
figs., 51 tables.
319. Bureau of Commercial Fisheries Great Lakes
Fishery Laboratory, Ann Arbor, Michigan. By
Bureau of Commercial Fisheries. March 1970,
8 pp., 7 figs.
330. EASTROPAC Atlas: Vols. 4, 2. Catalog No.
1 49.4:330/ (vol.) 11 vols. ($4.75 each). Avail-
able from the Superintendent of Documents,
Washington, D.C. 20402.
331. Guidelines for the processing of hot-smoked chub.
By H. L. Seagran, J. T. Graikoski, and J. A.
Emerson. January 1970, iv + 23 pp., 8 figs.,
2 tables.
332. Pacific hake. (12 articles by 20 authors.) March
1970, iii + 152 pp., 72 figs., 47 tables.
333. Recommended practices for vessel sanitation and
fish handling. By Edgar W. Bowman and Alfred
Larsen. March 1970, iv + 27 pp., 6 figs.
335. Progress report of the Bureau of Commercial
Fisheries Center for Estuarine and Menhaden
Research, Pesticide Field Station, Gulf Breeze,
Fla., fiscal year 1969. By the Laboratory staff.
August 1970, iii + 33 pp., 29 figs., 12 tables.
336. The northern fur seal. By Ralph C. Baker, Ford
Wilke, and C. Howard Baltzo. April 1970, iii +
19 pp., 13 figs.
337. Program of Division of Economic Research,
Bureau of Commerecial Fisheries, fiscal year
1969. By Division of Economic Research. April
1970, iii + 29 pp., 12 figs., 7 tables.
338. Bureau of Commercial Fisheries Biological Lab-
oratory, Auke Bay, Alaska. By Bureau of Com-
mercial Fisheries. June 1970, 8 pp., 6 figs.
339. Salmon research at Ice Harbor Dam. By Wesley
J. Ebel. April 1970, 6 pp., 4 figs.
340. Bureau of Commercial Fisheries Technological
Laboratory, Gloucester, Massachusetts. By Bu-
reau of Commercial Fisheries. June 1970, 8 pp.,
8 figs.
341. Report of the Bureau of Commercial Fisheries
Biological Laboratory, Beaufort, N.C., for the
fiscal year ending June 30, 1968. By the Lab-
oratory staff. August 1970, iii -f 24 pp., 11 figs.,
16 tables.
342. Report of the Bureau of Commercial Fisheries
Biological Laboratory, St. Petersburg Beach,
Florida, fiscal year 1969. By the Laboratory staff.
August 1970, iii + 22 pp., 20 figs., 8 tables.
343. Report of the Bureau of Commercial Fisheries
Biological Laboratory, Galveston, Texas, fiscal
year 1969. By the Laboratory staff. August
1970, iii + 39 pp., 28 figs., 9 tables.
344. Bureau of Commercial Fisheries Tropical Atlan-
tic Biological Laboratory progress in research
1965-69, Miami, Florida. By Ann Weeks. Oc-
tober 1970, iv + 65 pp., 53 figs.
346. Sportsman's guide to handling, smoking, and pre-
serving Great Lakes coho salmon. By Shearon
Dudley, J. T. Graikoski, H. L. Seagran, and Paul
M. Earl. September 1970, iii + 28 pp., 15 figs.
347. Synopsis of biological data on Pacific ocean perch,
Sebastodes alutus. By Richard L. Major and
Herbert H. Shippen. December 1970, iii + 38
pp., 31 figs., 11 tables.
Continued on inside back cover.
ATMOSAl,
Iff NT of c
U.S. DEPARTMENT OF COMMERCE
Frederick B. Dent, Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
Robert M. White, Administrator
NATIONAL MARINE FISHERIES SERVICE
Philip M. Roedel, Director
NOAA Technical Report NMFS CIRC-371
Ocean Fishery Management:
Discussions and Research
ADAM A. S0K0L0SKI (Editor)
9-
o
u
o
a
0)
Q
Proceedings of a workshop
sponsored by the Division
of Economic Research, National
Marine Fisheries Service,
November 5-6, 1970
This report contains a group
of independent research studies
published on their own merits.
They do not necessarily reflect
the policies or intentions of
the National Marine Fisheries
Service.
SEATTLE, WA
April 1973
PARTICIPANTS
Paul Adam, Organization for Economic Cooperation
and Development
Harold B. Allen, National Marine Fisheries Service
Frank M. Anderson, Oregon State University
Frederick W. Bell, National Marine Fisheries Service
Daniel W. Bromley, University of Wisconsin
Ernest W. Carlson, National Marine Fisheries Service
Donald P. Cleary, National Marine Fisheries Service
Francis T. Christy, Jr., Resources for the Future
James A. Crutchfield, University of Washington
John P. Doll, University of Missouri
Richard F. Fullenbaum, National Marine Fisheries
Service
John M. Gates, University of Rhode Island
Thomas Geer, International Bank for Reconstruction
and Development
William G. Gordon, National Marine Fisheries Service
Loren Grant, Canadian Department of Fisheries and
Forestry
John E. Greenfield, National Marine Fisheries Service
Murray L. Hayes, National Marine Fisheries Service
Andreas A. Holmsen, University of Rhode Island
Paul Hooker, University of Florida
A. M. HuQ, University of Maine
Harvey M. Hutchings. National Marine Fisheries
Service
A. D. Insul, British White Fish Authority
Joshua John, Canadian Department of Fisheries and
Forestry
Milton G. Johnson, National Oceanic and Atmospheric
Administration
Edward Kane, Boston College
E. A. Keen, San Diego State College
Richard K. Kinoshita, National Marine Fisheries
Service
Jukka A. Kolhonen, National Marine Fisheries Service
Richard J. Marasco, University of Maryland
Bruce Mattox, University of Rhode Island
Morton M. Miller, National Marine Fisheries Service
Paul Mlotok, University of Rhode Island
Darrel A. Nash, National Marine Fisheries Service
Bruno G. Noetzel, National Marine Fisheries Service
Virgil J. Norton, University of Rhode Island
Frederick L. Olson, National Marine Fisheries Service
Erwin Penn, National Marine Fisheries Service
Giulio Pontecorvo, Columbia University
Frederick Prochaska, University of Florida
R. Bruce Rettig, Oregon State University
Jack Rich, Oregon State University
Jack A. Richards, Kansas State University
Jon C. Rittgers, University of Rhode Island
Richard Roberts, Fisheries Service of Canada
Michael A. Robinson, Food and Agriculture Organiza-
tion of the United Nations
Edilberto L. Segura, Columbia University
Frederick J. Smith, Oregon State University
Adam A. Sokoloski, National Marine Fisheries Service
Miller Spangler, National Planning Association
Joe B. Stevens, Oregon State University
David A. Storey, University of Massachusetts
John K. Sullivan, National Marine Fisheries Service
William M. Terry, National Marine Fisheries Service
B. G. Thompson, National Marine Fisheries Service
Russell G. Thompson, National Water Commission
Lawrence W. Van Meir, National Canners Association
John Vondruska, National Marine Fisheries Service
Hoyt A. Wheeland, National Marine Fisheries Service
Donald R. Whitaker, National Marine Fisheries Service
Frederick E. A. Wood, Fisheries Service of Canada
Robert Wilson, Texas A & M University
PREFACE
Until recent years only biological or technical aspects of fisheries con-
servation have advanced beyond esoteric professional journals or smoke-
filled back rooms to be given serious consideration when formulating work-
ing management programs. In recent years the social sciences, especially
economics with its emphasis on rational management, have gained some
respectability beyond mere conceptual discussion.
With the mounting urgency of fishery management problems serving
as a catalyst, the National Marine Fisheries Service has multiplied its
research in this area, aided in part by the rapidly growing Sea Grant pro-
gram formerly within the National Science Foundation and now incorpor-
ated within the National Oceanic and Atmospheric Administration.
Within the past two years much progress has been made. To aid in
assimilating these results and to provide some sense of a proper future
direction for both research and the design of management programs, the
National Marine Fisheries Service convened a Workshop on November
5 and 6, 1970. Invited were virtually all known researchers in Fishery
Economics throughout the world, many administrators, and researchers
in related disciplines.
What follows in this circular are the papers presented at this workshop,
with an introduction which makes a first attempt at distilling the combin-
ed impact of these papers.
As editor I wish to thank all the authors for their diligent cooperation.
The services of the secretarial staff of the Division of Economic Research,
especially Miss Carol Reese, are gratefully acknowledged. The generous
support of the many institutions that absorbed the financial burden of
travel from distant geographic regions was necessary for the ultimate
success of this workshop.
The National Marine Fisheries Service sponsors the publication of
these papers, as it sponsored the workshop itself, to crystallize the issues
relating to fishing management and to stimulate further debate. As such
the papers present the views of the individual authors and none of the
material contained herein should be construed as reflecting official policy
statements of the National Marine Fisheries Service.
Adam A. Sokoloski
in
CONTENTS
Page
INTRODUCTION 1
The Status of Fisheries Management Research: An Overview.
Adam A. Sokoloski 1
ISSUES IN FISHERY MANAGEMENT 7
Problems in Implementing New Fishery Management Programs.
Lawrence W. Van Meir 9
On the Utility of Bioeconomic Models for Fisheries Management.
Giulio Pontecorvo 12
Multiple Objectives for Marine Resource Management.
R. Bruce Rettig 23
Economic, Political, and Social Barriers to Efficiency in
Selected Pacific Coast Fisheries. James A. Crutchfield 28
PRODUCTION FUNCTIONS AND BIOECONOMIC MODELS:
RESEARCH IMPLICATIONS 39
Cross Section Production Functions for North Atlantic Groundfish and
Tropical Tuna Seine Fisheries. Ernest W. Carlson 42
Optimal Fishing Effort in the Peruvian Anchoveta Fishery.
Edilberto L. Segura 57
Natural Resources and External Economies: Regulation of the
Pacific Halibut Fishery. Jack Rich 65
Production from the Sea. Frederick W. Bell, Ernest W. Carlson,
Frederick V. Waugh 72
Some Suggestions for the Development of a Bioeconomic Theory of
the Fishery. Russell G. Thompson 92
Practical Problems of Constructing Bioeconomic Models for Fishery
Management. Paul Adam 96
(Continued)
ISSUES RELATED TO FISHERY MANAGEMENT:
RESEARCH RESULTS : 104
Management of the Peruvian Anchoveta Resource.
Andreas A. Holmsen 106
A Stochastic Investment Model for a Survival Conscious Fishing
Firm. Russell G. Thompson, Richard W. Callen, Lawrence C. Wolken ... 112
Simulation Experiments to Evaluate Alternative Hunting Strategies
for a Deer Population. F. M. Anderson, G. E. Connolly, A. H. Halter,
W. M. Longhurst 121
Augmentation of Salmon Stocks through Artificial Propagation:
Methods and Implications. Joe B. Stevens and Bruce W. Mattox 133
Limited Entry: The Case of the Japanese Tuna Fishery. E. A. Keen 146
A Study of the Socioeconomic Impact of Changes in the Harvesting
Labor Force in the Maine Lobster Industry. A. M. Huq 159
VI
INTRODUCTION
The Status of Fisheries Management Research
An Overview
Adam A. Sokoloski1
All disclaimers to the contrary, there is one
research area near and dear to the hearts of
virtually all economists conducting research
on marine resources: measuring the gap be-
tween the "optimum" management solution for
a given fishery and current management arrange-
ments. This is not to say that this gap has ever
been successfully measured for a fishery.
In recent years some first approximations
have been made, however. These have been
reasonably consistent with a body of economic
theory which has existed in one form or an-
other for several years. This theory is the
original source of suggestions that the gap
existed, as casual observation of practice re-
vealed inconsistencies with "proper" theory.
Initial empirical works unearthed several
critical components which are currently com-
plicating the issue. These are both empirical
and conceptual in nature and multidisciplin-
ary in scope. These may be listed as follows:
1. Existing yield functions need to be ex-
panded and alternative functions need
to be specified, both with respect to such
factors as diminishing returns (success
probabilities for effort on a fixed bio-
mass) and multispecies interrelation-
ships.
2. The appropriate emphasis for economics
and biology in bioeconomic models.
3. The correct theoretical and empirical
components of effort series are needed
1 Formerly of Division of Economic Research,
National Marine Fisheries Service; present address,
Environmental Protection Agency, Division of Water
Quality Standards, Arlington, VA 22202.
to construct indices of fishing power as
utilized in management programs.
4. More effort is needed in the design of
"correct" operational management plans:
the choice between variations of licens-
ing, quota, auction and/or leasing
schemes.
5. A resolution of the choice between long
run versus short run "optimal" solutions.
6. An evaluation of the appropriateness of
directly applying theoretical models to
fisheries for the purpose of deriving im-
plied net gains from the practical appli-
cation of identical working models.
7. The role of social transfer costs in the
evaluation of benefits from new manage-
ment programs.
8. The desirability of an incentive (pull)
approach versus a limited entry (push)
form of management program.
9. The place of jurisdictional consideration
in program design and operation.
10. The desirability of a multidisciplinary
objective simulation approach to the
measurement of management ramifica-
tions as contrasted to simultaneous
equations with maximization and other
limiting assumptions.
11. The role of artificial propagation in the
design of total management plans.
12. The role of competing uses for the re-
source base.
Virtually all of these items reflect the fact
that as economists begin to penetrate the sur-
face of the management issue they gain a
greater appreciation of the vital role to be
played by the physical scientist, usually a
biologist who has become a population dy-
namics expert. This involves more than just
using the output of the population dynamics
expert; it entails understanding the intricacies
of this work so that it won't be misused.
Here is where the first problems arise. Once
familiar with population dynamics models the
economist falls prey to the temptation to alter
components which may not be ideally suited to
his needs. What results is two versions of
population dynamics with one being the result
of both explicit and implicit imperfections in
the other.
From this point several ramifications may
develop, depending on how far each conceptual
base may have been developed toward an actual
working management program. If this has oc-
curred original differences in population dy-
namics models will have been magnified. These
resultant differences generate a debate, and a
portion of this debate, as currently stated, is
contained in the following papers. To amplify
let me refer in greater detail to the twelve
points mentioned above.
(1) The Need for New Yield Functions: The
biologist's yield function is the analogue of
the economist's production function. Produc-
tion economics focuses upon the allocation
of inputs to achieve production goals designat-
ed as optimum, this proper allocation being the
most efficient (least cost) combination of these
inputs. Partial derivatives, giving the incre-
mental contribution of each unit of a particu-
lar input, may be used to construct efficiency
indices roughly equivalent to the biologist's
measures of the fishing power of a vessel.
These derivatives are obtained from general
form equations of a linear, Cobb-Douglas (con-
stant elasticity of substitution equal to one)
or C.E.S. (any constant elasticity of substitu-
tion) type. Contained within these general types
are certain assumptions concerning constant,
increasing, or decreasing returns (output) from
increasing increments of a particular input.
Critical here is an appreciation of the fact
that these are fundamental calculations which
would be carried out whether or not any re-
lated biological work existed. When this work
does exist it serves as a reference point to the
economist as he proceeds systematically
through a series of steps dictated by the classi-
cal scientific method which has evolved for
his profession.
When, therefore, an economist specifies a
function implying diminishing returns to ad-
ditional inputs, we have the potential for debate
when the biologist has diminishing returns
due to population dynamics but constant re-
turns from a fixed biomass. These two differ-
ing approaches will lead to different evalua-
tions of the historical effort being exerted on
a fishery, to different estimations of the actual
yield curve, to different calculations of MSY
(maximum sustainable yield) and then to dif-
ferent management solutions.
The issue becomes further complicated when
many species intermix and then must be con-
sidered simultaneously when designing and
operating a management program. The eco-
nomic portion of this analysis is actually more
readily solved in this case via a standard
analysis of the joint product case, whereas the
biological literature still carries a debate con-
cerning the proper use of Beverton-Holt dy-
namic pool models as opposed to the Schaefer
logistic approach. This issue is becoming more
critical as the trend in the technological capa-
bility of harvesting units is leading toward
some point in the future where the flexibility
and maneuverability of these units will make
all management considerations multispecies
to correctly reflect actual harvesting practices.
(2 & 3) Economics and Biology in Measures
of Fishing Power: For management purposes
what is the appropriate emphasis of economics
and biology in bioeconomic models? One ex-
treme suggests that it is necessary to under-
stand the complete microdynamics of all stages
of the food chain, an ecological approach, and
all forces that act upon these stages, to proper-
ly specify the results of variation in fishing
effort and, therefore, to suggest the optimum
dimensions of that effort. This would confine
economists to a role of evaluating the economic
costs and benefits of the program suggested
by this detailed formulation.
The opposite extreme finds the economist
placing the fisherman in an active role, where
he responds to various market incentives,
these responses subsequently becoming an in-
tegral step in determining variations in fishing
effort and resultant success. Some would
suggest that prices, quantities landed, and a
statistically acceptable production function
are all that is needed to derive the functional
relationship necessary for management, as-
suming that this production function can be
used to determine the continuing relationship
between effort and landings.
This last phrase is important, for it may
well be that a final decision in the allocation
of research and management resources will
depend upon the spin-off, or secondary benefits
from certain research endeavors above and
beyond their direct contribution to manage-
ment. This would, of course, be especially true
with regard to the extreme of the broad-based
ecological approach, with both long run and
short run considerations, as suggested in some
biological circles. The recent reorganization
of certain agency functions within the National
Oceanic and Atmospheric Administration
may have some bearing here.
(4) Using Research Results to Design Oper-
ational Schemes: Existing schemes which may
be actually classified as direct measures to
limit entry have established certain precedents.
Canadian programs in Atlantic lobsters and
Pacific salmon have emphasized licenses for
principal capital inputs, a limited entry pro-
gram. The Inter-American Tropical Tuna
Commission (IATTC) has utilized quotas and
area restrictions while International Commis-
sion for the Northwest Atlantic Fisheries
(ICNAF) has utilized mesh size and area re-
strictions and recently seasonal closures for
overfished species. The Union of South Africa
regulates via licenses issued through proces-
sing plants which allocate these among vessels.
The debate concerning these existing types
and many other hypothesized forms may be
divided into two subject areas, one concerning
whether the plan will actually lead to an al-
location of resources which approaches some
predetermined optimum and the other whether
the plan is operationally realistic, which may
invoke social and political considerations as
well as those of biology, technology, and eco-
nomics. Mr. William Terry, in his opening
statements to the participants of this work-
shop, emphasized the urgent need to begin
evaluating the proper mix of all possible inputs
into fisheries management. He asserted that
we must begin now to define the components
of the interface, looking beyond the immediate
problems of each discipline. Consistent with
this he suggested the possible need to develop
new, broader objectives of fishing management.
Presently discussed mechanisms have two
principal components, a way of limiting entry
into fisheries and a means of allocating the
quasi-property rights which result. Some
mechanisms perform these functions simul-
taneously, such as an auction system, whereas
others, such as a licensing system, require the
administrator to make some judgement as to
the number of licenses as well as how they
shall be allocated. Relating to some historical
experience in the oyster fishery, there are
many unexplored questions regarding the ap-
plicability of leasing schemes for sessile re-
sources.
The message here is clear. We have devoted
considerable effort to developing sophisticated
conceptual constructs for fisheries manage-
ment. Regarding actual operational alterna-
tives we have confined our efforts to informal
and often exclusively internal debates. Re-
sponsible researchers must soon assume the
task of a thorough evaluation of the many
suggested working plans. This evaluation
must be subsequently exposed to discussion
via the professional journals so that all the
preceding work can truly be productive.
(5) Long Run Versus Short Run Solutions:
The fruits of economic modeling are proposals
involving changes in the quantities of labor
and/or capital in commercial fishing, often
reductions of both in addition to increases
in the capital/labor ratio. Capital inputs are
usually quite fixed; indeed, this may be true
of labor inputs also. What results is the quite
obvious conclusion that achieving these opti-
mum solutions will in all cases involve extend-
ed time periods. What then is the preference
ranking for the many plans which may have
to be initiated in the interim?
The truthful answer to these questions is
that we really don't know. We have not fully
designated the compromises which would be
necessary, much less made a careful evaluation
of which would be preferable. This suggests
two immediate tasks to be undertaken by
the economist.
The first of these relates to the fact that
the responsible administrator will not wait
for the perfect solution to be formulated for
each time horizon. He must formulate plans
and action programs on a continuing basis.
In this instance the economist can indicate
those steps which can be taken which will
lead toward the optimum solution, or at least
toward some "better" solution in the tradition
of the theories of second best as discussed
in the literature on welfare economics.
Simultaneously the economist can perform
a second function, which would be to construct
detailed interim plans and test and evaluate
these. These could be constructed for alterna-
tive time periods and based upon restrictions
suggested by the administrator or the other
disciplines where additional intermediate term
planning and research was also being con-
ducted.
The result would be an array of economic
research considering time horizons from the
present to the long run optimum solution.
With such an array it would be easier to in-
corporate the interdisciplinary (especially
social) aspects of the overall problem as sug-
gested in the other points I am presenting
here. Most critical is the fact that immediate
steps are necessary if there is to be a fishery
to optimize in the long run.
(6) Theoretical Versus Working Models:
This point relates to the previous issue and
also to several of the following. Brieflly the
question here is whether theoretical models,
confined solely to a select number of variables,
and seldom involving more than two disci-
plines, can be utilized directly for generating
a stream of benefits to be included in the calcu-
lation of a benefit/cost ratio for a particular
management program. Some have argued that
there is not sufficient realism in theoretical
models for these to be applied directly. Con-
versely, it may also be argued that many of
the imperfections of this approach are not so
much inherent in the theoretical models them-
selves, but rather stem from the use of com-
plementary information when performing B/C
analysis. Such errors may be found principal-
ly on the cost side, where not all indirect pro-
gram costs are included, especially when
these costs may exceed the actual flow of
benefits in the short run. What may be the
most significant of these cost components is
discussed next.
(7) Social Transfer Costs: Not long ago it
was not possible to discuss limited entry ex-
cept under the most constrained circumstances.
Now, with the development of more forceful
arguments and with a growing urgency in
certain fisheries, limited entry plans are re-
ceiving wider consideration. As this occurs
the operational elements of alternative plans
are being formulated and new questions are
resulting. The most prominent among these
relates to the magnitude of the social transfer
costs which may result from either the direct
or indirect reduction of the fishing labor
force in a fishery.
If displaced labor must be retrained and
relocated, or absorbed on the welfare rolls,
it may be wise to develop programs based
exclusively on excluding excessive new entry,
with input balances to be attained via attri-
tion. This is tantamount to concluding that
the short run solution must be contrary to
the suggested long run optimum. Neverthe-
less, it is the desired means of achieving the
long run optimum. Research is now begin-
ning on this issue both within the National
Marine Fisheries Service and the Office of
Sea Grant Programs. The results will play
a critical part in determining the character
of future management plans.
(8) Encouraging Exit Versus Limiting Entry:
Virtually all discussions of management plans
emphasize licenses, or quotas, or some form
of right which will accrue to a reduced num-
ber of harvesting units. The mechanics of
reducing these units involve some form of
exclusion. Seldom has a plan been suggested,
however, which emphasizes a program where-
by excess inputs would be attracted away
from the fishery by a more rewarding altern-
ative.
To my knowledge such a program has been
attempted once, a recent attempt to divert
excess capacity from the overfished haddock
resource of the Northwest Atlantic to the
underutilized pollock resource. As a limited
short term program it met with only limited
success. This is not inconsistent with other
similar non-fishery programs, such as the more
substantial effort designed for Appalachia. In
all instances, unanticipated attractions, pic-
turesquely described as "psychic income," re-
sulted in a greater amount of labor immobility
than original calculations suggested. Program
costs had to be adjusted accordingly.
Nevertheless, if transfer costs, as discussed
in the previous point, can be reduced by some
increment by an incentive program costing
less than this increment, then the overall costs
of the total management program may be
reduced sufficiently to result in a favorable
B/C ratio. These calculations would be over
and above the more favorable political re-
sponse to a program which considered these
transfer costs as opposed to one which did not.
The problem of response to incentives may
be reduced in multiple species fisheries where
we wish to reduce pressure on one of the
species and this is technically possible. Hard-
ships resulting from restrictions on the king
crab resource were reduced by the ability of
the harvesting units to adapt to alternative
species. Indeed, New Bedford scallopers, 13
vessels in all, journeyed to Alaska when that
resource appeared (somewhat falsely) more
profitable than their traditional fishery. If
they could have been induced to leave earlier
then perhaps the degree of depletion in the
Atlantic could have been reduced.
(9) Jurisdictional Issues: Fisheries research-
ers interested in formulating management
plans usually focus on specific fisheries in
their entirety. This is appropriate for every
"discipline" except one, the area of legal-
political considerations. Fish do not respect
jurisdictional boundaries and this has long
been a critical operational issue in fisheries
management.
Resolution of these jurisdictional issues will
involve more lead time than biological and
economic questions. Developing interstate co-
operative mechanisms and widely accepted
international arrangements which will be polit-
ically acceptable while incorporating biologic-
al, economic, and social factors will be a
herculean task, witness the slow progress of
developing a national quota system in ICNAF
and the 200-mile dispute with countries
bounding the yellowfin tuna fishing areas.
The individual disciplines can contribute to
solving this problem by orienting their work
so that" the trade-offs between alternative juris-
dictional arrangements can be readily assessed
in each disciplinary dimension. As the U.S.
develops new coastal zone and contiguous
zone legislation and as all nations prepare
for another Law-of-the-Sea conference it be-
comes increasingly necessary that these trade-
offs be specified in the near future.
(10) The Potential of Simulation Models:
Much of the population dynamics research
done to date has involved single or multiple
equation regression techniques of constrained
maximization. Within the capabilities of these
techniques one (biology) or at most only two
(biology-economics) disciplines would be con-
sidered, and even then only a limited number
of variables in each. Many of the twelve points
discussed here are not included within these
analyses. At best they are appended on an
ad hoc basis.
To formalize this ad hoc process one would
set out specifically to systematize these multiple
considerations via a simulation model, where
each consideration would appear sequentially
leading to outputs which would represent
many combinations of these interactions. With-
in this framework each specialist would not
be trying to extend his own area to include
other disciplines in the process of specifying
optimum solutions. Rather he would merely
be characterizing his own special consider-
ations, which might be one of several sub-
routines in the entire simulation program.
The proper manner in which these inputs
would be combined would be a joint responsi-
bility of all researchers providing the principal
inputs.
To be feasible, each separate input area
must have reached a sufficient stage of sophisti-
cation and accuracy to be of use in a simula-
tion model. I believe this judgment can now
be made. This suggests that the work that
has been initiated at the University of Wash-
ington and Massachusetts Institute of Tech-
nology should be expanded to encompass all
major fisheries. Work on other water resource,
game resource management problems provide
an additional base of expertise to facilitate
development of these models.
Initially these models would include: (1) an
assessment of the resource base, (2) a popu-
lation dynamics model, (3) cost and earning
functions, (4) demand functions, including
provision for foreign trade flows, (5) exit-entry
functions based on profitability, (6) character-
ization of existing and alternative regulatory
constraints, and (7) a depiction of the social
response function, with some reference to
transfer costs. These models would originally
be constructed for each of the principal fish-
eries of Alaska, the Pacific Northwest, the
tuna fisheries, the shellfish and menhaden
fisheries of the Gulf and the Middle Atlantic
and lobster and groundfish in the North
Atlantic. Ultimately multispecies regional
models would be developed, leading to a
national model which would characterize the
entire U.S. fishing industry.
Initial failures in the construction of these
models will suggest immediate research needs.
The output of each model will indicate the
sensitivity of each component of the model
for each fishery.
(11) Artificial Propagation and Fishery
Management: With few exceptions, when we
identify a fishery which has excess capital
and/or labor in relation to the sustainable
resource base we recommend reduction in
these inputs. A Canadian fishermen's group
has eloquently phrased another course, that
is, expand the resource base. This would
especially be recommended if the incremental
returns from dollar expenditures on expansion
exceeded the incremental benefits from dollars
spent withdrawing inputs.
At this time such a possibility could only
be anticipated for Pacific salmon. Several
factors could enhance these trade-offs, among
these being the possibility that demand rising
faster than costs would bring the cost of
hatchery production into a more favorable
light and a full realization of the political
resistance to withdrawing excess inputs. With
the further development of hatchery tech-
nology other fisheries, perhaps shellfish, may
be supplemented by artificial propagation and
rearing. As this occurs it will be necessary
to include the dimensions of this alternative
as an new subroutine in the simulation models
discussed in point 10 above.
(12) Competing Uses: A new dimension, an
additional complication, has entered upon the
scene of fishery management, suggesting new
priorities here as it has elsewhere. It comes
under the banner of ecology, an old word
with new urgency. With the scarcity of natural
resources increasing relative to multiple de-
mands, and with the new insistence upon
quality in addition to (or rather than) merely
quantity, the management of coastal resources
has suddenly taken on a new dimension. Man-
agement of commercial fisheries will be obliged
to reflect this trend.
Coastal fisheries must now be managed as
part of the total coastal resource. No sug-
gestion has yet been made as to how this
will be done. Suffice to say that such critical
issues as fishery tolerances to certain water
quality levels and the interrelationship be-
tween sports and commercial fisheries will
be critical issues. I will forego an attempt to
treat this issue in a few brief paragraphs
here, acknowledging the likelihood that the
next fisheries workshop will certainly treat
this area as one of its principal topics.
With this general background on the princi-
pal issues in fisheries management we can
now look briefly at the workshop contributions
to summarize their contents. With these papers
serving as the stimulus the discussions at
the workshop inevitably revolved around two
related issues: (1) the necessity for develop-
ing short term models due to the extreme
urgency of resource management problems
in many fisheries and (2) the need to assume
the full responsibility for measuring all social
costs associated with alternative resource use
plans and to suggest ways by which these
social costs can be minimized.
At the conclusion of this workshop one
was definitely left with the impression that
if significant steps cannot be made in both
of these areas in the near future (2-4 years)
then serious questions will have to be raised
about the utility of the bioeconomic,
socio-political research and planning which
we are conducting. In this light much of the
work reported at the workshop provides some
encouragement that progress will be made
on these issues.
ISSUES IN FISHERY MANAGEMENT
The opening paper by Van Meir appropri-
ately cites the Burkenroad observation that
fisheries should be managed for people not
fish, a trite, but occasionally overlooked ad-
monition. He emphasizes that the critical
element now is that time is running out in
many fisheries. The solution is to replace com-
mon rights with private rights, these rights
to be consistent and in balance with allow-
able yield. The program should not only
permit, but also promote economic efficiency
both in the short run and in the long run.
To begin limited entry programs we must
emphasize three areas: (1) a resolution of
jurisdictional conflict, (2) an educational
program which will communicate the poten-
tial benefits and dispel the idea that the
scheme is to be a government monopoly and,
(3) trial programs which will demonstrate
how limited entry operates in practice.
Van Meir suggests that in practice we must
be willing to accept a second-best solution,
i.e., agree with biologists on harvesting maxi-
mum sustainable yield (MSY) and proceed to
specifying the most efficient way of doing this.
We must develop a system which will insure
that fishing rights will be allocated to the
most efficient producer at any point in time.
Van Meir concludes by suggesting a system
for doing this. It is here that he introduces
the first real element of controversy. He sug-
gests a licensing mechanism. Licenses would
be allotted so as to include all grandfather
rights. They would be reduced by attrition
with the total number changing as technology
changes. Monopoly powers would be restricted
and rents would be redistributed via license
fees or taxes.
Undoubtedly this is a reasonable step toward
a politically palatable solution. Others would
argue that there are other schemes that
would be more appropriate for certain fish-
eries. They would argue that this proposal
contains the same faults as U.S. agricultural
programs of the past decade, where a central
authority is granted the right to determine
the number of licenses. To do so it must use
existing measures of technological capacity
and technological change, when both of these
may change substantially under the exogenous
influence of a newly introduced licensing
scheme. Some alternative suggestions would
allow both the rate of technological change
and the size and number of property rights
to be determined within the market mechan-
ism. The paper presented by Holmsen refers
briefly to one alternative. Also the paper by
Carlson could serve as a basis for preliminary
calculations of the appropriate number of
licenses in the tuna and groundfish fisheries.
Pontecorvo introduces several broad con-
ceptual issues, among these being the need
for short run models which can be utilized
directly in resource management. If the short
run is critical we should examine those models
which appear to be more satisfactory for the
short run.
Pontecorvo focuses upon the difficulties of
choosing biological models and combining
these with economic models for both short
run and long run analysis — to determine
optimum solutions. He cites violent fluctua-
tions in the Pacific red salmon resource as
a characteristic which militates against the
use of even short run models, and also where
the costs of improving the information flow
may exceed the benefits. Further complica-
tions arise due to instability on the economic
side (demand and the general state of the
economy) and changing political and social
considerations. One suggestion here is that a
program geared to catch some average level,
less than the allowable yield during the
highest year, may be the desirable economic
solution — one case where we would suggest
taking less than MSY.
Pontecorvo's position on social and political
issues is that these are fully accounted for
(albeit incorrectly) in the economists' assump-
tions of full employment and factor mobility.
The economists' assumption of human rational-
ity forces the social-political ordering into
the same ordering as economics. The more
reality deviates from this ordering the more
the economic conclusions must be altered by
subsequent ad hoc social and political con-
siderations.
This can be extended to multiple use issues
as well. Often we treat the fishery as if it were
the only user of the resource. Future regu-
latory organizations will have to incorporate
such considerations directly and this will affect
the design of these organizations.
Rettig adds to the mounting chorus warning
of the social implications of certain fisheries
management plans. He suggests that these
may lead us to actually restructure the ob-
jectives of these plans. Absence of these con-
siderations may be one reason for our failure
to initiate revised management programs.
Other reasons for failure may be the present
existence of a severe divergence between the
objectives of administrators and researchers,
incompletely specified models, or the mere
absence of sufficient educational programs.
Regarding the incorrectly specified models
Rettig makes the intriguing observation that
market imperfections on the buyers' side could
alter the optimum solution. Ignoring this
fact would actually result in a further mis-
allocation of resources. He suggests a further
evaluation of inter-market linkages before
making irreversible management steps.
Additional issues which must be faced are
the multispecies management problems and
the absence of a reasonable discount rate in
the sustainable yield curve. This relates to
some degree to his final conclusion that we
must include so many diverse factors that in
the end our "theory" may be useless. Never-
theless, like many others as well as partici-
pants at this workshop, he can see no other
alternative but to follow this course unless
we intend to ignore realism and the needs
of fishery administrators.
In the last of four general papers on the
issues in fishery management, Crutchfield re-
views the inputs to fishery modeling work
now developing for four Pacific Northwest
fisheries: anchovy, salmon, king crab, and
halibut.
These models have three basic components:
economics, biology, and law. In the economics
portion the cost and earnings and profit and
loss statements for representative vessels are
developed, related to certain catch rates, tech-
nological factors and market conditions (pro-
duct price, interest rate, alternative employ-
ment). By this manner the complete operation
of vessels in the selected fisheries can be speci-
fied and from this it is possible to construct
an exit-entry function which would relate to
changes in these economic variables, indepen-
dently, or as affected by biological and/or
legal variables.
The biological elements of this model include
gross stock parameters and a yield-effort
function which generates catch rates, these
serving as direct input into both the economic
model and into population dynamics compo-
nents of the biological model. In the case of
the salmon fishery separate, though similar
models, are developed for five different stocks
at ten locations, a 50-cell matrix. Any per-
tinent species interactions are also included.
The legal portion of this model specifies
the existing regulatory structure which may
determine the components of both the biological
and economic models, determining what is fished
for, when, how, and to what extent. As in
other portions of the model, alternative legal
structures will be posited to allow for alterna-
tive patterns of resource utilization.
The ultimate purpose of this model is to
take a complete interdisciplinary approach
to fisheries management. Alternative manage-
ment programs will be specified. Among these
an optimum plan will be identified, with the
sequence of steps which would most effective-
ly lead toward this plan. In its most extensive
form this model will consider multiple species
management cases such as anchovy-mackerel-
tuna in California and salmon-tuna-crab-
halibut of the Pacific northwest. As empha-
sized by Crutchfield, in its present form the
model emphasizes the multidisciplinary nature
of the management problem and will readily
incorporate many of the suggestions made
at this workshop.
A. A. S.
Problems in Implementing New Fishery Management Programs
Lawrence W. Van Meir1
ABSTRACT
Even though an "optimum" management program, in an economic sense, may
never be achievable in the management of commercial fisheries, changes can be
initiated which will allow individual governments to realize economic gains over
the status quo in harvesting common property fishery resources. These changes
primarily involve jurisdictional issues; country quotas for international fisheries;
accord between the Federal government and the states; and a within-industry
system for allocating fishing rights. A system of vessel licensing is described with
reference to the ultimate use of licenses on units of fishing effort.
The management of fisheries is intended
for the benefit of man, not fish, therefore,
effect of management upon fish stocks cannot
be regarded as beneficial per se.
Martin D. Burke?iroad
These words by Burkenroad were published
almost 20 years ago. This statement is a par-
ticularly cogent phrasing of the crux of the
question of fishery management for it raises
both the question of what benefits will be
sought in managing fisheries and the question
of to which men will these benefits accrue.
These are the two 64 dollar questions in the
area of fishery management policy.
In spite of Burkenroad's admonition that
the conservation of fish stocks per se cannot
be regarded as beneficial, and articles and
studies on the economic aspects of fishery
management that have appeared in the last
decade, most fishery management programs
remain oriented to the conservation of fish
stocks with no consideration of the economic
results that may be obtained. We still resort
to practices that either encourage the ineffi-
cient use of vessels, gear, and labor, or that
limit and impede the efficient use of these
economic inputs as a means of conserving
fish stocks. This does not mean that we do
not advocate conservation, but rather that
Staff Economist, National Canners Association.
we state more completely our objectives for
the conservation program.
Time is running out on us. With technologi-
cal development yielding a 3.5 to 4.0% annual
increase in the productivity of labor in the
economy, the fishing industry will find itself
in an ever increasing economic squeeze if
positive steps cannot be taken to include eco-
nomic objectives in fishery management. We
may continue to conserve fish stocks but it
will not be for the benefit of U.S fishermen or
fishing communities.
The entire problem of fishery management
of course stems from the common property
status of fishery resources. In the past, when
scientific evidence indicated that a particular
fish stock was being overfished, or in danger
of being overfished, the solution was to place
a quota on the fishery and/or add regulations
that either impaired the efficiency of fishing
gear, or in some cases required the use of
inefficient gear and fishing methods. The con-
sequences of such programs have been com-
pletely discussed in other articles and are
not the purview of this paper. Instead I want
to concentrate on the question of what must
be accomplished to change the situation, and
how it is to be done.
Obviously, common property status must
be replaced by explicit fishing rights. More-
over, in the process of conserving fish stocks
we must do so by bringing these specific fishing
rights in balance with the allowable yield of
the resource in a manner that not only permits
efficiency but also actually promotes economic
efficiency and technological development in
both the short and the long run. In short,
some system must be developed to limit the
amount of labor and capital employed in har-
vesting the allowable catch and at the same
time assure that the labor and capital is used
in an economically efficient manner.
Various economic advantages can result
from limited entry. Catch per vessel and fisher-
man employed will increase, thus increasing
wages and return on investment. The overall
value of fish landed may be increased in some
cases if the fishery management program re-
sults in a better marketing pattern. Labor
and capital employed in processing and dis-
tribution can be brought into better balance
with the volume of fish processed, thus realiz-
ing economic gains in these sectors.
A number of complex problems must be
overcome in order to realize the fruits of
limited entry. In the first place, the question
of fishery jurisdiction must be solved. In the
case of international fisheries, the solution to
jurisdiction will necessitate some system of
national quotas. Once national quotas have
been agreed to and established, then each
individual nation can institute its own pro-
gram for harvesting its quota. Jurisdictional
problems also exist between States and be-
tween the Federal government and the States.
Many fish stocks are fished by fishermen from
more than one State. In the case of pelagic
fish, the fish may migrate through the waters
of several States or between international
waters and waters under the jurisdiction of
several States. Moreover, a specific fishery
may involve waters under the jurisdiction of
several States and the Federal government.
Consequently, no one jurisdiction or authority
by itself can come to grip with the problem.
Certain enabling legislation will be needed at
both the Federal and State levels of government.
An important prerequisite to solving the
jurisdictional and legal questions will be a
thorough understanding of the concept of
limited entry, and the need for limited entry,
on the part of the fishermen, government of-
ficials, and congressional representatives.
Many individuals in commercial fishing today
are convinced of the necessity for a limitation
on the entry of labor and/or capital in those
fisheries that are fully exploited. However,
these individuals are still in a minority. Some-
how, the problems -we are facing in many of
our fisheries, and the effectiveness of limited
entry in dealing with these problems, must
be brought to the attention of the rest of the
commercial fishing industry in a meaningful
way.
One reason why many commercial fisher-
men are wary of limited entry proposals may
be because they have not been presented a
specific proposal to study and, hence, are
understandably cautious about embracing a
new concept without having some idea of
how they might fare under the new regime.
Thus, specific proposals likely will have to be
worked out and presented to industry as a
step in overcoming their resistance to the idea.
A second reason why some people are sus-
picious of the concept of limited entry is be-
cause they have formed the opinion that
limited entry is a scheme to put the govern-
ment in monopolistic control of fisheries to
enable them to extract either the monopoly
profit or economic rent from the fishery. Econo-
mists may have contributed to this image in
their writings on objectives of fishery manage-
ment.
From the viewpoint of the economist, pro-
gress toward a more rational fishery manage-
ment program will be a process of accepting
second best solutions. One of the first instances
in which we need to be willing to accept a
second best solution, at least initially, is on
the objective of a fishery management pro-
gram. If we accept the historical precedence
of "maximum sustainable yield" (MSY) and
seek agreement with the biologist on the im-
portance of harvesting the MSY as efficiently
as possible, it should be possible for the econo-
mist and the biologist to approach industry
with a common argument. The improvement
in returns to capital and labor in moving
from present management methods to a method
which achieves a reasonable degree of effi-
ciency in harvesting the MSY, will represent
the major share of total improvement in re-
turns that might result from any other man-
agement objective. As a starting point I would
suggest that the emphasis be placed on im-
proving the returns to labor and capital in
the fishery management program while de-
leting the argument for either seeking to maxi-
10
mize net economic return or economic rent.
If we are going to accept MSY as the basis
for managing a fishery, what economic ob-
jectives should we try to build into a new
fishery management program? As I mention-
ed earlier, the first objective would be to
seek the optimum amount of effort to harvest
the MSY, or at least a reduction in effort,
as a means of improving the catch and re-
turn per unit of effort in the fishery. In ad-
dition to this, we should also seek to build
into the management program some means of
insuring continued efficiency through time.
This means that over time the management
program must allocate the fishing right to
those economic resources that are most ef-
ficient in fishing. Thirdly, the program must
stimulate the development and adoption of
technological advancements in fishing. How
can these objectives be attained in fishery
management?
The system I foresee consists of a com-
mercial fishing license issued either by the
Federal government or by joint agreement of
the Federal government and the individual
States concerned. Each license issued would
represent a specific amount of fishing effort.
Initially, the number of licenses and fishing
effort would have to accommodate all vessels
and crews that have historically been employ-
ed in the fishery. However, as vessels were
retired from use, licenses would be cancelled
until normal attrition reduced the amount of
effort to the desired level. When the number
of licenses have been reduced to the optimum
number, a market for the licenses would be
allowed to develop. Licenses could be sold or
leased. Thus, the more efficient manager of
a fishing enterprise would be given the op-
portunity to lease or buy fishing rights from
the less efficient operator. The total number
of licenses could be adjusted over time as
productivity of fish stocks and technology
merited. In this manner, the management
program would work toward allocating the
limited fishing rights to those fishing firms
that were most efficient. Moreover, it would
now be advantageous to the fishing firm to
seek means of improving its efficiency and to
adopt new technology that improved efficiency.
Limitations could be placed on the number of
licenses that any one company could own or
control as a means of preventing someone
from developing a monopoly over the fishery.
A licensing scheme of this nature would
generate a certain amount of economic rent
in the fishery. This economic rent could either
be taxed away in the form of the license fee
or could be allowed to accrue to the resources
employed in the fishery. If the rent is taxed
away, it should be used for administration of
the management program and for research
on and development of the resource itself. If
the rent is allowed, either in total or in part,
to accrue to the resources employed in the
fishery, it would in turn be redistributed in
the economy through taxes, and would ulti-
mately be built into the cost of production
through the cost of fishing licenses. In either
case, this should not be a serious deterrent
to initiating a licensing system.
One of the most difficult problems to cope
with in the licensing system will be the adop-
tion of the effort base for the licenses. Licenses
could be based on vessels, tonnage of fish, or
an index of fishing effort. Vessels would be
the least adequate base for issuing fishing
licenses because of the variation in fishing
ability from vessel to vessel. It would certainly
seem technically possible to develop an index
of fishing effort and to define licenses in units
of fishing effort. Ideally the units of fishing
effort would be the link between the economic
aspects of harvesting and the biological model
used in assessing stock and yield character-
istics.
The steps toward a more rational system of
managing fisheries will no doubt be a system
of compromises. Perhaps the way to facilitate
these compromises with the least loss of time
and effort will be to include the commercial
fishing industry in the actual development
of the specifics of new management programs.
The situation is sufficiently urgent that this
should be given the highest priority in our
"new look" toward the oceans.
11
On the Utility of Bioeconomic Models for Fisheries Management '
GlULIO PONTECORVO2
ABSTRACT
Short run and long run biological and economic models are inevitably bound to-
gether in any comprehensive plan to manage commercial fisheries. While these
disciplines can be treated rigorously, political and social considerations can be
considered only generally and therefore on an ad hoc basis. Within this framework
long run models are useful primarily for goal setting. More work must be done in
developing short run models which will measure the immediate biological and economic
impacts of alternative management steps in addition to immediate political and social
ramifications. Emphasis would then be placed upon the economic sources of short run
instability, with an initial economic rationalization of the fishery providing the funds
for subsequent management and biological forecasting which will concentrate on ex-
tending management from a rationalized fishery at a given harvesting level to rational-
ized fishing at some optimum level.
BIOLOGICAL MODELS
The Yield from Ocean Resources
A 19th century view was that the high seas
fisheries were inexhaustible. We who are pre-
occupied by our failure to control our own
numbers and the possibility of worldwide eco-
logical disaster can at best regard such notions
as quaint. Nevertheless, certain implications
of the 19th century view of the oceans, with
its infinitely elastic aggregate supply curve
for fish, are worth considering here in our
attempt to understand the biologists' concept
of the maximum sustainable physical yield.3
The need for biological regulation of fish-
eries arose because the economic interests
involved in exploiting these resources became
aware, through rapidly declining yield-effort
relationships, that the supply of any particular
species was limited.4 Recognition of the exis-
1 I wish to express my thanks for helpful sugges-
tions to Dr. Brian Rothschild and Dr. EdilbertoSegura.
2 Professor of Economics, Graduate School of Busi-
ness, Columbia University.
3 The elasticity of the supply function may be with
respect to the inexhaustibility of one stock of fish or
as indicated below it may arise from a process of sub-
stitution of one stock for another.
4 The concept of limit is highly ambiguous. It may
be thought of as the maximum sustainable yield; the
tence of these limits led the biologist to in-
vestigation of the characteristics of particular
populations in order to find, if possible, a
level of exploitation that would be safe. A safe
level is defined as that maximum rate of ex-
ploitation that would preserve the stock, and
yet allow the catch to continue at the maxi-
mum rate through time.5 The imposition on
the stock of the appropriate (to achieve the
maximum sustainable yield) level of fishing
mortality involves the movement from one
long run equilibrium condition of the fish
population to another, with both equilibriums
considered stable.6
The development of the argument thus far
level of scarcity of fish beyond which it would not
pay to fish, i.e., an economic limit, the level that maxi-
mizes the net economic yield from the resource, or in
complete destruction of the stock.
5 For the development of an alternative view of
the relationship between the maximum sustainable
yield and the net economic yield, see (Schaefer, 1970a).
Schaefer also develops the necessary conditions for
adequate biological management (p. 9ff.)
H The first equilibrium is the natural or unexploited
state of the stock. The second is the condition of the
stock being exploited at the maximum sustainable yield.
There is a question about the relative stability of
the two equilibriums both per se and also because
of the effect of fishing effort on ecological conditions.
More simply, populations that are heavily exploited
by fishermen may show greater fluctuations in stock
size.
12
has already led to a conceptual difficulty.7
There is some evidence that in certain of the
populations which historically were overfished
the costs to society of rehabilitation of the
populations exceeded the benefits from the
subsequent higher yields which resulted from
the imposition of the biological control pro-
gram required to obtain the maximum sus-
tainable yield.
The debate between economists and biolo-
gists over the "success" or "failure" of the
International North Pacific Halibut Commis-
sion as an instrument in fisheries manage-
ment is an illustration of this type of difficulty.8
Aggregate Yield and Biological
Models of Particular Species
A more subtle set of difficulties is involved
in the interrelationships between an aggregate
yield function and the partial yield functions
derived by the biologists for particular popu-
lations. In the recent development of the eco-
nomic literature on fisheries the economists
7 More precisely, if biological overfishing has occurred,
and if a population is pushed well beyond the second
equilibrium point, does it enhance the material well-
being of the society to spend time and money (labor
and capital) to restore it to the second equilibrium?
8 It is also appropriate to point out that the Halibut
Commission has been an important training ground
for biologists interested in population dynamics and
fisheries management problems, a benefit not included
in the economic calculus. And also it is important
to note that there should be a clear distinction between
population dynamics as part of an academic discipline
and the administrative process of a social institution
such as a commission. The two activities have differ-
ent goals, but in the field the practitioners interact so
closely it is difficult for outsiders to observe the dis-
tinction. For a biologist's view of the Halibut Com-
mission's work, see Schaefer (1970a, p. 14). Another
illustration may be in the sea lamprey control pro-
gram.
Since the economist, like St. George, traditionally
defends the general welfare — the maximization of
Gross National Product — of the society against the
onslaught of particular interests it seems appropriate
to me to make this argument an economic one. What
is omitted, however, from the economic analysis is
adequate consideration of the place (role) of any par-
ticular species in the ecological structure. This omis-
sion, together with the usual inadequacy of the defini-
tion of the appropriate rate of social discount, is
probably sufficiently important to make one want to
proceed with care with a decision to fish out a resource.
It does not follow, however, that such a resource
should be "saved" by a costly biological rehabilitation
program.
have looked to the biologists' yield function
for a particular species as the source of the
production function upon which subsequent
economic analysis of the fishery can rest.
This development is logical in that the most
money for biological research has been spent
(for the most part) on those populations that
are economically interesting. More directly
put, the market demand for fish has been an
important determinant of the direction of
application of biological work.
In the historical development of ocean fish-
eries the interaction between market forces
and biological limits on the supply represented
by specific fish populations has been a typical
case of exploitation at an extensive margin. In
the long run, for certain fisheries (given a posi-
tive income elasticity of demand) operating on a
particular stock of fish, there has been a ten-
dency for the fishery to extend itself both geo-
graphically and temporally. If, after this ex-
tension has taken place, we assume that through
the imposition of biological regulation the
supply function for the stock in question be-
comes infinitely inelastic, economic adjustments
will take place and the fleet will tend to move
on to another similar stock or perhaps to a com-
pletely differentiated stock.
Thus we have the development of fisheries
management essentially on an ad hoc basis
as a response, often belated, to the expansion
of fishing effort against a finite supply of fish.
The continuous expansion of fisheries at the
margin (taken collectively) has resulted in an
aggregate supply curve which has been elastic.
World production of protein from the oceans
has risen and is expected to continue to rise.
The ultimate limit will be determined by a
trade off between the capacity of the basic
chemical biological processes of the oceans
to produce protein and the cost of collecting
it.9 At the same time the expansion in world
fisheries has tended to conceal the condition
of the specific stocks already exploited and
9 The inputs, the labor and capital utilized to bring
about the increase in aggregate output of fish are
not on the average highly specialized. Both are able
to shift from one -fishery to another. Vessel construc-
tion and reconditioning is a relatively easy process.
Labor immobility is a larger problem but it is less
acute in high seas than in inshore fisheries.
L3
complicate the problem of management (Stew-
art and Pontecorvo, 1970).
For a first approximation we may visualize
the fisheries management apparatus as re-
quiring several aggregate yield functions as
operational concepts. The plurality is neces-
sary because our concepts of aggregate yield
are ambiguous: it may be a regional concept,
a concept associated with a particular level
of the food chain, a concept that involves a
set of stocks that are either economically or
ecologically consistent in some way, etc. The
ambiguities involved derive from the inade-
quacy of biological knowledge of aggregate
yield, what the economists might call macro
biological ocean processes, and also from the
open access common property status of the
stocks. This latter situation permits each
nation state to define its output goals in terms
of its own tastes (exploitation of alternative
species) and then to proceed to bargain for
its share in purely nationalistic terms.
In the absence of adequate goals at the
level of aggregate yield, fishery administrators
are left dealing with partial equilibrium
systems, i.e., yield functions for particular
species, a circumstance which makes them
particularly vulnerable to pressure from eco-
nomic interests, fluctuations in the stocks and
the interaction between the two.
Population Dynamics
The population models developed by the
biologists are basically consistent with economic
models. Difficulties in the process of data col-
lection, statistical problems in fitting functions,
and the development of accurate forecasts are
all familiar ground. The question of the ade-
quacy and the cost of basic data does require
further comment. The observation of wild
populations is a time-consuming and costly
process. The fishermen are close observers of
the behavior of these populations and the com-
mercial catch is therefore an important data
source. Several types of bias may be involved
in using catch data, the most obvious being
that the data are restricted largely to what the
fishermen want to catch when they want to
catch it.
Perhaps more important, however, are
certain problems inherent in the structure of the
biological models. Biologists distinguish a
number of types of biological models, among
which are the logistic and the dynamic pool-
type.
The logistic model results in a parabolic
yield curve with a well-defined maximum,
and this has been utilized by many economists.
The maximum point on the yield curve repre-
sents the maximum sustainable yield.10 At
this point the stock is roughly half as abundant
as in its initial state or maximum size. Two
assumptions of interest to economists lie be-
hind this model, the first "that the rate of
increase in the stock responds immediately
to changes in population density; second,
that the rate of natural increase at a given
weight of stock is independent of its age (or
size) composition."
Naturally the adequacy of the assumptions
and the intrusion of exogenous forces affect
the adequacy of the model. But the matter
of greatest concern to the economist lies in
the first assumption. If there is not, as the
biological evaluation of this type of model
suggests there is not, an instantaneous ad-
justment between changes in population
density and population rate of growth the
economist for one becomes immediately inter-
ested in the time dimension of the adjustment
mechanism and the lag function that may be
utilized to describe it.11 Unless we limit our-
selves to consideration of long run equilibrium
solutions the integration of the biological
yield function into an economic system will
require the specification of the lag function.
For many purposes, particularly exposition,
it may be adequate to define the biological
system in terms of equilibrium points. How-
ever, if, as appears to be the case, the time
lags are significant, i.e., if they are of such
duration as to influence economic variables
(price, fishing effort, entry and exit), then
10 Holt, (1962 p. 141-142), has suggested that this
particular function may be flat topped which "simply
means that the biological facts are not very relevant
to determining where fishing becomes stabilized over
quite a range of variations in the situation." See
alsoGulland (1968).
11 Prof. G. Paujik has pointed out to me that the
time lags involved are a function of the species to
which the model is applied. In general tropical species
fit the first assumption fairly well but those in temperate
zones show much slower time rates of adjustment.
11
equilibrium models of this type lose a great
deal of their utility as a basis for regulation.
The short run characteristics of the economic
adjustment process will be discussed later in
this paper but at this point it should be clear
that the short run economic and biological
adjustment processes are inexorably bound
up with each other.
A second type of model, distinguished with-
in the biological literature, the dynamic pool,
presents an even thornier set of difficulties
for fisheries management.12 There are two prob-
lems inherent in dynamic pool models. The first
is that the maximum sustainable physical yield
is defined as a limit that can be reached only
by the expenditure of infinite fishing effort
(infinite cost). The second difficulty is more
analogous to those found in trying to maximize
the net economic yield from the resource. Vari-
ous degrees of overfishing and underfishing are
quantities of output which are deviations from
the eumetric yield curve or curve of best yield,
i.e., catching the appropriate size of fish for
that level of fishing effort. Deviations from the'
eumetric curve are controlled by making
changes in the selectivity of the gear utilized.
The necessary conditions for making these
gear adjustments is a knowledge of the con-
dition of the stock and a reasonable degree of
flexibility in the regulatory process. The ab-
sence of an operationally definable maximum
sustainable yield plus the necessity of adjust-
ing regulatory technique is a requirement on
management that is similar to the adjustments
in output level and inputs that would be re-
quired by changes in price and cost under
economic regulations.
Forecasting with Biological Models
In the commercial fisheries the forecasting
problem is a mix of the complexity of the life
cycle of the individual species and the avail-
ability of resources to carry on the necessary
biological research programs. For the bulk
of the populations fished the effort devoted
to biological research is simply insufficient
to provide sophisticated forecasts. And while
at first glance the cure for this particular in-
adequacy would appear to be simple it is not.
In general it seems unlikely that sufficient
funds for meaningful broad-based biological
programs can be obtained except from the
income generated by the fisheries themselves.13
If this hypothesis is approximately valid then
it suggests that economic rationalization
(realization of the potential net yield) is a
necessary condition for achieving the level of
funding of biological research sufficient to
allow the development of dependable forecasts.
The type of forecast made depends upon
the behavior characteristics of the species.
The Bristol Bay red salmon fishery has been
studied intensively by three agencies; the
National Marine Fisheries Service, the Alas-
kan Department of Fish and Game, and the
Fisheries Research Institute of the University
of Washington.
Table 1 presents a summary of salmon runs
and a rough measure of the accuracy of the
forecasts for the recent decade. Despite the
investment in research and the heavy payoff
for accurate forecasts in the Bristol Bay
fishery it is clear that the existing forecasts
are not completely satisfactory.14 Furthermore,
even if forecasting in this fishery was 100%
accurate the instability on the supply side
would (does) cause severe economic problems.
Few other species present the forecasting
problems of the red salmon.15 In the simpler
cases it is possible to estimate the stocks,
and then assign under biological regulations
catch limits in some form. In subsequent
time periods the limits may be adjusted to
allow for errors in estimation of stock size.
12 From the viewpoint of the analysis of the biological
condition of the stock it appears to have certain ad-
vantages over the logistic model.
13 This follows from the common property status of
the resource which means that the rate of return to
the firm or the nation on investment, in research is
zero in long run equilibrium.
14 Crutchfield and Pontecorvo, (1969), especially
Chapter 7, develop the rationale for the high payoff
for accurate forecasts in Bristol Bay.
15 Schaefer (1970a), discusses the approach to the
maximum sustainable yield that may be utilized with
species such as the Peruvian anchoveta, halibut, etc.
Even within these, more stable populations there is
room for substantial disagreement about the appropri-
ate level of yield. For further examples see (Schaefer,
1967; 1970b and Segura, 1972).
15
Certain deductions can be made about the bio-
logical and economic implications of this lagged
response approach to maximum sustainable
yield. A priori, it appears that its economic
viability is a function of the stability of the
population. And since we have already sug-
gested that those populations which are fished
heavily, i.e., those exploited close to or even
beyond the maximum sustainable yield, tend
to show greater fluctuations, it follows that
accurate forecasts of short run supply changes
will require a continuous extensive biological
research program.16
Political Science and Sociology
The discussion thus far has been aimed
at understanding the nature and limitations
of the maximum physical yield as a biological
construct and as a tool in fisheries manage-
ment. The technical side of the problem is
however just one part of it. As one distin-
guished fisheries administrator put it:
I wish to inquire whether social and
political problems are included within the
scope of fisheries economics. If so, we may
be able to arrive at a fairly broad and com-
prehensive view on matters of fishery regu-
lation. If not, then I think they must be treated
as separate aspects (McHugh, 1962).
Neither the economist nor the biologist
will, based on what can be learned from the
individual disciplines, accept responsibility
for the social and political problems associ-
ated with the fisheries. They have both been
guilty of implying that social and political
objectives will best be met by choosing the
alternative they espouse. However, since social
and political objectives themselves are apt
to be as disparate as are the biological and
economic, the debaters have grasped at only
those aspects of social and political policy
that have best fit their needs at the moment.
Table 1. — Run of sockeye salmon to Bristol Bay,
1960-1970.*
16 One alternative would be to limit fishing effort
to that sufficient to harvest only the lower portion of
the range of variation in the stock. This would give
a small output at low cost with little or no require-
ment for investment in biology. Since, however, this
would be a disequilibrium situation with long run
excess profits in all probability it could not be sus-
tained in the face of the economic pressures to expand.
Millions o
ffish
1960
36.3
1961
18.0
Range 7.7-53.1
1962
10.4
Median 17.5
1963
6.8
Mean 20.8
1964
10.7
Coefficient
of variation 74%
1965
53.1
1966
17.5
Approximate forecasting
errors:
1967
10.3
1960-1970+40%
1968
7.7
1969
18.5
Anticipated accuracy of
forecasts in near future:
1970
39.6
± 20% in 4 years out of 5
±50% in 5 th year
I am deeply indebted to Dr. Donald E. Rogers of the Fisheries
Research Institute of the University of Washington for assem-
bling the complex data on the runs and forecasts of Bristol
Bay and Western Alaska Sockeye (from which Table 1 is ex-
cerpted). As noted in the text, the purpose in presenting these
figures is to emphasize the year-to-year variations in supply.
They have not dealt in a rigorous analytical
way with these problems.
In this it seems fair to say that the biolo-
gists have been the political realists while
the economists, to the extent that they have
dealt with the question of labor mobility,
income distribution, and the impact of barriers
to entry on the scale of enterprise, have been
closer to social realities.
Economists have insisted correctly, in my
judgment, in their discussion of the problems
of fisheries, that the general economic wel-
fare of the state and the individuals in it are
best served by maximizing the net economic
yield from the resource. Their occasional
willingness to temporize their position arises
for the following reason. For species of fish
with a high unit value such as lobster, red
salmon, etc., the discrepancy between the
maximum sustainable physical yield and the
net economic yield is not apt to be very
large, the former being a second best solu-
10
tion that does not deviate significantly from
the economic optimum.17
Of greater significance is the political appeal
of the idea of achieving the maximum sus-
tainable physical yield. The economist cate-
gorically rejects the idea that it is "good" to
maximize the output of any commodity just
because it is physically possible. To the poli-
tician negotiating fisheries agreements both
nationally and internationally it is important
to be able to state that the agreement makes
possible the utilization of all the fish avail-
able for all time, none will be "wasted." The
simplistic political argument runs as follows:
the production of food, particularly protein
food, is good. The maximum sustainable
physical yield is the most food that can be
obtained. Any other definition of optimum
output such as the net economic yield would
either represent less food (a waste) or if it
was greater than the maximum physical yield
it would be a threat to the stock.
The danger in this political exposition of
the problem is, of course, that it conceals the
underlying complexities of the biological pro-
cess as well as the interaction between those
processes and economic variables.18
THE NET ECONOMIC YIELD AND LONG
RUN PARTIAL EQUILIBRIUM MODELS
Economic and Biological Models
As a first approximation we may assert
that the utility of economic models in fisheries
management is symmetrical with the biolog-
ical counterparts upon which they rest.19 The
17 There is for the noneconomist a possible con-
fusion here. Both the output that will maximize the
net economic yield from the resource and the output
that will maximize the physical yield in the long run
are points derived from the same biological yield
function. However, the maximization of the net eco-
nomic yield requires, given the common property status
of fish stocks, an economic control mechanism as well.
18 -\ye have ignored any discussion of what has
been referred to as social problems. A good illustration
of the interaction of all the forces can be found in the
Norwegian coastal fisheries. For reasons that are
political, social, and national, the Norwegian govern-
ment has seen fit to subsidize coastal fisheries. These
subsidies have been indirect: education for dependents,
health care, transportation, etc., and direct: price
supports for raw fish, vessel construction subsidies,
etc. A key objective of this policy is to maintain the
population living along the coast of western Norway,
argument developed about the inadequacy
of partial models and the difficulty of utilizing
long run equilibrium systems for manage-
ment decisions are applicable to both biology
and economics.
By ignoring underlying definitional prob-
lems as well as those which result in short
run fluctuations in output, the biological con-
cept of the maximum physical yield can
present a facade of stability. No such facade
exists with the net economic yield. The ap-
propriate level of output is defined by the
interrelationship between market price and
costs. Since the price of most fish products
is determined in markets that are describable
as workably competitive, it is clear that when
the physical yield function is transformed
into a revenue function the appropriate level
of output will shift in response to price
changes.20
The voluminous literature on the economics
of uncertainty is suggestive of the magnitude
of the administrative and political problems
it creates for the regulatory mechanism, and
also the lengths administrators will go to
minimize it. But how much uncertainty actu-
ally would be created by the imposition of
regulatory practices aimed at maximizing
the net economic yield? The major sources
of disequilibrium in most fisheries appear to
be attributable to two forces, short run vari-
ations in the supply of fish and shifts in the
demand function. In this context it is impor-
tant to distinguish between stability in the
demand function and stability of price. Within
the economic models price changes are caused
by short run variations in supply which cause
the traditional farmer fisherman. Other considerations
are military, economic (balance of payment), and
political in that it is important to participate in the
exploitation of stocks as a claim against any future
regulation that might involve national quotas, etc.
19 There are of course important differences. For an
exposition of certain properties of variable propor-
tions diminishing returns unique to the fisheries, see
F. W. Bell, and E.W.Carlson (1970).
20 See Crutchfield and Pontecorvo, (1969, pp. 28-
88). The usual assumption is that cost functions are
linear and stable. This follows from the size of labor
markets, the general availability of the type of capital
instruments required in most fisheries, and especially
from the small scale of most fisheries encompassed by
the partial equilibrium systems analyzed.
17
a change in price and are therefore the pri-
mary source of uncertainty. This effect will
be dampened if the demand function is highly
elastic, a condition that seems applicable to
many fisheries.
In the long run, however, the situation is
different. The dynamics of short run supply
changes continue through time but in the
long run the demand function tends to be
responsive to income changes and therefore
shifts to the right, adding to the degree of
uncertainty.
Given the high level of uncertainty inherent in
fisheries, the workability of fisheries as in-
dustries is highly dependent on their economic
efficiency, i.e.. their ability to operate profit-
ably and not dissipate the rent from the re-
source among redundant inputs.
Sources of Short Run Economic Instability
If, for the moment, however, we limit our
argument to the short run and we hypothe-
size that the source of instability is on the
supply side then it is correct to say that
regulations aimed at either maximizing phys-
ical yield or net revenue are not significantly
different from each other in terms of the level
of uncertainty involved. An analogy may be
useful in putting this problem in better per-
spective.
Academic economists are virtually unani-
mous in the opinion that some degree of flexi-
bility in exchange rates is desirable. Central
bankers and to a lesser extent businessmen
are generally opposed to the creation of the
uncertainty that flexible rates would bring.
At the heart of this debate are two different
views of the stability of the underlying system.
If equilibrium not disequilibrium is the norm
then problems involving the short run adjust-
ment process are minimal. But if short run
variations are inherent and important in 'the
system then the regulatory process must be
flexible to be consistent with the dynamics of
the short run.
The short run economic adjustment process
in fisheries, unlike certain industries, is par-
ticularly responsive to changes in market con-
ditions. Common property and easy entry, a
relatively low level of specialization of inputs,
and the possibility of shifts of economic units
between fisheries, have combined to create
this condition.
Let us define the short run to be a period
sufficiently brief to exclude new entry. By
new entry we mean that fishing effort would
be carried on by units of capital and labor
with no previous experience in the fishery
in question. Fisheries (except in initial growth
stages or periods of significant technological
change) tend to be characterized by excess
capacity. In these circumstances short run
shifts in the price/cost ratio brought about
by changes in either supply or demand can
generate wide swings in fishing effort. This
effort may come from greater productivity
(longer hours, better organization, harder
work), by the activation of units previously
participating in this fishery but currently "on
the beach," or by the response of economic
units that work in several fisheries on a part-
time basis to the enhanced profit position in
this one.21 If the excess profits continue in
the face of the increase in fishing effort, new
entry will take place fairly rapidly.
The labor and capital utilized in many
fisheries may be characterized as having three
elements, one is a core of labor and capital
that is primarily identified with the particu-
lar fishery in question and this core may ex-
pand and contract its efforts in response to
market conditions but it lacks the mobility
required to shift rapidly to alternative fisheries.
The second element is a stock of standby
capacity, either currently employed elsewhere
or unemployed, that can and does respond to
changes in profit prospects. These two com-
ponents, possibly each individually if restric-
tions on productivity are considered, represent
more capacity than is needed to harvest any
average level of catch and perhaps even more
capacity than is required to land the upper
limits of the frequency distribution of the
abundance of the stock.
21 These responses are not symmetrical, i.e., fishing
effort increases more rapidly in response to profit op-
portunities than does exit to a reduction in earnings.
Inertia, the possibilities of windfall gains and the
inadequacy of the forecasts of short run supply all
contribute to the asymmetry. It is particularly im-
portant to note that we have not mentioned changes
in technology. In many fisheries the relationship be-
tween technological change and fishing effort is cir-
cumscribed by the regulatory process.
IK
The third element is the continuous threat
of entry. If the excess profits observed in one
time period continue, or if they are expected
to continue, entry will take place. Expecta-
tions and the competitive illusion play a role
in all industries but the great flexibility in
the capital instruments employed in fisheries
tend to make the interaction between market
conditions, expectations, and capacity par-
ticularly close (Pontecorvo and Vartdal, 1967).
Supply fluctuations, excess capacity, the
rapidity of responsiveness to changes in the
market, and the influence of expectations all
contribute to short run instability in fisheries.
Control of capacity, improvements in fore-
casting supply in order to reduce uncertainty,
plus recognition that capacity sufficient to
capture some average level of catch less than
the maximum sustainable yield may be ap-
propriate, are all elements in a management
program geared to meeting the conditions im-
posed by the short run dynamics of fisheries.
Long Run Equilibrium
If short run economic objectives are defin-
able in the terms indicated above it is ap-
propriate to inquire next about the long run
equilibrium conditions. Economic analysis of
fisheries has accepted as given the biological
yield function for the species in question, as
well as the usual assumptions of static equi-
librium analysis of full employment and factor
mobility. In these circumstances the condition
of Pareto optimality is roughly fulfilled if the
policy recommendations (essentially creation
of a set of regulations aimed at maximizing
the rent of the resource and in all probability
requiring barriers to entry) required to ration-
alize the fishery are met. Within the frame-
work of economic analysis (maximization of
Gross National Product) this is a necessary
and sufficient condition for making the maxi-
mization of the net economic yield the ap-
propriate goal of fisheries management. Any
alternative is less satisfactory in that it will
result in a lower level of material well-being
(GNP).22
22 A crucial assumption is that the opportunity
cost for labor is positive. Most of the attacks on the
concept of economic regulation of fisheries assert the
contrary. Perhaps the needed empirical investigation
of this point could start with a classification such as
suggested in Approaches to Fisheries Management.
Attacks on this goal have come from two
sources, biologists and fisheries administra-
tors, and also from within the economics pro-
fession. The position of the former group rests
in large part, in my opinion, on a funda-
mental misconception concerning the mean-
ing of economic optimization. Economics is
not sufficient to explain (or optimize), par-
ticularly in the short run, the entire set of
variables involved in a fishery. The economist
accounts for the social problems by his as-
sumption of full employment and factor mo-
bility. He does not normally account for po-
litical factors except indirectly in his under-
lying assumption of human rationality which
tends to force the political preferences into
the same ordering as the economic.
A bioeconomic position dominates thinking
about fisheries management simply because
there is no body of social or political theory
sufficiently powerful (relative to welfare maxi-
mization in economics or population dy-
namics in biology) to force a modification of
either the biological or economic position.
In these circumstances, which appear unlike-
ly to change in the foreseeable future, political
and social considerations can only be con-
sidered on an ad hoc basis. More specifically,
it is normally true that the biological optimum
and economic optimum are consistent with
each other in that both will protect the stock.
The economic goal is more general and there-
fore preferable in that in addition to protect-
ing the stock it also provides the maximum
economic benefits to society. Any deviation
from the economic maximum involves there-
fore a cost, a cost measurable in terms of
output foregone. Nothing in this argument
suggests that ad hoc reasons are not sufficient
grounds (given the weakness in the two under-
lying assumptions in certain circumstances)
to make an alternative objective (political,
military, social, etc.) either the primary or a
subsidiary goal of fishery management.
In this circumstance the economist's primary
concern would be to calculate the cost of the
alternative. The latter calculation presupposes
that the alternative can be specified, a con-
dition that is seldom met. What tends to
emerge as the management goals in fisheries
under the long run equilibrium condition that
dominates today's thinking is an unspecified
19
mix of all factors including purely adminis-
trative considerations.
The attack within the profession has raised
an appropriate question about the implica-
tions of limits on entry for Pareto optimiza-
tion. Another position also has been advanced
based on an assumption questioned through-
out this paper that the maximum physical
yield is so fundamentally sound in the oper-
ational sense that its utility as a tool out-
weighs its defects.23
In the competitive model there is no limit
on entry beyond that provided by what Knight
has called the "social function of ownership."
In the fishery with the resource being common
property the objective of the economist in
calling for a limit on entry is to provide the
ownership function while retaining the Pareto
optimum conditions inherent in the competi-
tive model.
Two questions may be raised about this
goal and the procedures necessary to achieve
it. Will the barriers to entry result in a situ-
ation that goes beyond competition, i.e., does
the creation of property rights just restore
the conditions that would be found with pri-
vate property operating under competition
or does it also imply the creation of monopoly
power so that buying and selling is no longer
on a competitive basis? The second question
is an integral part of the first. Does the estab-
lishment of barriers to entry and the subse-
quent economic regulation of the fishery in
the public interest require the creation of a
regulatory mechanism so costly and complex
as to be self defeating?24
Economic theory does not provide an
23 See Wantrup (1970, p. 18): "While maximum sus-
tainable yield constitutes a relevant, operational, and
noncontroversial objective of conservation policy, this
is quite different for the objective of 'maximum net
economic yield' • even if its realization through
limitation of entry could be agreed upon by the fishing
industry." Also (Wantrup, 1962, p. 292): "My approach,
therefore, would be set more modest regulation goals
which would concern themselves more with the re-
source base than with rent. We are dealing then
with matters we can measure. If we try to maximize
rent as a policy goal, then we get into an area where
I for one would put out a 'caveat' sign."
24 Virtually the entire literature on the economics
of fisheries has commented on these two questions
albeit in a not very satisfactory manner. For the
details on control plans to limit entry see Sinclair
(1962) and Royce et al. (1963).
answer to these questions. It is possible, how-
ever, based on experience with the social con-
trol of industry, to advance certain tentative
hypotheses. No control mechanism based solely
on biological considerations is workable in
the long run in the face of economic and
other pressures. Therefore, the cost of the
control mechanism is ultimately a joint bio-
economic cost. Even in these circumstances
it may be that the costs of control are dis-
proportionately large relative to the value of
the resource left unprotected.25
A second hypothesis is that if the cost of
bioeconomic control is not excessive then the
capacity to regulate the fishery in the public
interest, i.e., to preserve the stock and prevent
the emergence of significant monopoly power,
is well within the power of regulatory pro-
cesses and tax arrangements that have proved
themselves workable in other circumstances.
APPROACHES TO FISHERIES
MANAGEMENT
A major constraint in the development of
a consistent monetary policy is that the mone-
tary system itself is continuously evolving.
The analogy seems applicable to regulatory
problems involving ocean resources. The pat-
tern of resource exploitation in the oceans
and the law of the sea are changing rapidly.
In addition, military uses of the oceans, while
not a new phenomenon, are being transformed
and at the same time the very existence of
the ocean in the way we have known it is
threatened by the effects of the population ex-
plosion and the rising level of real income.
Furthermore, our capacity to deal effectively
with ocean living resource problems is limited
by the inadequacy of our scientific knowledge
of life processes in the ocean, the generally
weak economic condition of fisheries, and the
nationalistic interests involved.
It is beyond the scope of this paper to go
into these structural questions. It is clear,
however, that in the future the organizations
by which fisheries are to be regulated must
be prepared to negotiate the basic issues of
control of the environment and the priorities
25 It might be desirable to protect the resource for
other reasons, i.e., because it was unique, etc.
liO
appropriate to multiple use situations in the
oceans with external forces. In these activities
the strength of their bargaining position will
depend heavily on their having rationalized
both the' economic and biological sides of the
fisheries. Regardless of what potential eco-
nomic yields may be or what social pressure
for employment is present, a realized net
economic yield of zero from fishing does not
provide an adequate base for defending ocean
space for commerical fisheries in competition
with the oil industry, recreational use, power
generation, etc.
In recent years progress has been made
by governments, fisheries commissions, and
academic researchers in the analysis of fish-
eries problems. What are the elements of this
analysis that may be utilized to help reorient
our approach to fisheries management?
The long run partial equilibrium systems
constructed thus far make a major contribu-
tion by an exposition of the problems in static
terms. It is clear, however, that they are in-
adequate for resource management. In most
circumstances they do provide limits within
which the regulatory process may operate.
In the analysis of particular species the dis-
tinction between the net economic yield and
the maximum sustainable physical yield is
subject to empirical verification depending on
the unit value of the species, but in any case
it is a second order question. In any set of
priorities established for fisheries manage-
ment the first is to move toward meeting the
criteria of economic efficiency, probably by
establishing limits on entry. Once the fishery
is rationalized then the solution to the prob-
lem of the appropriate level of output should
be greatly simplified. This follows from the
nature of the adjustments that must be made
in the process of economic rationalization.
The administrators will be forced to consider
simultaneously the appropriate amount of
fishing effort (amount of inputs) relative to
the forecast of the frequency distribution of
supply and the impact of productivity changes
on fishing effort. Once the fishery is defined
in this way, the economic implications and
advantages of various levels of output will
be more apparent to all and self interest,
which today drives producers toward over-
fishing the resource, will move them toward
limiting the catch to maximize the net yield
from the resource.26
Given recognition of the long run bio-
economic limits, the adequacy of the regula-
tory mechanism may be evaluated in terms
of how well it handles the short run maxim-
ization problem and its success in restructur-
ing the fishery from a disequilibrium position
(excess capacity) to one of equilibrium. This
latter will require evaluation of the possibil-
ities inherent in aggregative yield functions
and clarification of the goal of a workably
competitive structure for the fisheries. The
rents captured in the rationalization process
are available to finance this transition. In
these Utopian circumstances the essence of
the internal regulatory mechanism will be
found in the interaction between changes in
biological supply, prices, and technology.
LITERATURE CITED
BELL, F.W., and E. W. CARLSON, 1970. The Pro-
ductivity of the Sea and Malthusian Scarcity. Work-
ing Paper Number 48, National Marine Fisheries
Service. Draft Manuscript.
CRUTCHFIELD, JAMES A., and GIULIO PONTE-
CORVO, 1969. The Pacific Salmon Fishery: A Study
of Irrational Conservation. Published for Resources
for the Future, Inc., The Johns Hopkins Press, Balti-
more.
GULLAND, J. A., 1968. Population Dynamics of the
Peruvian Anchoveta. FAO Fisheries Technical Paper
Number 72. Rome.
HOLT, S. J., 1962. Comments made in discussion
of Dickie, L. M. Effects of Fishery Regulations on
the Catch of Fish. In: Economic Effects of Fishery
Regulation, R. Hamlich, ed. FAO Fishery Report
Number 5. Rome. pp. 141-142.
McHUGH, J. L., 1962. Comments made in discussion
of Dickie, L. M. Effects of Fishery Regulations on
26 This statement is more than a pious hope but
less than a certainty. Its validity depends in part on
the nature of the frequency distribution of catch among
the participants in the fishery, i.e., if the fishery were
the property of a monopolist he would operate at the
level of the net economic yield. Only under certain
assumptions will this be true of the behavior of a set
of competitors. Their recognition of the desirability
of maximizing aggregate net revenue will come as a
process of education as studies of the characteristics
of the fishery reveal the advantages inherent in various
alternatives.
21
the Catch of Fish. In: Economic Effects of Fishery
Regulation, R. Hamlich, ed. FAO Fishery Report
Number 5. Rome. p. 147.
SEGURA, EDILBERTO E., 1972. Optimal Fishing-
Effort in the Peruvian Anchoveta Fishery. This publi-
cation.
POXTECORVO. GIULIO, and K. VARTDAL, JR.,
1967. Optimizing Resource Use: The Norwegian
Winter Herring Fishery. StatsjeSkonomisk Tidsskrift
Number 2.
SINCLAIR, SOL, 1962. License Limitation — British
Columbia. In: Economic Effects of Fishery Regula-
tion, E. Hamlich, ed. FAO Fishery Report Number 5.
Rome. pp. 306-328.
ROYCE. W., et al., 1963. Salmon Gear Limitation in
Northern Washington Waters. University of Washing-
ton Publication in Fisheries. New Series. 11(1). Seattle.
SCHAEFER, M. B.. 1967. Dynamics of the Fishery for
the Anchoveta off Peru. Boletin, Instituto del Mar
del Peru, Callao.
SCHAEFER, M. B., 1970a. Investigation, Conserva-
tion and Management of the Fisheries of the High
Seas. Paper presented at the Preparatory Conference
on Ecology and Science Policy, April 20-26. The
Center for the Study of Democratic Institutions, Santa
Barbara, California.
SCHAEFER, M. B., 1970b. Men, Birds and Anchovies
in the Peru Current — Dynamic Considerations. Trans-
actions of the American Fisheries Society. 99(3). p. 461.
STEWART, C. S., and GIULIO PONTECORVO, 1970.
Problems of Resource Exploitation: The Oil and Fish-
ing Industries. Chapter I in Ocean Enterprises, The
Center for the Study of Democratic Institutions, Santa
Barbara, California.
WANTRUP, S. V. CIRIACY, 1962. Comments made
in discussion of Pontecorvo, Giulio, Regulation of
the North American Lobster Fishery. In: Economic
Effects of Fishery Regulation, R. Hamlich, ed. FAO
Fishery Report Number 5. Rome. p. 292.
WANTRUP, S. V. CIRIACY, 1970. The Economics
of Environmental Policy. Paper presented at the
Preparatory Conference on Ecology and Science
Policy, April 20-26. The Center for the Study of Demo-
cratic Institutions, Santa Barbara, California.
22
Multiple Objectives for Marine Resource Management1
R. Bruce Rettig2
ABSTRACT
Management decisions suggested by recent bioeconomic models have been largely
disregarded by fishery managers. This negligible impact may be due to error on the
part of management, an incomplete grasp of the role of noneconomic objectives, and/or
the possibility that more sophisticated economic models might yield markedly different
results. More sophisticated models are suggested which consider the problem of second
best, risk and uncertainty, transaction and adjustment costs, and income redistribution.
Creation of analytical systems amenable to treatment of noneconomic variables along
with economic variables is suggested.
During the past two decades, a growing body
of economists has been articulating a rationale
for management of ocean fisheries which is
based upon the principle of maximum sustain-
able net economic yield. The usual paradigm
emphasizes the lack of clearly denned property
rights and arrives at a conclusion of a need for
limited entry, most commonly suggested through
a system of licensing and/or taxes. While the
better analyses have often hedged their con-
clusions with a set of qualifications, even these
balanced policy programs are rejected by
authorities actually responsible for fishery
management.
That articulate arguments from a respected
cross-section of the economics profession con-
tinue to carry only minor weight with their
intended audience is quite disconcerting. This
paper consists of an examination of two possible
reasons for the treatment of the bioeconomic
models to date. The first possibility is a diver-
gence between what public authorities consider
appropriate objectives to pursue and the assump-
tions of goals implicit or explicit in the bio-
economic models. The second possibility is that
bioeconomic models are incompletely specified
and that more complete models would be better
received. A third possibility that will not be
handled in this paper is that existing analysis
is correct and that all that remains is to
1 Oregon Agriculture Experiment Station, Technical
Paper 2996. Research for this paper was supported by
N.S.F. Institutional Sea Grant. GH-45.
2 Department of Agricultural Economics, Oregon
State University.
educate the resource managers on the merits of
implementing the correct suggestions already
available.
THE ELUCIDATION AND
LEGITIMIZATION OF SOCIAL GOALS
The characteristic of the fishery which lies
at the root of the problem is the lack of clearly
defined property rights over the fishing ground.
The severe depletion of such fisheries as the
Pacific halibut fishery and the sardine fishery
off the California coast stand as stark testimony
to the value of property rights. That the problems
associated with fisheries can be easily related to
the absence of property rights is seen by con-
sidering the central economic functions of
property as set forth recently by Bjork (1969,
p. 65):
First, it provides incentives for the creation and
improvement of assets. Second, it provides incen-
tives for efficient control of existing assets. Third,
it rations the use of scarce assets to ensure that
they will be used for those purposes which society
values most highly.
Bjork argues that property rights exist largely
because stable market-oriented societies value
the performance of these functions so highly.
Investment incentives, efficient allocation of
fishing effort and living fish, and distribution to
him who values the resource most highly are
indeed the central objectives behind the bio-
economic models of current interest. These are
not apparently the sole objectives of the societies
23
whose mandates the resource managers must
have in order to perform their duties viably. It
is useful to lump these other objectives into
two categories — equity in the distribution of
income, and noneconomic objectives. Equity of
income distribution is relevant both in con-
straining changes from the status quo and in
defining acceptable distributions of income. This
latter can be illustrated by the popularity
among some authors of giving some supra-
national agency control over all marine resources
and by current claims of U.S. nationals to the
natural right of the citizens of a country over
fish which swim over "their" continental shelf.
Noneconomic objectives are important in the
arguments of some concerning the inherent evil
of blocking entry to a fishing ground. Of course,
several objectives can be classified more than
one way. The repercussions of sudden, unexpected
unemployment of fishermen can be called adjust-
ment costs or could be appropriately titled
sociological phenomena.
In any case, it is not appropriate for econo-
mists to identify conclusions of their positive
models with normative policy proposals. Society
is not composed of economic men (Boulding,
1969). Rather the economist must first try to
maximize economic gain subject to noneconomic
constraints. However, when noneconomic ob-
jectives are not postulated at unique target
levels, the tradeoffs between economic and non-
economic objectives must be considered. I will
return to this practical problem after consider-
ing some problems of analyzing economic
objectives.
SECOND-BEST FISHERIES, OR
WHEN IS AN OPTIMUM NOT OPTIMAL?
The application of the theory of common
property resources to ocean fisheries leads
inexorably to two conclusions. First, it is
possible to observe an allocation of human and
capital resources in fishing with social marginal
products which are negative. Second, the optimal
allocation of resources is one in which social
marginal revenue product equals social oppor-
tunity cost of factors used to catch fish. This
is to say that less fishing effort should be
employed than would be required to harvest
the maximum sustainable yield unless the
alternative use of the marginal resources is
valueless.
Crutchfield and Pontecorvo (1969, p. 35) have
pointed out that the need for intervention in
fishery management hinges upon the assump-
tion of competitive behavior by the downstream
purchasers of the resource.
A monopsonist would impose a rational solution
on the fishery, i.e., he would capture the rent by
offering sellers a price that would permit only the
most efficient exploitation of the resource to take
place, and the malallocation of resources, which
results from the combination of free entry and
common property, would be avoided. If, in turn, the
product market in which he sells is highly competi-
tive, monopsony could provide a near-optimal level
of output and real costs.
While Crutchfield and Pontecorvo go on to
point out that the industry which purchases
fish from fishermen is a competitive oligopsony
and does not lead to a socially optimal solution,
their argument still holds qualitatively. When
an -oligopsonistic industry faces a group of
competitive sellers the price paid for the output
of the competitive sellers is less than the social
valuation for an incremental unit, i.e., the social
marginal revenue product exceeds the market
price. This divergence alters the optimal inter-
vention in the market structure of fish buyers
and alters the optimal fishery management
scheme, assuming control over the two cannot
be coordinated.
If the market between fishermen and fish
processors is not perfectly competitive, the
correction of fisheries resource allocation ignor-
ing this fact could actually misallocate resources.
Equating the social marginal product of fishing
effort multiplied by observed market price to
social opportunity cost of factors used to catch
fish would secure a level of fishing effort less
than the one where the true value of social
marginal product equals social opportunity cost
of factors. If the demand for fish is elastic
throughout the relevant range, the correct
solution would still occur prior to attainment
of maximum sustainable physical yield. The
simple maximum sustainable physical yield
criterion is still in error and the error is still in
the same direction, but the error is smaller than
the simpler analysis. Thus the efficiency loss
from using the physical rule is smaller than
previous analysis has suggested.
24
On the other hand, if a social suboptimum
does not exist in an ocean fishery, the dis-
solution of oligopsonistic structure in the fish
processing industry would further misallocate
resources. Assume that effort had entered until
the effort level which would secure maximum
physical sustainable yield had been exceeded.
Breaking up the monopsony and allowing the
higher social marginal valuation to be revealed
would further overcapitalize the fishery and lead
to a lower physical sustainable yield.
While the argument has not proven that
suboptimization in either fisheries or fish
processing industries will lead to a misallocation
of resources, the possibility of such an event
may lead to a desire to gather more information
about upstream-downstream linkages in these
industries before taking large irreversible policy
actions. It also may comfort those who wish for
maximum physical sustainable yield in the
fisheries and those who are hesitant about
breaking up what appears to be an oligopsonistic
fish distribution chain in the near future.
An extension of the preceding analysis may
lead one to observe that imperfectly competitive
factor markets from one side allows the theoreti-
cal possibility that the unorganized fishermen
may be able to organize and bargain collectively
for higher prices without third-party effects.
Nevertheless, the tenor of this piece has sug-
gested that third-party effects may possibly be
involved and that the public interest may imply
that this bargaining should be observed by
representatives of the third parties, such as
the Government.
Directly parallel to the problem of second
best in the economy is an ecological second best.
If two species are competitors in the ocean, an
increase in the sustainable yield of one may
reduce the sustainable yield of the other. Like-
wise, increasing the sustainable yield of a
species may increase the sustainable yield of
its predators and/or decrease the sustainable
yield of its prey.
RISK, UNCERTAINTY, AND
INTERTEMPORAL CHOICE
Anthony Scott (1962) has pointed out and
Plourde (1970) has recently given a concise
proof that looking for solution values on steady-
state curves (which these sustainable yield
curves are) is akin to ignoring the existence of
a positive rate of time discount. This becomes
obvious when one assumes an infinite rate of
time discount and immediately finds the in-
stantaneous yield curve to be the only relevant
one for consideration.
When one backs away from steady state
solutions, tries to pose the relevant horizon
curve, and tries to determine the role of time
discount, one realizes that he is postulating
expected values of a probability distribution of
possible yield curves with only a vague aware-
ness of the yields in short and long run which
will occur. It may be useful to separate problems
of uncertainty into two categories. On the one
hand, demand for particular fish species is
uncertain. It is one thing to extrapolate desires
for particular fish species into the near future.
It is quite something else to fail to realize that
current demand is dependent upon current
techniques of processing and marketing fish.
The rapid rise of consumption of frozen fish
steaks and fillets in recent years is only sug-
gestive of changes which we can expect in
the future.
Major research programs, such as those
supported by the Sea Grant college system, are
currently attempting to reduce uncertainty
about fish supply. A number of important areas
need to be resolved. To manage the supply of
anchovy, one needs to know the biological
production function of anchovy. As already
suggested, ecological parameters are needed to
manage both independent fisheries and biologi-
cally interdependent species. Thus we need to
understand the nature of supply of all possible
species which might occupy an ecological niche.
It is interesting to examine the controversy
over total yield of food from the sea. As Chapman
has pointed out frequently in recent years,
the wide divergence in estimates really depends
upon the trophic level assumed. Consequently,
not only is there uncertainty about the supply
of any particular species of fish, but there is
uncertainty about the relevant definition of
supply offish.
THE COST OF MOVING TO A PARTIAL
EQUILIBRIUM POINT
Equilibrium points in static analysis are
illuminating for recommendations concerning
25
direction of change. There are several reasons
for realizing, however, that one may not choose
to move to the point of maximum rent. This
reservation is strongest in short-run analysis,
but several parameters in a bioeconomic model
can be expected to shift in the long run.
Before proposing that a fishery should be
managed at the point of maximum net economic
yield, one must first show that the present value
of the fishery at maximum sustainable net
economic yield is greater than the value of the
status quo by more than the transactions costs
of moving to the new point. This was brought
out dramatically by Wantrup (FAO 1962) and
also in a comment by Crutchfield to the effect
that not reducing the existing level of fishing
effort could conceivably be a country's cheapest
unemployment or welfare policy in cases where
the excess number of fishermen truly had no
viable alternative to fishing.3
However, even if the maximum sustainable
net economic yield point is greater in value than
status quo by all relevant costs of change, this
does not preclude the possibility that some
other point, intermediate to those two, might
be more desirable. In summary, the proof of
superiority of a theoretical optimum over a
status quo position leads to an argument for
direction of change in effort, but does not show
the magnitude of change until the costs of such
a change are themselves considered.
PROBLEMS POSED BY
REDISTRIBUTION OF INCOME
Such a recommendation as moving to the
point of maximum net economic yield is roughly
akin to the statement that a readjustment is
recommended whenever the dollar value to
potential gainers is greater than the dollar
value to potential losers. This is the famous
Kaldor-Hicks criterion for an improvement' in
social welfare. The criticisms of this criterion
are now well-known (Rothenberg, 1961) but of
them all, the most commonly cited is the
inability to judge among different income dis-
tributions. To say that one state is better than
3 The comment was made in August 1969, during
discussions after a panel presentation given at Oregon
State University.
another state, when even one individual is worse
off in the former state, is to make those inter-
personal utility comparisons which the eco-
nomics profession has largely disavowed.
To confess that economists have no straight-
forward technique for judging among alternative
income distributions does not alter the fact that
judgments can and will be made. It does mean
that economists should try to describe the
effects of alternative management decisions
upon the distribution of income. In addition to
this, economists can realize that the general
interpretation of equity seems to frequently
preclude drastic changes in the distribution of
income. Management schemes in bioeconomic
models should include systems which compen-
sate losers whenever possible. It may be wise to
attach grandfather clauses, unemployment relief
funds, and the like to licensing or other limited
entry proposals. The costs of preserving stability
and the existing distribution of income should
not be overlooked, creating a parallel to the
awesome headaches of our contemporary farm
program.
AN OPERATIONAL PROPOSAL —
THE USE OF TARGET VARIABLES
Academics who would wish to have a voice
in public policy cannot devote themselves to
being solely naysayers. Decisions must and will
be made. While there are many reservations
which must be made about bioeconomic models,
there is also something intrinsically appealing
about them. How can one use the information
concerning net economic yield and still consider
other objectives?
One possible technique is to simply array a
group of options for the authority who has
received society's mandate. However, it is likely
that the fishery management body itself will be
somewhat removed from direct interaction with
the society. Hence, they will probably need to
infer relative values from another source. Our
experience with many operating agencies has
tended to show that "the wheel that squeaks
gets the grease." Thus, the use of operating
agency discretion may not reflect the values of
the underlying society.
A second technique is that proposed in most
bioeconomic models. Namely, net economic
26
yield would be maximized subject to certain
constraints on acceptable rules for redistributing
income and constraints with respect to minimum
levels of the assorted noneconomic objectives.
There is no denying that this system solves the
problem of weighting the various objectives
by simply avoiding the problem. While weights
are not explicitly chosen, they are implicit in
the levels of the noneconomic variables selected.
Consequently, some technique will have to be
devised for continuous reconsideration of non-
economic objectives with periodic adjustments
in the level of the constraints being made by
an authority responsible to society.
The technique of constrained maximization
will operate best where clear threshold levels
of other variables can be designated. In cases of
international fisheries management, it will
operate best when the parties to the international
agreement can agree on the objectives other
than net economic yield and when relative
weights of more than one species can be specified
where more than one commercially important
species is affected by the management decision.
When this is not true, a third technique of
explicitly agreeing on relative weights of several
objectives and maximizing the weighted function
may be superior.
It may well be that developing a general
theory of fishery management is to develop an
empty theory. Special consideration will be
needed for different species of fish and different
groups of nations. Nonetheless, it is apparent
that management of ocean fisheries is desired.
It is also apparent that biological criteria are
not sufficient to manage a resource in a world
in which there are more goals than merely
consuming one particular species of fish. It is
thus incumbent upon us to try to specify public
policy actions which public authorities can
undertake to achieve the best possible mix of a
large assortment of goals.
LITERATURE CITED
BJORK, GORDON C, 1969. Private Enterprise and
Public Interest, Prentice-Hall, Inc., p. 65.
BOULDING, KENNETH, 1969. Economics as a Moral
Science, American Economic Review, 59( 1) : 1-12.
CRUTCHFIELD, JAMES, and GIULIO PONTECORVO,
1969. The Pacific Salmon Fisheries, The Johns Hopkins
Press, p. 35.
FAO, 1962. Economic Effects of Fishery Regulation.
FAO Fisheries Report No. 5. Rome.
PLOURDE, C. G., 1970. A Simple Model of Replenishable
Natural Resource Exploitation, American Economic
Review, 60(3): 518-522.
ROTHENBERG, JEROME, 1961. The Measurement of
Social Welfare, Prentice-Hall, Inc., pp. 80-103.
SCOTT, ANTHONY, 1962. The Economics of Regulating
Fisheries, FAO Fisheries Report No. 5, Economic-
Effects of Fishery Regulation, p. 32.
21
Economic, Political, and Social Barriers to
Efficiency in Selected Pacific Coast Fisheries
James A. Crutchfield1
ABSTRACT
Multidisciplinary models are being developed for the salmon, halibut, king crab and
anchovy fisheries as an aid in fisheries management. These models will provide estimates
of economic rent in these fisheries, with an evaluation of alternative management
structures available to capture these net benefits. The character of the models for each
of these differing fisheries is described, including reference to the nature of the products,
markets, processors, harvesters, regulators, stocks, and locations sectors of these fisheries.
Introductory observations are made on the future role of multifishery modeling studies.
INTRODUCTION
In June 1970 the University of Washington
and the University of Rhode Island were funded
by the National Marine Fisheries Service to take
a first step in identification and quantification
of the economic costs of institutional barriers
to the efficient use of commercially fished marine
stocks. Anyone familiar with the American flag
fisheries will recognize that the time and financial
limitations of these one or two year studies
preclude any definitive findings applicable on a
broad scale. Nevertheless, first steps must some-
how be taken, and the two university teams,
together with their NMFS counterparts, share
the view that a convincing demonstration of
substantial economic gains from the elimination
of obvious sources of inefficiency is one of the
most important of these steps. Hopefully, it will
represent one phase of a broad-based attack on
the problems of modernizing the American
fisheries and rationalizing the objectives and
techniques of management.
This paper presents a summary progress
report of the Pacific Coast studies. The project
has two objectives. In the short run it is intended
to provide reasonable estimates of potential net
economic rent in representative Pacific Coast
fisheries, and to explore the feasibility of alter-
native management regimes to realize at least
a portion of these net benefits. The importance
University of Washington.
of this objective is underscored by increasing
pressure for tangible evidence that the overall
activities of the National Marine Fisheries
Service can be translated into economic benefits:
an outcome that is anything but likely under
present institutional arrangements in the fish-
eries. In the face of increasingly insistent de-
mands on the inshore waters of the United
States, and the likelihood of severe budget
stringency for an indefinite period, a convincing
demonstration of the net benefits that can be
generated by the elimination of unnecessary
barriers to efficient harvesting of marine stocks
may well determine the future existence of a
strong federal fisheries function.
The longer term objective of the study is to
develop primary data and modeling capacity to
test fully alternative management and develop-
ment regimes. Previous studies of individual
segments of the American fisheries (Crutchfield
and Zellner, 1962; Crutchfield and Pontecorvo,
1969; and Bell and Carlson, 1970) have been
concerned primarily with maximum potential
net economic rent in long run terms, with
varying assumptions as to acceptance or modifi-
cation of existing legal and other constraints.
It is clear, however, that a full reevaluation of
fishery management objectives requires a much
broader frame of reference and a larger kit of
tools. Since it is politically unlikely that all
barriers to efficiency will be removed simul-
taneously, it would be most useful to develop a
modeling technique that would permit us to
look at a wide variety of measures or combina-
tions of measures, at relatively low cost but with
2H
real numbers to provide real estimates of
economic and biological effects.
There is equally urgent need for a quantified
model that can be manipulated in terms of
multiple objectives: economic efficiency, income
distribution, structural unemployment, and per-
haps others. The modeling technique lends itself
well to assessment of a range of management
measures that might be undertaken to achieve
multiple objectives, or to maximize certain ele-
ments subject to constrained values for others.
There are both biological and economic reasons
for development of a more sophisticated model
than the long term equilibrium constructs used
in earlier work. Short term adjustments of both
fish stocks and fishermen to altered parameters
must be scrutinized much more carefully. Simi-
larly, the usual analysis of yield functions, and
of bioeconomic models based on them, is cast in
terms of a single fishery, while most American
fishing gear either exploits more than one species
or is capable of doing so. Even before the eco-
nomic numbers to be used in a more complex
process model of this sort can be developed, it
is possible to derive a great deal of knowledge
of immediate benefit in assessing alternative
management regimes by framing appropriate
functional relations in model form and testing
their sensitivity to various assumptions as to
quantitative values.
In short, it would be highly desirable to
develop a set of models specific to individual
fisheries but geared to a central common frame-
work that would permit comparison among
fisheries. Obviously, this will not be done in a
day or a year ; but if a good start can be made in
isolating the functions that must be quantified
and delineating data requirements, the ultimate
payoff in terms of flexibility and low operating
cost will make possible a dynamic concept of
fisheries management that can really utilize
increases in scientific knowledge, improved
technology, and more flexible administrative
arrangements.
THE FISHERIES
The Pacific halibut operation is a mature fishery,
relatively simple in economic structure, and
employing only a single type of gear. It has
been under a carefully conceived regulatory
program for a sufficiently long period to gen-
erate excellent data on both biological and
economic variables.
The Pacific salmon fishery stands at almost
the opposite extreme. It is complex in every
sense — biological and economic — that can
be imagined. It is subject to inherent data
limitations since it is based on populations that
are in constant short run disequilibrium, and
it is now regulated on such an irrational basis
that great improvement is possible with rela-
tively simple alterations in management tech-
niques.
The California anchovy fishery, barely ex-
ploited at the present time, represents one of
the largest single latent resources available to
American flag fishermen. On the assumption
that present legal limitations on commercial
exploitation are removed, the potential physical
yield from the fishery is almost as great as the
total United States landed catch. The possibility
of creating a new and highly attractive industry
under controlled entry conditions is intriguing,
to say the least. Data on the California anchovy
are still rather limited, but the basic stock
information is being developed rapidly, and
both the biological and economic analysis can
borrow extensively from the broad experience
of the Peruvian anchoveta fishery.
The king crab fishery of the North Pacific
presents a classic example of the speed with
which modern technology, under conditions of
open entry, can lead to overinvestment, over-
fishing, and potential economic disaster. In
addition, the hastily conceived regulations now
in effect present some of the worst examples of
efficiency-reducing techniques, coupled with
obvious efforts to redistribute income from one
set of fishermen to another. Data are woefully
inadequate in this fishery, but its economic
value and potential make it an excellent
case study.
The four fisheries chosen for analysis were
selected for characteristics which make them
broadly representative of the kinds of problems
to be faced in future fishery management pro-
grams geared more closely to economic objectives.
THE MODELING FRAMEWORK
The simulation approach which serves as the
basis for the longer run aspects of this project
is hinged on a general model which is adaptable
2!)
to each of the specific fisheries to be studied.
It involves a large computer program which
provides a framework for studying short term
and long term effects of alternative regulatory
policies on economic and biological performance
of various sectors of a fishery. The basic program
is written in a version of FORTRAN IV. A
schematic of the basic model is shown in Figure
1. The sectors simulated by the program are:
products, markets, processors, harvesters, regu-
lators, stocks, and locations. General operation
of the model is as follows. The stock sector
"grows" the resources and determines the
amount of each stock which is available for
harvest in each location. Harvesters operate in
locations of their choosing, catch a portion of
the available stock, and sell it to processors.
Processors convert their purchases into finished
goods and offer them for sale in the markets
which are available. Demand (and the marketing
activity of the processors) determine sales by
each processor of each product in each market.
The regulators are free to impose restrictions
of various types on the activities of both
processors and harvesters. In operation the
program compiles statistics on the operation of
the system and prints out monthly and annual
summaries of these statistics. The detail of the
printout is optional.
Each harvesting group operates as a semi-
independent unit, constrained only by links to
a location and one set of processors. At the start
Product
Sector
Consumer
Market Sector
Processor
Sector
Product
1
Harvester
Sector
Location
Sector
Stock
Sector
Regulator
Sector
Harvester
1
Stock
1
Regulator
1
Market
Harvester
2
Harvester
Stock
2
Regulator
2
Regulator
Figure 1. — Structure of fishery simulator.
30
of each month a group of fishing units moves
from its initial location to the harvesting location,
operates there for a specified number of days,
and sells its catch to the processor which will
produce the maximum profit to the group. If
the processor cannot absorb the group's total
supply, the group sells what it can and moves
to the next most profitable processor and so on.
A group may only sell to the set of processors
linked to the harvester that owns the group.
Any catch unsold at the end of the month is
recorded and discarded since it has no economic
value. The harvesting time for each group is
limited not only by harvester and regulator
decisions, but by the group's harvesting capacity.
Primary operating costs for the harvester built
into the model are distance costs, time costs,
harvest-proportional costs, and license fees.
A processor is a managerial entity that
operates in one physical location, buying stocks
from harvesters, transforming them into finished
products, and selling them in markets. As the
program is now set up, each processor's share
of the market can be made to depend on his
previous market share and on marketing ex-
penditures and product price relative to those
of other processors linked to the same market.
The cost structure of the processors is in standard
accounting terms.
A regulator is an agency that imposes restric-
tions on the activities of harvesters or processors
in any of the variety of ways now employed or
discussed in the literature. These include:
(1) license fees; (2) size limits; (3) gear efficiency
limits; (4) effort limits; (5) operating limits for
processors; (6) seasonal closures; (7) monthly
quotas; (8) annual quotas.
The model can be run in either of two basic
operational modes: as a conventional computer
simulation model, with built-in decision-making
algorithms specifying the behavior of processors,
harvesters, and regulators; or with human inter-
vention at intervals to allow for intuitive and
heuristic decisionmaking.
A stock is any type of renewable marine
resource. It is treated in the model as linked to
a given location, and the quantities available
to harvesters at any given time are computed
continuously.
It should be stressed that each entity is a
subroutine, and can be designed to any degree of
complexity warranted by the purpose of the
routine and the adequacy of the data base.
Similarly, the degree of detail for a readout on
monthly or annual bases can be predetermined.
The model can be programmed not only to
maximize specific objective functions, but can
accommodate dynamic feedback factors in assess-
ing different kinds of management alternatives.
It can also handle a wide range of spatial
distributions of stocks and harvesters without
difficulty.
THE ANCHOVY FISHERY
Figure 2 shows, in schematic form, a pre-
liminary version of the model of the California
anchovy fishery. This model reflects the activities
of the types of vessels presently exploiting the
fishery, and therefore attempts to deal with the
complications imposed by their harvesting of
bluefin tuna and mackerel as well as anchovy.
It is also complicated by the interaction between
the markets for sport fishing bait and for meal
and oil, both of which now absorb considerable
quantities of anchovy. Sufficient data are avail-
able to permit some preliminary conclusions as
to the economic return from this limited fishery,
which is now prosecuted at a level so low that
the more fundamental problems in the stock
sector are not really involved. These preliminary
findings suggest, as one might suspect, that the
return to vessels fishing for anchovy on a full-
time basis during a nine months open season
would be substantially more attractive than the
returns from mixed operations.
Accordingly, the model which will be used to
test regulatory alternatives will probably be
based on the assumption that a specialized
fleet of vessels optimized for the anchovy fishery
will develop once catch quotas are established
at levels sufficiently high to induce long term
investment in fishing and processing equipment
by major firms in the meal industry. Preliminary
work in a dissertation by Dr. Dennis Paulaha
(1970) provides excellent data on the type of
vessel and gear best adapted to the fishery.
Work on the complex stock model is reasonably
well advanced, and it is expected that a fairly
sophisticated and realistic approximation to the
behavior of the anchovy stocks under various
rates of exploitation can be developed.
The economics of the anchovy operation are
relatively simple to simulate, since total produc-
31
Meal
Canned
JM
Live
Bait
Monterey
Reducers
Los Angeles
Canners 5.
Reducers
Calif.
Fish &
Game
Monterey
Fleet
Wetfish
Fleet
Northern
Anchovy
Jack
Mackerel
Calif.
current
System
Bait
Processors
*-->--> -
Special
Anchovy
Fleet
Rait
Fleet
Figure 2. - - Graphic representation of the logical re-
lationships between sectors of the Simplified Northern
Anchovy Fishery System. (Area quotas of total system
are considered to be levied as production quotas in
the simplified system.)
tion from the fishery, even at a catch level of
one million tons, would still produce only a
small fraction of total American fish meal
consumption. Market price can thus be taken
as given to the California meal producer, and
the estimated net economic rent available under
various assumptions as to management regime
can then be calculated on the basis of alternative
forecasts of the time-path offish meal prices.
THE PACIFIC SALMON FISHERY
In Appendix I the general format of a pre-
liminary program for modeling the Pacific
salmon fishery is presented. The objectives in
modeling this extraordinarily complex opera-
tion are partly methodological and partly aimed
at answering specific management problems of
real significance. The complications are ap-
parent. Five separate species of salmon are
involved, and since each river usually contains
more than one species (and separate races of
the same species), the number of "management
units" which should, in theory, receive separate
treatment in modeling the stock sector is
probably from eight to ten thousand]
"The salmon fishery" is actually a large num-
ber of geographically separate operations, linked
in varying degrees by the mobility of the gear
involved. Several types of gear are used, and the
relative importance of each type varies from
area to area. Finally, salmon deteriorate very
rapidly unless processed soon after being cap-
tured, which creates a large number of primary
markets in which processors generate several
different end products from each of the types
of salmon purchased.
Even with the prodigious capacity of the
32
modern computer, overall modeling of a fishery
this complex is obviously severely limited by
available data, and the marginal cost of generat-
ing the necessary data is very high. In one
sense, then, the broader modeling exercise is
intended to provide some guidelines to the
limitations on the technique in dealing with
highly complex fisheries.
On the other hand, the program is flexible
enough to permit specific consideration of im-
portant policy questions in separable segments
of the salmon fishery. For example, the Columbia
River, Puget Sound, and British Columbia
fisheries are plagued by serious problems arising
from the spectacular growth of the ocean troll
fishery. Since the trollers take large numbers
of immature fish, they do a considerable, though
unknown, amount of damage in returning under-
sized fish to the water. The troll fishery is
inherently highly inefficient from both biological
and economic points of view. A substantial
part of its catch is made up of two and three
year old chinooks and two year old cohoes
which would almost certainly gain substantially
more in body weight than would be lost to
natural mortality if allowed to mature another
year. It is possible, with a restricted model, to
test the biological and economic impact of the
elimination or limitation of the troll fishery in
specified areas under varying assumptions as
to the resulting net increment in weight and the
distribution of the troll catch among other
types of gear. This analysis is, incidentally,
crucial to another public policy issue of major
proportions — the allocation of chinook and
coho salmon among commercial and recreational
users.
Earlier work by a University of Washington
team on the Puget Sound salmon fishery (Royce,
et al., 1963) indicates that modeling permits
surprisingly accurate prediction of the net
economic benefits and catch distribution effects
by area of different techniques for reduction of
gear and expansion of intraseasonal fishing
time. The earlier study was, for strategic rea-
sons, constrained by the assumption that any
reduction of gear must be proportional for each
type of gear. It is obviously desirable to develop
the capability to test quickly and inexpensively
the effects of altering gear mix by area and by
time period. Since any gear reduction program
in the salmon fishery will inevitably involve
intensely partisan political negotiations, a
display of the impact of a wide range of alterna-
tives is essential if any progress is to be made.
Finally, the model can be used to predict the
impact of recent court decisions requiring that
Indian fishermen must be granted a "prior
claim" on any total catch permitted under
regulation.
THE PACIFIC HALIBUT FISHERY
This fishery presents a far simpler set of
modeling problems. The stocks have been under
intensive study for more than 40 years, and a
wealth of reliable statistical information is
available on both the stock and harvesting
sectors. In addition, the use of standard gear
(whatever its economic validity) makes analysis
much simpler, as does the widespread use of
a standard accounting system for halibut vessels
devised by the Fishing Vessel Owners Associa-
tion. The fact that halibut is marketed almost
entirely in fresh and frozen form further
simplifies the analysis. The principal gain in
the modeling exercise for this fishery will be
the ability to incorporate badly needed studies
of the effects of introducing different types of
gear, potentially much more efficient, if and
when limitation of the number of operating
units becomes possible. It will also be possible
to introduce into the analysis the effect of
shifting many of the halibut vessels from their
present multipurpose form into larger, specializ-
ed units — a process which would almost
certainly follow any effective gear limitation
program.
THE KING CRAB FISHERY
It is doubtful that the king crab fishery will
be amenable to very effective empirical work
in the near future. Not only are data extremely
limited, but the fishery is based on a relatively
long-lived, slow-growing animal, and it is
currently in a state of disequilibrium. Con-
sequently, the fragmentary statistical informa-
tion on catch, effort, and economic returns
from the fishery cannot be considered representa-
tive of long term equilibrium values. Neverthe-
less, the situation in the king crab fishery with
respect to stock depletion is so serious, and the
33
regulatory methods already adopted so question-
able, that some analysis of this fishery, even
with very limited data, is clearly necessary if
we are to avoid serious and perhaps irreparable
mistakes.
CONCLUSIONS
If the approach embodied in this study proves
to be as useful as expected, it is considered
possible that the techniques could be extended to
provide a broader approach to multifishery
cases. The seasonal nature of the availability of
fish and of weather conditions on the Pacific
Coast suggests that an optimal harvesting
technique for virtually all species (with the
possible exception of halibut and some other
bottomfish) will involve multipurpose gear
exploiting multiple species in different geo-
graphic locations. For example, salmon, crab,
and albacore fishing by combination units may
be significantly more attractive economically
(always assuming some control over entry)
than the present hodgepodge of vessels involved
in each. We do have combination vessels at
present, of course, but they are not designed to
any set of specifications that present data would
make available if an integrated view of the
fisheries available to each type of gear were
taken as the frame of reference.
The discussion above suggests the nature of
the outputs to be expected from these models
in the short run. We are still limited to syn-
thetic numbers in many of the sectors at present,
but these are being systematically whittled
down. It cannot be stressed too strongly that
the whittling down process can be done far
more economically and effectively once the
sensitivity of the desired outputs to the various
parameters involved has been established by
the model. Moreover, some of the fisheries and
some elements of the model have now reached
a point where reasonably hard data are avail-
able which can be manipulated to provide at
least rank-ordering of a number of management
options. While the overall program is clearly
geared to longer range objectives, short run
outputs of real usefulness in management
planning can be expected, and will increase in
number and predictive value as the work
progresses.
Members of this workshop and other interested
scientists and economists are urged to com-
municate to us the nature of their interest in
the problems addressed by the University of
Washington team. In addition, it might be
mutually advantageous if visits to the University
of Washington could be arranged to permit
actual operating experience with these models.
LITERATURE CITED
BELL, FREDERICK W., 1970. Estimation of the
Economic Benefits to Fishermen, Vessels and Society
from Limited Entry to the Inshore U.S. Northern
Lobster Fishery, Marine Technology 1970 Preprints —
Vol. 1. Marine Technology Society, 6th Annual Con-
ference and Exposition June 29-July 1, 1970, Wash-
ington, D.C.
CRUTCHFIELD, JAMES and GIULIO PONTECORVO,
1969. The Pacific Salmon Fisheries. The Johns Hopkins
Press, Baltimore, Maryland.
CRUTCHFIELD, J.A. and A. ZELLNER, 1962. Economic
Aspects of the Pacific Halibut Industry, Fishery In-
dustrial Research. United States Department of Interior,
Fish and Wildlife Service, Bureau of Commercial
Fisheries, U.S. Government Printing Office, Washington,
D.C.
PAULAHA, DENNIS E., 1970. A General Economic
Model for Commercial Fisheries and its Application to
the California Anchovy Fishery. Unpublished Ph. D.
Dissertation. University of Washington.
RICKER, W.E.. 1958. Handbook of Computations for
Biological Statistics of Fish Populations. Bulletin of
the Fisheries Research Board of Canada, No. 119,
pp. 1-300.
ROYCE, W., D. BE VAN, J. CRUTCHFIELD, G. PAULIK,
R. FLETCHER, 1963. Salmon Gear Limitation in
Northern Washington Waters. University of Washington
Publications in Fisheries, New Series, Vol. II, No. 1.
34
APPENDIX I: PACIFIC SALMON SIMULATION MODEL COMPONENT
The proposed simulation model will treat the five species of Pacific salmon in the North American
fisheries as five separate stocks:
s,
Chinook
s,
Chum
Sa
Pink
s,
Sockeye
s.
Coho
The location sector in the model will be based on the areas for which statistical information is
available in published data sources:
L,
L2
L3
U
I,,
L6
U
L8
L9
Lio
Western Alaska
Central Alaska
Southeastern Alaska
Northern B.C.
Southern B.C./Fraser River
Puget Sound
Washington Coast
Columbia River
Oregon Coast
California
Stock/location interaction will be as follows:
Chinook Chum Pink Sockeye Coho
S\ 02 03 04 05
Li Western Alaska
L2 Central Alaska
L3 Southeastern Alaska
L4 Northern B.C.
L5 Southern B.C./Fraser River
L6 Puget Sound
L7 Washington Coast
L8 Columbia River
L9 Oregon Coast
Lio California
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
—
X
—
X
X
—
—
—
X
and these interactions reflect the distribution of species in the actual fishery. Data will be collected
to permit segregation of stocks into age groups, with age specific weights and spawner-return curves
(Ricker, 1958) developed for each stock/location. In effect, this scheme results in a total of 45 separate
stocks, since each species in each location will be treated separately with respect to spawner-return
characteristics.
Regulators will be based on locations, with one regulator in each location.
The principal types of fishing gear in the Pacific salmon fisheries are as follows:
Gill nets, drift
Gill nets, anchor
Seines
35
Troll lines
Reef and Pound nets
In order that each harvester be able to fish for each species with each type of gear, it is necessary
that the harvesters be denned as follows:
Hi Seine, Western Alaska (Li)
Ho Seine, Central Alaska (L2)
H3 Seine, Southeastern Alaska (L3)
H4 Seine, Northern B.C. (L4)
H5 Seine, Southern B.C./Fraser (L5)
H6 Seine, Puget Sound (L6)
H7 Anchor Gill Net, Western Alaska (Li)
H8 Anchor Gill Net, Central Alaska (L2)
H9 Anchor Gill Net, Southeastern Alaska (L3)
H10 Anchor Gill Net, Puget Sound (L6)
Hn Anchor Gill Net, Wash. Coast (L7)
H12 Anchor Gill Net, Columbia River (L8)
H13 Drift Gill Net, Western Alaska (Li)
Hi4 Drift Gill Net, Central Alaska (L2)
H15 Drift Gill Net, Southeastern Alaska (L3)
H16 Drift Gill Net, Puget Sound (L6)
Hi7 Drift Gill Net, Wash. Coast (L7)
His Drift Gill Net, Columbia River (L8)
Hi9 Gill Net, Northern B.C. (L4)
H20 Gill Net, Southern B.C./Fraser River (L5)
H2i Troll, Central Alaska (L2)
H22 Troll, Southeastern Alaska (L3)
H23 Troll, Northern B.C. (L4)
H24 Troll, Southern B.C./Fraser (L5)
H25 Troll, Puget Sound (L6)
H26 Troll, Wash. Coast (L7)
H27 Troll, Columbia River (L8)
H28 Troll, Oregon Coast (L9)
H2g Troll, California (Li0)
H30 Reef & Pound Nets, Puget Sound (L6)
The above table indicates that there is no fleet for a particular species in those cases where the
annual catch for that species in that location by that gear type is less than one percent (1%) of the
total annual catch of that species in that location by all gear types.
There will be one processer in each location, with processer locations defined for distance
computation purposes as follows:
Pi Bristol Bay (Western Alaska)
P2 Cook Inlet (Central Alaska)
P3 Yakutat (Southeastern Alaska)
P4 Prince Rupert (Northern B.C.)
36
Chinook
Chum
Pink
Sockeye
Co ho
Si
s2
s3
s4
S5
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
—
—
—
X
X
X
X
X
—
X
X
X
—
X
X
X
—
—
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
—
—
X
X
X
—
X
X
X
X
X
X
X
X
X
X
X
X
X
—
—
—
X
X
—
—
—
X
X
—
X
X
X
X
—
X
X
X
X
—
X
—
X
X
—
X
—
X
X
—
X
—
X
X
—
X
—
X
X
—
—
—
X
X
—
X
X
X
P5 Vancouver (Southern B.C./Fraser River)
P6 Seattle (Puget Sound)
P7 Westport (Washington Coast)
P8 Astoria (Columbia River)
P9 Newport (Oregon Coast)
Pio San Francisco (California)
Markets will be synonymous with products, with demand relationships developed for each
product as follows:
I),
Fresh/frozen
I).,
Salted or pickled
D3
Mild cured
I),
Smoked or kippered
D5
Canned
D6
Roe (cured)
Products will be produced by processers as follows:
Chinook Chum Pink Sockeye Coho
Si S2 S3 04 05
D2
D3
D5
D6
Pi Bristol Bay
Di Fresh /frozen
Salted or pickled
Mild cured
Canned
Roe (cured)
P2 Cook Inlet
D,
D2
D5
D6
P3 Yakutat
Di
D3
1),
D6
P4 Prince Rupert
P5 Vancouver
X
X
X
—
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
—
—
—
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
—
—
—
X
X
X
X
X
X
X
X
X
X
X
P6 Seattle (data available for Washington as a whole)
P7 Westport
D3
D4
I),
D6
P8 Astoria (data available for Oregon as a whole)
37
X
X
X
X
X
X
X
X
—
—
—
X
X
X
X
X
X
X
X
X
X
Chinook Chum Pink Sockeye Coho
Si S'2 S3 S4 S5
P9 Newport
D4 X — — — —
D5 X X X X X
P10 San Francisco
D, X — — — X
D3 X — — — X
D4 X — — — X
38
PRODUCTION FUNCTIONS AND BIOECONOMIC
MODELS: RESEARCH IMPLICATIONS
Against the broad background of these four
introductory papers we can proceed to some of
the more specific research which will constitute
the principal inputs into the broader manage-
ment process. The first of these papers relates
the results of an extensive effort by Carlson to
specify production functions for the North
Atlantic groundfish and tropical tuna fisheries.
In each case the research is designed to identify
the most significant determinants of vessel
productivity, with some of the investigation
devoted to the question of a proper measure
of productivity.
Using existing data series on the area and
time patterns of fishing activity, landings
statistics on species, quantity and value, and
other sources of data on vessel characteristics,
specific effort combinations are related to produc-
tivity. The "best" measure of productivity was
found to be value in groundfish and a weighted
combination of species landed in tuna.
This research output has many possible uses,
among these being the suggestion of the
optimum input package to maximize output
and the development of a fishing power index
which could be used to measure effort, a critical
input into those types of management plans that
require the administrator to develop seasonal
or sharing arrangements based on the fishing
capabilities of the fleet. This is the case for the
Inter-American Tropical Tuna Commission.
Here a technique of measuring fishing power
has evolved which is somewhat different from
the Carlson approach. Future investigations
will determine the advisability of each approach.
Indeed, if differences and difficulties cannot be
resolved, this may have some effect on the choice
between management plans which require this
type of calculation and types which do not.
The paper by Segura relates part of his broad
investigation into the world supply and demand
for fish meal. His efforts for this paper have
concentrated on a measure of fishing power in
the Peruvian anchoveta fleet for the purpose of
determining the optimum harvest level. His
focus is upon the role of technological change
as this relates to time series calculations of
effort indices.
In his paper, Segura points out the differing
results which will be forthcoming if you use
the most recent years' measure of yield-effort,
the index of vessel productivity, to calculate
changing pressures on the resource, the response
of the resource to that pressure, and use these
relationships to determine an optimum catch
quota for the coming year. He compares these
results to calculations now used where these
interrelationships are all derived based upon
some earlier base year. The results are sub-
stantially different, resulting in a suggested
catch of 16.2 million ton trips derived via the
existing method.
The work done by Segura relates closely to
that of Carlson in that a method of cross-
sectional analysis of recent years' data is being
developed which obviates the need to use
standard vessels from some base period, supple-
mented by ad hoc measures of technological
change. These considerations are in addition to
the question of diminishing returns as intro-
duced by the Carlson-Waugh-Bell function.
The research reported by Rich is an extension
of a generalized model to be applied to the
Pacific halibut fishery. The purpose is to evaluate
possible losses which may have resulted in the
fishery from the use of MSY as a regulation goal.
Consistent with the Carlson-Waugh-Bell
exposition, the function developed incorporates
short run diminishing returns. When combined
with a fish growth function it is possible to
measure the long run externalities associated
with this alternative specification of the yield-
effort function.
This approach is the antithesis of that sug-
gested by Pontecorvo in that it is explicitly
structured upon the classic assumptions of full
employment and complete labor mobility, both
in the short run and the long run. Political
and social questions are definitely excluded
and would have to be appended on an ad hoc
basis to determine if there was any cause for
modifying the constrained results. The work
39
done by Rich would serve as but one component
in the simulator described by Crutchfield,
albeit possibly the dominant component.
Bell, Carlson, and Waugh focus on the issue
of diminishing returns in fisheries, relaxing a
strong assumption of fixed proportionality
utilized by most writers in the existing literature.
The motivation for this exercise is the ap-
preciation that we are rapidly approaching total
utilization of the world's fish resources. As this
point is approached, demand pressures and
considerations of maximum efficiency dictate
the need to make maximum use of these
resources consistent with any overriding con-
servation objectives. The work done by these
authors, though preliminary, suggests that
some degree of diminishing returns can be
identified for the fisheries studied: Chesapeake
Bay menhaden, Atlantic and Gulf blue crab,
Atlantic longline tuna, Bering Sea king crab,
and Cape Flattery sablefish.
As with the other five papers in this section,
this paper modifies existing biological functions.
The modified logistic introduced here is the
author's candidate for a "better" function,
based primarily on the inclusion of diminishing
returns in the logistic specifications. As with
the other contributions this paper suggests
an area meriting further discussion in the near
future, with our best use of marine food re-
sources being the stake.
Thompson continues the parade of alternative
functions with his concern being the absence
of a proper dynamic component within the
prevalent fisheries models. To correct this
error he proposes the marriage of the Schaefer
model and the Thompson-George (TG) produc-
tion-investment model. He also suggests some
alterations in the Schaefer model.
The TG model replicates the sequence of
investment-production decisions which are in-
volved in the operation of the individual fishing
firm (vessel). Pertinent stocks and flows are
specified with elaborate preconditions for entry,
though there are no provisions for entry within
the decision period, an interesting trait in light
of the Johnson fixed asset theory as referred
to by Stevens and Mattox subsequently. By
adjoining this model to the Schaefer biological
fluctuation we have a bioeconomic model which
is uniquely micro in character; the dynamics of
change in the fishery stock (and hence fishing
success) will be reflected in the investment
decision of the sole owner as the limiting case,
and vice versa.
This method avoids the critical use of static
methods prevalent in economic literature. In-
herently, the adjustment mechanism in the
individual owner also facilitates the modification
of the Schaefer function to incorporate decreas-
ing returns to effort, as discussed by Bell,
Carlson, and Waugh and by Rich and increasing
returns to scale. Relaxation of the sole owner
condition further amplifies the critical nature
of these alterations and within the confines of
standard economic assumptions reaffirms the
desirability of limiting entry and suggests an
additional method of measuring the critical
management variables.
The final author in this section addresses
the problem of multiple species fisheries — or
combination vessels. In this regard three issues
are of prime importance to Adam. The first of
these relates to the existence of yield curves for
fisheries. Adam views most of these curves as
average curves, pointing out that for many
fisheries this average curve will be bounded by
upper and lower curves which are usually the
result of substantial fluctuations in either
effort and/or recruitment. The average curve
is essentially a product of a stable fishery where-
as the boundary curves are the result of a
rapidly growing fishery. In his opinion we do
not move along the average curve as a fishery
rapidly develops. We move from one curve to
another, somewhat erratically as the fishery
develops. He looks to the economist, via a
function akin to Carlson, where effort is value-
dependent, to indicate what effort will be in
subsequent years, as the fisherman's response
to his monetary success is one of the few
reliable variables which can be presented to a
biologist in such a dynamic situation.
His second point extends this argument to
multiple species. If a vessel has the capability
to adjust his harvesting pattern based upon
conditions in the fishery or the market, this
would preclude estimation based solely on
biological factors. It suggests that many of
these calculations must be made instantaneously,
at that time each year when a fishery is being
initiated. It suggests also that this must be
done for several fisheries simultaneously if
those fisheries are interrelated. For the North-
40
east Atlantic this is increasingly the case. be closely examined, however, so that we may
Adams's final related point concerns the maximize their comparability and/or ascertain
measurement of fishing effort. Simply stated, which measure would be most appropriate for
he concludes that there is no single measure each circumstance,
which can unequivocably serve the needs of all
the disciplines. These different measures should A. A. S.
41
Cross Section Production Functions for North
Atlantic Groundfish and Tropical
Tuna Seine Fisheries
Ernest W. Carlson1
ABSTRACT
This paper explores the use of cross section production functions to estimate the
fishing power of individual vessels. The problems addressed are: The proper measurement
of output; the measurement of technological change, and the effect of location, crew
size and important vessel characteristics.
Regression analysis upon data from the North Atlantic groundfish fishery and the
tropical tuna seine fishery yielded highly significant results. Many of the hypothesized
relationships are measurable and stable with relatively small errors. The tests indicate
that: there are better measures of output then total pounds; fishing time is measured
better using days absent rather than days fishing; the use of more vessel characteristics
improves explanatory power; crew size can be an important variable; the effects of
location can be measured; and technological change can be measured.
The production functions measured can then be used as inputs in devising
management schemes.
INTRODUCTION
One of the more difficult problems in the
management of fisheries has been the measure-
ment of vessel productivity. If the vessels in a
fleet were physically homogeneous and utilized
for the same amount of time and if no learning
took place, the problem of measuring productivi-
ty indices would be less difficult. The problem
does exist, though, because vessels are far from
homogeneous. For example, a typical fleet may
have vessels that are 10 or more times larger
than the smallest vessels in a fleet. Obviously,
under such conditions there will be serious
errors introduced if attempts are not made to
measure the productivity of different vessels.
To handle this and related problems, econo-
mists have developed techniques of measurement
that fall into a general category called production
functions. One of the important attributes of
using a production function is that it allows
the simultaneous measurement of as many
1 Economist, Economic Research Laboratory, National
Marine Fisheries Service.
parameters of fishing power as may be thought
to be important in its determination. According-
ly, production functions were estimated using
data from the New England trawl fleet and the
tropical tuna seine fleet. Many problems were
considered in arriving at a "best" production
function for these fisheries.
THE PRODUCTION FUNCTION FOR
A FISHERY
The basic assumption of this paper is that a
production function can adequately describe
the relationship between inputs and outputs in
a fishery. The production function is a technical
or engineering relation between inputs and
outputs and is the base upon which the economic
theory of supply is built. Since it is an engineer-
ing relationship, considerations such as prices
and costs are not relevant to the production
function itself. The schedule of maximum output
for given inputs is the production function we
are trying to measure.
The classical production function for the
individual firm is usually presented as follows:
42
x =f(l,k,t),
where x = output,
I — labor,
k = capital,
t = natural resources.
Output (x) is measured as the flow of goods and
services during an accounting period. The input
variables (I, k, t) are the various kinds and
qualities of labor, capital, and natural resources
that go into producing the output. It is assumed
that a given set of inputs produces as much
as possible.
The estimation of the parameters of the
production function is accomplished by running
a regression upon a cross section of fishing
vessels. A cross section is a sample of the vessels
in a fishery for a fixed time period. The para-
meters estimated from the cross section will give
the marginal contribution to output of each
variable being used to explain output.
We will discuss the variables that will be
used in the production function in the following
section.
Output in a Fishery
Most systems for measuring relative vessel
productivity have, ultimately, related output to
some fishing vessel characteristic. The basic
problem with this is that output, when using
commercial landings statistics, is a very complex
concept. Except in extremely simple fisheries,
fishermen do not ordinarily attempt to maxi-
mize pounds of fish landed. One working
hypothesis is that in all fisheries, the fishermen
attempt to maximize their profits. This is not
necessarily the same as maximizing total pounds
of fish landed. Using total pounds as a measure
of output would be an acceptable measure of
output (1) where there is a single species
fishery or (2) if, in a multispecies fishery, the
prices of the target species are approximately
the same and the species are equally catchable.
In the general case, these conditions are not met.
How do the fishermen decide where to go and
what to catch when there are multiple species
in a fishery? Again, the answer to this question
is difficult. Let us consider two models of
behavior that might help answer this question.
In the first type of fishery, the vessel captains
take into account the species that are available,
the grounds where they are available, the prices
for which they can be sold, and the expected
catch rates for their vessels on the grounds.
Integrating all this information, the captain, if
he is a profit maximizer, will decide to go to
the grounds and fish for the species which pro-
vide the highest net profit. His decision may or
may not be to fish where the catch rates are
highest or for those species that bring the
highest prices.
We have been discussing this as if the choice
were always between species. The choice can
also be made within a species, such as a decision
to fish on local grounds rather than on distant
grounds where the catch rates are higher. In
this case, the higher catch rates may not offset
the extra running time necessary.
If this abbreviated discussion is an adequate
description of how fishermen behave in one
type of fishery, then it follows that we may not
be able to estimate relative vessel productivity
with total pounds, but must rely on some higher
order measure such as the value of catch.
Value was considered by Gulland (1956) as
a measure of output and rejected because of the
variability of prices. A large part of the vari-
ability of fish prices is due to the seasonal
availability of the fish themselves with prices
moving inversely to availability. We can lessen
the objections to value at least partially by
using annual data so that the interseasonal
effects of availability average out. Another
alternative would be the estimation of relative
efficiency on a quarterly basis.
The second type of fishery is one where the
location of the fish by species is generally
known, but where there is considerable mixing
of single species schools in the same area. If
locating any school has a low probability per
unit time, the fishermen will attempt to catch
all that they can of those they do locate. In this
case, the fish will be joint products of the fishery.
If the fish are equally catchable and their prices
are not too different, then total pounds could
be the measure of output. If they are not equally
catchable, it would take more fishing power to
catch one than the other. In such a case, we
might have to utilize a modified estimation
scheme to arrive at a proper weighting for
output. One such scheme will be discussed under
the statistical section on tuna.
41]
Inputs in a Fishery — Fishing Time
The abstract production function refers to
outputs and inputs per unit of time. The unit of
time is undefined. When using annual vessel
data, we have to note the fact that the vessels
are not utilized for the same amount of time
and standardize for this.
In the simple case, an economist would prefer
to use days absent from port as a measure of
fishing time rather than days fishing. If a
fisherman is an economic maximizer, he will
attempt, ceteris 'paribus, to maximize his gross
revenue per day at sea and will plan his fishing
strategies accordingly. Under this assumption,
the fisherman may or may not fish when or
where his expected catch is higher.
The theory is not clear as to how time should
enter the production function. Two basic
specifications are possible:
(1) x=Dccf(l,k,t), or
(2) x=Dif(l,k,t)
There are theoretical reasons that could justify
the use of either. Equation (1), with D°; can be
justified if we hypothesize that the fishermen
makes trips of varying length. Therefore, we
would want to find the marginal contribution
of an extra day at sea. Equation (2), with
D1, can be justified if we hypothesize that all
inputs are being used to produce output all the
time, so that the relationship is strictly linear.
Experiments were run initially in both forms,
but the second form was abandoned for what
may have been specious reasons. If further work
is done the alternative specification will be
tested more fully.
Capital — The Vessel Characteristic
Variables
The abstract production function has a vari-
able called capital. This represents the di-
mensions of the equipment being utilized. In
fishing, the individual firms and many of the
characteristics of their capital are identifiable
and measurable.
Vessel size has been recognized as a deter-
minant of catch and is explicitly recognized in
most of the productivity measures in use.
Beverton and Holt (1957) related gross tonnage
to fishing power, and the Inter-American Tropi-
cal Tuna Commission (IATTC) focuses on the
capacity of a vessel's freezers (Shimada and
Schaefer 1956).
Other researchers have noted that there are
other measures of vessel size that are correlated
with output, among them horsepower and length.
Gulland (1956) and Noetzel and Norton (1969)
experimented with production functions that
included both tonnage and horsepower. Their
results showed that these variables may make
an independent contribution to output. In
fisheries, the possibility of independent con-
tributions should not be overlooked because
there may be a tendency for vessel configurations
to be changed in such a way that fishing power
is increased. This happens especially with horse-
power relative to gross tonnage as old engines
are replaced and also as new vessels are built.
The role of horsepower in the trawl fleet
appears to be that the larger the engine, the
larger the net that can be dragged, the faster
the net can be dragged, or the deeper the water
that can be fished. In this type of fishery, the
profit-maximizing skipper will adjust his net
to obtain the "best" results. Although it has
been noted that trawlers do not often use the
full power of their engines, a larger engine
increases the number of possibilities a skipper
can consider when deciding where to fish and
what to fish for.
In a seine fishery, the role of horsepower is
less clear, except that, ceteris paribus, higher
horsepower increases the "search power" of
the vessel. A better measure of this search
power than horsepower would appear to be
running speed. The only way to obtain this
information is by interview or sea trials.
Hull construction is an identifiable parameter
of a vessel. Throughout the U. S. fisheries, there
has been an increasing tendency to build new
vessels of steel rather than wood, in spite of the
extra initial cost. One would presume, then,
that there are lower operating costs for steel,
or that it is more "productive." It is possible
to test for the effect on productivity of a wood
hull by creating a dummy variable that takes
on the value "one" if the hull is wood and
"zero" otherwise.
The last capital input variable that was con-
sidered was age of the vessel. Most people would
consider older vessels less productive, ceteris
44
paribus, than newer vessels. It is rather simple
to test this hypothesis by including in the tests
the age of the vessels.
Hence, the dimensions of the capital input
will be measured by (1) gross tonnage, (2) horse-
power, (3) construction materials, and (4) age
of the vessel.
Labor — The Crew
Crew size could also be tested as an input
variable in the production function. It seems
reasonable that a larger crew would produce a
higher output, and this should be tested.
One need not work in fisheries very long
before he is made cognizant of the "good captain
hypothesis." That is, the catch of a vessel depends
as much upon the managerial skill of the captain
and crew as it does upon the characteristics
of the vessel. As such, there is no way to test
this hypothesis.
One might attempt to test the good captain
hypothesis by using the years of schooling or
the years of experience of the captain to arrive
at a proxy for his skill. One may suspect on
economic grounds that the best captains would
gravitate to the best vessels because they would
be able to buy the more productive vessels or
be hired away from the poorer vessels. In other
words, part of the higher output of a larger
vessel may not be due to its hardware but to
the superior men running it. In this analysis
we are restricted to crew size as one measurable
variable.
Location
The production function provides for the dif-
ferential productivity that could be due to
location with respect to the fishing grounds
through the variable called land. Vessels from
some ports could have higher productivity than
vessels from other ports by being located closer
to the better grounds. Since these locations
cannot be appropriated, the vessels will allocate
themselves between ports so that effects on net
profits will be dissipated. It is possible to test
whether certain locations are more productive
by creating dummy variables that correspond
to home ports. If their coefficients are statistically
significant, then a location may be either more
or less productive than the average location.
Technological Change
One of the major problems encountered in the
management of fishing power has been the
difficulty in adjusting for technological change.
Attempts have been made to adjust for techno-
logical change, but on the whole they have been
less than satisfactory.
The test for the added productivity of an
innovation should be done when the fleet is
in a period of transition from the use of the
old to the new technique. This method will
hold abundance and availability constant and
therefore, all vessels will have the same op-
portunities. Bell (1966) used a dummy variable
to measure the increased productivity due to
stern trawling. He created a variable that was
1 if a vessel was a stern trawler and 0 if it was a
side trawler. The coefficient of the dummy
variable was the added productivity due to
stern trawling.
This technique can be used to test the added
productivity of any innovation, for example, a
new electronic instrument or the use of spotter
planes or maybe even the use of a radically new
technique such as switching from bait boats to
purse seining. The added productivity of a new
technique would thus become a permanent
attribute of the vessels even after it was no
longer possible to measure the contribution of
the technique, i.e., even after it was universally
adopted.
THE DATA
The New England Trawl Fishery
The National Marine Fisheries Service
(NMFS) has collected comprehensive data on
the landings of the New England trawl fleet
for many years. The data consist of landings
information by trip. The following information
is noted for each trip:
1. Official number
2. Departure date
3. Arrival date
4. Number of days fishing
5. Grounds fished
6. Pounds landed, by species
7. Price/pound by species
The data are stored on magnetic tapes and
can be manipulated with a digital computer.
45
The data used were for the years 1964, 1965,
and 1967. The data were aggregated by vessel
for the whole year. For each vessel, the following
information was produced:
1. Days at sea
2. Days fishing
3. Total trips
4. Days at sea by calendar quarter
5. Days fishing by calendar quarter
6. Trips to major areas: offshore, inshore,
off Canada
7. Pounds caught, by major species
8. Value, by major species
9. Total pounds caught
10. Total value
This information was augmented by the
addition of information from the Merchant
Vessels of the United States (1965), including:
11. Gross tons
12. Horsepower
13. Hull construction
14. Year built
Information from National Marine Fisheries
Service files was added on:
15. Crew size
16. Home port
Vessels with total landings valued at less
than $10,000 were excluded from the sample;
we made the assumption that these were casual
fishermen. There were about 120 vessels excluded
per year, accounting for 3% of New England
landings. Otherwise, no editing was done;
therefore, the sample contains all trips, in-
cluding brokers. Thus, the estimates have built
into them all conditions that vessels from this
fleet experience on the North Atlantic. The total
sample consisted of about 383 vessels per year
or 1,149 vessel years.
The Tropical Tuna Purse
Seine Fleet
The Inter-American Tropical Tuna Commis-
sion (IATTC) kindly let us transcribe landings
data from their files for the years 1966,1967,
and 1968. The data were for the whole year
for the full-time purse seiners. The data trans-
cribed were as follows:
1. Official number
2. Days at sea
3. Landings by species
4. Major area fished: Atlantic or Pacific
This information was supplemented by the
addition of information from the Merchant
Vessels of the United States (1965) including:
5. Gross tons
6. Horsepower
7. Length
8. Year built
Finally, the following information was added:
9. Capacity (American Tunaboat
Association)
10. Crew size (NMFS files)
The total sample consisted of 89 vessels per
year or 267 vessel years. The data were divided
into two periods: (1) when there was unrestricted
fishing for yellowfin and (2) when yellowfin was
restricted to 15% of the total catch. The data
from the restricted season were not used in the
analysis because of the different conditions
following the season closure.
THE STATISTICAL RESULTS
Overall Results
The statistical results of these experiments
are quite encouraging. It is possible to explain
very high variations in catch with a minimum
of information. In the tropical tuna fishery we
can explain approximately 70% of the variation
in the dependent variable, and in the New
England trawl fishery, approximately 84% .
Tests for heteroscaedasticity showed that it
existed in the linear equations. When it is
present, we have inefficient estimators. Log-
arithmic transformation of the variables in both
fisheries removed this problem. Results in both
forms are reported, but only the logarithmic
results are suitable for analytical work.
Several regression experiments were run
46
using a single year's observations in both
fisheries on the same variables. The results
were very encouraging in that there was a high
degree of stability in the coefficients and their
t ratios. These stable results were obtained in
fisheries which, if anything, are notorious for
their variability in almost all aspects: biological,
economic, atmospheric, and oceanographic. Some
results illustrating this stability for the trawl
fishery are shown in Appendix 1.
The New England Trawl Fleet
The statistical results for the New England
trawl fishery were very good. The overall "fit"
of the data in the equations was very high,
especially when one considers the heterogeneity
of this fleet. The equations are rich in informa-
tion in that many of the variables about which
hypotheses were made were statistically sig-
nificant with the right signs.
Because of the unclear nature of variables
discussed, the equations were run using the
alternatives for the same variables where pos-
sible. This will allow direct comparison of the
results. In a sense, we shall permit the data
to decide which are better variables. We will
briefly run through the results according to the
topics covered in the theoretical section.
The following general production function
was established for the New England trawl
fleet:
(3) 0 = /(FT, GRT, HP, CR, AGE, C, PT)
where O = output, either total pounds
or total value,
FT = fishing time, either days
fished or days absent,
GRT = gross registered tonnage,
HP = horsepower,
CR = crew size,
AGE = age of the vessel,
C = construction, 1 if wood, 0
otherwise,
PT = homeport dummy variables.
The equations providing the best results are
shown in Table 1. These equations will be
discussed below. A more complete set of regres-
sions is shown in Appendix Table 1.
The tests of whether total value or total
pounds was the better measure of output in this
fishery are shown in Problems 1 through 4. The
measures of overall fit (R2) are lower in Problems
1 and 2, which use total pounds as the dependent
variable (0.40 and 0.54), than in Problems 3 and
4, which use total value as the dependent
variable (0.83 and 0.83). Thus, the fishermen
appear to have implicitly taken into account
expected prices, expected catch rates, and
steaming time to the grounds and made deci-
sions as to where to go and what to fish. Hence,
relative total revenue appears to reflect the
fishing power of New England vessels. The
more fishing power, the higher revenues are
expected to be.
The most powerful explanatory variables
for either total pounds or total value were the
fishing time variables. That is, the more days
fished or days absent, the higher the total value
and total pounds. On the basis of contributions
to the overall goodness of fit, there is no way
to choose between these two variables. Our
choice, therefore, will have to rest upon their
effects on other variables and on the cost of
gathering the information.
In Problem 3, using total value as the
dependent variable and days fishing as the
measure of fishing time, crew size becomes
statistically nonsignificant and negative. In
Problem 4, when days absent is used, crew size
becomes statistically significant and a very
powerful explanatory variable. Days fishing
appears to be a less desirable measure of fishing
time in that: (1) It is theoretically inferior on
economic grounds as discussed previously;
(2) it causes other important variables to have
the wrong sign; (3) it costs more money to
collect this information; and (4) it is probably
more subject to error.
The vessel size variables used were gross
registered tonnage (GRT) and horsepower (HP).
GRT was the more powerful of these variables
as it was statistically significant in all equations
and explained a large part of output. HP was
not as powerful a variable in terms of its partial
correlation coefficient. However, it was statisti-
cally significant when total value was the de-
pendent variable, indicating that it made an
independent contribution to fishing power.
The variable that indicated the age of a
vessel had a negative coefficient and was sta-
tistically significant in most cases. There are
at least three hypotheses why older vessels
47
Table 1. — New England trawler production functions: alternate specifications.
INDEPENDENT VARIABLE
Dependent variable
LOG DAYS
LOG DAYS
LOG
LOG
LOG
LOG
CONSTRUC-
DUM
DUM
Y
n 2
ABSENT
FISHED
GRT1
HP2
CREW
AGE
TION3
654
674
INT
R
F
Problem 1
Log total pounds (All years)
Reg. Coef.
/.649
.409
.038
-.410
-.240
-.138
-.018
-.084
4.69
.405
98.70
t ratio
18.300
6.340
.525
5.160
4.540
3.780
.776
3.420
Part. Cor. Coef. s
.477
.184
.016
-.151
-.133
-.111
-.022
-TOO
Problem 2
Log total pounds (All years)
Reg. Coef.
1.060
.429
.002
-2.66
-.207
-.024
.011
-.059
3.39
.542
170.28
t ratio
27.800
7.580
.037
4.040
4.470
.752
.533
2.750
Part. Cor. Coef. 5
.636
.219
.001
-.119
-.131
-.022
.015
-.081
Problem 3
Log total value (All years)
Reg. Coef.
.886
.365
.113
-.002
-.107
-.043
-.024
.0006
2.43
.834
724.34
t ratio
47.900
10.800
2.980
.062
3.860
2.280
1.920
.0500
Part. Cor. Coef. s
.817
.305
.088
-.001
-.113
-.067
-.057
.0010
Problem 4
Log total value (All years)
Reg. Coef.
1.080
.373
.074
.347
-.129
.095
.023
.010
1.44
.833
718.97
t ratio
47.600
11.000
1.940
8.830
4.660
5.000
1.790
.855
Part. Cor. Coef. 5
.815
.309
.058
.253
-.136
.146
.053
.025
'Gross registered tonnage.
2 Horsepower.
Construction; equals one if wood, zero otherwise.
4 Dummy variables for year of observation.
5 Partial correlation coefficient.
may be less productive: (1) Older vessels might
tend to have more breakdowns and equipment
that was not in the best working order; (2) older
vessels might have poorer working conditions
and accommodations and, therefore, attract less
able crews; (3) older vessels may embody older
technologies. If the last hypothesis is dominant,
vessels do not become less productive as they
get older, rather old vessels are less productive.
This would have different implications than
the first hypothesis when fishing power factors
are computed.
The dummy variable created for hull con-
struction took on the value 1 if the hull was
wood and 0 if steel. The results using this vari-
able were mixed. In Problem 4, using total
value and days absent, it was positive and
significant. This may mean that ceteris paribus
wooden hulls are 25% more productive.2 There
is no theoretical reason why these results
2 The antilog of 1 is 10. We have 10. 095 which equals
1.25. Therefore, a wooden hull is 25% more productive.
should be obtained. The data in Appendix Table
3 show that the large vessels in the fleet are
steel and the small ones wood, with a very small
overlap. We may be observing an upward
adjustment for the wood vessels because they
fish many fewer days during the most productive
portion of the year.
The tests for locational differences in produc-
tivity were made by creating an array of six
dummy variables, one for each of the major
ports in New England. A "one" was placed in
proper location in the array corresponding to a
vessel's home port and a "zero" in all the
others. Equations showing the results of these
tests are given in Appendix Table 1. In the
logarithmic forms of the equations, there are
no consistent differences between ports when
total value is the dependent variable, the ports
designated "Maine" appear to catch significantly
more and "Boston" significantly less (Problem
10). These differences appear because Maine
specializes in low value species and Boston in
48
high. When weighted by value, these differences
disappear.
On the basis of these statistical tests, we
conclude that the best specification of the pro-
duction function for the New England ground-
fish fleet is shown in Problem 4, where total
value is the measure of output and days absent
is the measure of fishing time. Good descriptions
of the capital variable are given by gross
registered tonnage, horsepower, vessel age, and
construction materials. The contribution of
labor is measurable and important.
The Tuna Seine Fleet
In fisheries such as the tropical tuna fishery,
the species are, in the jargon of the economist,
"joint products." That is, the fishermen take as
much of both species (yellowfin and skipjack)
as they can in an effort to fill their holds as
quickly as possible. They are essentially indis-
criminate between tunas in that they do not
appear to pass up any that they sight solely
because it is the less desirable species, although
such behavior was noted up to about 1950
(Shimada and Schaefer, 1956).
According to IATTC records, the probability
of a successful set on yellowfin is higher than
on skipjack. This leads one to hypothesize
that a ton of skipjack represents in some way
more input than a ton of yellowfin because it
takes more work to catch skipjack. There are at
least two techniques that might be used in this
fishery to determine a weighting system for
output. One technique (which is not used here)
is canonical regression which was developed by
Hotelling and described by Tintner (1952). In
a sense, it is a search technique that "weights"
the dependent and independent variables in
such a way that the sum of the squares of the
unexplained variance of all the variables is
minimized. The second technique3 is to sys-
tematically try different weights (whose sum
is one) for the dependent variable and run a
series of regressions using a common set of
independent variables. The regression that
maximizes the coefficient of determination would
have the weights, which are, in a sense, best.
3 Suggested by Henri Theil during a discussion of
this problem with the author.
The following regression was run in an at-
tempt to arrive at the best weighting system
for output:
(4) Q = f(D, T, CAPAC, GRT, ND, PR,
CR,AGE,HP)
where Q = (aY + |3S + 5B) and (a + 0 + 5) = 1
and Y is tons of yellowfin landed,
S is tons of skipjack landed,
B is tons of bluefin landed,
D is days at sea of each vessel,
T is the number of trips of each vessel,
CAPAC is the capacity of each vessel,
GRT is the gross registered tonnage,
ND is a dummy for new design,
PR is 1 for Puerto Rico home port,
zero otherwise,
CR is the crew size,
AGE is the age of the vessel,
HP is the horsepower of each vessel.
The results of this experiment are shown in
Table 2, where the left hand column shows the
different weights applied to each species. The
column headings are for each year's observations
and for pooled observations. Tests using the
H statistic show that the observations are not
random. Weights of .3 for yellowfin, .4 for
skipjack, and .3 for bluefin are best. This fits our
a priori expectation that a vessel exhibited
more productivity when it caught a ton of
skipjack than a ton of yellowfin. The statistical
results indicate that a vessel does one-third
more work to catch a ton of skipjack than a
ton of yellowfin.
The above experiment presents one approach
to the determination of output in a fishery.
Three alternative specifications of output in
the tuna fishery were used in estimating the
production function. These specifications were
as follows: total value, total pounds, and
weighted total pounds using the weights
determined above.
Selected results of the regression experiments
run are shown in Table 3 and in Appendix
Table 2. The various specifications of the
dependent variable could be explained with
varying degrees of precision. As expected,
weighted total pounds had the highest coef-
ficient of determination, followed by total
pounds, total value, skipjack and yellowfin, in
that order. The actual difference between co-
41)
Table 2. — Regression results using various weights for
tuna species holding independent variables constant.
Weights of yellowim. 1966 1967 1968 All_years
skipjack, and bluefin Rz R2 R2 R2
".
1..
.2
.559
.332
.697
.486
.6.
.1.
.3
.573
.351
.701
.505
.6.
.2.
2
.650
.542
.731
.612
.5.
.1.
.4
.588
.380
.705
.531
.5.
.3.
.2
.730
.785
.758
.757
.5.
.2.
.3
.677
.622
.739
.652
.4.
.1.
.5
.598
.426
.711
.565
.4.
.4.
2
.772
.873
.775
.779
.376.
.286.
.344
.756
.837
.767
.763
.4.
.2.
.4
.703
.711
.748
.698
.4.
.4.
.3
.756
.837
.767
.760
.3.
■5,
.2
.770
.884
.778
.776
.3.
2
.5
.707
.790
.757
.740
.3.
4.
.31
.775
.883
.778
.785
.3,
■ 3,
.4
.764
.868
.774
.783
.2.
•3,
.5
.723
.875
.774
.775
.2.
.5,
.3
.744
.877
.769
.757
■ 2,
4.
.4
.745
.879
.774
.769
•2,
■ 2,
.6
.646
.833
.762
.748
-3,
• 1,
.6
.584
.494
.715
.603
• 2,
.1,
.7
.523
.572
.713
.619
'The "Best" solution.
Source: Economic Research Laboratory, National Marine
efficients of determination in the weighted
total pounds equation and the total pounds
equation is not statistically significant (0.70
vs. 0.68).
The total pounds variable has, of course,
almost the same weights (Vb, Vs, V3) as the
weighted output variable so that, ultimately,
it may be of marginal significance to distinguish
between them in this fishery. Nevertheless,
we cannot know this before further experiments
are conducted.
Total value as a dependent variable is inferior
to total pounds. This tends to confirm our
hypothesis that yellowfin and skipjack are joint
products in this fishery. The weight of skipjack
in total value is less than the weight for
yellowfin and bluefin.4 Therefore, it appears
that the amount for which skipjack can be sold
is not reflected in the extra work done in
catching it, at least relative to yellowfin and
bluefin.
The best production functions for the tuna
fishery are shown in Table 3. The only fishing
time variable available for this fishery was
Fisheries Service. 1970.
4 The relative price weights are .286 for skipjack,
.376 for yellowfin, and .344 for bluefin.
Table 3. — Tuna purse seine production function: alternate specifications.
INDEPENDENT VARIABLE
Dependent variable LOG CAPACITY LOG DAYS LOGH.P.1 66 DUM 67 DUM YINT. R
Problem 1
Log total value
Reg. coef.
.365
.310
.368
.067
.044
t ratio
5.14
3.32
4.66
2.08
2.21
Part. Cor. Coef.
.303
.201
.277
.128
.136
Problem 2
Log total pounds
Reg. Coef.
.438
.373
.339
-.024
.049
t ratio
7.39
4.79
5.15
.914
2.94
Part Cor. Coef.
.416
.284
.304
-.056
.179
Problem 3
Weighted total pounds
Reg. Coef.
t ratio
Part Cor. Coef.
.520
8.41
.462
.416
5.12
.302
.328
4.77
.283
026
.065
946
3.71
058
.224
.196
.587
76.17
.453
.680
113.84
.168
.704
127.07
1 Horsepower
Source: Economic Research Laboratory, National Marine Fisheries Service, 1970.
50
days absent so that alternative specifications
of the equations could not be run. Days absent,
however, was not as important a variable in
this fishery as in the trawl fishery. The reason
for this may, be that there is a basic difference
in the way the vessels in these fisheries operate.
The trawl fishery is a wetfish fishery so that the
vessels are constrained by time when they go
to sea, whereas the tuna boats are freezers,
and they stay at sea until their holds are filled;
hence, there is a different connotation to the
fishing time variable.
The vessel size variables used in the final
equation were capacity and horsepower. Capacity
was the more important of these variables. This
indicates that the industry is justified in using
capacity as an index of a vessel's fishing power.
Several tests were run with gross tonnage in
place of capacity, but the results were not as
good, although they were still meaningful.
Horsepower makes an important independent
contribution to explanation of output. The con-
tribution of horsepower to the increase in the
coefficient of determination, though small at
any point in time, may be important in the
maintenance of an effort series as the com-
position of a fleet changes.
Tests were run using crew size but results
were poor, presumably because there is such
small variation of crew in this fleet (12-14
men). In addition, crew size is defined by custom
and union contract according to the capacity
of a vessel, hence crew size does not give
additional information.
The tuna fleet has two main bases: Puerto
Rico and southern California. To test whether
vessels located in Puerto Rico were more pro-
ductive, a dummy variable was created that took
the value one if a vessel's home port was Puerto
Rico and was zero otherwise. The results were
generally positive but not statistically signifi-
cant. This indicates that the fleet's shift toward
Puerto Rico is because of reasons other than
catching more fish (see Appendix Table 2).
Tests to see if the age of the vessels could
explain some of the variation in output generally
showed that older vessels were less productive
in the linear forms of the equations. When the
logarithmic transformations were made, the
age variable became nonsignificant; hence, it
is not included in the final equations.
The original purse seine fleet consisted of
vessels converted from either military craft
or bait boats. There has been a major expansion
of this fleet since 1963 with vessels designed
specifically for purse seining. To see if these
vessels were superior in a way that could not
be accounted for by either horsepower or capaci-
ty, a dummy variable was created that took the
value one if a vessel were built after 1962 and
zero if built before 1963. It was hoped that
this would pick up technological change. The
results using this were generally positive and
sometimes statistically significant, but the
dummy variable is not included in the final
equations because it was not statistically sig-
nificant in them.
We conclude that for the tuna fishery the
best production function is given by Table 3,
Problem 3, where weighted total pounds is the
dependent variable, days absent is the measure
of fishing time, and capacity and horsepower
are measures of the capital used.
CONCLUSION
The basic assumption underlying this work
is that a production function can adequately
describe the productivity of vessels. The stability
of the estimates arrived at using this technique
rely most upon the constant patterns of economic
behavior. The coefficients would have to be
re-estimated if the ratio of days absent to days
fishing changed significantly in a fleet, or if
the form of regulation changed the pattern of
fishing. Pattern changes are undoubtedly taking
place in the tuna fishery where the quota system
of regulation makes it imperative for vessels
to leave the home port the day the season opens
and to fish as intensively as possible. This makes
vessel utilization in the first part of the year
much higher than it has been historically or
would be without the quota regulation system.
It has probably had the effect of changing the
effective productivity markedly by putting a
premium upon running speed.
Once an estimating equation has been deter-
mined suitable, it should be used as long as
possible, say up to 10 years to provide continuity.
Checks should be made periodically to see if
the equation being used is still appropriate.
The technique of using dummy variables to
51
measure technological change can be a very
powerful means of keeping productivity indices
up to date. Any new device, strategy, or vessel
design can be tested for its ability to increase
productivity as it is being introduced and
therefore, can be permanently built into the
vessel productivity indices.
One of the more important attributes of these
production functions is that they provide a
simple way to test whether information being
gathered is relevant to the task at hand. For
example, fishing days are collected in New
England. Upon further testing it may be
decided that this information is not worth
its cost.
The technique can also provide a way to
handle some of the causes of secular changes in
the productivity of a fleet. For example, in
both of the fleets considered, both vessel size
(GRT and capacity) and horsepower made
significant contributions to the determination
of productivity. Thus, as new vessels are added
to a fleet, their productivity can be estimated
even though they have larger engines relative
to vessel size than other vessels in their size
class. It is also possible to keep estimates of
productivity current as the engines of old
vessels are replaced or upgraded and changes
in crew size are made.
LITERATURE CITED
BELL. F. W. 1966. The Economics of the New England
Fishing Industry: The Role of Technological Change
and Government Aid, Research Report to the Federal
Reserve Bank of Boston, No. 31. Boston: Federal
Reserve Bank of Boston. 216 pp.
BEVERTON, R. J. H., and S. J. HOLT. 1957. On the
Dynamics of Exploited Fish Populations. Ministry of
Agriculture, Fish and Food (U. K.), Fishery Investiga-
tions, Ser. II (19): 533 pp.
GULLAND, J. A. 1956. On the Fishing Effort in English
Demersal Fisheries. Ministry of Agriculture, Fish and
Food (U. K.), Fishery Investigations, Ser. II (20):
44 pp.
Merchant Vessels of the United States. 1965. U. S.
Department of the Treasury. Washington, D. C: U. S.
Government Printing Office.
NOETZEL, B. G., and V. J. NORTON. 1969. Costs and
Earnings in the Boston Large-Trawler Fleet. Economics
of Marine Resources, University of Rhode Island
Agricultural Experiment Station, Bulletin 400.
SHIMADA, B., and M. B. SCHAEFER. 1956. A Study
of Changes in Fishing Effort, Abundance, and Yield
for Yellowfin and Skipjack Tuna in the Eastern Tropical
Pacific Ocean. Inter-American Tropical Tuna Commission,
Bulletin 1(7): 348-469.
TINTNER, G. 1952. Econometrics.
Wiley and Sons.
New York: John
52
Appendix Table 1 — New England production function.
INDEPENDENT VARIABLE
Dependent variable
DAYS ABSENT
DAYS FISHING
CRT
i
CREW
YEAR BUILT
H P
Reg Coef.
t val
Reg. Coef.
f val.
Reg. Coef.
(sal
Reg. Coef.
(val.
Reg Coef
(val
Reg Coef
(val.
f'< <htr"i !
lot.il pounds 64
2677.
2.47
6334.
6.19
•33074.
1.42
10899.
2.64
1140
2 58
Total pounds 65
2215.
2.02
6282.
6.13
-30615.
1.31
7291
1.75
1256
2.79
Total pounds 67
■150.6
.13
6198.
6.87
-33566.
2.70
11076.
2,95
607.6
1 59
Pooled total pounds
2498.
4.15
6395.
11.41
■38441.
3.83
8561.
3.73
896.8
3.67
Log total pounds 64
.4914
8.87
.9850
11 59
- 1912
1.62
.2665
2.18
.7188
12.76
Log total pounds 65
.2530
4.48
9640
12.56
-.0500
.47
.2510
2.30
6430
12.38
Log total pounds 67
.1032
1.85
.9800
11.84
.0206
.18
.5560
4.67
91111
15.78
Problem 2
Total pounds 64
5687.
8.01
5415.
5.74
■42049.
2.26
6252.
1.62
951 6
2.31
Total pounds 65
5455.
7.28
5375.
5.62
^4 1627
2
22
3234.
.83
1081.
2.54
Total pounds 67
5037.
6.40
5153.
5.94
-45400.
4
07
5828.
1.63
372.8
1.02
Pooled total pounds
5796.
13.87
5365.
10.18
-44789.
5
24
4395.
2.05
718.9
3.15
Log total pounds 64
.8527
11.08
.7662
8.95
-.1051
98
.2256
1.93
.4669
7.18
Log total pounds 65
.5070
6.00
.8320
10.45
-.0200
20
.2170
2.02
.5050
8.39
Log total pounds 67
.3263
4.07
.8940
10.55
-.0330
32
.5140
4.41
7750
11.43
Problem 3
Total value 64
KK9 '
24.77
214 2
6.32
1939.
2.51
1284
.94
103.7
7.08
Total value 65
884.3
22.61
200.5
5.47
2082.
2.50
41.75
28
117 7
7.31
Total value 67
728.2
21.30
204.4
7.43
-380.6
1.00
8299
73
72.24
6.20
Pooled total value
889.1
4 2 ss
223.2
11.47
416.2
1.19
26.93
34
94 ss
11 19
Log total value 64
.9603
25.20
.5459
9.34
.0497
6 1
1779
2
12
.0927
2.39
Log total value 65
.7880
23.95
.4990
1118
1714
2.75
.2060
3
23
.0830
2.75
Log total value 67
.6848
22.52
.5530
12.22
.045
.73
.2620
4
03
2730
8.66
Problem 4
Total value 64
566.1
19.78
30.36
km
8706.
11.61
148.3
.95
89.94
5.40
Total value 65
582.1
18.24
11 16
.27
8778.
11.00
119.1
.71
103.6
5.73
Total value 67
538.5
21.31
154.0
5.52
1324.
3.69
90.76
79
77.16
6.62
Pooled total value
607.8
34.90
100.4
4.57
4476.
12.58
69.60
.78
96.03
10.03
Log total value 64
1.349
23.08
.2402
3.69
.3465
4.26
.1633
1.84
2122
4.29
Log total value 65
1.140
21.38
.2290
4.53
.4190
6.69
.2017
2.95-
-.1560
4.08
Log total value 67
.9900
21.81
.3930
8.19
.2200
3.65
.2860
4.34
.0360
.94
Problem 5
Total value 64
896.5
24.43
206.7
5.95
1759.
2.21
133.1
.97
101.7
6.88
Total value 65
8940
22.44
190.9
5.10
1854.
2.17
48.61
.33
114.3
7.01
Total value 67
730.7
21.13
201.1
7.04
-426.9
I MS
80.81
.71
71,4
6.06
Pooled total value
896.1
42.24
214.3
10.69
262.3
.73
28.70
.36
92.45
10.78
Log total value 64
.9690
25.22
.5154
8.38
.0369
.45
.1855
2.21
.082
2.09
Log total value 65
.7960
23.92
.4750
si
.1600
2.57
2090
3.29
.0740
2.40
Log total value 67
.6877
22.36
.5450
11 58
.0390
.63
.2600
4.00
.2710
8.54
Problem 6
1 .i.il * dur (.4
563.6
19.92
61.4
1.58
9046
12.06
114 0
.74
96.96
5.84
Total value 65
581.0
18.35
36.43
.88
9040.
11.33
87.10
.52
111.2
6.12
Total value 67
535.6
21.18
165.5
5.74
1460
3.96
94.66
.83
79.59
6.78
Pooled total value
605.1
34.94
122 5
5.44
4729.
13.16
55.37
.62
101.4
10.64
Log total value 64
1.352
23.26
.2846
4.24
.3536
4.37
.1469
1.66
-.2040
4.15
Log total value 65
1.140
21.61
.2700
5.17
.4240
6.83
.1890
2.79
-.1460
3.83
Log total value 67
.9890
21.91
.4200
8.56
.2330
3.87
.2880
4.39
.0390
1 04
Problem 7
Total value 64
921 1
24.25
198.5
5.49
2203.
2.47
63.16
.45
102.1
6.86
Total value 65
927.3
22.57
187.0
4.84
2102.
2.20
-21.7
.14
116.0
7.10
Total value 67
865.3
22.92
168.0
6.05
-279.3
.72
-.4334
.003
68.96
6.15
Pooled total value
951.57
43.00
200.8
9.92
502.3
1.33
-25.40
.32
92.60
1092
Log total value 64
.9626
23.87
.4611
6.80
.2064
2.24
.1467
1 'i
TOO
2.29
Log total value 65
8039
22.93
.4610
8.90
.2760
3.91
.1610
2.53
09526
2.79
Log total value 67
.7550
21.93
.4660
9.35
.1969
2.95
1749
2.73
2.557
7.24
Problem 8
Total value 64
561.1
19.55
63.58
1.56
8686.7
9.68
132 7
.85
100.4
5.98
Total value 65
578.9
17 91
40.95
.93
8806.1
9 12
7642
.45
115.3
6.24
Total value 67
554.6
21.05
162.5
5.55
1308.
3.33
143.3
1.23
79.43
6.78
Pooled total value
606.3
34.49
130.1
5.61
4157.
10.30
96.70
1.07
103.4
10.79
Log total value 64
1.344
21.93
.2518
3.39
34 1 3
3.59
.1953
2.20
-.1900
3.43
Log total value 65
1.130
19.79
.2830
4.78
4050
5.40
.2000
2.90
■ 1240
2.80
Log total value 67
1.006
19.51
.3920
7.14
2880
4.12
.2780
4.12
.0340
.75
Problem 9
Total pounds 64
3982.
3.53
.6973
4.44
-10062.
.38
10111
245
754.2
1.71
Total pounds 65
3546.
3.09
4981.
4.62
-17193.
.65
7576.
1.80
867.6
1.90
Total pounds 67
2983.
2.32
5251.
5.56
-26509.
2.02
8165
2 14
512.3
1 14
Pooled total pounds
4386.
6.90
5180.6
8.92
-27597.
2.55
7471
3.26
639 2
2.63
Log total pounds 64
.5389
9.57
.6974
7.37
.2183
1.70
.2400
2.03
1,1.111,9
10.85
Log total pounds 65
.2827
4.82
.7771
8.96
.3120
2.64
.2310
2.17
.6550
11.47
Log total pounds 67
1420
2.21
.8430
9.06
.2820
2 27
.4879
4.08
9330
14.17
fr,,blem f"
Total pounds 64
5955.
8.47
3774.
3.78
-10501
.48
6796.
1 77
613.4
1.49
Total pounds 65
5828.
7.80
3893.
3 84
-5828.
.71
4626
1.18
731 9
1.71
Total pounds 67
6698
8.50
3959
4.52
-26853.
2.28
3868
1 1 1
371.3
1.06
Pooled total pounds
6402
15.50
4062.
7 45
■2.16 14.
2.49
4586
2 16
528 1
2.34
Log total pounds 64
8643
10 94
5303
5.53
24291
1 98
.25445
2 22
4233
5 93
Log total pounds 65
4KI4
5 52
.6810
7 52
3260
2.84
.2320
2 21
.5340
7 90
Log total pounds 67
3600
404
.7540
7.95
2370
1.96
.4810
4.13
7800
9 86
:Hor>
registered tonnage
53
Appendix Table 1. — New England production function. — (Continued.,
INDEPENDENT VARIABLE
. _ — _
Dependent variable
CONSTRUCT.3
MAINE
4
GLOUCESTER4
BOSTON4
NEW BEDFORD4
RHODE ISLAND4
Reg. Coef.
fval.
Reg. Coef.
rval.
Reg. Coef.
rval.
^eg. Coef.
rval.
*eg. Coef.
rval.
Reg. Coef.
rval.
Y INT
R*
R1
F
Problem 1
Total pounds 64
-90900.
.368
362
44.37
Total pounds 65
74900.
.387
381
46.096
Total pounds 67
11300.
.285
277
29.541
Pooled total pounds
38300.
.339
337
117.5
Log total pounds 64
1.83
.650
646
141.2
Log total pounds 65
2.35
.592
588
105.85
Log total pounds 67
1.52
.568
563
101.06
Problem 2
Total pounds 64
-355000
.451
445
62.48
Total pounds 65
-222000.
.460
453
61.915
Total pounds 67
-209000
.354
347
42122
Pooled total pounds
-261000
.426-
424
169.7
Log total pounds 64
1.70
-681
677
162.0
Log total pounds 65
2.22
.608
604
113.29
Log total pounds 67
1.49
.578
528
107.10
Problem 3
Total value 64
48000
.877
876
543.1
Total value 65
46200
.871
870
494.17
Total value 67
-18600
.814
813
338.33
Pooled total value
-35000
.846
845
1260
Log total value 64
1.56
.845
843
415.0
Log total value 65
1.86
.876
878
516.19
Log total value 67
1.63
.845
843
419.96
Problem 4
Total value 64
-82900
.842
840
404.6
Total value 65
-85500
.838
846
378.53
Total value 67
48100
.814
812
336.77
Pooled total value
-70700
.808
807
960.3
Log total value 64
1.28
.828
826
365.3
Log total value 65
1.54
.858
857
442.78
Log total value 67
1.26
.839
838
402.3
Problem 5
Total value 64
-5419
.96
41800
.878
876
452.6
Total value 65
-7455
1.24
-37700
.872
870
412.89
Total value 67
-2020
.44
-16200
.815
812
281.3
Pooled total value
■5982
1.81
-28000
.847
846
1053.
Log total value 64
-.0581
1.58
1.67
.846
844
347.6
Log total value 65
-.0420
1.56
1.94
.877
875
432.17
Log total value 67
-.0180
.66
1.67
.845
843
349.59
Problem 6
Total value 64
19489
3.16
-105000
.846
844
346.8
Total value 65
17085
2.61
-104000
.841
839
321.68
Total value 67
7184
1 -f,
-56600
.815
813
282.14
Pooled total value
14256
3.95
-87100
.810
809
813.1
Log total value 64
.0923
2.40
1.12
.830
828
309.3
Log total value 65
.0780
2.74
1.40
.861
859
376.95
Log total value 67
.0670
2.35
1.14
.842
840
340.43
Problem 7
Total value 64
-690.9
.12
11319.
1.55
-808.7
.12
4635
.62
1 1 74
.17
11044
1.52
51000
.881
878
251.3
Total value 65
-1323
.21
12207
1.55
-853.1
1 1
8115
1.00
1470
20
14986
1.91
49700.
.876
873
232.05
Total value 67
-382.1
.09
3708
.61
-5542
.94
-7683
1.23
19939
3
42
7313.1
1.19
-13300.
.839
835
178.81
Pooled total value
-1253
.45
8604
1.99
-2798
.67
2738
.62
7520
1
86
11149
2.57
-34700.
.854
852
605.9
Log total value 64
-.0529
1.36
.1055
1.95
.0044
.02
.0535
.97
0251
50
.0363
.67
1.67
.853
849
197.7
Log total value 65
-.0310
1 09
.0790
1.96
-.0007
.002
-.0280
.68
0320
87
.0820
2.02
1.88
.884
881
248.8
Log total value 67
-.0164
.57
.0110
.25
-.0960
.67
.1620
3.54
.1060
2
54
.0320
.74
1.827
.862
858
214.64
Problem 8
Total value 64
18171
2.79
-7228
K8
-7694
.98
6246
75
2402
.32
-14404
1.77
-96600.
.848
844
190.2
Total value 65
18068
2.59
-5969
.68
-8943
1.04
4459
.47
.3492
.42
-9282
1.05
-99100.
.842
838
174.48
Total value 67
7666.
1.64
-14675
2.31
-8656
L39
7234
1.10
16873
2.77
-9428
1.45
48900.
.822
818
159.45
Pooled total value
14241
3.78
-12459
2.56
-8300
1.75
2039
41
8035
1.75
-12687
2.59
-78900.
.812
811
449.3
Log total value 64
.0581
1.43
-.0034
.06
.0268
.49
.0283
.49
0046
.09
-.1293
2.30
1.15
.838
834
175.9
Log total value 65
.0600
1.93
-.0160
.39
.0037
.09
.0200
.45
0120
.30
-.0510
1.17
1 40
.863
859
206.2
Log total value 67
.0634
2.08
-.0260
.57
-.0370
.81
.0740
1.53
.0560
1.28
-.0220
.49
1.19
.843
839
185.77
Problem 9
Total pounds 64
-339024
1.96
562668
2.60
351898
1.69
59089
.27
29683
.15
358050
1.66
51500.
.408
392
23.4
Total pounds 65
-380585
2.19
479210
2.18
332023
1.59
103437
.45
51156
.30
355627
1.62
189000-
.418
401
23.393
Total pounds 67
-102455
.67
85775
.42
-124072
.62
354221
1.67
409081
2.06
215326
1.03
257000.
.330
313
16.979
Pooled total pounds
-272358
2 8 1
386757
3.12
181225
1.51
52889
42
135014
1.16
321650
2.58
196000.
.378
372
62.83
Log total pounds 64
-.0678
1.25
.2383
3.16
.0319
44
.1261
1.63
.0587
.84
.0658
.87
2.15
.694
686
77.14
Log total pounds 65
-.0280
.60
.1960
291
.0079
.12
.1300
1.86
.0190
.31
.0190
.28
2.41
.638
628
57.558
Log total pounds 67
-.0560
of,
-.0330
.42
-.1400
1 75
.3240
3.80
.1670
2.13
-.1470
1.81
1 Sll
.598
587
51.203
Problem 10
Total pounds 64
-265530.
1.66
475561.
2.37
319786.
1.65
(7985.
.23
45782.
.25
205331.
1.03
-312000.
.487
473
32.25
Total pounds 65
-298694
1 85
429528.
2 10
302920.
1.53
S2596.
.39
20693.
1 1
230513.
1.13
-187000.
489
475
31.19
Total pounds 67
-106493.
.76
-52962.
28
-314526
1.69
532042.
2 71
655419.
3.60
50766.
.26
105000.
430
415
25.934
Pooled total pounds
-209963.
2 3 7
278594.
2.43
103667.
.93
118571.
101
250792.
2.33
159885.
1.38
-130000
965
460
89.87
Log total pounds 64
-.0017
.03
1834
2.53
.0584
.82
0926
1.23
.0722
1.07
-.01951
.27
1.88
711
704
83.81
Log total pounds 65
.0050
.12
.1670
2 52
.0170
.27
.1160
1.66
.0290
.47
-.0240
.37
2.22
645
635
59.29
Log total pounds 67
-.0330
.64
-.0390
.50
-.1300
1 65
.3040
3.62
.1910
2.50
-.1570
1.96
1.72
610
599
53.770
3£quak 1 if wooden vessel. 0 otherwise.
4 Equals 1 if vessel's homeport Puerto Rk
54
Appendix Table 2. —Alternative specifications of production functions for vessels in the eastern tropical Pacific tuna fishery, 1966, 1967, 1968.
Dependent variable
INDEPENDENT VARIABLE
CAPACITY A^T HORSEPOWER CRT' CREW ™f*J2° Jgg gfjfj DUM 66< DUM 67« VINT F
Problem 1
Total value
Linear -
Reg. Coef.s
.407
.594
.124
f ratio
7.590
.285
4.820
Log
Reg. Coef.5
.365
.310
.368
t ratio
5.140
3.320
4.660
Problem 2
Total pound
i
Linear -
Reg. Coef.5
3.840
5.620
.751
f ratio
1.030
3.890
4.180
Log
Reg. Coef.5
.438
.373
.339
t ratio
7.390
4.790
5.150
Problem 3
Weighted total pounds
Linear - Reg. Coef.5
4.810
6.740
.605
f ratio
11.800
4.270
3.080
Log - Reg. Coef.5
.520
.416
.328
t ratio
8.410
5.120
4.770
Problem 4
Total pounds
Linear - Reg. Coef.5
f ratio
Log - Reg. Coef.5
t ratio
3.930
5.780
179.000
10.700
3.960
2.880
.410
.061
.242
1.67
4.870
5.860
3.330
2.87
37.300 - 6.920 36.890 96.28 .643
1.350 2.440
.067 .044 - .196 76.17 .587
2.080 2.210
582.000 202.000 - 660.400 121.20 .694
2.750 1.030
.024 .049 .453 113.80 .680
.914 2.940
255.000 732.000 - 116.300 132.00 .712
1.180 3.500
.026 .065 .168 127.00 .704
.946 3.710
414.000
209.000
540.000
1.980
1.060
2.810
.010
• .037
.051
.304
1.290
2.870
-1735.000 102.00 .697
.559 70.74 .649
Problem 5
Weighted total pounds
Linear - Reg. Coef.5
t ratio
Log - Reg. Coef.s
t ratio
4.440
6.300
8.030
3.940
.448
5.030
.217 1.230 -283.00 370.00 13.30 -143.000 -668.000 -479.000 1856.000 69.68 .648
.829 1.750 1.54 1.97 1.22 .458 3.290 3.890
.065 .317 1.56 - .010 - .039 6.580 - .762 75.87 .653
.586 4.111 2.53 .287 1.300 3.530
'Gross registered tonnage.
2 Equals one if vessel's home port is Puerto Rico, zero otherwise.
3Equals one if vessel was built after 1962, zero otherwise.
4 Dummy variables for year of observation.
5 Regression coefficient.
Appendix Table 3. - New England trawl fleet: average vessel data by tonnage class, 1964, 1965, 1967.
Number of
observations
GRT
Days
absent
Days
fishing
Trips
Horsepower
Year
built
Crew
Construction
(percent wood)
Measures of
output
Tonnage class
Thousands of
pounds
Total
value
($1000)
0-50
492
30
118
4S
87
163
42
3.6
98
808
37
51-100
354
7(1
149
89
36
253
43
5.9
93
1086
83
101-150
147
120
162
104
24
349
44
7.9
88
1225
118
151-200
57
170
168
96
20
479
44
8.6
24
1142
114
201-250
33
229
235
155
24
604
45
14.4
0
2672
242
251-300
15
271
224
152
2 3
630
38
13.7
0
2591
253
301-400
15
313
235
141
17
623
36
9.0
0
4942
191
400
6
495
221
126
24
503
44
12.7
0
3439
260
55
Appendix Table 4. - Tropical tuna seine fleet: average vessel data by tonnage class, 1966, 1967, 1968.
Number of
observations
Capacity
Days
at sea
GRT
Horsepower
Year
built
Measures of ot
tput
Capacity class
Total
value
($1000)
Total in
thousands of
pounds
Weighted
total in
thousands of
pounds
100-199
47
173
152
210
508
46
236
1504
1388
200-299
83
251
168
370
731
48
292
2542
2401
300-399
62
346
172
421
908
51
360
2550
2461
400-499
24
453
182
482
1100
50
389
2765
2766
500-599
19
537
162
619
1281
56
523
3749
3719
600-699
5
650
133
673
1649
59
448
3166
3319
700-799
4
793
180
856
1589
63
817
6447
6946
800-899
6
811
191
804
1600
64
781
6016
6479
900-999
12
924
161
793
1850
53
637
5092
5492
K
5
1067
171
855
1600
43
687
5751
6454
56
Optimal Fishing Effort in the Peruvian Anchoveta Fishery
Edilberto L. Segura1
ABSTRACT
This paper introduces a new approach to measuring technical change, increased
skills of the skipper and the fishermen, water temperature, etc., to obtain a better measure
of fishing effort and therefore a revised estimate of the optimum quantity to be landed.
The revised technique used adjusts the level of landings to an index rather than the
level of fishing effort, indicating the level of landings that would have resulted in
previous periods if the current landings/effort relationship is used.
The revised yield/effort relationship which results yields 16.2 million ton-trips as
the optimal fishing effort, as opposed to the 23 million ton-trips which were obtained
without this measure of technical change.
INTRODUCTION
During the last decade the Peruvian fishing
industry has become one of the most important
elements of the Peruvian economy. In 1969
Peruvian exports offish meal and fish oil reached
U. S. $195 million, or 30% of total Peruvian
foreign exchange earnings during that year.
Almost all fish meal and fish oil production has
utilized "anchoveta" {Engraulis ringens) as raw
material. Total landings of anchoveta have
increased from 1.9 million metric tons in 1959 to
8.9 million metric tons in 1969. This increase in
landings represents an average annual rate of
growth of 18% .
In recent years, due to the rapid expansion
of the industry, its importance to the Peruvian
economy, and the size of the landings, several
studies have been made to determine the maxi-
mum sustainable yield of the Peruvian fish
stock and the optimal level of fishing effort
(Boerema et al., 1965; Schaefer, 1967, 1970;
Gulland, 1968). Although these studies contain
extensive discussions of fishing effort, there
remain some doubts about the adequacy of the
measures used to evaluate fishing effort. As a
result, the estimation of the optimal level of
fishing effort has been biased. In this paper I
attempt to estimate the optimal level of fishing
effort taking into consideration the effect of
input variables not previously included, such as
technological change, increased skills of skippers
and fishermen, water temperature, etc.
CONCEPTUAL ISSUES
In a bioeconomic model, fishing effort is an
index or proxy for several inputs that partici-
pate in the fishery, including capital, labor,
management, technological change, and other
variables. Although the fishing effort index
might vary for different fisheries, it can be
generalized as being the product of fishing time
(number of days in grounds, number of trips
made, number of hours fished, etc.) multiplied
by some measure of fishing power (gross tonnage,
length, engine horsepower, etc.). This measure
should be a proxy for capital and labor. The
resulting measure of fishing effort should be
corrected by such factors as technological
change (introduction of power block, echo
sounder, steel vessels, etc.), changes in manageri-
al and fisherman skill, and other variables that
represent changes in fishing power. To determine
the optimal fishing effort in the fishery and the
maximum sustainable yield, most of the studies
of the Peruvian stock have utilized the following
Schaefer production function:
1 Ph.D. Candidate, Columbia University and Economist
of Bailey, Tondu, Warwick and Company, Ltd., New York.
(1) CIE = a
- bE or, C = aE- bE2
Where:
C
= Total landings of anchoveta
E
= Fishing effort
a,b
= Parameters
57
Schaefer (1967) used as a measure of fishing
effort the average number of boats during the
period, adjusted for changes in the size composi-
tion of the fleet. In 1970, Schaefer, recognizing
that this measure could generate some bias,
utilized as a measure of fishing effort the number
of trips made by the fleet, times the average
vessel capacity (ton-trips). This unit was also
utilized by Boerema et al. (1965). However, all
these studies ignored the effect of technological
change and increased fisherman skills on the
level of landings. This neglect arose from the
difficulty in quantifying these variables.
Although for several advanced fisheries these
two variables can indeed be ignored, such
neglect is questionable in the Peruvian anchoveta
fishery. The importance of such variables was
recognized by Gulland (1968).
The size of the Peruvian fleet has increased
from 462 units in 1959 to 1,064 in 1962 and to
1,836 in 1964. After 1964, fleet size began to
decline, reaching 1,308 units in 1969. From
these figures it is clear that up to 1964 a large
percentage of the skippers and fishermen were
fishing for the first time. However, after 1964,
with the reduction of the fleet size, only the
most efficient skippers remained in the fishery.
This situation and the experience gained by the
fishermen after several years of operations, have
served to increase the average skill of the
fisherman.
In addition to increased labor skills, during
the last decade several technological innovations,
such as power block, echo sounder, steel vessels,
and pumps for transferring the fish from the
net to the hold, have been gradually introduced
into the fleet. In 1969, 92% of the fleet had at
least three of these items of gear, as opposed
to 79% two years before. If a measure of fishing
effort omits the effect of increased labor efficiency
and technological innovations, then the most
recent estimates of fishing effort will be biased.
The estimation of the optimal fishing effort
will also be biased.
The type of bias that will be introduced by
omitting the increased efficiency of the fleet can
be deduced from Figure 1.
In Figure 1, if the efficiency of the fleet
increased during periods 1 to 3, the observed
data for catch and effort will produce curve A.
However, the relationship of catch to effort in
terms of efficiency in year base "0" is given by
curve B. The effect of ignoring increased ef-
ficiency would be to underestimate the most
recent measures of fishing effort. If the observa-
tions of fishing effort, unadjusted by efficiency,
are consistent from year to year, they still will
give a correct measure of the maximum sus-
tainable yield, as it is shown in Figure 1.
However, the determination of the optimal level
of fishing effort, in terms of some constant level
of efficiency, will be biased. Usually, one is
interested in obtaining the optimal level of
fishing effort in terms of efficiency during the
current period. This relationship is given in
Figure 2, where "period 3" is the current period.
Since vessels are more efficient during period
3, to obtain the maximum sustainable yield
C-2, the industry will require a smaller effort
in terms of number of ton-trips than the effort
used in period 2. In fact, instead of requiring
an effort E2, the industry will require only an
effort E'2, considering the higher efficiency of
vessels in period 3. It is obvious that to obtain
an unbiased optimal level of fishing effort at
current efficiency, it will be necessary to adjust
the index for fishing effort to reflect technological
change, changes in fishermen skills, and other
variables.
Although the construction of an index for
fishing effort that includes technological change
and other such variables is the ideal method to
determine an unbiased level of optimal fishing
effort, usually it is not easy to construct such
an index. This is because several of the above-
mentioned variables are difficult to quantify.
When this is the case, an alternative approach
has to be devised.
The alternative approach that is used in this
paper is to adjust the level of landings obtained,
rather than the level of fishing effort, for changes
in efficiency. That is, given the observed un-
adjusted fishing efforts and the landings in
several periods, the problem is to obtain a catch-
to-effort relationship that will show that level
of landings that would have been obtained in
the several periods if vessels of efficiency of the
current period would have been used. This
adjusted curve and the actual observed curve
are shown in Figure 3. The optimal level of
fishing effort in terms of vessels of current
efficiency (E*2 in the figure) will be obtained by
maximizing catch in curve A.
The difference between this approach and the
58
MSY
"~ Curve Bi Catch-effort
relationship in terms of
efficiency of period 0.
Curve Aj
Catch-effort
relationship
as observed
Figure 1. — Biased estimate of fishing effort due to an underestimate of increased efficiency.
Catch
"2 2
Level of effort required Actual level of
to obtain the catch of period 2 effort observed
using vessels of effic. 3. in period 2.
Effort
(ton-trips)
Figure 2. — Optimal fishing effort based on current efficiency levels.
59
MSY
Actual level
of landings
observed in
period 2
Level of
landings
that would
have been
obtained in
period 2, if
vessels had
efficiency
of period 4.
Effort
(ton-trips)
Figure 3. — Actual and adjusted catch effort curves.
first one is that in the first approach the fishing
effort is adjusted for efficiency changes and
catch remained at the observed levels; in the
second approach the catch is adjusted for ef-
ficiency changes and the fishing effort remains
at the level observed. The second approach has
the advantage that it can be more easily handled
with statistical techniques. It should be noted
that curve A in Figure 3 gives the level of
landings that would have been obtained in past
periods if vessels at that time had the efficiency
of the current period. Actually, this curve has
not been observed; and the maximum of the
curve, although it indicates the optimal level
of fishing effort in terms of current efficiency,
will not give the maximum sustainable yield
of the stock. The maximum sustainable yield
will in fact be given by curve B, as it was
shown in Figure 1.
Since curve A in Figure 3 is actually the
relationship of effort to catch keeping all other
variables (including efficiency) constant, the
multiple regression technique can be applied.
In fact, the statistical meaning of a partial
regression coefficient is that it measures the
effect of the independent variable on the de-
pendent one, keeping all other variables constant.
The use of the regression analysis to obtain the
optimal fishing effort is presented below.
The logistic model as presented by Schaefer
(1957) and reproduced in equation (1) is a
stochastic rather than an exact relationship:
(2) C = aE - bE2 + e
Where "e" is an error term.
In this model, if the measure of effort used
were a proxy for all the several inputs utilized
when fishing and affecting catch, then the error
term "e" should be randomly distributed. That
is, no other input variable, when added to
equation (2), should be statistically significant
in explaining changes in the level of catch. In
fact, if no variables have been omitted in
equation (2) (all of these are represented in the
proxy fishing effort), then no sign of auto-
correlation of the error term should exist. If
this is the case, one could conclude that the
measure of fishing effort used is adequate and
that it can be reliably used to estimate both the
maximum sustainable yield and the optimum
level of effort.
We can further test if the measure of fishing
effort is adequate by introducing into equation
(2) input variables such as technological change
(JO
and crew size. If we did this the following
equation would result:
(3) C = axE - bxE2 + cL + dT + e
Where:
C — Total landings
E = Fishing effort
L = Labor employed or crew size
T = Technological change expressed
as T = 1 in period 1, T = 2 in
period 2, T = 3 in period 3, etc.
If the coefficients of "L" and "T" are statisti-
cally significant (as given by their t-values),
it means that the measure of fishing effort used,
"E," did not adequately include the effect of
these variables on catch. Consequently, the use
of equation (2) alone would produce biased
estimators of the coefficients "a" and "6" of
"E" and "E2," respectively. In this case we
can either correct the measure of fishing effort
used (which is the first procedure indicated in
Figure 1) or we can isolate the effect of other
variables on catch using a multiple regression
equation that would include these variables
(which is equivalent to the second approach
indicated in Figure 3).
If the second approach is used, technological
innovation and crew size must be kept at a fixed
level in equation (3). Usually this would be at
the current levels. After this is done we can
obtain the true value of fishing effort by maxi-
mizing catch in equation (3). Keeping the effect
of "T" and "L" on catch at some constant level
K, equation (3) would become
C
axE - bxE2 + K
or
(4) (C - A') = aiE - bxE2
Which is the model as developed by Schaefer
(1957) after the effect of technological change
and crew size is removed. The optimal level
of fishing effort, at constant vessel capacity
and crew size, that will maximize catch is
given by equating zero to the first derivative
of equation (4) as follows:
d(C-K)
de
a, -2 6, E = 0
or
(5) Optimal fishing effort = E* =
ax
2 6,
STATISTICAL RESULTS
Using the data presented in Table 1, several
regressions were made to test for the adequacy
of the measures of fishing effort available to us.
In Table 1, total landings is defined as the catch
by the fishermen in thousands of pounds. The
unit used for fishing effort is the number of
trips made times the average vessel capacity.
Data on fishing effort was compiled by the
Instituto del Mar del Peru, and it is supposed
to be adjusted for the effect of closed seasons,
strikes, and for some changes in gear efficiency.
Other variables included in the analysis are the
number of fishermen employed in the industry,
the size of the bird population (which is supposed
to be an important element in fishing mortality),
and veda (closed) seasons.
As has been recognized by Gulland (1968)
and by Schaefer (1967), because of the rapid
growth of the Peruvian fishery, it has not
remained in steady state equilibrium in every
year. Under these circumstances, the use of a
relationship of catch to effort will produce too
high an estimate of steady state abundance
and catch for a given fishing effort. One way to
correct this situation is to use the "Gulland
Method" (Gulland, 1961) in which the total
landings are related to the average effort exist-
ing during the life span of a fish in the fishery,
which is approximately two years. This method
has been used in this paper.
Schaefer (1970) used the same data presented
in Table 1 to estimate the maximum sustainable
yield of the stock and the optimum level of fishing
effort. I have added observations for the year
1968-1969. The regression equivalent to the
one used by Schaefer in 1970 is as follows:
(6) C = 0.7769 E - 0.1706 E2
(8.6) (-3.8)
Coefficient of Determination {R2) = 0.84
Durbin-Watson Statistic (D-W) = 0.7
Standard Error of Estimate (SEE) = 813
Figures in parentheses are £ -values.
Equation (6) is useful for finding the maximum
sustainable yield of the fishery. The estimated
MSY is given at 8.8 million metric tons. This
value is very close to the value of 8.5 million
metric tons obtained by Schaefer (1970). By
observing the data of total landings in Table 1
we cannot appreciate the danger of overfishing
61
Table 1. — Catch and Effort Data for the Peruvian
Anchoveta Fishery, 1960-1969.
Catch by
Fishing
Adult
fishermen
effort
Number of
bird
Catch per
Fishing
103 metric-
103 ton-
fishermen
population
unit of
year
tons
trips
employed
103
effort
(1)
(2)
(3)
(4)
(5)
1960-61
3,934
6,367
8,800
12,000
0.551
1961-62
5,502
8,131
11,750
17,000
.603
1962-63
6,907
11,788
19,100
18,000
.478
1963-64
8,006
17,866
20,100
15,000
.376
1964-65
8,037
21,329
18,900
17,300
.376
1965-66
8,096
22,058
19,000
4,300
.356
1966-67
8,242
20,845
17,800
4,800
.435
1967-68
9,818
19,874
17,500
4,500
.472
1968-69
10,088
22,350
19,600
5,000
.421
Source: (1), (2), (4): Years 1960-1968, from Schaefer (1970)
Year 1968-1969, from Instituto del Mar del
Peru, Resumen General dela Pesqueria,
Lima, 1970.
(3): From Sociedad Nacional de Pesqueria, unpublished
materials.
in the Peruvian stock, since landings have
increased throughout the period. However, by
analyzing data for calendar years up to 1969
a different picture of the situation is observed.
During recent years annual landings have been
as follows:
Year
1961
1962
1963
1964
1965
1966
1967
1968
1969
Million Metric Tons
4.58
6.28
6.42
8.80
7.23
8.53
9.82
10.44
8.95
It is clear from these data that landings will
not continue to increase at the rates experienced
in the past, and that we can only expect to
see fluctuations in landings around the MSY,
if fishing effort is kept under control.
The result given by equation (6) as to the
optimal level of fishing effort is less than
satisfactory. The value given by this equation,
and which is close to that obtained by Schaefer
(1970), is 23 million ton-trips. Observing the
data in Table 1 we can see that this value of
fishing effort has not been obtained up to now.
This result is very unrealistic since it says that
the Peruvian fishery has actually surpassed
the MSY but has not yet reached the optimum
level of fishing effort. However, from the
discussion in the first part of this paper, it
seems that the measure of fishing effort used
is inadequate.
In equation (6) we can see that the value of
0.7 for the Durbin-Watson statistics indicated
that there is a strong autocorrelation of the
error term. This level of autocorrelation is an
indication that important variables have been
omitted from the equation. Using the procedure
indicated above, several input variables will be
introduced in equation (6), in order to determine
their significance and the bias of the estimation
of fishing effort. Some of the regressions that
were run are the following:
(7) C = 0.7022 E - 0.2167 E2 R2 = 0.97
(15.7) (-9.5) D-W - 1.8
+ 541.0 T SEE = 382
(5.2)
(8) C = 0.5225 E - 0.1722 E2 R2 = 0.98
(2.7) (-3.4) D-W = 2.2
+ 561. IT + 0.0884 L SEE = 384
(5.1) (1.0)
(9) C = 0.499 E - 0.1556 E2 R2 = 0.98
(4.2) (-4.0) D-W = 2.8
+ 733.7 T + 903.0 V SEE = 325
(5.2) (1.8)
62
(10) C = 0.4977 E - 0.1539 E2 R2 = 0.98
(4.3) (-4.0) D-W = 3.0
+ 690.9 T + 6.43 5 SEE = 322
(5.7) (1.8)
(11) C = 0.6584 E - 0.2129 E2 R2 = 0.98
(13.5) (-10.3) D-W = 2.5
+ 582.2 T + 0.215 °C SEE = 360
(5.9) (1.6)
Where:
C
E
T
L
B
V
°c =
= Total landings
= Fishing effort
= Technological change, labor
skills (1961, T = 1; 1962,
T = 2; 1963, T = 3; etc.)
= Labor employed in the fishery
= Adult bird population
= Dummy variable: closed sea-
son V — 0; open season V = 1
Temperature of water in
Trujillo, Peru
Due to the fact that the theoretical Schaefer
model does not include a constant term, the
estimations of the t-values of the coefficients
presented above are biased upwards. However,
in regressions having the constant term in it,
it happens that this constant term is not
significant in any regression (f-value around
0.2). The difference between coefficients of
regressions with and without the constant term
is not significant, since in all cases this dif-
ference is less than 0.4 standard deviations of
the coefficients.
In all regressions having the constant term,
the variable technological change (T) is sta-
tistically significant at the 1% level of signifi-
cance. In the equations presented above, even
though the ^-values are biased upwards, the
variables labor size (L), veda seasons (V), bird
population (B), and temperature (°C) are not
statistically significant. However, the impor-
tance of technological change (T) alone is such
that its introduction into equation (7) is suf-
ficient to improve substantially the coefficient
of determination of the equation from 0.84 in
equation (6) to 0.97 in equation (7). Also the
Durbin-Watson statistics (1.8) are now in the
acceptable range (1.6-2.4).
Using expression (5) on page 61 we can
obtain the optimal level of fishing effort in terms
of the efficiency of 1969 vessels. Equation (7)
gives 16.2 million ton-trips as the optimal level
of fishing effort. Equations (8) to (11) give the
following values for optimal effort in terms of
million ton-trips: 15.2, 16.0, 16.0, and 15.0,
respectively. All these estimates are in close
agreement, but differ markedly from the value
of 23 million ton-trips obtained by Schaefer
(1970), and from equation (6). However, because
of the statistical significance of the variable "T"
in equation (7), the high autocorrelation in
equation (6), and the theoretical appeal of the
procedure, it seems that the value of 16.2
million ton-trips is closest to the true optimal
level of fishing effort. Also, this value makes
more sense in terms of the data presented in
Table 1. In this table we can see that in 1962-
1963, with vessels of less efficiency than those
existing today, 11.8 million ton-trips produced
6.9 million metric tons of landings. A simple
extrapolation would indicate that 8.8 million
tons of fish could be landed by 18.3 million ton-
trips of vessels with 1963 efficiency levels.
CONCLUSIONS
The method presented here appears useful
in obtaining an unbiased estimation of the
optimal level of fishing effort in a fishery. It
adequately considers the effect of several signifi-
cant inputs that cannot be directly introduced
into the traditional measure of fishing effort.
Using this procedure, the optimal level of fishing
effort in the Peruvian fishery is 16.2 million
ton-trips, or only 68% of the level of effort used
in Peru in 1968-1969. This result has clear
implications for the management of the Peruvian
fishing industry.
LITERATURE CITED
BOEREMA, L. K., et al, 1961. Report on the Effects of
Fishing on the Peruvian Stock of Anchovy. FAO
Fisheries Technical Paper Number 55. Rome.
GULLAND, J. A., 1961. Fishing and the Stock of Fish
in Iceland. Ministry of Agriculture, Fisheries and Food.
United Kingdom.
GULLAND, J. A., 1968. Report on the Population
Dynamics of the Peruvian Anchoveta. FAO Fisheries
Technical Paper Number 72. Rome.
Instituto del Mar del Peru, Lima. 1970. Resumen General
de la Pesqueria, 1970.
63
SCHAEFER. M. B., 1957. A Study of the Dynamics of the Anchoveta, Engraulis Ringens, off Peru. Instituto
the Fishery for Yellownn Tuna in Eastern Tropical del Mar del Peru, Bulletin L-5.
Pacific Ocean. InterAmerican Tropical Tuna Commission
Bulletin Number 2. SCHAEFER, M. B., 1970. Men, Birds, and Anchovies
in the Peru Current-Dynamic Interactions. Transactions
of the American Fisheries Society, Volume 99, No. 3.
SCHAEFER. M. B., 1967. Dynamics of the Fishery for pp. 461-467.
64
Natural Resources and External Economics
Regulation of the Pacific
Halibut Fishery
Jack Rich1
ABSTRACT
In a static, long run competitive equilibrium framework, a catch function allowing
for short run diminishing returns is combined with a fish growth function developed by
Pella and Tomlinson which facilitates the derivation of an expression for the long run
marginal cost of "effort" in a common property resource such as a fishery. This
expression takes into account both "congestion" and "growth" costs. The diagramatic
technique of Crutchfield and Zellner is modified to take account of these externalities.
The modified Crutchfield-Zellner diagrams are used to illustrate the potential economic
losses from maximum sustainable yield regulation or other nonoptimal output.
INTRODUCTION
The task of the International Pacific Halibut
Commission, as established by treaty between
the United States and Canada, is to regulate
the Pacific Halibut Fishery at maximum sus-
tainable yield (MSY). The purpose of this paper
is to develop a model which will permit the
estimation of the economic losses which may be
associated with MSY regulation or other non-
optimal output levels. The model has certain
inherent limitations. It is static, deterministic,
partial equilibrium, and ignores income dis-
tribution and second-best effects. Still, it may
be useful in analyzing a fishery not much
affected by others, such as the Pacific halibut
fishery, and in focusing attention on the potential
magnitude of economic losses resulting from
the present type of regulation and from a
decentralized, unregulated fishery, although,
at least at present, it does not provide an
answer to the problem of how long run equi-
librium is to be attained.
THEORETICAL FOUNDATION
The starting point for the current model is
the Crutchfield-Zellner model (1962). Modifica-
1 Department of Economics, Oregon State University,
Corvallis, Oregon.
tions to this model are made which are designed
explicitly to account for technological exter-
nalities resulting from the common property
nature of the fishery, several of the modifications
having been developed by Smith (1969), Carlson
(1969), Bell (1969), and Worcester (1969),
among others. The present paper develops a
framework for the estimation of the rent and
consumer surplus losses (conventionally defined)
resulting from MSY regulation or other non-
optimal output in the static framework out-
lined above.
Figure 1 depicts the Crutchfield-Zellner
model. Growth of the fish stock biomass as a
function of stock size is illustrated in Part A,
and has the typical characteristics. The de-
centralized, competitive supply and demand for
fish are illustrated in Part B, where the in-
dividual "S" curves are "short run" supply
curves for fish and show how the amount
supplied varies with prices, increases in quantity
resulting from additional units of "effort"
entering the fishery at higher prices. Decreases
in fish stock, such as from OC to OB, result in
an upward shift of the S curves, from S-OC
to S-OB; hence, with fewer fish exposed to the
gear, the costs of catching any given quantity
of fish are increased. The curve XX "traces
out the locus of points on each of these supply
curves which are sustainable; that is, where
the catch at the corresponding population will
65
$/lb.
Part B: Demand and
Competitive Decentral-
ized Supply of Fish
M.-.Y
(lbs. per Unit Time)
Figure 1. — Industry demand together with competitive supply of fish.
leave population (biomass) constant over
time" (Crutchfield and Zellner, 1962).
Since the individual competitive fisherman
has no control over the size of the fleet or the
stock of fish, these factors do not enter the
decision making process of the individual fisher-
man, although they do enter the cost function.
Thus there are technological externalities
associated with a fishery — a "congestion"
cost, reflecting the decreasing catch per unit
effort from a given stock of fish as more vessels
enter the fishery, and a "growth" cost, reflecting
the decreased catch per unit effort by a given
number of units of effort from a reduced biomass
of fish, and represented by the upward shifting
of the S curves as the stock offish is reduced.
The curve XX is thus a long run average
cost curve. A regulatory agency which has as
its purpose the maximization of the net eco-
nomic benefits of a fishery will have to take
account of the technological externalities in-
herent in a common property resource, such
as a fishery.
Figure 2 adds Long Run Marginal Cost
(including congestion and growth costs) to the
Crutchfield-Zellner model. The LRMC curve
is the sum of the marginal congestion and
marginal growth cost curves, and is asymptotic
to MSY since, as sustainable yield harvest
increases, equilibrium fish biomass decreases
(from its maximum level WmaK ) until eventu-
ally a further increase in effort results in a
66
$/lb
m n
(lbs. per Unit Time)
Figure 2. — Longrun marginal cost added to Crutchfield-Zellner
model.
zero increase in sustainable yield. That is, the
marginal physical product of another unit of
effort (in terms of sustainable yield) is zero.
This occurs when sustainable yield is maximum,
and at this point the cost of an additional pound
of fish (in terms of sustainable yield) is infinity
(Carlson, 1969). In the static framework of
this model, the economic benefits from the
fishery are maximized when price is set equal
to long-run marginal cost, including congestion
and growth costs — that is, where the extra
costs of an additional pound of fish are just
equal to what consumers are willing to pay for
that additional pound.
Assuming a normal downward sloping
demand for fish, long run equilibrium under a
regulatory agency which sets price equal to
marginal cost can be determined, and this
equilibrium can be compared with that for an
unregulated, competitive regime, and with
MSY regulation.
Under a decentralized, competitive regime,
the fishery will be in long run equilibrium where
the long run average cost curve (including
normal returns) is equal to price — point A in
Figure 3 — with catch X0 and price P0. But, as
noted by Carlson (1969, p. 20), "the cost ... of
(harvesting) an additional unit of fish [X0B]
at this level is in excess of what consumers are
willing to pay for it" [X,A]. Since LRMC is
always above XX, a competitive fishery always
operates in long run equilibrium at a non-
optimal output, with too small a stock of fish,
although the harvest may be larger than
(Figure 3), equal to (Figure 4), or smaller
than (Figure 5) the optimum level.
Under the present assumptions (including
instantaneous transfer of resources to their
next best alternative use, and that demand
accurately reflects consumer preferences), the
"social" or "welfare" loss of a decentralized
as compared to an optimally regulated fishery
is the area ABE of Figure 3 — the excess of
the extra cost above what consumers are willing
to pay for the extra production of fish
Xo — Yi, beyond the level Xi. ABE is also the
extra value, above the gain in consumer surplus
PiEAPo the resources used to produce the
extra fish X0 — Xi could have produced had
they been used in their next best alternative
67
PRICE
$/lb.
(lbs. per Unit Time)
Figure 3. — "Deadweight" loss (Area ABE).
(lbs. per Unit Time)
Figure 4. — Identical competitive and regulated output rent
loss only (Area (P0 - C0) X0).
68
Price
(lbs. per Unit Time)
Figure 5. — Comparison of equilibria: competitive output lower
than regulated output. (If, with demand as given, output
is restricted to MSY, the welfare loss is area HJKL plus
the shaded area above the demand curve and to the right
of LRMS curve.)
use. In Figures 4 and 5, a rent loss (PiAGCi)
is also included. In Figure 4, the entire loss
consists of rent. That is, output under decentral-
ization and optimal regulation are identical.
However, that output would be produced with
a much larger stock of fish, and hence lower
costs, under optimal regulation than under a
regime of decentralization. Thus, all the extra
units of effort used to produce output X0 are
"wasted," and could better have been used in
other industires.
A MEASUREMENT MODEL
The derivation of marginal congestion and
growth costs can be expressed mathematically.
This will permit estimation of the production
function, once specific growth and catch func-
tions are determined. With the addition of
costs and demand, estimation of the welfare
losses discussed above may be achieved.
Summarizing all inputs under the umbrella
term "effort" (E), catch (X) is a function of
effort and the stock of fish ( W) :
(1) X = f(E, W).
Effort, catch, and stock can all be expressed in
terms of the long run equilibrium catch, X,
which will give us an expression in terms of
long run equilibrium catch alone:
(2) E(X) = g(X, W(X)),
where E(X) is the effort associated with a long
run equilibrium catch of X, and W(X) is the
stock of fish consistent with a sustainable catch
of X — i.e., one such that dW/dt = X. Since
cost is a function of effort, we have, for long
run equilibrium,
(3) C = C(X, W(X)).
From (3), we can obtain marginal congestion
cost, marginal growth cost, and long run
marginal cost:
(4) MCC = dc/dx = Cx
(5) MGC = dc/dw ■ ^ = Cw -^
ax dx
(6) LRMC = MCC + MGC = Cx+Cw
dw
dx
69
Equations (7) — (16) summarize the model
developed by J. J. Pella and P. K. Tomlinson
for the Inter-American Tropical Tuna Com-
mission (1969) (hereafter called the TC model).
The TC biological model results in a curve
relating growth of population to population
size. It resembles models previously used by
the International Pacific Halibut Commission
(Southward, 1968) although it is in terms better
suited for economic analysis. The biological
portion of the TC model will be used in what
follows for an unexploited fishery. In the dis-
cussion of an exploited fishery modifications
will have to be made to take account of the
congestion phenomenon, and this will be achieved
by use of the Carlson "engineering" function
for a fishery (1969).
In the TC model the growth of the fish stock
is
(7) dWt/dt = HW"' -KWt
where H, K, and m are constants. Limiting
population to some absolute maximum Wmax ,
and integrating (7) yields the population at any
time t:
\ —m \ — m l — mi
(8) Wt = [Wmax -(Wmax -Wo )
x e
K(\ -m )t ,1
where W0 is the population at time zero, and
(9) Wmax =(K/H)l/(m~1}
Further, Wmsy , the stock which yields the
maximum sustainable yield, can be expressed
as
(10) Wmsy = (K/mH)lKm~l)
The TC model for an exploited fishery hy-
pothesizes a constant "catchability coefficient,"
q, which is the fraction of the population caught
by a standard unit of fishing effort per unit of
time. The model assumes that the instantaneous
catch rate, dXJdt, can be expressed as:
(11) dXt/dt = qf,W,,
where fi is the number of units of effort applied
to the fishery at time t. It is the assumption that
qf varies in the same proportion as q or / that
must be modified to take account of the con-
gestion externality. The TC model implies a
constant short run marginal physical product
of effort, and hence a constant short run mar-
ginal cost of fish, at least until the stock of
fish is exhausted. This assumption does not hold
for the Pacific halibut fishery, and may not hold
for any fishery. However, maintaining the TC
assumptions for the moment, (7) for an un-
exploited fishery becomes
(12) dWt/dt = HW,m -KWt-qftW,
for an exploited fishery.
With effort constant in the time interval (0, t),
and excluding those cases in which the stock of
fish is fished to extinction, integration of (12)
yields
(13) W,
\ H
-(* + <?/)(!■
m)t
[k + qf
1/1 -m
wn -m))
Eliminating the time variable, and considering
only those populations that have adjusted to
the given constant level of effort (i.e., as t ap-
proaches infinity), we have
(14)
W-i3£jfi
1 lm
Biological equilibrium when catch (X) is equal
to growth of the fish stock is
(15) X = HW'" -KW= qfW.
From (15) we can now express biological equi-
librium catch as a function of effort:
(16) X = qf(
Qf+k
H
i lm
To take account of congestion externalities
the Carlson "engineering" function will be used.
Let k be the fraction of a stock of fish caught
by the first unit of effort applied to the fishery;
assume that two units of effort catch not 2k
of the original stock, but only k + k(\ — k) of the
initial stock. That is, each unit of effort catches
a fraction k of the stock remaining after all
previous units of effort have been applied to
the fishery. For N units of effort the fraction,
F, of a fish stock caught is
d7) f= i-n-k)x
where total catch is
70
(is) x = a -a-k) )w
or, writing W in terms of N
(19)^-U-(l-»,']|1-(1-^' + V
That is, whenever we find qf in equation (16)
we replace it with equation (17).
Restricting ourselves to equilibrium values
(that is, where catch is equal to growth),
differentiation of (18) with respect to N yields
the marginal physical product of effort in long
run equilibrium:
(20) dX/dN = -(l-kfln(l-k)W + [1 -
(l-kf]dW/dN
The first expression on the right of (20) is the
short run marginal physical product of another
unit of effort, and is always positive for any
positive W, and declines as N increases, thus
illustrating short run diminishing returns. The
second expression on the right of (20) shows
the effect on long run equilibrium catch of
another pound of fish stock, and is equal to the
percentage of the stock caught by N units of
effort multiplied by the change in equilibrium
stock resulting from a marginal change in
equilibrium effort. Thus, (20) includes both
congestion and growth externalities.
Solving explicitly for dW I DN and rearrang-
ing terms can also yield the expression for
dX I dN in terms of W, N, and the parameters
k, m, and H:
(21) dX/dN = -[(1— fe) ln(l-fe)
-l-(l-fe)
(m— 1)H
.N
w2-m +W}.
If we assume that the cost per unit of effort
is some constant A, then the marginal cost per
pound of fish under biological equilibrium con-
ditions, and including both congestion and
growth externalities (the long run marginal
cost, LRMC) is
(22) LRMC = A KdX/dN).
Estimation of the parameters H, K, m, and
k, together with data on demand and the cost
of effort can be used to estimate long run
equilibrium catch and the welfare losses in any
year associated with MSY regulation or other
nonoptimal output.
LITERATURE CITED
BELL, FREDERICK W. 1969. Estimation of the Econ-
nomic Benefits to Fishermen, Vessels, and Society from
Limited Entry to the Inshore U. S. Northern Lobster
Fishery, National Marine Fisheries Service, Working-
Paper Number 36.
CARLSON, ERNEST W. 1969. Bio-Economic Model of
a Fishery. National Marine Fisheries Service, Working
Paper Number 12.
CRUTCHFIELD, JAMES and ARNOLD ZELLNER.
1962. Economic Aspects of the Pacific Halibut Fishery,
Fishery Industrial Research, 1(1).
PELLA, J. J. and P. K. TOMLINSON. 1969. A Gen-
eralized Stock Production Model, Inter-American Tropical
Tuna Commission, Bulletin 13(3).
SMITH, VERNON. 1968. Economics of Production from
Natural Resources, American Economic Review, 58(3):
409-431.
SOUTHWARD, G. MORRIS. 1968. A Simulation of
Management Strategies in the Pacific Halibut Fishery,
International Pacific Halibut Commission, Report
Number 47.
WORCESTER, DEAN A., JR. 1969. Pecuniary and
Technological Externality, Factor Rents, and Social
Cost, American Economic Review, 59(5): 873-885.
71
Production from the Sea
Frederick W. Bell, Ernest W. Carlson,
and Frederick V. Waugh1
ABSTRACT
The sea constitutes a common property resource which causes factor productivity to
be heavily influenced by technological externalities. The sea is also subject to the spectre
of Malthusian scarcity since man cannot manipulate the ocean environment (Barnett and
Morse, 1963). We estimated the parameters using ordinary least squares of the dynamic
Schaefer production model of the intervention of man into the oceanic ecosystem. A
second production model for the sea to specify diminishing returns to capital and labor
for any fixed biomass was developed. The parameters of the latter model were estimated
by a computer search technique. The results indicate that the industry production
function for marine life is subject to diminishing physical returns to capital and labor. For
the cases considered in this study it also appears that the parabolic yield function
developed by Schaefer, assuming constant returns to factors inputs, is not as realistic
as a production function with diminishing returns to inputs with a given biomass.
INTRODUCTION
After explaining the principle of diminishing
returns in agriculture, that great economist,
Alfred Marshall (1920, p. 166) wrote:
As to the sea, opinions differ. Its volume is vast,
and fish are very prolific; and some think that a
practically unlimited supply can be drawn from the
sea by man without appreciably affecting the numbers
that remain there; or in other words, that the law of
diminishing returns scarcely applies at all to sea-
fisheries; while others think that experience shows a
falling-off in the productiveness of those fisheries
that have been vigorously worked, especially by steam
trawlers. The question is important, for the future
population of the world will be appreciably affected
as regards both quantity and quality, by the available
supply offish.
We have waited 50 years to answer Marshall's
question. We must not wait much longer. The
world's population will double by the year 2000.
What will happen to the production, prices,
and consumption offish (Bell et al., manuscript)?
As in Marshall's day, some doubtless still
think that the future supply offish is practically
1 The authors are respectively Chief and Economist,
Economic Research Laboratory, National Marine
Fisheries Service, and Professor, Department of Agri-
cultural Economics, University of Maryland. The ideas
expressed in this article do not necessarily reflect the
official position of the National Oceanic and Atmospheric
Administration (NOAA).
unlimited. But those biologists and economists
who are studying fisheries doubt this. They
know that some species offish have already been
"overfished"; that is, increased inputs of capital
and labor have actually reduced yields. Examples
are menhaden and haddock in the Atlantic
fisheries. Biologists have found that the catches
of eastern tropical Pacific yellowfin tuna and of
northeastern Pacific halibut have reached their
"maximum sustainable yields." International
controls have been found necessary to prevent
depletion of the aforementioned species.
Of course, these are only a few of the many
species of commercial fish. But we doubt if any
fishery biologist today would be among those
who Marshall said, "... think that a practically
unlimited supply can be drawn from the sea."
To be sure, the sea is vast, but Ryther (1969), a
prominent biologist, says:
The open sea — 90% of the ocean and nearly three-
fourths of the earth's surface — is esentially a
biological desert. It produces a negligible fraction of
the world's fish catch at present and has little or no
potential for yielding more in the future.
Upwelling regions, totaling no more than about
one-tenth of 1% of the ocean surface (an area roughly
the size of California) produce about half the world's
fish supply. The other half is produced in coastal
waters and the few offshore regions of comparably
high fertility.
We could cite many other fishery biologists
72
to indicate that the potential supply of fish
from the sea is limited. But, even if there were
no fixed limit to fish production, we believe that
diminishing returns would apply to fisheries at
least as much as to agriculture; perhaps more.
This has important implications to public
policies, as Marshall noted. Hence, the purpose
of this article is to explore the production
function for the sea.
DIMINISHING RETURNS OF FISHERIES
Marshall's (1920, p. 150) first statement of
the law of diminishing returns in agriculture
was:
An increase in capital and labour applied in the
cultivation of land causes in general a less than
proportionate increase in the amount of produce
raised, unless it happens to coincide with an im-
provement in the arts of agriculture.
In the case of fisheries, indices of capital and
labor inputs are known as "effort." Diminishing
returns from fishing means (paraphrasing
Marshall) that an increase in effort results in
less than a proportionate increase in the yield
of fish, assuming no change in technology.
Thus, if effort were doubled, the yield would be
less than doubled.
But if we are to manage the world's fisheries
well, we need more than general comments
about diminishing returns — we need usable
estimates of the effort-yield functions for the
major species of fish. Schaefer (1954) wrote a
pioneering paper on the theory and measurement
of such functions. In recent years, many bi-
ologists have added to the theory in this area,
and have presented important statistical veri-
fications and measurements (Pella and Tomlin-
son, 1969; Fox, 1970).
The necessary theory is in two parts: (1) the
theory of biological growth, and (2) the theory
of yield from a given biomass.
Theory of Biological Growth
First, consider biological growth — for
example, the growth of "biomass" or the total
weight of marketable fish. Schaefer (1954),
hypothesized that if there were no fishing, the
growth curve of the biomass would look some-
biomass
Figure 1. — Growth with no fishing.
thing like that shown in Figure 1. The species,
in each region, would tend to approach some
maximum biomass, M. Here natural mortality
would just offset recruitment (from young stock)
and growth in body size.
A curve commonly used to represent such
growth is the logistic,2
(1) m,
M.
1 + be
where mt is the biomass at time t, M is the
potential maximum biomass, e is the base of
natural logarithms, t is time, and a and b are
parameters. (We shall generally measure time
in years.) Davis (1941) discussed the proper-
ties of this curve in detail, and gave many
references to its uses in biology and in the study
of growth of human populations. Its derivative
is:
2 Most work using the logistic has been done with
numbers in populations, here we are applying it to the
total weight of the population. Tomlinson and Pella
(1969) have suggested that the following function be
used to approximate biological growth:
' HP'" It) -KP(t).
When m = 2, the growth function becomes the well-
known logistic or as used by Gulland, an autocatalytic
equation. Fox (1970) has suggested a Gompertz function
to approximate biological growth.
73
(2) dmr/dt= am
■(i-jt)-
So the proportional rate growth (with no
fishing) is:
dmt _ n /-, mt\
The second derivative of (1) is:
d ~mt 2
(i-SO^)
Maximum absolute growth occurs when (4)
equals zero; that is, when mt = V2M (when
current biomass is one-half the potential
maximum). At that point, equation (2) shows
that the maximum growth, dmt/dt = aM/4.
Suppose aM/4 were taken from the biomass
each year by fishermen: each year, the biomass
would grow by aM/4; biological growth would
just offset the amount taken by fishermen; and
there would be a steady-state equilibrium.
The Theory of Yield
from a Given Biomass
We now consider how yield responds to
effort when we abstract from changes in biomass.
Schaefer (1954) made the simple assumption
that the catch mt would be proportional to
effort, k is the constant of proportionality, and
xt is effort:
(5) y,/m, = kxt.
Schaefer assumed that, with a given biomass,
there would be constant returns to effort; dou-
bling the effort would double the yield, tripling
the effort would triple the yield — and so on.
As a first approximation, this may be adequate
in many cases within the observed range of the
data. Schaefer and others have used it to make
many important estimates of maximum sus-
tainable yield; and as a basis for economic
controls.
But we think that a more realistic catch
function is:
(6) y,/m, = (l-zXr),
with 0 <z < 1, and with m, fixed.
The rationale of (6) was explained by Carlson
(1969). Briefly, assume that the original biomass
is m, and that one unit of effort will catch pmr,
leaving (1— p)mT\ assume that the next unit
of effort will catch the same proportion of the
remaining biomass — that is, it will catch
p(l—p)mr,leaving(l—p)2mr. The same reasoning
shows that n units of effort will catch (1 m ,) .
In equation (6), we simply let z = 1— p. We
believe that on an a priori basis (6) is more
realistic than is (5). But probably there is no
magic mathematical formula that is exactly
right for all species and for all amounts of effort.
yield, y
Figure 2. — Two yield functions. (Based upon equations
5 and 6, assuming that one unit of effort yields one-
half of the existing biomass.)
Figure 2 compares the growth functions
represented by equations (5) and (6). Each
assumes that one-half the existing biomass was
caught with one unit of effort in some base
period. (The units are arbitrary. We find it
desirable to "normalize" both yield and effort by
dividing by the base-period data.) Note that
equation (5) would indicate that the entire
biomass would be caught with two units of
effort. But equation (6) would indicate that if
effort were increased indefinitely, the existing
biomass would be approached as a limit, but
never quite reached. Within the observed range
of historical data, it may not be easy to choose
between the two curves in Figure 2. But they
give far different results when they are ex-
trapolated to estimate the effects of large in-
creases in effort. This is especially critical where
one must make forecasts of the likely effect of
the expansion in fishing effort.
71
STATIONARY STATE EQUILIBRIUM
The stationary state equilibrium is found
bj' letting annual yield equal annual growth :
where y is the equilibrium yield and mt is the
corresponding biomass. Thus, Schaefer let:
(8) kxr = a{l-^L)
Solved for mt
(9) mt=M(l-k-f)
and got the equilibrium yield as a function of
effort:
(10) yr = rhtkxt = Mkxt (l - -^) ■
This is a simple quadratic. To estimate it
from statistical data using ordinary least-
squares, we write:
(11) yt=Axt-Bxi
where A = Mk and B =
Mk'
The graph of (11) is shown in Figure 3. Note
that while Schaefer assumed constant returns
from a fixed biomass, his curve of equilibrium
yield indicates decreasing returns. In fact,
equilibrium yield
Figure 3.
effort
Equilibrium yield-effort with constant
returns.
average yield per unit of effort is easily seen
to be (by dividing (9) by x),
(12) yt/xt =A -Bxt .
If we use (6), instead of (5) as an estimate
the response of yield to effort with a fixed
biomass, we have:
(ia)i-<"-.(i-Sf).
Solving for m, we find :
(14) mr=M[l-(l-~)]-
So the steady-state equilibrium yield is:
(15) y,=mt(l-zXt) =
m[(i-/-)-±(w;')1
that is,
(16) yt = C(l-zXf)-D(l-zXtf
where C = M and D = Mia.
This is not as easy to fit statistically as is the
Schaefer function (11). It can be handled without
undue difficulty on a computer by a "search
method," trying a series of values for z; in each
case computing R2, the Durbin-Watson statistic
(D-W), and the t values of the two regression
coefficients; then by interpolation we find the
"best" fit.
Equations (15) and (16) indicate decreasing
returns to effort. Their graph is like that in
Figure 4. In this case — which we think is
more realistic — we get diminishing returns
for two reasons:
1. Because annual growth declines as the
fish population increases, and
2. Because the yield-per-unit-of -effort de-
clines with effort; that is, doubling the effort
will result in less than doubling the yield, even
with a fixed biomass. The net result is a much
flatter curve after MSY is reached.
STOCK ADJUSTMENT MODEL
So far, we have considered only the steady-
state equilibrium. This assumes that full adjust-
75
equilibrium yield
MSY
Figure 4.
effort
Equilibrium yield-effort with diminishing
returns.
ment is made instantaneously, thus the present
catch is a function of the present effort only.
This may give a satisfactory approximation for
some species. But in other species, several time
periods may be required to establish a new
equilibrium. In such cases, current yields are
affected not only by current effort, but also by
the efforts of several past periods.
That is, annual observations on catch and
effort do not represent equilibrium observations.
To remedy this situation, biologists have sug-
gested various adjustments to the data (Appen-
dix I).
In reality, the observed catch in any given
year may be the result of effort expended in
previous periods; i.e., the observed catch is
some kind of weighted average of catch produced
by fishing effort in previous periods. The Gulland
procedure employs a similar assumption in that
it assumes that this year's observed catch is
parabolically related to a simple average of
previous effort. An alternative specification of
the yield effort-relation for many stocks of fish
may take the following form (assuming for
example a logistic and constant returns equi-
librium relation):
(17) yt - axt — bxt + ax xt-\
+ . . .€t ■
2
Let us now make the classic assumptions about
the disturbances, et, of constant variance and
zero covariance.
Although (17) is a general specification of the
yield-effort relationship, its estimation presents
obvious difficulties. Since our sample will be
finite in size, the infinite set of lagged regressors
must be terminated at some point. Also, there
is likely to be colinearity among the successive
regressors.
One way of solving the problem is to hypothe-
size that the coefficients on the lagged variables
diminish in size as the time period is more
distant from the present observation on catch.
Put differently, let us hypothesize that the
coefficients on successive x's decline systemati-
cally as we go further back in time. This was
suggested by Fisher (1925); more recently it has
been revived and extended by Koyck (1954) and
by Nerlove (1958). We shall call this a Koyck
specification. Koyck hypothesized that a useful
approximation would be that the coefficients
of (17) decline geometrically:
(18) ak = a\k (k = 0, 1, . . . ) and
(19) bk =b\ (fc-0, 1, ...).
(17) may be rewritten as the following: 3
(20) yt = axt — bxt + Xaxt- 1
— Xbxt- 1 + . . . et .
If we lag (20) by one period and multiply by A.
we obtain
7
(21) Xyt_ i = Xaxt_ ^ — Xbxt_ ]
2 2 2
+ X axt_ 2 —X bxt_ 2 + . . . Xet _ i
Now, subtract (21) from (20) and rewrite:
(22) y, = ax, —bx, + Xyt- i + e,
where
(23) et = et — Xet-\ .
3 Equation (20) may be interpreted to mean that
observed catch depends on this year's effort (a common
assumption used by many population dynamicists) plus
effort expended in previous periods. This is merely a
hypothesis that can be tested empirically.
76
Equation (22) may be estimated using ordinary
least-squares.4
Nerlove provides an alternative theory to
justify (22). Suppose that xt determines v,* , the
"equilibrium value" of catch,
(24) yt* = axr - bx] ,
but that the adjustment to the equilibrium value
in one period is only gradual (i.e., not complete):
(25) yt—yt-\ = b(yt* —yt-i)
where 0 < 6 < 1 is the coefficient of adjustment.
Inserting (24) into (25) and rewriting gives
the same form as (22):
(26) yt = abxt — bhx] + (1—5) yt_ i
where (1—5) = X.
Using (26) or (22), we may also compute
how many periods it takes one-half the gap to
be filled. If yt- i is in equilibrium, then the gap
at period t(Gt) is equal to the following:
(27) (yt*-yt-i) = Gt.
Each period a constant percentage of the re-
maining gap is filled; so that at time t + k
the remaining gap is
(28) Gt+k =Gt (1-5 f .
If K = 0, (29) indicates that all the gap remains
to be filled. When will one-half of the initial
gap be filled? This may be found by substituting
y2Gt for Gt+K , or
(29) Gf (1-5)*' =1/2.
Hence,
4 In essence, a researcher attempting to estimate the
parameters of the yield function can ran the following
regressions: y, - ax, - bx? ■,
x, — bx,
+ Xy,-
or y, =
fc+x,.
i +...
♦ «, .1
L
n + t
j
t[" +
t, i *
..♦«,. „-],
L
n *
i J
where (n + 1) is the number of years the fish are in
the fishery. The latter is the Gulland technique where
the first two specifications are with and without the
Koyck formulation respectively. Equation (20) may be
specified as the following:
(y/x)t = a — bxt — \bxt_ ! - ... -X bxt_k.
With this form, the final estimating equation will have
(yjx) as a lagged independent variable.
(30) (1-6)A =i/2,
or
(31) K = log 1/2 * log (1-5) = log 2
log(i^5)
K is the "half-life"; that is, the number of
periods required to cut the gap in half. In 2K
years, the gap will be reduced to V4 ; in 3A'
years to Vb . . . and so on. It would never com-
pletely disappear. In theory, K should be
related to the following biological factors:
(1) Fertility of the species (i.e., number of
eggs laid and reaching full term);
(2) Rate of growth of the species (i.e., how
many periods it takes to reach maturity). K
should be large for relatively unfertile and
slowly growing species and small for very
fertile and rapidly growing species.
In sum, we are interested in eight estimating
equations. First, a group of four equations based
upon the assumption of constant returns from
a fixed biomass; these are all designated LCR
(logistic constant returns). LCRa is the static
function with total yield, yti dependent. LCRb
is the same with average yield per unit of
effort, yt/xt, dependent. Then LCRaS and
LCRbS are lagged or stock adjustment models.
This gives us four functions. There are four
more (designated LDRa, LDRb, LDRaS, and
LDRbS) based upon the assumption of decreas-
ing returns from a fixed biomass. Finally, we
have included an estimate of the parameters of
LCRa using the Gulland technique for adjusting
the effort series.5
RESULTS OF THE ANALYSES
In order to illustrate the applicability of our
theoretical yield functions, we selected five
species for consideration: (1) Chesapeake Bay
menhaden; (2) Atlantic and Gulf blue crab;
(3) Atlantic longline tuna; (4) Soviet and
Japanese king crab fishery in the eastern
Bering Sea; and (5) Cape Flattery sablefish.
5 For the five fisheries studied (below) the fish are
in the fishery about two years. Therefore, a two-year
moving average of effort was computed.
77
Chesapeake Bay Menhaden
Table 1 shows the empirical results for this
fishery. Based upon the R2 criterion, LDRa
represented the "best" function where total
catch was used as an independent variable.
There is no doubt from the statistical analysis
that the Schaefer function (LCRa) is definitely
inferior when compared to the LDRa model in
its ability to describe the catch-effort relation
in the menhaden fishery. No evidence of auto-
correlation was detected in the LDRa function.
As shown by LDRaS, there seems to be no
stock adjustment effect as the coefficient on the
lag variable is not statistically significant.
Among all the functions, LDRb shows the best
fit when catch per unit of effort is used as the
dependent variable. From a theoretical point of
view, there should be no difference between the
"a" and "b" functions. However, the statistical
estimation procedure does yield two estimators
for each parameter. LDRa and LDRb do yield
similar estimates of y* and a;*. Also, the
Gulland -LCRb equation yielded very similar
estimates of y* and x* as the LCRb (unadjusted
data). Further the choice between the "a" and
"b" functions should be made on the basis of
just what one wants to predict — catch or catch
per unit of effort. The LCR and LDRa functions
are shown in Figure 5. It should be noted that
in equation (16) M = A. Thus, the LDRa
equation estimated by least-squares will also
yield the maximum biomass without fishing.
That is, MSY = M/4 = 158.7 thousand tons. M
is therefore equal to 634.8 thousand tons. The
logistic function can be directly computed since
a = AIB and a = 1.1512, and if t - 0 at the
point of maximum growth, then
m,
M_ 634.8
2 1 +be
0 , so b = 1, or
m, =
634.8 thousand tons
1 + e-i.isi2f
This is one additional advantage of the LDRa
over the LCRa function.
Atlantic and Gulf Blue Crab
Table 2 shows the empirical results for this
fishery. Based upon the R2 criterion, it would
seem that we have little basis on which to
choose between the LCRaS and the LDRaS
models, each having an R2 of 0.94. Both show a
strong stock adjustment effect. The half-life
for the adjustment process was 0.57 years. In
this case, the data cannot adequately distinguish
between the two functions. The MSY ranges
from 129.6 million pounds in the LCRaS model
to 189.0 million pounds in the LDRaS model.
The autocorrelation test for the two functions is
inconclusive. Hence, the choice between the
functions must be made on a priori grounds.
Since the LDRaS model seems more plausible
on a priori grounds, it would seem that this
function should be selected for fishery manage-
ment purposes. As the fishery expands, addition-
al data will be generated to verify the existence
of one or the other function. This general
prescription will probably apply to many
fisheries where data are only available in the
upward expansion phase (i.e., catch is below
MSY). Finally, as with Chesapeake Bay men-
haden, there seems to be little difference between
Gulland LCRb and LCRb unadjusted. Figure 6
shows the two functions discussed above.
Atlantic Longline Tuna
Table 3 shows the results for the Atlantic
longline tuna fishery. On the basis of R2, the
LDRa model is superior in predicting changes
in catch in response to effort. The stock adjust-
ment coefficient was not statistically significant.
The autocorrelation test is inconclusive for
LDRa. The MSY for the LDRa function is 106.7
thousand metric tons with 140.1 million hooks
of effort. Notice that the MSY's associated with
the LCRa and LDRa functions are not appre-
ciably different; however, the number of hooks
necessary to harvest MSY is vastly different.
This is due to the flatness of the function
generated by the LDRa model. The Gulland-
LCRb gives a much higher estimate of y* and
a lower estimate of x* than the unadjusted
LCRb. Figure 7 shows the LCRa and LDRa
functions.
Bering Sea King Crab
Table 4 shows the results for the Bering Sea
king crab fishery. On the basis of R2, the LDRa
model is the best in "explaining" the catch-effort
78
182.6
158.7
LDR„
LCR„
o.oo 50.00 loo.oo 150.00 200.00
EFFORT: NUMBER OF UNITS
250.00
rf
• •"
0.00 50.00 100.00 150.00 200.00
EFFORT: NUMBER OF UNITS
250.00
Figure 5. — Chesapeake Bay menhaden fishery, 1946-68: Catch, effort,
and catch per unit of effort.
79
r
M
I
"u ~*
<— D
X =>
q
no
■tf
oo
■*
■*
—
NO
rt
00
00
»N
in
d
oo
00
00
OC
o
(N
-H
r-j
-^
ON
ON
ON
o-
3|
c
o
00
C-)
m
oo
in
m
c
OC
N
rr
00
NO
oo
00
q
IO
ON
t--
o
NO
r-4
*t
VI *>
CN
<N
en
r-'
_'
oo'
r^"
oo'
fs
^ 0
CN
00
~m
oo
m
m
NO
in
NC
{N
•— *
CM
<•■ i
~*
-H
•-H
•^
.
o
!s<2
VO
ON
<
<
00
V")
Z
Z
"3
^
u
o
2
£
<
00
>
>
o
O
<N
m
NO
r-
r~)
. — 1
C-"
t
"*
Q
m
q
•H
oo
1;
ON
rn
00
<^~
C-)
t-;
o>
C-
■*
00
m
(^1
m
rs
Qi
(N
r^
■*
p»
NO
00
NO
OC
O)
—
I
'
>,
~-
>,
II
X
4t
r-
+
-1
r^j
q
q
1
OJ
r-l '
X
X
Cn]
>> Ix
o
00
m
X
in
Cn!
X
X
in
X
o
m
X
m
V0
00
fNj
01
tt
On
ON
i
E
1
tS ^j
•*
■*
t—
©
ro
2
NO
<S
CnI
in
ON
in
a
+
m
ON
OO
—
ON
X
+
NO
.
m
00
C3
cs
(S
m
TT
i —
rn
'-»
NO
m
X
^r
in
o
o
X
nf
■^;
E
•<*
q
X
m
ON
oo
r—
m
00
rn
X
o
CN,
x
w
M
i
X
1
m
1
X
in
ro
X
X
o
"
i
1
X
1
'
■*
—
00
*-•
c—
^H
(^1
00
T
(N
^H
x
00
00
no
in
On
o
cs
r^
rn
f>l
in
-i-
+
m
ON
00
ON
■<t
M
NO
X
m
■*
oo
ii
NO
ON
m
ii
"*
-1-
II
~
II
n
II
II
_ -
II
_
-
>,
>.
>TTx
>,
>TTx
>.
>.l X
>>l .
_,
^
ai
5
U
Vi
n
as
O0
O0
oo
-=
_i
-3
71
as
tA
as
aS
2
(5
ai
e
o
U
u
o
Q
Q
Q
Q
• T*
ii
_J
_i
_J
_i
_)
_)
_J
_J
o 2
c as
a;
3
>
- h
a>
5 B
— c
fc^ SO
0) — '
R x,
3 2
-c >,
E b
*J 3
x «
ca T3
E c
C 3
2 o
it «
JC
a. w
^"S
.52 ^J
-a 2
C
- c
C a)
T3 —
O c
to >
- tiljO
o —
c
HI
s^
0
,. -o
*- 2 3
o Jj
' 1
c c<*-
n.2
.a -o
c
3
u « o
»H > <*-
T3 C
^
bug
C 3
u
'I
in ^ 3
ST s
■^
c o 2
5 sf
« 8 -a
Q 8
■J
o «
G
E
DC
W5 cd
0)
>
c
ifl
W <u
3
E
-r
C JS
a-O'S
On
o ■"
=
3
•T1
5S as
3: £
o
O
B
TJ
CO 5* tfl
C M
Ul
«n ^ W
u
£ E
"-»
'J
C/5
c aSI
u
3 to
Q U
Q
II
n ii
E
<
>
oo tt — . oo
E 2
o J3
iota
•a H
X —
'— o
0 "
5S&
i NO &
X J2 CO
— ?5 n ..
'S e B '
o o d
0 >
5S1
-I "
< > ■—
ii ii ii ii ii
as as
80
100.00 180.00 280.00 3U0.00
EFFORT: POTS (X103)
1
U20.00
500.00
100.0
180.00
200.00
300.00 1+20.00
500.00
EFFORT :POTS (XLO )
Figure 6. — Atlantic and Gulf blue crab pot fishery, 1950-67: Catch, effort
and catch per unit of effort.
81
2
-.
2
2
3
C
<
I
s
©
o
CT-
I/O
i/o
00
o
o
o
c
r*i
ro
r*N
no
00
NO
■<*
NO
■£ o.
o
i-
cm
VO
^H
CN
>/->_
■rr
or
ts
r~
^T
NO
NO
fN
r~J
. c
cs
c-i
CI
r—
ON
"*
r-i
r-4
oo
X .
NO
r-
NO
NO
NO
00
VO
Tf
CN
03 O
■— '
(N
m
iA
o
O
no
ON
00
o
*
r*".
o
J3
r>
NO
o
00
NO
r^-,
l^)
>■ —
r-~
!-H
CN
vq
ri
CI
1/1
O
23 s
S o
fN
I/O
fsf
ON
r-*
"*"
u->
ON
r-f
ts
r*l
ON
<N
r-
■*
NO
00
o
x:
Oi
[_
0
tw
3
r-
o
NO
00
J^c
uo
oo
u->
t~-
03
X
W9
<D
O
Z
>
>
<
>
<
>
>
CO
00
*—
C—
o
r-
<N
n
00
o
~^
ON
i/o
00
~*_
ON
wo
c-
Q
rl
ro
ON
■<*
o
m
■*
o
BS
lo
ON
"Of
ON
C~
ON
lO
ON
r—
1i *•
t
3*
I"
to
ro
r-
■*
■ — '
ON
tJ-
n
cs
+
^_
+
oi
,— «
cs
t^l
_£-
1
CN
X
00
X
■-1
■-H
<j
""*<
X
X
♦J
x
X
oo
00
r
X
oj
+
>>
>,
I/O
ON
r
x"
1
03
X
NO
f»
.'
i— «
' — '
E
' — '
r)
</0
1
—
• — *
*
r-
>/0
o
T_H
NO
%
r-
to
o
ON
fN)
ON
o
oi
*— i
OJ
wo
n
■*
cn
lO
uo
m
—
(N
rt
NO
00
O
03
r
w
+
+
<~l
r-
00
3
X
X
CM
X
1
1
£
ro
r-
, 1
i/o
1
j^*
^-^
u
II
on
ro
00
[ —
oi
X
X
00
-X
X
X
oo
-
1
I/O
1
00
1
r^l
I
ON
1
no
1
CN
1
o
X
ON
1
**
c—
X
o
CN
ON
oo
Tt-
ON
T
cs
tn
•<*
r--
ON
I/")
/
Tf
oj
r-
on
00
l/^
+
r )
1 '
C*-
o
1/0
^f
r-i
II
ON
oo
II
«*
u")
~~;
CO
X
— '
^r
,*?
fN)
II
"3-
II
II
X
II
X
II
^^"
II
^~
>v
~5
>.|X
>^_x
>.
>>
s^
>*
—
XI
^
0*
o
u
-3
03
tt5
C/3
03
B<
00
Pi
03
0C
X)
00
03
oo
as
c
3
*0
u
u
U
U
Q
a
a
2
r3
-J
-J
-J
-J
J
j
—i
_J
c
§ 2
a e
2 |
>. 3
-g S
•a
c >
O OJ
Is
^ c
■e,a
a a
-§ E
O w
.SP a
E^
T3
1>
SB S'
2 SN
oo c oa
ii ii n
00 * — i
.2P-2
E E
^3 OJ
,3 c
c- U
^> «
<u ""
•£ —
03 TO
B 3
rv w5
"" U
o^ u
C
C 03
■2 a
03 >
o
2 s
3X>
K o
M C u
° ° 5 u
o
u
03
3 2
C 3
3 ^~
u
Q S
o 1
*3 c
4-> o. >
0 *' o
«°°
^S =
C ed «
II?
QWQ
< D>
T3
U
■d
^
u
—
o
0)
a.
0^
>
0>
C
trj
0
03 T3
■o g
c o
O uo
• - ON
c —
300
XJ -
.2 S3
2».a -
ii ii n ii
JU Q 03
82
108.5
106.7
LDR
0.00
1;0.00 80.00
120.00 160.00
EFFORT: HOOKS (XLO )
200.00
1
Uo.oo
0.00
80.00
120.00
160.00
200.00
EFFORT: HOOKS (XLO )
Figure 7. — Atlantic tuna longline fishery, 1956-67: Catch, effort,
and catch per unit of effort.
83
-
a
Taj
a
z
-
CO
0!
Q -*
r-
t
r~
in
00
o
o
M-
c: o
~-
c~t
NO
r~i
fN|
■"3-
in
CO
^f
V, 0
IT)
in
00
■*
r—
*
vq
Tf
°.c
00
<N
©
(nJ
■*"
o
ON
cs
(S
S§
■*
oo
NO
00
NO
■*
00
in
ON
OJ
si
~
j
^1
C/3 c
cc
vq
r-
r-
r-i
r-
m
r^
Tf
o
r^
CO
oo"
c-4
r-i
NO
r-
00
NO
CN
o
o
i — i
**
o
ON
o
ON
~*
*~ <
—
— ■!
*—
^^
•— 1
o
«g
— i
<
C<1
<
"3
vq
Z
r^,
Z
JS
—
fl
o
Z
cs
£
1/1
On
00
r-
C--
"9-
rj
r~
r-
Q
NO
<N
O
m
o
"*
r-4
t*~
o
as
NO
in
c-)
on
■*
00
'O
ON
mj
r-
r-
00
r~
00
00
ON
00
ON
r~)
n
"^i
t
r-4
x
n
rl
X
>>
X
o
in
X
X
X
in
+
in
IT)
1
in
•3-
*-
~
in
ON
i
1
•*
1
X
X
V,
o
|
CO
+
X
>> 1 X
%
in
in
00
00
o
r-
rn
00
NO
ON
00
"rt
O
ci
*
00
ON
C*"l
1
1
-'
-Si
ON
X
^_,
1
, — .
1
'
.13
m
©
^H
X
oo
-w
^
*-•
^
^
a-
ON
in
X
**rf
X
4_i
«^
c
c^
r-
*t
ON
X
in
^
X
X
w
II
1-
o
Ol
n
ON
o
m
m
X
in
mi
<-
es
1
no
1
00
00
1
1
i 3!
1
X
^
x
1
X
1
C
~
I
X
*
X
NO
O]
i£>
r-
CI
ON
lO
oo
o
o
r-
ON
+
00
m
^f
•— '
vq
ON
rl
r-
NO
NO _
x
—'
rJ
-"
r-i
<*>
ri
r-i +
II
II
II
ii
II
II
ii
II
>.
>.
>CTx
>,
>Ax.
>>
>CTx
>,
>,|X
^_
^
as
5
_j
oo
■s.
to
C/}
en
'^
rz
p
S3
_£<
"
Si
n
5
(3
■a
0i
a;
OS
a:
as
CC
-7*
^
_J
_J
_1
_j
_j
_j
_l
j
O
o
>
c
i/i
-
IM
0
n
E
£
■a a
9 E
.a £
o jj
£ -°
3 °
C «l 2
^:
ft
1)
c
>.
O
i/)
a
■a
E
3
^
^
o
r
0
-,j
ft
-3
0
E
E
b
o
c
£=5 S
.2,tS «
to '5
2,« a as >
o o
OS I Q
ii n ii
C/N H —
o o
JO!
■o
T3
g .a
a T3
1.1
_ £ t- C NO ft^
— C 3 o m x
2 K «,? 5 - 5
Si «. C J . - — ' X
•-i e '7 ~" n
■2 a a .= it. ii 'S ■
~ " ° — rt " n
'~t- r- • ■ rz i- 3^
ss- <.
II II 5" "
-jit/1 -*—
84
Pi
10.00
8189
si- . •
60.00
110.00
160.00
210.00 260.00
EFFORT: TAN DAYS (X105)
10.00 60.00 110.00 160.00 210.00
EFFORT: TAN DAYS (X105)
260.00
Figure 8. — Soviet and Japanese king crab fishery in the Eastern
Bering Sea, 1955-67: Catch, effort, and catch per unit of effort.
85
IE
v
>.
02
e
1)
X
X)
T3
C
O
—
— c M
fi
on
ON
NO
rN
rN
0
00
00
"*
i/->
r-j
NO
r-
■*
00
0
■<t
CI
1/1
vn
10
rN
m
■•Oa
■*
■*
*fr
0
•*
00
NO_
O
00
x _ n
CNJ
«c2
0
■*
•*
■*
u->
CI
<♦
io
S^
~
""
^H
~
^
-
™h
v> <r
x>-o
M C
vO
ON
00
rN)
fS
■*
NO
00<_ 3
0
On
00
tn
t-
ON
ON
00
ON
O
O
500
ON
00
ro
in
CM
00
ON
NO
r~
ro
m
r-r
lO
SO
90
0*-1
vD
00'
00*
00'
•rf
r—
00*
t~
r^
Z.E
^^
0
—
m
<
(N
<
^Ss
1;
Z
I~-
Z
3
£
0
z
01
>
O
<^l
ON
**
00
■*
0
ON
r<>
OO
00
rsi
ON
NO
<-h
CNI
"*
«o
a
(N
"*
l/N,
oj
lO
r»
00
ON
r~
a:
r»
00
00
00
00
00
00
00
00
~H
1
*j
>T
X
_
*
r-l
1
CNl
ro
NO
00
NO
■*
00
w
rN
00
1
x
. — 1
■ ,
rn
CNl
+
1
2
n
X
lO
+
*x
ON
ed
x
£
<^l
00
CO
X
X
NO
1"
ON
l"
x"
r-i
X
in
X
00
0
00
10
l"
*i
QO
rs
Tj-
^^
ON
rH
rs
0
+
00
NO
-d-
' — '
m
_
1
r~
06
m
10
r3
r-1
O
N <-
1
V>
r^
r~
&
Ol
X
X
X
r-
X
ci
vn
1
1
ro
t1
r
.7^
PS
nO
m
r--
on
^^
X
1
X
II
f*"l
("*N
ON
00
10
10
rs
ON
NO
On
O
1
p
X
NO
ON
r
X
X*
ON
X
^^
.
1
1
WO
tH
1
1
x
X
NO
X
#
CO
NO
00
rs
CNl
ro
1
ON
ON
NO
10
00
ci
NO
+
O
0
00
t»*
•*
00
ON
CN
Tfr
r-4
■*
0
(N
w
c*l
r~i
00
m
q
Ti-
NO
r~
X
rN)
II
— i
II
NO
ll
r^
II
II
II
11
>>
+_,
II
^*~>
^
'Z.'
>.
€~
>J><
>,
>>|x
2
ai
C
0
C/)
C/N
tSl
W5
~
n
X)
n
X>
a)
JO
a
X)
u
■a
OS
tC
OS
OS
OS
OS
ai
Oi
E
O
U
U
U
a
Q
a
Q
— "
*-*
J
_)
_1
_l
_i
-J
_)
—1
o
■S cn|
a e
o E
c x
I!
*T3 o
3 cfl
^3 T3
5 s
o t: *a
"" c o
3 —
.§ 1
x «
e g
'J
0i =
11 11 11 11
c/3, #— 1 rN
E
"3
>
a
u
13
-0
a
id
U
M
■^
U
S3
O «j
o c
■ — n
tc c
o o
-JO
3 O no
2 § o£
"5 3 u
u fl i-
i — <U II
J3 E
11 11 11 11 11
— J U Q rs X) C/5
86
2bh-3
1 1 1 1
0.00 20.00 1(0.00 60.00 80.00
EFFORT: NUMBER OF SKATES (XLO3)
100.0
^ \.
0.00
20.00
1(0.00
60.00
80.00
EFFORT: NUMBER OF SKATES (X10 )
100.00
Figure 9. — Cape Flattery sablefish fishery, 1917-52: Catch, effort,
and catch per unit of effort.
87
4)
O
M
— .*
S o
CJ ■*"
o
m
O
ON
o
lO
ro
ro
in
NO
NO
ro
in
NO
ro
oo
ro
a 0
ON
r—
NO
ON
•*
ON
ON
NO
ro
ON
O
00
ON
**■
NO
ro
lO
ro
ON
■*
•"•
lO
V.
-a
s
3
C"~
NO
ON
r-
ON
■*
>",?
rh
r-
ro
lO
l>
m
NO
r-
&oa-
■*r
•*
NO
00
NO
m
©
m
00
S 3
cm
■*
O
r-
00
OO
NO
O
o
■a-
(S
NO
in
ro
00
NO
o
O
rs
cm
cm
CM
CI
CM
CM
£
H
.*
Cm
ON
lO
lO
o
^<A
•w
""
o
■*
ca
X
Q
z
<
<3
>
<
>
5S
ro
ON
t—
o
t-
co
(—4
r-~
vo
vq
00
CM
00
q
NO
■*
q
a
cm
CM
CM
i/i
t~
ON
ON
■*
o
r~
•<t
no
Cl
m
<*
r-
>n
OO
io
00
i
>,
ft
>.
X
CO
N-^
ON
r-H
ON
NO
CO
00
ts
— ;
— ;
^H
CM
CM
+
CM
+
CM
*j
+-•
X
^^
•^"N
X
X
O
X
o
CM
X
CD
+
„
>T K
>n
l"
X
I/O
WO
ro
ca
X
i
1
»-H
•
|"
E
1
CM
>>
00
CM
•*
CO
^
!/5
m
ON
CM
lO
w
CM
CM
NO
in
00
^H
r~
"c3
O
ON
1
cm
CM
+
00
CM
+
ON
ro
m
-1
CM
r~
CM
ro
00
O
'ft
tN
X
X
O
oo
NO
00
CM
x"
NO
o
^H
1
r^
1
E
ON
q
II
NO
NO
ro
•ti-
ro
X
ON
r~
NO
o
CM
1
X
o
m
X
o
CM
r
X
1
X
■JO
lO
X
in
ro
f
c^l
1
1
1
|"
1
X
00
ro
00
ON
1
X
OO
CM
OO
m
ro
tn
o
00
o
X
+
00
ro
OO
ON
NO
ON
1^
O
if
00
ro
NO
m
1/0
ON
ON
x
_;
II
ON
II
CM
II
CM
II
—
ii
^
II
^ *■'
II
^*
II
^~>
>>
>T
>>|X
>i
>>|X
>T
>.|X
>,
>;|X
_
X
c
0
oi
-1
c/3
on
w
V5
eg
X
ca
XI
ca
XI
ca
X
o
T3
OS
a
at
OS
OS
CA
OS
OS
E
U
u
U
U
Q
Q
Q
Q
—
,£3
J
_1
-j
-J
J
-J
J
_)
3
w
c ■-
.2 5
*- ca
,0 (-4
E*
3 E
11
ca x
a ca
fc E
2 w
a; ca
T3 u
•3 ~
|s
_-c 0
cu *- ■ —
^ ^ O n
.2" S §
O S ^ h,
Sao
S S i "-
^ Cl u g
■~ c o «
«gaE
flCr- 0
E ia-s
g XT >> <u
£-£*> "S
t«.a ^ CM
3 «S as
o
CL* "^
T3 • J3
ca ca
^ ts
00
o o" 3
tS c as I
c^ *.
ca "O
C =>
cu cu
u ca
B-S5
- c a
S o -a
3 3 ca
o •£
o
S 3
3 X)
"•a
O
o g
w5 ca
ca
r3
W cu
ex:
o "
CA —
« °
^ s
c «
S E
Uh ■ —
5 «
<
Q
>
u 5
U o.
3 i_
2^
ca
cu C
3-2
71
b
H
0
O ca
o
X
X
-3
T3
u
X
X
.s^
mi
0
o
cs
Gfl "
>n
O -
09
3
C r~
B.£
XI t
3
c
_0
ON
0
u
— 1
o S
c
u
C
3
in
^e
ac-w
3
u S
3
Lm
611
ii
i- ■ —
o g
U
■a
"5
0
73
u
>
"3
0
ii ii
ii
II
II
E
n
a:a:
-J o
U
«
X OO
-i—
relationship (the LDRaS model gave a larger
R2, but y,_ i was not statistically significant).
However, the LDRa model was marginally
significant over the LCRa model (R2 of 0.88
versus 0.85). There is evidence of positive auto-
correlation for the LDRa function. The Gulland-
LCRb does give somewhat different estimates of
y* and x* than unadjusted LCRb. Figure 8 shows
the LCRa and LDRa functions.
So far, we think that the logistic-decreasing-
returns function has considerable merit. It
should, of course, be tested further. Other func-
tions should also be tried, including those
assuming a Gompertz growth function and the
more generalized function used by Tomlinson
and Pella. It is also hoped that this effort by
economists will be reviewed by people in the
field of biology.
Cape Flattery Sablefish
Table 5 shows the results for the Cape Flattery
sablefish fishery. Again, the LDRa model is
superior in explaining the catch-effort relation
with an R2 of 0.54. The stock adjustment co-
efficient was not statistically significant at the
5% level. Positive autocorrelation was found for
the LDRa function. There does not seem to be an
appreciable difference between the Gulland-
LCRb and the unadjusted LCRb.
On the basis of the sample fisheries it would
seem that the LDRa function is a more realistic
description of the catch-effort relation than the
LCRa model employed by Schaefer. In addition,
it is apparent that for the above species the Gul-
land method of adjusting this data yields very
similar results to the unadjusted. Catch-effort
data have been gathered on 49 stocks of fish by
the Economic Research Laboratory. We plan to
carry out similar investigations for the other
stocks since the basic computer programs have
been written. Figure 9 shows the LCRa and
LDRa functions.
CONCLUSIONS
We do not claim to have discovered the "true"
relation between effort and yield for the stocks
of fish discussed in this paper. We have no
guarantee either that biological growth is exactly
a logistic function, or that yr = m,(l — z,x )
is exactly the relation of effort to yield from a
fixed biomass. But we believe that (1) the
decreasing-returns functions yt = m, (1 —zt t)
is theoretically better than the constant-returns
function y, = km, employed by Schaefer; and
(2) the decreasing returns function also gives
better statistical results as shown graphically
in Figures 6 to 9 and is confirmed by the
correlation coefficients.
LITERATURE CITED
BARNETT, H. J., and C. MORSE. 1963. Scarcity and
Growth: The Economics of Natural Resource Avail-
ability, Baltimore.
BELL, FREDERICK, DARREL NASH, ERNEST
CARLSON, FREDERICK WAUGH, and RICHARD
KINOSHITA. Manuscript. The Future of the World's
Fishery Resources: Forecasts of Demand, Supply and
Prices to the Year 2000 with Recommendations for
Public Policy, U. S. Department of Commerce, National
Marine Fisheries Service, Economic Research Laboratory.
CARLSON, ERNEST W. 1969. Bio-Economic Model
of a Fishery, Economic Research Laboratory, U.S.
Department of Commerce. Working Paper 12.
DAVIS, HAROLD T. 1941. The Theory of Econometrics,
Bloomington, Indiana: Principia Press, Chapter 11.
FISHER, IRVING. 1925. Our Unstable Dollar and the
So-Called Business Cycle, Journal of the American
Statistical Association, pp. 179-202.
FOX, WILLIAM JR. 1970. An Exponential Surplus-
Yield Model for Optimizing Exploited Fish Populations,
Transactions of the American Fisheries Society, No. 1.
GULLAND, J. A. Manual of Methods for Fish Stock
Assessment. Part 1. Fish Population Analysis. FAO
Manuals in Fisheries Science No. 4. Rome.
KOYCK, L. M. 1954. Distributed Lags and Investment
Analysis, Amsterdam, pp. 9-14.
MARSHALL, ALFRED. Principles of Economics, Mac-
millan, 8th edition 1920 and reprints to 1930, p. 166.
London.
NERLOVE, MARC. 1958. The Dynamics of Supply. The
Johns Hopkins Press, Baltimore.
PELLA, JEROME J., and PATRICK K. TOMLINSON.
1969. A Generalized Stock Production Model, Inter-
American Tropical Tuna Commission, Bulletin 13(3).
89
RYTHER. JOHN H. 1969. Photosynthesis and Fish
Production from the Sea, Science, Vol. 166, pp. 72-76.
SCHAEFER, MILNER B. 1954. Some Aspects of the
Dynamics of Populations Important to the Management
of Commercial Marine Fisheries, Inter-American Tropi-
cal Tuna Commission, Bulletin 1(2), 27-56.
SCHAEFER, MILNER B. 1956. Some Aspects of the
Dynamics of Population Important to the Management
of Commercial Marine Fisheries, Inter-American Tropi-
cal Tuna Commission, Bulletins 1 and 2.
APPENDIX I. METHODS OF ADJUSTING CATCH AND
EFFORT DATA TO REPRESENT EQUILIBRIUM OBSERVATIONS
The Schaefer (1957) Method
The Schaefer analysis (using his notation) is
based on the assumption that the rate of
population change can be represented by the
equation
(l)^ = klP(L-P)-k2FP
where k\ is the rate of population increase,
k2 is the catchability coefficient, L the maximum
population size, F is fishing effort, and P is
the current population size. Further, it is
assumed that at level P, in year i, equilibrium
yield, Ye is estimated by P + Catch, and that
(2) AP
Pt + 1- Pr-
Ct*i/Ft+i -Ct-i/Ft-
where C is catch. To use these equations it is
necessary to relate P and u, catch per unit effort,
that is
(3) P=k2u.
If P in equation (1) is replaced by P, then all
three parameters k, k2, and L can be estimated
from a series of data on catch and catch per unit
of effort. This 1957 procedure of Schaefer's was
first tried as a basis for a decision rule.
Initially a 15-year series of data was divided
into three equal parts, that is, 1 to 5, 6 to 10, and
11 to 15 years. The three parameters were
estimated from the three sets of data by solving
the simultaneous equations of the form
nt
11 ri
A Ui = k] X
r " — ■
L i u]
y=l n.
-2
-k2 v nt Iil
7=1 n,
where k\, and k2, and L are parameters, A u, is
the change in catch per unit effort, u, is the
average catch per unit effort u,~ is the average
catch per unit effort squared, f, the number
of units of effort and n, the length of the period
in years.
Pella and Tomlinson suggested that the series
of data be divided into periods with the greatest
differences in stock levels to avoid absurd results.
They also pointed out the lack of a unique
solution, since different partitioning of the data
may give different results. There is also no
statistical basis on which to infer properties of
the parameters such as bias, consistency, or
efficiency, etc.
Gulland (1961, 1968b) Method
This method involves relating the mean
annual catch per unit of effort in a given year
to the fishing effort, averaged over that year
and a certain number of previous years cor-
responding to the mean number of years that a
year-class contributes to the fishery.
For example, the catch in period t would be
related to the average effort over the last 3 years
for the yellowfin tuna since a year-class con-
tributes to the fishery for about 3 years. We are
doubtful of the validity of this since it gives
equal weight to each year of effort in computing
the average effort. We feel the hypothesis ex-
pressed in this paper is more realistic. In addi-
tion, the statistical properties of a moving
90
average of effort as used in regression are not in the Eastern Tropical Pacific Ocean," Inter-
well known. Finally, the technique is not as Amer. Trop. Tuna Commission. Bull. 2(6) 1957,
direct a test for adjustment of the population pp. 245-285 and Gulland, M. Manual of Methods
to effort as the one used in this paper (see for Fish Stock Assessments. Part 1. Fish
below). See Schaefer, M. B. "A Study of the Population Analysis. FAO Fish Technical
Dynamics of the Fishery for Yellowfin Tuna Paper, 1968, FRs/T40 (Rev. 2), 97 pp.
91
Some Suggestions for the Development of a
Bioeconomic Theory of the Fishery1
Russell G. Thompson2
ABSTRACT
In this study, the fundamental characteristics of the Schaefer model and the
Thompson-George (TG) production-investment model are reviewed, and extensions of
the TG model are discussed. It is then indicated how a bioeconomic model for the
sole ownership fishery may be obtained by adjoining the Schaefer model to the TG
model (or any of the extensions). This leads into a discussion of the fundamental variables
in a dynamic analysis of the fishery problem and the limitations of published bioeconomic
analyses. It is further pointed out that further work needs to be directed to the
formulation of catch functions allowing for varying marginal returns with respect to
fishing effort, in particular.
INTRODUCTION
In 1954 Schaefer used the first-order terms of
the sigmoid growth law to describe the dynamics
of an unexploited fish population and assumed
the catch to be proportional to effort3 to describe
the exploitation by man. The catch function was
subtracted from the natural growth law to
obtain the following model (which is commonly
referred to as Schaefer's model):
(1) x(t) = rx(t) (v—x(t)) - ay(t)x(t)
where x is the fish biomass, y is fishing effort,
t is time, x(t) = dx(t)ldt, and the remaining
symbols are parameters.
In 1968 Thompson and George formulated a
production-investment model for the firm in-
volving stocks and flows. Less than full use of
the capacity was allowed for by introduction of
a production scale variable. Short- and long-run
distinctions in economics were thus possible.
The firm could increase the capital stock by the
1 Partially supported by the National Science Founda-
tion as a part of the Sea Grant Program for 1970.
2 Russell G. Thompson is Professior of Quantitative
Management Science, University of Houston.
3 As indicated by Schaefer and Beverton (1963), this
assumption is common to the Beverton-Holt approach
as well.
purchase of capacity in excess of attrition. None
of the capital stock could be sold within the
decision interval of finite length; it could only
be sold at the end of the interval. Therefore, the
problem was irreversible during the finite period.
Extensions to allow for increasing marginal
costs are straightforward and were left to the
reader. The decision rules for the optimal
production and investment controls were derived
by use of control theory methods. An algorithm
was developed by which to compute solutions to
the controls so that the model had practical as
well as theoretical value.
In 1970 George showed that solutions to
the optimal controls for a cash flow form of
analysis (as used by Thompson and George) were
identical to those for a discounted form of
analysis. That is, in reference to the TG model,
the optimal controls are the same for the case
where b(t)>o and D(t) = o as for the case b(t) = o
and D(t) is evaluated at the market rate of
interest i(t). George further showed that one
model or the other must be used (in an exclusive
sense).
In 1971 Thompson, Hocking, and George
showed how the initial values for the physical
and money capital accounts can be derived
optimally as a part of the solution to the
investment-production problem (as well as the
values for the controls during the decision-
making period). In 1970 Proctor studied the
investment problem for the firm in a reversible
and also in an irreversible setting (where the
92
firm may buy and sell its capital stock during
the period as well as at the end). He further
derived the demand functions for capital in each
case and deduced their economic characteristics.
CONCEPTUAL MODIFICATIONS
By adjoining the Schaefer model to any one
of these formulations, a production-investment
model for the sole ownership fishery is obtained.4
Such a formulation has a number of distinct
advantages: First, the inherently dynamic prob-
lem of the fishery is formulated accordingly in a
mathematical sense, second, the model (since it
encompasses the economic and biological rela-
tions) is bioeconomic in form; third, given mean-
ingful expressions for the functions involved,
decision rules for the production and investment
controls (and hence the basis for a bioeconomic
theory) may be derived by the straightforward
use of published mathematical methods.
Lack of such a methodology may be the reason
for the historical development of the bio-
economic theory for the fishery. For example,
virtually all economists who have published
in the professional journals (or by the way of
Resources for the Future) have commonly as-
sumed the inherently dynamic problem of the
fishery to be static at the outset of their
analyses (cf. Smith 1969), Christy and Scott
(1965), Gordon (1954), and Crutchfield and
Pontecorvo (1968).
Another example is provided by the form of
the catch function used. Until recently, econo-
mists have not seriously questioned the form of
the catch function introduced by Schaefer, oyx.
This formulation implies constant marginal
returns with respect (w.r.) to effort and in-
creasing returns to scale.
Crutchfield and Zellner (1962) made static
and dynamic analyses of the fishery problem
(with this catch function) and found different
constant solutions! They failed to note that a
capacity limitation must be imposed on fishing
effort. The problem is similar to maximizing
the function y = x in which the domain must be
4 Any of these forms of the problem are consistent
with Turvey's formulation (1964). Variations in mesh
size would be associated with different capital character-
istics, and require the introduction of more than one
capacity variable and possibly functions relating vessel
types and mesh size.
bounded from above for the problem to have
finite solution.
Following this analysis, Crutchfield and Zell-
ner introduced a Cobb-Douglas form for the
catch function and made a partial analysis of
this case. This problem also requires a capacity
limitation on effort to be well posed. In addition,
increasing returns to scale in capacity for
sufficiently small expenditures may be neces-
sary as well as decreasing returns beyond some
point. This is particularly relevant when the
competitive model is desired for a reference
framework. Decreasing returns everywhere are
inconsistent with the market requirements for
a competitive structure (Proctor, 1970).
Still another example of the unusual approach
used to date is the specification of an infinite
horizon for the completely irreversible invest-
ment problem. The optimal length of the horizon
in a common property resource problem might
well be one of the fundamental results being
sought in the analysis, and not an input to the
analysis, as specified by Crutchfield and Zellner.
There are no transferable rights to the fishery
resource; and hence, the entrepreneur might
desire to take all of the resource within a finite
period of time. Thus, the optimal solutions to
the investment and production controls and the
length of the decision horizon would be expected
to be the fundamental variables for a bioeconomic
theory of the fishery.
For the case of the Schaefer model, the decision
rules for the production-investment controls
follow immediately from the TG model. The
necessary condition for the optimal length of
the decision interval, if one exists, follows as
an immediate extension of their results. In
fact, the decision rules for investment and
production are particularly straightforward and
easy to state. Let v — investment, m = the
investment upper-bound, 7 = the fish price,
6 = production cost per unit of effort, f = in-
vestment cost per unit of capacity, <P = the
discount function, z = fishing capacity, p\ —
the marginal value of the fish per unit weight,
and p2 = the marginal value of capacity. Then
the decision rules are:
(2 =0iip: -0f <0,
m if p: — 0f >0,
93
(3) v,,
0 if 0ioxo — pi ox,, — <f>0 <0,
z,t if6yox0 —pi axo — <pO >0.
with the subscript on v, y, x and z denoting
optimum values.
method may be further enhanced considerably
by the development and estimation of more
robust forms of the catch function.
The sole owner firm invests the maximum pos-
sible amount if the marginal value of capacity
is greater than the discounted marginal cost of
capacity and does not invest at all if the opposite
is the case. The firm uses all of its capacity if
the discounted net marginal revenues from
fishing effort, <p( ■yox,,— 0), are greater than the
marginal value of the fish resource, p/ ox,,, and
the firm does not fish at all if tht marginal value
of the fish resource is greater than the net
marginal revenue from fishing.
The difference between a sole owner firm and
a competitive firm is immediate. In the latter
case, the effects of fishing on the resource are
ignored; and hence, the marginal value of the
fish resource is always zero (since px (t) = o). It
can further be shown that pi^(t) for all £. Thus,
the marginal value of the fish resource reduces
the value of the decision rule for fishing effort.
If the Schaefer model is augmented to allow
for a Cobb-Douglas type of catch function, for
example, then an interior solution (in the
interval [o, z0]) for fishing effort is possible.
Similarly, an interior solution (in the interval
[o, m]) for investment costs is possible if
increasing marginal costs of capacity are
specified.
The main difficulty in applying the TG model
(as first developed) is specification of the invest-
ment upper-bound. It is clearly a proxy for
various limitations on investment. For instance,
there might be borrowing limitations imposed
by the financial community. If so, Rahman's
extension (1970) of the TG model may be ap-
propriate. On the other hand, the investment
upper-bound may be superfluous if the catch
function is of a traditional production function
form. Few serious efforts have been directed to
investigations of alternative forms for the catch
function. Further efforts of the type being
pursued by Carlson (1969) surely need to be
given top priority in fishery research.
In summary, an operational methodology for
the management of a fishery is available by
adjoining the Schaefer model to the TG model,
or one of its extensions. The potential for this
LITERATURE CITED
CARLSON, ERNEST W. 1969. A Bio-economic Model
of a Fishery. Working Paper No. 12, Division of
Economic Research, National Marine Fisheries Service,
U.S. Department of Commerce.
CHRISTY, F. T., JR., and A. SCOTT. 1965. The Common
Wealth in Ocean Fisheries. Published for Resources
for the Future, Inc., by the Johns Hopkins Press,
Baltimore, 281 pp.
CRUTCHFIELD, J. A., and G. PONTECORVO. 1969.
The Pacific Salmon Fisheries, A Study of Irrational
Conservation. Published for Resources for the Future,
Inc., by the Johns Hopkins Press, Baltimore.
CRUTCHFIELD, J. A., and A. ZELLNER. 1962.
Economic Aspects of the Pacific Halibut Industry.
Fishery Industrial Research. Vol. 1, No. 1.
GEORGE, M. D. 1970. Discounting and Cash Flow
Analysis in Investment Problems. Unpublished manu-
script available from author on request.
GORDON, H. S. 1954. The Economic Theory of a
Common Property Resource: The Fishery. Journal of
Political Economy, 62(2): 124-142.
PROCTOR, M. S. 1970. Investment Theory for the
Firm: Deterministic and Stochastic Models. Unpub-
lished Ph.D. dissertation, Texas A&M University.
RAHMAN, QUAZI MD. MAFIZUR. 1970. An Optimal
Investment and Financial Control Model: Theoretical
Solutions and an Application. Unpublished Ph.D.
dissertation, Texas A&M University.
SCHAEFER, M. B. 1954. Some Aspects of the Dynamics
of Populations Important to the Management of the
Commercial Marine Fisheries. Inter-American Tropical
Tuna Commission, Bulletin, 1(2): 26-56, La Jolla,
California.
SCHAEFER, M. B., and R. J. H. BEVERTON. 1963.
Fishing Dynamics — Their Analysis and Interpretation.
In M. N. Hill (editor) The Sea, pp. 464-483. Interscience,
Vol. 2, New York.
SMITH, V. L. 1969. On Models of Commercial Fishing.
Journal of Political Economy. 77(2): 181-198.
THOMPSON, R. G., and M. D. GEORGE. 1968. Optimal
94
Operations and Investments of the Firm. Management 767-772.
Science, 15(1): 49-56.
THOMPSON, RUSSELL G., R. R. HOCKING, and TURVEY, R. 1964. Optimization and Suboptimization
MELVIN D. GEORGE. 1971. A Nonconvex Control in Fishery Regulation. American Economic Review,
Problem for the Competitive Firm. Econometrica. 39(5): 54(2, 1): 64-70.
95
Practical Problems of Constructing Bioeconomic
Models for Fishery Management
Paul Adam1
ABSTRACT
In many practical cases it is impossible to construct a complete bioeconomic model
of a given fish stock, such as when one or several fleets move irregularly from one stock
to another, or when fishing effort increases so rapidly that it is not possible to
accurately specify a reliable yield/effort relationship. A continuing bioeconomic model
is proposed here which will allow inclusion of these dimensions while allowing both
for year-to-year fluctuations in managed effort and also for gradual adjustment of labor
and capital to those levels designated as optimal within the broad ranges of this
continuing model. Year-to-year re-evaluation offish stocks and capital-labor requirements
is stressed.
INTRODUCTION
This paper is devoted to the problem of mixed
fisheries. Few fish stocks are exploited by one
fishing fleet only and few fishing fleets are
dependent upon only one fish stock. In the rare
cases of isolated fisheries (one main species,
one fleet, one market) there are often incidental
catches which, although they may be relatively
small, are important for the overall profitability
of the fleet. It can be said that in most fisheries
the rule is to switch from one type of fishing to
another or from one stock to another, according
to the seasons or to the variable fish abundance
in the different stocks. These continuous adjust-
ments, occurring irregularly, make the problem
of fishery management a most complex one.
Furthermore, it must be added that in the
last 10-15 years the techniques used in some of
the most important world fisheries have been
considerably improved. These developments
include: long distance stern trawling associated
with freezing at sea, purse seining for pelagic
species in the North Atlantic, purse seining for
tuna species in the Central Pacific and Atlantic,
double beam trawling in the North Sea, etc. As
a consequence of these recent developments, it
is more difficult to study those fisheries which
1 Head of the Fisheries Division, Organization for
Economic Cooperation and Development. The author is
solely responsible for the ideas and information presented
in this paper.
are the most advanced and consequently the
most interesting.
The study made in this paper will obviously
be economic, but no serious or complete eco-
nomic study of any fishery can be undertaken
without consideration of the available resources.
In other words, the work of the economist in this
context cannot begin or would have no solid
basis without starting with the findings of
marine biologists. It is therefore indispensable
to examine the nature, the scope and especially
the shortcomings of the biological findings inas-
much as they have to be used by the fishery
economists.
SHORTCOMINGS OF THE
BIOLOGICAL MODELS
The whole process of the fishing operations
is expressed in Figure 1. The arrows indicate
the basic components of an operating fishery.
It makes it apparent that any research which
would isolate either biology or economics would
be cut off from the feedback occurring in reality.
Any model used to describe reality will be false
if it is divided into two isolated parts.
The traditional catch curve derived from the
biological findings on one fish stock cannot be
directly used by the economists. In fact, this
curve, which is an average catch curve, should
be supplemented with two curves indicating the
maximum and minimum yields according to the
96
BIOLOGICAL RESEARCH
ECONOMICAL RESEARCH
Fishing Effort
(vessels + gear)
expressed in
physical terms
Cost of Fishing Effort
+
Profit
Fish stocks
Catches
(in weight of fish
for each stock)
Returns
from the sales
of the landings
Figure 1. — The basic components of the fishing process.
fluctuations of abundance. As shown by Figure
2, these curves of maximum and minimum yields
accentuate departure from MSY as compared
to the average curve with increasing fishing
effort (and, after the point of MSY, increasing
overfishing). The reason is that the more in-
tensive is the fishing effort, the faster the year
classes are exhausted, as there are often rather
wide fluctuations in the strength of the succes-
sive year classes. The fluctuations of the catches
can only be increased with a faster exhaustion
of the best year classes.
As shown by Figure 2, it is difficult to
evaluate the social cost of fishing effort unless
we have the simple case of a given fleet exploit-
ing a given fish stock. In such a case, the losses
of years of bad catches are compensated by the
profits made in better years. Or, if the market
for the landings is also isolated, it might be that
the returns are more or less equalized by higher
prices when there is a scarcity in landings and
lower prices when the landings are more
abundant.
For mixed fisheries, Figure 2 should be
transformed into Figure 3, thereby taking into
account the fact that the fishing fleet exploiting
a given stock at a given average level is maxi-
mum when the abundance in the given stock is
maximum and when the abundance in the other
stocks that can be fished by the same fleet is
minimum, and vice versa. No stock can be
subject to a stable fishing effort. It will vary
97
Catches
Average cost curve
Average yield curve
Fishing effort
Figure 2. — Maximum, average, and minimum catch curves for a single fish stock.
Average yield curve
Fishing effort
Figure 3. — Maximum, average, and minimum catch curves for a multiple stock fishery.
between two extremes determined by the
abundance in the stock considered but which also
depend upon the abundance in the neighboring
stocks. It should be noted that the shape of the
resulting curve and the location of the point of
equilibrium would have to be determined for
each particular case. Each case would not only
be the result of the structure of the given fish
stock and of the exploitation borne by this
stock, it would also be the result of the structure
of the other stocks which would be more or less
attractive, i.e., profitable. The findings of the
98
biologists should therefore cover all the stocks
which are exploited by the fleets considered by
the economist, otherwise there would be a
substantial gap in an essential part of the
needed information.
The previous paragraphs were based on the
assumption that the pattern of the recruitment
to the fish stocks remains unchanged whatever
the size of the stock and the level of the fishing
effort. In practice, this assumption is certainly
not realistic. But the opposite assumption that
the level of recruitment is linked solely to the
size of the stock is certainly equally erroneous.
These two remarks oblige us to enter some-
what into the intricacies of the computations
made by the marine biologists. When these
scientists are examining the past catches they
proceed along analytical lines which are cor-
rected every year according to what has hap-
pened. Their analyses are summarized and
systematized with the help of mathematical
functions. These functions can serve the addi-
tional purpose of making forecasts about the
effect of a diminishing, sustained, or increased
fishing effort in the years to come, ceteris
paribus.
Among these other factors the main one is
the pattern of recruitment. When a constant
rate of recruitment is assumed, the mathematics
lead to a curve tending asymptotically to a
minimum yield equal to an exploitation level
associated with average yearly recruitment.
When recruitment is assumed to be aligned with
the size of the stock, mathematics lead to a
curve asymptotic to the X axis or to a parabola.
In fact, both assumptions are false and known to
be false; the real curve for each stock is in
between these two different mathematical
formulations, but present scientific knowledge
in marine biology does not allow us to know
when the pattern of recruitment becomes
different.
The resulting margin of error is of course
without practical importance when there is a
stable fishing effort. When the increase of fishing
effort is slow, the impact can be surveyed step
by step and the margin of error remains small.
But when the increase of fishing effort is fast
and furthermore when fishing effort is, as is
true in complex fisheries, significantly varying
from one year to the other, the margin of error
is bound to be as large as the distance between
the two curves. This precludes an accurate
forecast. In any case, it seems that most often
the yield curve is relatively flat around the
maximum. The Schaefer model tends to exag-
gerate the sharpness of the turning point at
MSY, whereas the Beverton and Holt model
may tend to exaggerate the flatness after MSY.
Let us imagine a fish stock exploited as in
Figure 4 at a variable level of fishing effort,
with fluctuations stabilized at maximum and
minimum levels unchanged for a number of
years. The calculations of the biologists lead
to a derivation of a yield curve as drawn in
Figure 4. The margins of error in the calculations
are such that, if there were a change in recruit-
ment function around the point of average yield,
it could not be easily seen; the actual average
yield curve could well be drawn by the dotted
lines and no one could prove which is the real
one. This is not critical if the fishing effort is
not increased, but assuming, as it is often the
case at present, an increasing demand for
protein and improved productivity due to tech-
nological change, the only practical problem
would be the problem of an increased fishing
effort . . . for which, with such data, no forecast
at all could be made before a new stabilization
of fishing effort for a subsequent number of
years. Before such a stabilization, the most
detrimental consequences could have materi-
alized (cf. the California sardines). The faster
the increase in fishing effort, the more difficult
are the assessments.
PARTIAL BIOECONOMIC MODELS
While initially I attempted to prove that
biological models cannot be complete, at least in
the most important cases of increasing fishing
effort, it is not necessary to stress that com-
plete bioeconomic models cannot exist. It is an
obvious fact that in bioeconomic models biology
comes first; they are fully dependent on the
reliability of the basic biological data. This is
a very big drawback which would well render
the whole exercise of very little practical help
in managing fish resources. But it should not be
forgotten that, in most cases, the biologists can,
with reasonable accuracy, indicate the level of
maximum sustainable yields. This limit gives
a very important and solid basis for assessment
99
M.S.Y.
Fishing effort
Figure 4. — Alternative yield curves for a fish stock exploited at variable levels
of fishing effort.
as such a limit cannot be overstepped without
economic losses.
It could also be added that the impossibility
of constructing complete bioeconomic models is
not as harmful as might be thought. In many
cases of advanced overfishing, complete bio-
economic models would not necessarily supply
practical management policies. In a situation
of advanced fishing effort, the benefits to be
expected from fishery management are benefits
which could not be reaped before the stock is
rebuilt to its MSY level. In the meantime the
reductions likely to be made in fishing effort
would cause problems of de-investments (e.g.,
scrapping premiums . . .) and of employment
(re-employment of the fishermen concerned).
Furthermore, a reduced and less costly fishing
effort exploiting a rebuilt stock would give
rents; it is possible to imagine regulatory means
by which such rents would be at least partly
taken from the remaining fishermen, but this
could only be made on the basis of the fishing
techniques prevalent at the time of making the
regulation. It would often be difficult to find
the regulations which would result in the desir-
able aggregate effort while permitting new
technological developments at the same time.
Some success has been achieved in the Canadian
salmon program toward attaining both of these
ends. In other words, even if complete bio-
economic models would exist they would not
as such provide complete solutions to the
problems of re-establishing overfished stocks to
the ideal situation of MSY.
Before going further it is necessary to say a
few words about the techniques of communica-
tion between biologists and economists. In fact,
there is not much difficulty with the basic
Schaefer model which is widely used in the
United States. Its mathematical expression is
as follows:
(1) Y = aE + bE^
where
Y =
E =
a,b =
yields, expressed in weight of
catches
fishing effort, expressed in
number of given vessels during
given times
parameters characterizing
each particular stock.
The economist has little difficulty in following
and utilizing biological results from this model.
100
Unfortunately the Beverton and Holt model
is not so easy to handle. In its simplest expres-
sion, it reads:
(2) Z = M + F =
log(, number of fish at beginning of year
number of fish at end of year
where
Z — total mortality
M = natural mortality
F = fishing mortality.
If calculated on a weight basis instead of a
number basis, account should also be taken of
the rate of growth of live fish.
With such a model converting the figures of
the biologists into units which can be utilized
by the economists is most often impossible.
No mathematical barrier exists as long as it is
understood that the natural logarithm of a
ratio between the catches or the stocks of two
years is, in fact, a percentage. However, an
important part of the data utilized by the
biologists, when it is all published, is scattered
in many different publications. It is not suf-
ficient to know the ratio of abundance derived
from fishing effort (F) and the ratio of natural
mortality (M); the ratio of the growth of the
fish and the assumed recruitments are also
indispensable but not easily available. Further-
more, the relationship between ratios and actual
figures are too often summarized to an extent
which forbids reconstruction of the details of
the computations and of the results.
While the present paper is mainly directed
toward an improvement of the cooperation
between biologists and economists, it should
be stressed that a prerequisite is to have access
to the results of the computations of the other
discipline. Cooperation does not require working
at the same desk, but it would ask for this
minimum of understanding.
Unfortunately, the facility with which the
Schaefer model can be used by the economists
does not always mean that there is a perfect
and total understanding between fishery biolo-
gists and economists. More important perhaps
than the unit of measurement are a few basic
concepts which are commonly used with different
meanings. The fishing effort concept is by far
the most important one.
Fishing effort is in fact usually expressed in
many different ways: either by its physical
characteristics or by its returns in weights of
different fish species or in money values (either
returns or costs, or profits). The usage of these
different units should be systematized, other-
wise the concept of fishing effort would be
misleading as is too often the case when so
many researchers use it with different and
implied assumptions on the way it should be
expressed. In fact, there could not be one single
way of expressing fishing effort; fishing effort
considered in its full and general meaning is a
combination of the different units by which it
could be expressed.
Physical Characteristics of Vessel and Gear
This could include any kind of measure
describing the characteristics of the vessel:
GRT, power, length . . . also taking into account
items like the number of berths (which might
be significant for pole-and-line techniques), or
the sonar (for purse seining), or the number of
pots (for crab or lobster, etc.). Obviously, for
each specific case the most important character-
istic^) to be used as a measure of the impact of
the fishing on the stock or as a measure of the
fishing power in relation to a given fish stock
will vary. Therefore, a multipurpose vessel has
a different fishing power according to the gear it
is utilizing; it might even be that the fishing
power has to be different when the same vessel
with the same gear is exploiting different stocks.
As a result the fishing effort of the same boat
would have to be expressed differently for each
type of exploitation, each season, each year,
each stock, etc.
Cost of Fishing Effort
Building costs and operating costs which
could be combined by using operating costs
including depreciation plus overhead are a more
permanent type of unit. First, the costs of a
given boat are not so much changed when it
changes gear. Secondly, many boats have been
built for a definite type of usage. The costs of
a boat will be easily defined by so much per
day at sea.
101
Unit of Time
The biologist and the economist will be
naturally inclined to use different units of time
(time at sea for the second and time fishing for
the first). Anyway, the distance to the grounds
will have an opposite effect for both researchers,
the longer the distance, the higher the costs or
the fishing effort for the economist; the shorter
the distance the higher the impact on the fish
stocks, or the fishing effort for the biologist.
The conclusion is obvious. There cannot be
such a unit as a unit of fishing effort. Fishing
effort is a complex concept; it is a ratio or a
relationship between different units. To assume
that it can be defined once and for all and be
used indifferently by researchers of both dis-
ciplines, economics and biology, is a complete
mistake. Each time that the concept of fishing
effort is utilized it should be made clear what
it really means. Attached to a stock or fishing
technique, its value is limited to this stock or
technique. Given in money terms its compar-
ability is attached to the economic systems of
which it forms part.
CONCLUDING REMARKS
A substantial complexity is the consequence
of the impossibility of building up a complete
bioeconomic model, of the difficulty of converting
to economic measurement the ratios used by a
number of biologists, of the lack of a clear
understanding of what fishing effort is, of the
impossibility of forecasting the pattern of
recruitment of the fish stocks. To overcome this
complexity it does not seem that one can in-
definitely rely upon equations which, whether
they are Schaefer's or Beverton and Holt's, are
mostly used analytically to give account of
past developments but cannot make apparent
the mechanisms through which future develop-
ments are taking place. Figure 5 shows that
these biological equations only concern the
squares 1, 2, and 3 when a complete simulation
BIOLOGICAL RESEARCH
(by fish stock)
ECONOMIC RESEARCH
(by fishing fleet)
Actual vessels
catch rates
Direct assessments
e.g. by accoustic
methods .
Corrected for
standard vessels
Assessments of
mortalities and of the
size of the fish stock
Growth and recruit-
ment observations
Forecasts of the
future stock and
mortalities
Market
assessments
Existing
fishing fleets
Economic assessment
of the past and
present situation
Expected
changes in the
markets
Expected
changes in the
fleet
Necessary
adaptation
leading to
New market
conditions
Redevelopment
of the fleet
J
Figure 5. — A simulated flow chart of a fishery.
102
model should incorporate the 14 squares includ-
ing independent measures of the size of the
stocks and of recruitment and the feedback
from the economic side.
It is often said in international fishery dis-
cussions that no regulation should be adopted
or even proposed before it can be justified by
sufficient "scientific evidence." Nobody is fooled
any more by this sophisticated expression which
means that national economic short term
interests should prevail as long as there is no
definite proof that such national interests are
leading to detrimental international economic
consequences. It is obvious that such scientific
evidence has often been supplied by the
biologists, if only when they stated that numer-
ous stocks are exploited beyond the point of
MSY. But the precise economic consequences
of these statements are very rarely available;
and there is practically no case where the
economic consequences of the cuts to be made in
the fishing effort have been evaluated (short
term costs or losses and long term benefits
according to the possible regulations to be
adopted). It is obvious that such "practical"
evidence will never be supplied without a close
cooperation between biologists and economists.
The possibility of successful fishery management
is entirely dependent on such bioeconomic
research work.
103
ISSUES RELATED TO FISHERY MANAGEMENT
RESEARCH RESULTS
In the final section concerning other issues
related to fishery management, the first paper
by Holmsen summarizes the results of his study
of the Peruvian anchoveta fishery. His is very
much an applied study, for he is interested in
indicating the critical components of what they
have done in the past, the faults that may exist,
and an evaluation of alternative management
programs.
By his measure the current excess capacity
in the fleet should be reduced by 14-38% depend-
ing upon the biological or social constraints
imposed (length of closed season). Alternative
plans which might correct this situation are
reviewed, including:
(1) restrictions on fleet size.
(2) government purchase of scrap fleet, the
cost to be covered by an assessment on the
remainder of the fleet; new entry would be
restricted simultaneously.
(3) require private scrapping to permit new
private construction — a scrapping ratio.
(4) tie fleet size to licensed capacity of fac-
tories.
(5) a quota system with variable, long-lived
shares allocated via an auction system.
As there is excess capacity at the processing
level also this becomes part of the consideration.
Possible controls here would be (1) reducing
licensing capacity leading to forced insolvency,
(2) government purchase of plants, or (3) trans-
ferable factory quotas.
Holmsen recommends a combination program
including both levels. Emphasized would be a
high scrap/rebuild ratio and lifting the debt
moratorium on plants.
In the paper by Thompson, Callen, and Wolken
the Thompson and George model, as previously
referred to, is expanded to account for income
taxes and depreciation. Emphasizing the desire
for survival as a key decision element the
authors apply this model to sample firms in the
Gulf shrimp fishery, using alternative sets of
price and landings data. The critical nature of
each decision variable is noted for each set of
inputs.
Anderson, Connolly, Halter, and Longhurst
present another version of a simulation approach
to evaluation of management alternatives,
relating experience in the management of deer
population subject to different hunting strategies
defined by alternative sets of regulations.
Some interesting general methodological
points are made in this paper. Among these is
the stress on the iterative-feedback elements of
the simulator. By stressing this mechanism in
fisheries we could obtain a continuing evaluation
of the quality of the input in addition to the
quantitative dimensions of alternative programs.
Thus, a type of continuing sensitivity analysis
can be performed on such items as estimates of
MSY, alternative measures of fishing power,
the existence of diminishing returns, social
transfer costs, and alternative discount rates.
As does Adam, the authors consider biological
issues to be the essence of first generation models.
Second generation models would include eco-
nomics and other considerations. This differs
somewhat from Pontecorvo, who would have
biology and economics as first and second
generation models, respectively, and other con-
siderations as part of third generation models.
A final element of general interest is the use
of a random number generator to create an
array of "forage factors." This would be a
method of considering the many combinations of
environmental factors that affect recruitment in
fish stocks. In particular, as Pontecorvo suggests,
there may be tradeoffs between levels of accuracy
and the costs of these levels. This analysis
could be performed within a complete simulated
fishery system with the aid of this generator.
The paper by Stevens and Mattox is actually
a report on two separate, but related studies,
one on the economics of salmon hatchery opera-
tions and the other on the supply response of
fishing vessels (boats) to changes in catch/effort
ratios and market conditions. The hatcheries
issue is one which has achieved little attention
in the economics literature and is timely con-
sidering the growth in salmon hatcheries and
the increasing research and development work
being conducted for other species.
That these hatcheries programs are critical
to the overall management plans is a patently
104
obvious, but seldom mentioned, fact. As pointed
out by the authors, with hatchery fish ranging
from 30-80% of all fish caught from hatchery
streams and 20% of all Pacific salmon, no
management program could be successful with-
out explicit Consideration of the hatcheries. In
this examination of 15 Oregon hatcheries pro-
duction functions were estimated which indicated
fixed input proportionality, constant returns
to size and substitution between the fixed
proportional input and water temperature.
In the study of entry and exit an irreversible
function was found to exist. Entry followed
good years, but exit did not follow bad years
to the same degree. Thus, successful "hatchery
years" would lead to entry and expanded fleet
size which could not be justified by lesser, even
average years. This is a further enforcement of
the argument for limited entry as the effective-
ness of hatcheries programs in raising fisher-
men's incomes will be mitigated unless the
countervailing tendency to overcapitalize is
restricted. Part of this restrictive element may
include a deliberate effort to increase opportunity
costs, as discussed previously.
Keen is the only author here reflecting on a
historical system used to limit entry, the
Japanese experience. When reviewing this work
it is necessary to recall that the principal
objective of the Japanese program has always
been "to maintain the viability of the individual
enterprise." As this objective is somewhat akin
to "maintaining the family farm" it differs from
the objective held by most economists to be
desirable. If the Japanese program can be judged
successful in meeting its own objective, it may
still not be suitable to our purposes. Neverthe-
less, we can proceed to evaluate the components
of the program to determine its failure and
successes and to gain an appreciation of the
critical decisions which need to be made in a
management program as it evolves over time.
The Japanese system began in 1946 when all
craft greater than 10 tons had to be licensed. It
evolved to include area restrictions and to be
divided into tonnage groupings, with different
restrictions for distant-water fisheries as these
developed. Its principal overall characteristic
was its pliability. When pressures for additional
development of certain fisheries mounted, ad-
justments were made to allow for some of this
investment. In some instances, when certain
fishing operations were no longer viable, attrac-
tions to divert excess effort to other fisheries
were established. The principal thrust of these
regulations was to modify the tendency to over-
invest and dilute capital values. In some in-
stances, the growing value of fishing licenses
attest to the success of this program.
Critical is the effect of these programs on the
development of technology. It can be shown that
in some cases technology took some strange
courses because of the regulations, somewhat
akin to our own Alaskan limit seiners. This and
other elements of an existing scheme could
prove a fruitful area of examination in the
future, now that substantial progress has been
made in theoretical studies.
The final paper by Huq is so timely as to
appear to be at the unanimous request of the
other authors and participants in the workshop.
This is because the subject is labor mobility and
social transfer costs, with the study reported on
being confined to three representative com-
munities in the Maine pot-lobster fishery.
In this study the goal is to evaluate such
measures of labor mobility as age, level of
education, income levels, technical skills, other
employment, time in present occupation, invest-
ments in the fishery, attitudes toward fishing
as an occupation, and attitudes toward certain
elements of the harvesting process so that alter-
native forms of limited entry would be evaluated.
Results indicate that immobility is substantial,
but that this may not be a problem as the
limitation may successfully be applied to capital
inputs with little reduction in the labor input
for much of the sample examined in the three
communities. For the remainder, some form of
an adjustment assistance program may be
necessary, particularly since a portion of the
labor force in the fishery is currently supple-
menting public assistance or social security
incomes with its lobstering activity. These
members of the labor force truly have limited
opportunities. Restricting their participation
would place a greater burden on other family
members, who may also be in the lobster fishery.
A.A.S.
105
Management of the Peruvian Anchoveta Resource
Andreas A. Holmsen1
ABSTRACT
The best available estimate of the maximum sustainable yield of the Peruvian
anchoveta resource is 9.5 million metric tons ( ± 1 million). The productive capacity of
the purse-seine fleet and the fishmeal factories far exceed this tonnage with the result
that the open season is becoming shorter year by year. This paper describes the current
fishery management program in Peru and the degree of overinvestment in the industry.
It further outlines the alternative methods which can be used to reduce excess capacity
in the catching and processing phase and the advantages and disadvantages of the
various alternatives.
INTRODUCTION
It is well known among fisheries people that
Peru is the leading fishing country of the world
in terms of tonnage landed. About 97% of the
catch is anchoveta, which is used strictly for
production of fishmeal and oil. Besides the
employment and earnings derived from the
harvesting and processing of this resource, fish-
meal makes another valuable contribution to the
economy of Peru. Like many other less developed
countries, Peru has balance of payments prob-
lems and exports of fishmeal and oil account for
approximately one-third of foreign earnings.
With the exception of Iceland, I doubt that any
other country is as dependent on its fishery
resource as Peru, and few are so concerned
about it.
To protect the resource Peru claims a 200-
mile fisheries limit which may be twice as much
as is necessary. Seventy miles is the maximum
distance from shore that anchoveta fishing
takes place. The stock is concentrated in the
waters off the southern two-thirds of the country,
so except for some mixing on the Chilean border
it is entirely a national resource.
Peru's emergence as a fishing nation began
in the 1950's, but most of the growth of the
industry has taken place during the last decade.
During the 1960-61 fishing season (September-
August) Peru's landings of anchoveta were
about 4 million metric tons. During the 1969-70
1 Department of Resource Economics, University of
Rhode Island.
season, landings reached about 11 million metric
tons, and every season during the decade land-
ings were higher than the previous year.
During the early years of the decade, the
rapid development of the industry took place
with little planning, basic knowledge, and
experience. As a result, overexpansion, particu-
larly in processing capacity, has plagued the
industry ever since.
The number of vessels in the fleet reached a
high of 1,778 vessels during the 1963-64 season,
but later gradually declined to the current
size of about 1,400. The vessels have become
bigger every year, however. While 5-6 years
ago, a vessel with 180-ton hold capacity was a
large vessel, the smallest built today has a
capacity of 275 tons and most vessels built
during the last 2 years have a 350-ton capacity.
Thus, the fleet capacity has increased from
about 180,000 tons capacity in the mid-60's to
somewhat above 200,000 tons during the 1969-70
season.
A large part of the fleet is considered obsolete,
consisting of wooden vessels built from 1962
to 1964 (in Peru, 7 years are considered the
economic life of such vessels). In recent years,
most vessels have been built of steel and con-
struction of fiberglass vessels has started. Echo
sounder, powerblock, and fish pump are standard
equipment in the fleet, and the most modern
vessels also have sonar. A fishing trip normally
is a day trip, the vessel leaving early in the
morning and returning with or without catch
in the afternoon.
106
Most of the Peruvian fishing fleet is owned by
firms who also own factories and only about 20%
of the fleet is owned by independent vessel
owners. A fair number of these are tied to a
particular factory, however, and have to deliver
their catch there, owing to financial help
rendered when buying the vessel or for similar
reasons.
As the number of vessels has declined so has
the number of processing plants. A consolidation
has taken place into fewer and larger units.
Currently, Peru has 127 fishmeal factories with
a total capacity of close to 8,000 tons of fish
per hour. About 10 of these plants did not
operate last season. While most firms own only
one factory, a number of larger firms own
several each. These are generally located in
different ports or geographic regions as a hedge
against poor fishing in one particular area.
CURRENT MANAGEMENT PROGRAMS
Both the Peruvian authorities and the Peru-
vian fishing industry have for several years
been aware of the danger of overexploiting the
anchoveta stock, and have taken steps to reduce
the pressure on the resource. Fishing effort
expanded quickly until the 1963-64 season when
the total catch reached a level of 8 million tons.
Thereafter, first closed seasons and then overall
catch quotas were established. At the present
time, the following programs or restrictions
are in force:
1. The fishery is closed on Saturdays and
Sundays.
2. The fishery is closed about 1 month in
summertime during the "peladilla"-season.
That closure ("veda") takes place when
there are large amounts of small fish
(peladilla) in the catch. The time of the veda
varies from year to year. In 1970 the
closure was from mid-February to mid-
March, which was too late.
3. During the fishing season, after the pela-
dilla have entered the fishery and explora-
tory cruises to assess the recruitment have
taken place, an overall quota is established
for the season. When this quota is reached,
the fishery is closed.2
2 Except from the port of Ilo close to the Chilean
border.
4. Each factory has been given a license for
a certain daily input of raw material. The
license capacity is stated in terms of tons
per hour. This quantity multiplied by 24
is the maximum quantity a factoiy is
permitted to accept in one day. Due to the
fact that both the licensed and the technical
capacities of the fishmeal factories have far
exceeded landings, factory licenses have
not been effective in reducing fishing
pressure.
THE CURRENT SITUATION
The Anchoveta Resource
Anchoveta generally spawns in late winter
(August) and reaches a harvestable stage in
midsummer (December-February). It has a life
span of 2 to 3 years. In the early and middle
60's, fish 1-year old or more contributed to most
of the catch, while later the zero year class has
become dominant in the annual catch and its
percentage of the total catch is increasing. This
is considered a warning signal. Actually at the
beginning of last season, September-November
1970, the catch was lower per month than in any
month in 1965, five years ago. The rich 1969-70
fishery did not perform well before the zero
year class came of size. An FAO panel on stock
assessment which met in Peru in January 1970
came to the conclusion that the maximum sus-
tainable yield of the Peruvian anchoveta resource
most probably was 9.5 million tons ( ± 1 million
tons). The experts recommend that the authori-
ties permit a 10-million ton catch coupled with
close observation of the fishery to see what effect
this fishing pressure would have. The authori-
ties, however, permitted 11 million tons to be
caught, which biologists think will significantly
hurt the fishery in 1970-71, both because too
much of the 1-year class already might have
been harvested and possibly also due to reduced
reproductive stock.
Fishing Pressure
While the summer veda is of biological sig-
nificance since it prevents the catching of large
quantities of very small fish, and while the
prohibition of weekend fishing might have some
107
social advantage, the long winter veda, which
has been increasing over time to reach SV2
months in 1970, is only due to an excessive
catch capacity of the fleet relative to the resource
available. Given a maximum sustainable yield
of 9X2 million tons, increases in capacity or
technological improvements of the fleet will
mean a longer winter veda.
During the two years from 1966-67 to 1968-69,
hold capacity increased by 16,000 tons per
year, or about equal to old tonnage leaving the
fleet. During 1969-70, however, about 32,000
tons of new construction entered the fleet and
according to interviews with the various ship-
yards that rate of construction has continued
for the remainder of 1970 (Holmsen, 1970b).
Thus, the fishing season (the number of fishing
days permitted) has gradually declined from
289 in 1963-64 to 166 days in 1966-67 and 155
days in 1969-70. To catch a quota of 9V2 million
tons, a 145-day fishing season would have been
sufficient in 1969-70. Due to the amount of new
construction, with the same abundance and
availability of fish as last season, the fleet
would be able to catch 9V2 million tons in less
than 140 days in 1970-71. As long as fishmeal
prices are high and factories have to pay con-
siderably more to independent owners per ton
of fish than the cost per ton for operating their
own vessels, construction will continue, resulting
in a shorter and shorter season, to the detriment
of the industry as a whole. There are similar
examples from other fisheries where overall
catch quotas have been established with no
limit to entry, such as Pacific halibut and
yellowfin tuna.
Peru is short of investment capital and par-
ticularly short of foreign exchange. In addition
to being a misallocation of capital, however, the
pressure of an excessive fleet poses the danger
of pressure on government to keep the season
open longer than the period recommended by
stock assessment experts.
Processing Capacity
The same problem of overcapacity is found in
the processing phase. Some years ago, the gov-
ernment prohibited the building of more fac-
tories and issued licenses restricting the input
to a specific tonnage per hour for the existing
fishmeal plants. The technical capacities of
various plants were increased, however, without
regard to the license. Last year, the government
started to enforce the law and several firms had
to buy plants to bring their own licensed capac-
ity up to their technical capacity, even when
they had no use for the purchased plant's build-
ings or equipment. Thus, some consolidation
took place and the total licensed capacity now
reasonably reflects the total technical capacity.
The licensed capacity is about 50% more than is
needed, however, even with the short season
now in effect, and the excess capacity would of
course be even greater if the fleet size were re-
duced so the season became longer.
The fishmeal industry as a whole is deep in
debt, liabilities about equal to assets. Since
some firms are in a good financial position, this
means that many firms are thoroughly insolvent,
and would have been bankrupt but for a mora-
torium on debt collection.
DESIRABLE OBJECTIVES
The problem facing the Peruvian anchoveta
industry is how to reduce the excess capacity
both in the catching and the processing phase,
so that excessive closed seasons can be prevented
and the productivity of the remaining production
units improved.
A reduction in capacity and lengthening of
the fishing season has a fourfold advantage:
1. Less pressure will be placed on the govern-
ment to exceed recommended levels of catch.
2. Fewer investment funds will be needed for
the industry.
3. The remaining units will be more produc-
tive and thereby, the economic situation
for the industry will improve.
4. The sustainable yield in the fishery will
increase, as more fish will be caught at a
higher age or larger sizes.
The cost savings which will accrue depend on
the percentage of fixed and variable costs in the
catching and processing phase. For the catching
phase, it will also depend on what percentage
of the variable costs are associated with volume
and what percent with time.
Based upon budgetary data for 1970-71 from
a handful of companies, the following break-
108
down might be a reasonable approximation.3
Forty-seven percent of the cost of harvesting
was found to be fixed and not related to the
number of fishing days, nor the size of catch.
Thirty-five percent of the cost was apportioned
to the size of the catch, of which 34% was the
crew share and social benefits. The remaining
18% was related to the number of days the
vessels were out fishing, catch or no catch.
For a fishmeal factory the cost of fish is a
variable expense and this item alone amounted
to 59% of total cost. The variable cost of pro-
ducing meal and oil amounted to 75% of total
costs and the fixed cost 25% . Excluding the cost
of the fish, the variable costs were 39% and the
fixed costs 61% (Holmsen, 1970a).
What the current overcapacity in industry is
depends on what kind of management program
one has in mind — whether one recommends a
1- or 2-month peladilla veda, whether one sticks
to the 5-day week rather than a 7-day week, etc.
Based on various alternatives from a 7-day
week and no veda to a 5-day week and a 2-month
peladilla veda, the fleet reduction necessary
was found to range from 38% to 14% (Boerema
and Holmsen, 1970). By using the coefficients
above, this would lead to savings ranging from
about $20 million annually in the first case to
about $6 million in the latter case. The savings
in the processing phase would also be significant.
An FAO management panel, which met in Peru
in June 1970, concluded that the technical
capacity of the factories could be reduced nearly
50% under year-round fishing, and that total
savings to industry from reduction of fleet size
and number of plants could perhaps run as high
as $50 million. No value can presently be put
on the lessened risk of overfishing and depletion
of the stock.
ALTERNATIVE CONTROLS
A fisheries management program should have
a double goal: 1) to protect the resource from
overexploitation, and 2) to prevent overinvest-
ment and economic wastes in harvesting and
processing. To achieve these goals in the
3 The percentages are median observations based
upon representative vessel size (140- to 220-ton
capacity) and plants with technical capacity of 60-90
tons per hour.
anchoveta fishery, restrictions can be put on the
fleet or on the factories or on both. Some
programs might achieve the desired result
rather fast, while others might take more time.
Alternative programs related to the catching
phase will first be discussed.
Restrictions on Fleet Size
(1) A reduction in the size of the fleet to the
desired level can be achieved by an embargo on
new construction. Despite the fact that a number
of vessels, which otherwise would have been
scrapped, would be repaired and remain in the
fishery, a fair number of vessels would disappear
from the fishery each year and the season for
those remaining would become longer. Argu-
ments against such a proposal would be that
older, smaller vessels in the fishery, which are
the highest cost producing units, would get an
additional "lease on life" and the fleet would
stagnate technically.
(2) Another possibility with immediate effect
would be for the government to buy up the scrap
part of the fleet (the high cost producer), and
assess the cost on the remainder of the industry,
preferably through a fee per ton of meal produced.
A large number of such vessels would have to
be bought since each contributes very little to
the total catch. The industry would be better
off, however, since the marginal cost of the
remaining vessels would be far below the average
cost of the vessels removed from the fishery.
Such a program would have no long run effect,
however, if restrictions on new construction
were not implemented at the same time.
A scrapping ratio would have to be intro-
duced limiting the annual output of productive
capacity to the amount of productive capacity
leaving the fleet during the year. Such a pro-
gram, which has some support in Peru, still
leaves a difficult question unanswered. Which
vessels should the government buy and scrap
and what would the prices be? Two 6-year old
150-ton vessels are not necessarily worth the
same price. Appraisal and judgment are called
for, which could easily result in kickbacks in a
country where civil service salaries are low and
where bribery has not been unfamiliar in
doing business.
109
(3) A third alternative would be to rely
entirely on a scrapping ratio. If a firm or in-
dividual wants to build a new vessel, he would
have to scrap a larger tonnage of old vessels. If
a firm has no vessels to scrap, it will have to
buy tonnage for scrapping. The time necessary
for an adjustment of the fleet size to the desired
level will be longer than under the previous
alternative. A scrapping ratio (based either on
gross tonnage or tonnage capacity) has to be
relatively high in the beginning, possibly three
to one, but will, over time, come close to one to
one, just sufficiently high to offset the effect of
technological improvements in vessels and gear.
If nobody is willing to scrap vessels at the
initial ratio (except for credits obtained when
vessels sink or burn) the effect will be the same
as an embargo on vessel construction. The
price of obsolete vessels will then be close to
zero, however, so some new construction will
surely take place. Some vessels, which ordinarily
would not have been removed from the fishery,
might be removed if the owner can sell them to
someone needing tonnage to scrap.
This program falls somewhere between the
two previously mentioned, but neither does it
involve government outlays nor does it prevent
technological improvements in the fleet during
the transition period. All these three programs
would necessitate a scrapping ratio when the
fleet is reduced to the desired level.
(4) Recommendations have been made to the
government of Peru to reduce fishing effort by
tying the size of the fleet to the licensed capacity
of the factories. The recommendations called
for a maximum of 1.4 tons of hold capacity per
ton of daily processing capacity.4 Even if a
ratio were imposed on a firm (some firms have
several factories) rather than on a factory so
that vessels can be used where fish are abundant,
there seems to be certain disadvantages with
such a program. While previous programs
mentioned have not differentiated between
factory owned and independently owned vessels,
the question now arises as to how to deal with
the 20% of the fleet which is independently
owned. Secondly, such a program would lessen
competition and freeze the industry in a given
pattern.
4 This ratio is too high, as few firms currently have
a higher ratio.
(5) To reduce the size of the fleet and expand
the fishing season, a quota system can also be
implemented. Catch quotas can be established
for individual vessels, factories, or firms. To
reduce uncertainties about investment, quotas
should be given for a number of years and not
for one season at a time. Further, due to changes
in recruitment and the amount of effort the
resource can bear, quotas should be allotted
as a percentage of the overall annual quota.
A quota system for the purpose of reducing
the number of producing units would most likely
have to be based on an auction system. Such a
system, whether introduced on the vessel,
factory, or company level, would tend to elimin-
ate not only the less efficient producers but
also those which are financially weak. Such a
program would transfer significant funds from
the fishing industry to the public treasury. Due
to the structure of the Peruvian anchoveta
industry, a company quota would seem prefer-
able as this would reduce the size of the fleet
(overhead costs) more than a quota on factories
or vessels. Under the two latter arrangements,
many vessels may be tied up because they have
reached this quota, while others still are fishing
because a factory may be located in an area
where availability of fish is low in a particular
season resulting in excessive steaming time by
the factory fleet. Even company quotas would
result in an excessive fleet, however, as each
company would keep a fleet big enough to be
sure it will catch its quota.
The various management alternatives so far
mentioned have been directed towards reducing
the capacity of the fleet and extension of the
fishing season and thus, reducing the size of
investment in the catching phase. Some of the
alternatives will have little or no impact on the
excess investment and low capacity utilization
of the fishmeal factories, while others will have
a significant impact.
Reduction in Processing Capacity
(1) Reduction of the total licensed capacity of
fishmeal plants will indirectly affect the fleet.
As indicated earlier, the industry as a whole is
in a poor financial position. By lifting the
moratorium on debt collection, many firms
110
would go bankrupt and this would improve the
situation for those remaining. Since most of the
debt is to the public sector, it would mean the
government would have to write off some bad or
uncollectible claims.
(2) Spokesmen for Sociedad National de
Pesqueria (a trade organization for the fishmeal
producers) are extremely concerned about excess
capacity and have indicated a willingness to
bail out the government through a program
where the government buys up the high-cost
plants and assesses the cost on the remainder
of the firms over 2-3 years by a fee per ton of
meal produced. Whatever methods are used
for eliminating the excess capacity, they will be
beneficial for the industry as a whole and reduce
the pressure on the government to increase the
overall catch quota.
(3) In addition to eliminating the insolvent,
high cost, or marginal producers, a further
reduction in the licensed processing capacity
will be needed. Capacity should be reduced to
a level just sufficient to process the catch over
an extended fishing season. Currently the
licensed capacity of a plant is for tons of fish per
hour, and only rarely does a factory produce at
full capacity. To encourage fleet reduction, the
license should be issued to companies rather
than on a factory basis and as previously
mentioned, should be a percentage of the overall
catch quota. A quota might be either on input
of fish or output of meal. The latter is easier
to control since the meal is exported through a
government monopoly. A quota on input, how-
ever, would give a strong incentive to increase
the yield (output per ton of fish) and improve-
ment in this respect is badly needed. Quotas
could be based on the company's current share
of the market, or be put up for auction.
Quotas or licenses to operate might be trans-
ferable or nontransferable. A transferable quota
could put large and small companies (one-plant
operators and multiplant operators) on a more
equal competitive basis. The author can see
little advantage in a nontransferable quota
except for the fact that it might prevent con-
solidation of the industry into too few hands.
CONCLUSIONS
To manage the anchoveta industry solely
through regulation of the processing phase
would very likely put the independent vessel
owners at a serious disadvantage. To prevent
this, a management program for the Peruvian
anchoveta industry should include both regula-
tions at the catching and the processing level.
Of the various alternatives available for manage-
ment of the Peruvian anchoveta industry, the
author would be in favor of relying on a fairly
high scrap and rebuild ratio to reduce the fleet.
Lifting of the moratorium on debt collection,
combined with transferable licenses for factories,
so market forces could be effective, might be
sufficient to reduce processing capacity to the
desired level.
LITERATURE CITED
BOEREMA, L. K., and A. HOLMSEN. 1970. Some
Economic Aspects of Management of Fleet Size in the
Peruvian Anchoveta Industry. Unpublished manuscript.
HOLMSEN, A. 1970a. Cost Structure in the Peruvian
Anchoveta Industry. Unpublished manuscript.
HOLMSEN, A. 1970b. Factors Affecting the Potential
Productivity of the Peruvian Anchoveta Fleet. Un-
published manuscript.
Ill
A Stochastic Investment Model
for a Survival Conscious Fishing Firm
Russell G. Thompson, Richard W. Callen,
and Lawrence C. Wolken1
ABSTRACT
In this study, the stochastic investment model for a survival conscious firm developed
by Thompson and George (1970) is extended to take into account income taxes and
depreciation of the capacity. This model is applied to shrimp fishing on the Texas Gulf
coast. Values of the parameters, as in the deterministic application by Thompson et al.
(1970), were based on proprietory information, current market conditions, and present
institutional restrictions. The effect of growth in real per capita income on shrimp
prices is estimated, and two different rates of income growth are analyzed. Solutions
to six problems based on two different sets of random sequences are computed and
discussed. The results indicate the effect of the survival constraint on investment
decisions, and the importance of revealed information in decisionmaking.
INTRODUCTION
In 1970, Thompson and George formulated a
stochastic dynamic investment model for the
survival conscious firm, derived the optimal
decision rules for investment, and computed
solutions to several problems. This model takes
into account the probability distribution of the
yield (output per unit of capacity) and output
price, as well as all of the information known to
the decisionmaker at the time of each investment
decision. The entrepreneur is initially assumed
to be in a financial position where a feasible
investment solution always exists if the lowest
output price and yield occur in every period of
the planning horizon. In the model, the objective
of the firm is to maximize expected net worth at
the end of the planning horizon. All production
expenses, investment outlays, interest costs,
1 Russell G. Thompson is Professor of Quantitative
Management Science, University of Houston; Richard
W. Callen and Lawrence C. Wolken are Lecturers in
Quantitative Management Science, University of Hous-
ton. This work was partially supported by the National
Science Foundation GH 59 as a part of the Sea Grant
Program for 1970.
and planned cash withdrawals must be paid
for as incurred (or scheduled).
In this study, the Thompson-George model is
extended to take into account income taxes and
depreciation. This requires the introduction of
another state variable to account for the value
of the firm's capital — the investment in
capacity. Straightforward extensions of the
fundamental constructs (developed by Thompson
and George) were required, and are available
from the authors upon request.
Because of the vagaries in fish prices and
catches, this model would be expected to be a
particularly appropriate decision aid for invest-
ments in fishing capacity. There are generally
few, if any, alternative uses for specialized
fishing equipment. Also, fishermen typically
have poor alternative opportunities by which to
earn a living. Low prices and small catches
would be expected, as a result, to be dreaded
much more than high prices and large catches
are desired. A sequence of worse than expected
net revenues (even in the case of a very favorable
expectation) could terminate the existence of
the fishing firm. This could well be an unaccept-
able risk of failure. Hence, survival of the
fishing firm would be expected to be a funda-
mental factor influencing the firm's investment
decisions.
112
DEVELOPMENT OF THE
SURVIVAL MODEL
In the survival model, the decisionmaker
evaluates the worst sequence of net revenues
that could occur in every year of the decision-
making period. This sequence, in conjunction
with the value of the initial investment in
capacity and the value of the money account,
determine the survivable set of fishing capacity
purchases at the beginning of the first year.
The decisionmaker selects from this set the
investment that contributes the most to his
terminal net worth. After the first year and
before the second operating year begins, the
output price received and the yield obtained in
the first year have been observed. This is now
a part of the information known to the decision-
maker for planning in the second year. The
decisionmaker again evaluates the worst se-
quence of yields and prices that could occur in
every remaining year of the decisionmaking
period. This abbreviated sequence is now
evaluated in conjunction with the capacity and
money position at the end of the first year. It
determines the survivable set of capacity pur-
chases for the second year. Again, as in the
first year, the decisionmaker selects from this
second set the investment that contributes the
most to his terminal net worth. This procedure
is repeated in every year throughout the
decisionmaking period. Investment decisions
are conditioned by experience, and are not based
solely on expected values.
By definition, the firm survives in a given year
if the value of the capacity exceeds the value of
the indebtedness. A survivable investment is
defined in the following way: the decisionmaker
has completed operations in year k-1 and is now
planning for year k. He wants to survive above
all else during the remaining N — (k-1) years
of the decision period, even if all future yields
and prices are the lowest possible. An invest-
ment decision in the fcth year, s^, is said to be
survivable if the value of the capacity in every
remaining year is never less than the indebted-
ness owed (with capacity not being purchased
in any of the years after the fcth 0ne and the
lowest net revenues being visualized in every
year of the yet undisclosed future).
Under these conditions, a survivable ca-
pacity purchase in year k is found to be
equivalent to the following one: the product of
the capacity units purchased in year k and the
marginal value of capacity calculated under the
assumption of the lowest net revenue occurring
in every forthcoming year — the marginal cost
of capacity visualizing the worst — is never
greater than the value of the money account in
year k — 1 plus the terminal value of the ca-
pacity in all of the remaining years (with the
lowest prices and smallest catches occurring)
minus any fixed cash withdrawals in the rest
of the planning period. (All money flows are
adjusted for the values of alternative oppor-
tunities, income taxes, and depreciation.) This
upperbound would be the value of the firm's
assets if the worst possible sequence of net
revenues occurred — the decisionmaker's final
asset position visualizing the worst.
To reflect the fear of low net revenues, revenue
per unit of capacity when the lowest price and
yield occurs is assumed to be less than the
operating cost per unit of capacity. It is also
assumed that per unit prices of capacity are
not increasing so rapidly that operating losses
per unit may be covered by value appreciation
in capacity. (Speculation is never a sure bet.)
This implies that the marginal cost of capacity
visualizing the worst is positive. Hence, dividing
the lower bound for the firm's final asset position
by this positive marginal cost, the upper bound
for a survivable purchase of capacity in a given
year is obtained. This represents the maximum
amount of capacity that the decisionmaker can
purchase and still insure survival of the firm
throughout the rest of the decision period. It
depends upon the value of the firm's money
account, the amount of capacity owned, and
the value of that capacity in the previous year.
This upper bound function in year k is denoted
by Hk(zk - 1 , Vk-it %k-i), where at the end
of the k — lst year zk _ / is the cash balance,
yk _ i is the units of capacity owned, and xk _ y
is the purchase value of the firm's capacity. The
firm is in debt if zk / is negative and has
savings if zk_1 is positive.
We will also introduce the following notation
now; St is the units of capacity purchased at
the beginning of the ith year (and used for the
first time in year i); r, is the operating costs per
unit of capacity in year i\ a, is the per unit pur-
chase price of capacity before the beginning of
the operating season in year i; A, is the cash with-
113
drawal in year i for sundry expenses; 7 is the
interest rate paid (or received) on the cash ac-
count z; co, is the unknown revenue per unit of
capacity in the ith year; N is the number of years
in the planning period; |3 is the fraction of the
value of the capacity recoverable at the end of
the planning period; 5 is the income tax rate;
and e is the straightline depreciation fraction.
Also E will be used to denote the mean of the
random variable go,-; and L will be used to denote
the smallest possible annual net revenue having
a positive probability of occuring. The symbol
a, is used to denote the output price where only
the yield is a random variable in the application
below.
Using the above development, the survival
model may be stated as follows:
Maximize E(zN + j3a/v + iy/v) over all n-tuples
of functions s,(coi , C02 , . . ., to,_i ), i = 1, 2, . . .,
N, satisfying the difference equations
(1) Xf—Xi-i =OiSj,x0 =o0yo,
where
y( — y,_, = S{, yo given and non-negative,
(2) zt — zt-\ = yzt-\ + y,- (co — 77) — 07 s/
A- - 5 |y, (co~ t{) + yzi-i - a,.
— exi], e = 0.091,
where z0 given, and i = 1,2,.
ing the inequalities
., N, and satisfy -
O^si^Hi (z/_i , y,_, , x/_i ), ; = 1,2,.. ., N.
In words, the decisionmaker desires to maxi-
mize expected net worth at the end of the
decision period where the purchases of capacity
are selected from the survivable set in each
year (delineated by the inequality restrictions).
Thus, in the maximization process, the decision-
maker, who takes into account all of the informa-
tion known at the time of the decision, selects
the investment from the survivable set of capa-
city purchases that maximizes expected net
worth at the end of the planning horizon.
THE DECISION RULE FOR INVESTMENT
By the use of dynamic programming
methods, the method developed by Thompson
and George was extended, as mentioned above,
to allow for depreciation and income taxes. The
extended rule for optimal investments is sum-
marized in the following theorem.
Theorem: Suppose H|(z0, yo, x0 )>0, i.e. the
upper bound for investments in the first year
is non-negative. Let Rk be the expected mar-
ginal value of capacity for survival investment
decisions— the marginal value of capacity vis-
ualizing the worst. Then the decision rule for
optimal survivable investment is as follows:
( 3 ) sk° = Hk (zk°. 1 , yfc°_, , xfc°_, ) if R , >0,
and sic ~ o if Rk<0
with the feasible value of Sk being immaterial
if Rk =0.
In other words, the decisionmaker buys the
survivable limit of capacity in year k if the
marginal value of capacity visualizing the worst
is positive in that year, and he makes no capacity
purchases if this marginal value is negative. It
also follows that the optimal purchase is im-
material in any year (because of the linearity
of the problem) whenever the decision rule is
zero. The upper bound for investments in the
first year insures the existence of a feasible
investment solution in each year of the planning
horizon.
An Application to Shrimp Fishing
To indicate how the model may be applied to
a shrimp fishing firm, parameters were specified
for a relatively small fishing firm operating
73-foot steel hull trawlers (see Table 1). In the
specifications, the values of the parameters were
specified to reflect prices, costs, and landings per
vessel as reported by the firms cooperating in
the study. There is an exception with regard to
Problem 3. Average landings per vessel which
were found to be 57,560 pounds of heads-off
shrimp per year in the years 1958 through 1969
were specified to be one standard deviation
above the mean to evaluate the effect of better
than average management. That is, in Problem
114
Table 1, — Values of the parameters for four survival problems: the Gulf shrimp fishery.
Parameters Problems
N -- number of years in planning period
Z ■- initial cash balance in dollars
o
y - initial number of boats in fleet
x - initial investment in dollars
o
J - annual interest rate per dollar
T - annual production cost per vessel in dollars
a - per vessel purchase price in dollars
e - annual depreciation fraction per dollar invested
t - annual income tax rate per dollar of taxable
income
p - recoverable fraction of investment in fishing
capacity
1
5
5
5
5
0
96,145
0
0
1
(t
1
1
100,000
(l
100,000
100,000
0.085
0.085
0.085
0.085
30,000 x
30,000 x
30,000 x
30,000 x
(1.03)r
(1.03)*
(1.03)'
(1.03)f
100,000 x
100,000 x
100,000 x
100,000 x
(1.03)*
(1.03)'
(1.03/
(1.03)'
.091
.25
.65
.091
.25
.65
.091
.25
.65
.091
.25
.65
annual cash withdrawal for sundry expenses in
3,600 x
3,600 x
3,600 x
3,600 x
dollars
(1.03)'
(1.03/
(1.03)r
(1.03/
owner's expected annual revenue per vessel in
49,790 x
49,790 x
54,400 x
49,790 x
dollars
p/1.03)'
p/1.03/
p/1.03)*
p/1.03)'
owner's lowest annual revenue per vessel in
dollars
22,500 x
22,500 x
22,500 x
22,500 x
(1.03)f
(1.03/
(1.03/
(1.03)'
3 landings per vessel were 63,291 pounds of
heads-off shrimp per year.
Since the real price of shrimp — the price
adjusted for the purchasing power of money —
is highly influenced by growth in real per
capita income, and since it appears that the
economy may be entering a period of modest
growth (possibly much like the late 1950's),
the real price of shrimp was specified to reflect
a 1.5% rate of growth in real per capita income
in Problems 1, 2, and 3, and to reflect a 3.3%
rate of growth (as observed in the mid 1960's)
in Problem 4.
To evaluate the economic attractiveness of
shrimp fishing versus the best alternative to
fishing (as reflected by the interest rate on
money), the decisionmaker in Problem 2 initially
has the approximate money equivalent of an
investment in one vessel. Recall that the en-
trepreneur is a profit maximizer, given that he
can survive. Thus, the decisionmaker would
opt for the savings alternative whenever the net
rate of return from a dollar invested in fishing
capacity is less than the interest rate on money.
That is, the second problem indicates the eco-
nomic advantage (or disadvantage) of investing
in fishing relative to loaning the money to
someone else.
Since the model takes into account the infor-
mation obtained through time as the values of
115
Table 2. — Solutions to four survival problems in table 1; landings per vessel are random.
Marginal value
Investment
Debt to
of another vessel
in boats
Boats owned
Cash balance
gross asset
Problem
Year
(dollars)
(number)
(number)
(dollars)
ratio
0
_
_
1.00
0
1
5,843
1.44
2.44
-146,356
.57
2
- 784
0
2.44
-127,678
.48
1
3
- 7,896
0
2.44
-116,862
.43
4
-15,474
0
2.44
-108,022
.38
5
-23,490
0
2.44
- 74,436
.26
0
-
-
2.44
96,145
_
1
5,843
2.44
2.44
-145,083
.57
2
- 784
0
2.44
-126,507
.48
2
3
- 7,896
0
2.44
-115,728
.43
4
-15,474
0
2.44
-106,908
.38
5
-23,490
0
2.44
- 73,534
.26
0
—
_
1.00
0
_
1
21,419
1.44
2.44
-136,487
.53
2
16,198
1.13
3.57
-216,534
.56
3
3
7,080
4.03
7.59
-581,958
.68
4
- 5,562
0
7.59
-511,662
.58
5
-18,570
0
7.59
-358,977
.40
u
_
_
1.00
0
_
1
10,655
1.44
2.44
-145,128
.56
2
9,943
.80
3.23
-240,502
.58
4
3
2,624
3.15
6.38
-503,596
.70
4
- 7,595
0
6.38
-462,898
.63
5
-19,119
0
6.38
-341,999
.45
the random variables are revealed, solutions to
two sets of problems were computed. In the first
set, the landing per vessel is random; whereas
in the second set, the price received is random
as well. The first set of results is presented in
Table 2, and the second set in Table 3.
It is important to note that this application
of the survival model is not exhaustive of the
many that could be made, or to imply that the
normative results presented are likely to occur.
This work is only meant to indicate how an
investor interested in shrimp fishing, who has
a limited amount of money capital, might
obtain bench marks (from the model) for in-
vestment planning.
Values of the Parameters
In this application, the firm's initial fishing
capacity was specified to be one vessel in Prob-
lems 1, 3, and 4. The values of the data (excluding
the basis for the expected shrimp price in the
first set of problems) are given in Table 1. The
initial purchase price of one vessel was taken
to be $100,000. In Problems 1, 3, and 4, the
firm is visualized as having an initial debt-free
investment of $100,000 with no savings. This
relatively large amount of initial equity was
necessary for the survival problem to have a
feasible solution. The minimum value for the
firm's initial equity in Problem 1 was found
to be $97,000.
In Problem 2, where the entrepreneur has
his equity in savings rather than invested in
fishing capacity, the initial value for savings is
$96,145. This is approximately equivalent to
owning one vessel initially because of the pro-
cedure used to calculate interest earnings and
tax allowances in the model.
There is only one money account in the model,
and accordingly one interest rate. This rate was
specified to be 8V2% per year.
116
Table 3. — Solutions to two survival problems in table 1 ; shrimp prices and landings per vessel are random.
Marginal value
Investment
Debt to
of another vessel
in boats
Boats owned
Cash balance
gross asset
Problem
Year
(dollars)
(number)
(number)
(dollars)
ratio
0
1.00
0
0
1
5,843
1.44
2.44
-156,026
.60
2
- 784
0
2.44
-126,538
.48
1
3
- 7,896
(l
2.44
-118,206
.43
4
-15,474
0
2.44
-118,517
.42
5
-23,490
0
2.44
-100,311
U
0
_
_
1.00
0
—
1
21,419
1.44
2.44
-147,856
.57
2
16,198
.69
3.13
-176,191
.52
3
3
7,080
4.22
7.34
-569,170
69
4
- 5,562
0
7.34
-533,307
.63
5
-18,570
0
7.34
-438,263
.50
To reflect inflation, the purchase price of new
vessels was specified to increase at 3% per year.
This rate is 2% below reported price trends,
which include costs of technological improve-
ments. Newer vessels have been powered by
larger engines. This has allowed for larger
trawls to be towed at faster rates. This rate of
improvement in technology is believed to have
increased investment costs by 2% per year.
From the cost records of the cooperating
firms, the annual cost of operating a 73-foot
trawler was found to be $30,000 in 1969. This
cost figure includes an allowance for overhead
and insurance. Representatives of the firms
interviewed indicated these costs have increased
3% per year in recent years. Thus, the annual
production cost per vessel, rf, was specified to
be 30,000 (1.03)f.
Straight line depreciation methods were used
for tax purposes with an 11-year depreciation
period being used for a fully outfitted vessel.
This average was estimated on a value weighted
basis from the records of a number of firms.
The reciprocal of this figure, 0.091, was the
value used in the depreciation function.
Income is the sum of the revenues received
(by the owner after the "lay") less operating
costs, interest costs (or plus interest earnings),
and taxes. The income tax rate, which is denoted
by £, was taken to be 25% of the taxable income.
This rate was paid in the late 60's by a number
of the firms studied.
In shrimp fishing, the captain and first mate
on a vessel are commonly paid on a "lay" basis
wherein they receive for services rendered a
percentage of the revenue earned by the vessel.
The header, who is the third crew member, is
typically paid on a per box basis; his wages are
included in the production cost per vessel. For
73-foot vessels, the "lay" for the captain and
first mate is commonly 35% (with the owner
getting in effect 65% of the ex-vessel price) ; they
typically pay for all of the groceries.
In interviewing the cooperating firms, the
relative resale value of the vessels sold was
found to be fairly well approximated (for vessels
5 to 6 years old) by summing the accumulated
depreciation fractions with an appropriate ad-
justment for technological improvement. This
procedure, which implies that the resale value
of a vessel 5 years old would be 65% of the
purchase price, was used as the basis for specify-
ing (3 to be equal to 0.652.
To project per vessel expected revenue re-
ceived by the owner, the log of the real shrimp
price received by the cooperating firms, Pt, was
regressed on the log of the index of real per
capita income (in the United States after taxes),
yt, and the log of per unit effort landings, h,
caught in depths beyond 10 fathoms off the
Texas coast. (See the earlier study by Thompson
et al. (1970, p. 12) for data.) The estimated
regression equation was as follows:
(4) In pt = -4.571 + 1.175 In yt -
(t = 3.6)
R2 = .748, oe = .0888.
-.379 In /,,
(*=3.5)
2 This approximating procedure was necessary, since
the vintage was not accounted for in the model.
117
Variations in landings per unit effort, which
were found to be highly correlated for the Texas
and Gulf-South Atlantic fisheries, are still
regarded by biologists as being largely random.
Thus, to remove the effect of landings on price,
landings were specified to be equal to the mean
value observed for the Texas fishery in the
period 1958 through 1967. Hence, the price
estimating equation with an adjustment to a
1969 base year was as shown below.
(5) lnpf =-1.332 + 1.175 In yt
To use this equation, the index of real per
capita income had to be projected for the years
1970 through 1974. This was done by regressing
In yt on time, t, for the years 1953 through
1960, and also for the years 1961 through 1968.
The following two income projection equations
were developed for the period t = 1970, 1971,
. . ., 1974.
Specification I: 1.5% rate of income growth
(6) lnyr = 4.94 + .015*
Specification II: 3.3% rate of income growth
(7) lnyr =4.94 + .033?
By substituting the desired specification
from (6) into (5), the price projection equation
was obtained. The effective expected real shrimp
price, ar, was 0.65 of the antilog of pt. To convert
to money terms, the projected prices were
multiplied by the consumer price index value
for 1969, 1.277, and by a price inflating factor
of 3% per year thereafter. In Table 1, fit denotes
the price reflecting the high rate of income
growth and pt the low rate.
For the first set of four problems, the estimate
of the owner's lowest annual revenue per
vessel, Lr, was found by taking the lay residual
of the product of the 1969 shrimp price, a69,
and the projected lower bound for landings per
vessel. This lower bound was taken to be 3.4
standard deviations (in t units for 11 degrees
of freedom) below the mean landing per vessel
of 57,560 pounds with the sample standard
deviation being 5,731 pounds. Thus, the prob-
ability of the landings per vessel being greater
than this lower bound (assuming this to be a
valid probability basis) is greater than 0.99.
Moreover, since the growth rate in real per
capita income is not taken into account in Lt,
the probability of revenue per vessel falling
below the implied estimate of the owner's lowest
annual revenue per vessel (where the price is
projected under either specification) decreases
steadily as the planning period unfolds. In
other words, the estimate of L, is very conserva-
tive for the year 1970 and becomes increasingly
conservative thereafter in the planning period.3
For the second set of two problems in which
the shrimp price is random as well as the
landing per vessel, the same value was used for
the owner's lowest annual revenue per vessel.
This resulted in a slightly smaller probability
of survival than in the first four problems
(because of the additional randomness in the
price), but one still greater than 0.99. Thus, in
the interest of simplicity, the same value of
Lt was used in both sets of problems.
Knowledgeable industry representatives (who
were consulted with regard to the above specifi-
cations) indicated a 5-year survival period
would be especially meaningful for firms operat-
ing the 73-foot trawlers. Accordingly, two 5-
year sequences of random revenues per vessel
were developed with only the landing per vessel
being random in the first sequence. Landings
per vessel were regarded as independent of
price, since the fishery is relatively competitive;
moreover, for the period studied, per vessel
landings for the cooperating firms were not
highly correlated with landings per unit of
effort in the Texas fishery4 (r2 = 0.16). Using
the regression estimate for price in each year
1970 through 1974 and the estimated standard
error of the regression, and also using the sample
mean and standard deviation for landings per
vessel of the cooperating firms, the random
prices and landings per vessels were calculated
as follows: (1) By use of the Box-Muller (1958)
method, normal random deviates for prices and
landings per vessel were independently gener-
ated ; and (2) the products of these two random
variables were adjusted for the lay and expected
changes in the purchasing power of money. The
following random sequences were accordingly
obtained and used in the analysis.
3 To have a probability support at Lf, this small
probability of non-survival is implicitly assumed to be
insurable.
4 Landings per unit effort in the Texas Fishery are
highly correlated with those for the Gulf and South
Atlantic.
118
Random Sequences of Revenues per Vessel
Seq
xence No. 1
Problems 1 & 2
Problem 3
Problem U
$30,741
$36,141
$31,413
42,572
48,233
44,457
39,859
45,795
42,531
39,797
46,020
43,393
50,784
57,308
56,583
Sequence No. 2
Problem 1 Problem 3
$25,450
47,261
38,810
36,077
44,747
$29,920
53,546
44,589
41,719
50,495
It may be helpful to recall that the decision-
maker is regarded as being a better than average
manager in Problem 3. The 1.5% rate of real
economic growth per capita is used in Problems
1, 2, and 3; and the 3.3% rate of economic
growth is used in Problem 4.
In evaluating the solutions to the first set of
four problems in Table 2, the results indicate
the profitability of investing in shrimp fishing
capacity during the 5-year planning period.
The model fisherman opted for investing in
fishing capacity in Problem 2, even though he
could have left his money in savings at 8.5%
interest. Thus, the rate of return over cost from
shrimp fishing was greater than 8.5% . In further
analysis, it was found to continue to be so until
the rate of interest reached 9.5% ; then the rate
of return over cost switched in favor of savings.
The value of better than average management
is indicated by the results in Problem 3. There,
the average landing per vessel was taken to be
one standard deviation (5,731 pounds) greater
than in Problem 1. The same amount was
invested in the first year; but in the second and
third years there were striking differences. The
model fisherman bought 5.2 vessels in Problem
3, while he did not buy any in Problem 1. He
chose to pay off debt in the first problem after
the initial investment, since that represented
a more profitable use of the money. It may be
noticed that the investment upper bound limited
the size of the purchases in the first 3 years of
Problem 3 (and the first year of Problem 1).
The marginal value of another vessel was
positive; however, the money was not available
for investment given the desire to survive.
Success in shrimp fishing is clearly influenced
by the rate of income growth in the economy —
compare Problems 1 and 4. In Problem 4, the
marginal value of another vessel is almost
twice as large in the first year as in Problem 1,
and remains large in the second year when the
value in the first problem goes negative. This
increased growth in per capita income results
in an increased ability to invest in the second
year in Problem 4 and still further increased
ability, at a lower marginal incentive, in the
third year. The model fisherman carries a con-
siderably larger debt load, as a result of the
increased profitability, in Problem 4 than in
Problem 1.
In evaluating the second set of results given
in Table 3 and comparing these solutions to
the ones in Table 2, only slight differences
between the results may be noticed. Somewhat
less is invested over the planning period in
Problem 3 in the second case than in the first.
Also, a slightly larger debt load was generally
carried in most of the planning period. Of
course, the marginal investment incentives
were the same in both sets of problems; they
are based on expected values. Vagaries in land-
ings seem to be much more important than
unexpected variations in price.
LITERATURE CITED
THOMPSON, RUSSELL G., and MELVIN D. GEORGE.
1970. A Stochastic Investment Model for Survival
Conscious Firm. Presented at Winter Meetings of the
Econometric Society, Detroit, December, 1970.
THOMPSON, RUSSELL G., RICHARD W. CALLEN,
and LAWRENCE C. WOLKEN. 1970. Optimal Invest-
ment and Borrowing Decisions for a Model Shrimp
Fishing Firm, Texas A&M University, Sea Grant
Bulletin No. 205, April.
BOX, G. E. P., and MERVIN E. MULLER. 1958. A
Note on the Generation of Random Normal Deviates,
The Annals of Mathematical Statistics, Vol. 29, June,
pp. 610-611.
119
APPENDIX
Appendix Table 1. — Values of projected index of real Appendix Table 2. — Values of projected real shrimp
per capita income. prices.
Year /Specification 1 Specification II Specification I, p Specification II. p
Year (cents per pound) (cents per pound)
1 136.98 139.52
2 139.06 144.27 1 85.68 87.56
3 141.17 149.19 2 87.22 91.07
4 14332 154^7 3 88-78 94-73
5 145.50 159.53
4 90.37 98.53
5 91.99 102.49
Appendix Table 3. — Values of landings per vessel for
random sequences 1 and 2.
Problems 1 , 2 & 4 Problem 3
Year (pounds) (pounds)
1 41,965 49,336
2 55,435 62,806
3 49,501 56,872
4 47,140 54,511
5 57,375 64,746
120
Simulation Experiments to Evaluate Alternative
Hunting Strategies for a Deer Population1
F. M. Anderson,2 G. E. Connolly3
A. N. Halter,2 and W. M. Longhurst3
ABSTRACT
A population dynamics model of the deer herd in Mendocino County, California,
is presented. Environmental influences are modeled as density dependent birth and
death rate functions. The computer program for this biomanagement model is outlined
and validity checks devised to improve the model are discussed. The output shows
the impact of selected hunting strategies on productivity, natural mortality, and
other population characteristics. Tests of hunting strategies related to alternative
management goals are summarized. Implications of computer simulation methodology
for the management of wildlife and fish populations are discussed.
INTRODUCTION
Management of a natural resource, such as
a deer herd or fishery, is the manipulation of
that resource and/or its environment in an
attempt to satisfy a set of objectives. The
Objectives can be economic or noneconomic.
They may or may not be quantifiable, and
hence, the management problem may or may
not be solvable in the framework of "extremum"
problems.
The management of a deer herd, like that
of a fishery, can be directed toward multiple
objectives. The deer herd may be maintained
at a particular level and age composition to
achieve a hunting kill having the greatest
value; alternatively, the herd may be main-
tained for purely aesthetic reasons. A multiple
objective of management may be to sustain
a certain deer density (deer per square mile)
at one time of the year to provide hunting,
or at another time of the year to provide
scenery for sightseers.
Under certain environmental conditions,
managers may be prevented from knowing
whether or not the objective(s) has (have)
been attained. In areas of dense ground cover,
1 Technical Paper Number 2998, Oregon Agricul-
tural Experiment Station, Corvallis, Oregon.
2 Department of Agricultural Economics, Oregon
State University, Corvallis, Oregon.
3 Department of Agricultural Zoology, University
of California, Davis, California.
managers must often resort to crude sampling
techniques to derive population estimates.
Other parameters can be readily measured.
For example, in a deer herd where hunting
is done only by license, the kill figures are
available soon after the hunting season, and
can be used in the formulation of subsequent
management strategies. It may be that certain
objectives will be satisfied if crucial parameter
values are between certain upper and lower
bounds. Alternatively, the objective of man-
agement may be to maximize the value of a
parameter. Examples of these two cases are
(1) to keep the average size of the herd be-
tween two values, and (2) to maximize the
annual hunter kill, respectively. Other paral-
lels to the objectives of management for a
deer herd can be found in the management
of a fishery resource.
Both deer and fish populations are members
of complex, dynamic ecosystems. For each,
the age composition changes over time due to
the changes of such parameters as birth rates
and death rates. In addition to relatively
simple variability about these parameters,
changes in the population are compounded by
environmental changes.
To illustrate, assume there is a functional
relationship between deer density and the
mortality rate of each age category. Further-
more, assume a fixed habitat structure and
that variability in the biosystem is introduced
only by changes in the weather. The effect of
these changes will usually be lagged. Other
121
relationships can be hypothesized to complete
the abstract model. For each time period, the
mortality rate in each age category depends
upon the density. Over time this density will
change, as will the inventory of deer in each
category. Hence, mortality rates will differ
over time, even if, the same functional re-
lationships are hypothesized.
Now, add in the complicating factor of
changes in some or all of these mortality func-
tions consistent with an improved habitat and
a higher plane of nutrition. In the real world,
changes in the deer habitat — and its counter-
parts in other fish and wildlife species — are
occurring continuously.
In making management decisions, some
knowledge is assumed of the structure of the
relevant biosystem. However, knowledge is,
at best, uncertain, and heroic assumptions
are aften made about the effect of a structural
change. Thus, decisions may be made which
move the biosystem toward the objectives
desired in an unpredictable manner. Manage-
ment is usually carried out within the
boundaries described by legally authorized
regulations, which are, hopefully, both con-
sistent with a set of objectives and flexible
enough to afford the on-the-spot manager some
discretionary action. When regulations are
for more than one distinct resource unit this
flexibility is desirable because each unit is
unique.
For example, regulations for deer hunting
in a particular state usually embrace more
than one herd. No two herds will be identical
at any point in time, and the regulations must
be sufficiently flexible to allow for these dif-
ferences. Regulations are ideally formulated
with regard to the structure of the relevant
biosystems, but knowledge of these biosystems
is not complete. The response of the biosystems
to particular management actions cannot be
predicted with certainty. Therefore, there is
a limit to the rigidity of the regulations. Be-
yond this limit, management will be ineffec-
tive in attempts to satisfy the set of objectives.
Thus far, we have briefly described three com-
ponents of the management system of a public
resource. These are the complex biosystem,
the set of objectives, and the set of regulations
relating to the particular resource. One more
component is necessary to complete a work-
able management system; that is, a means of
monitoring the system is required. For any
biosystem, the selection of the parameters to
be monitored is the result of experience and
expertise. However, to be useful to manage-
ment, the selected parameters must be indica-
tive of the performance of the biosystem so that
it can be determined whether, or to what
extent, objectives are being accomplished.
Typically, only relatively few parameters
can be monitored accurately and rapidly enough
to be useful. Information on the state of the
system is of most value when it is current.
The role of time in monitoring systems cannot
be overemphasized. Information on the state
of a biosystem at any time is usually incomplete.
For example, the total number of deer in a
herd is a useful parameter in developing man-
agement strategies. In most herds it is im-
possible to take an accurate annual census,
and estimates of the total population must
be based on samples, which often may be col-
lected only at certain times of the year.
Historically, researchers and managers have
been restricted to experimentation on the real
biosystem. However, with the advent of com-
puters and programming languages, it is now
feasible to perform simulated experiments on
biosystems that can be described by mathe-
matical equations. This paper is concerned
with the computer simulation of the deer
population in Mendocino County, California.
The model shows the population dynamics
and some of the economic and recreational
consequences associated with various hunting
strategies.
COMPUTER SIMULATION METHODOLOGY
Simulation involves building and operating
a model designed to represent those features
of the real system under study and to provide
information about the performance of the
system under assumed controlled conditions.
Three classes of simulation models can be
distinguished: (1) physical models, such as
scale models of river systems and planetar-
iums, (2) mathematical models where a set of
equations describing the system under study
is written and these equations are solved, per-
haps analytically, and (3) computer simula-
tion where the system is described and the
122
logic is programmed for computer calculation.
In the latter case, the intent is to simulate
complex systems which usually involve non-
linear relationships, random components, and
time varying events.
Computer simulations are applicable to prob-
lems of the type where management can influ-
ence the system's behavior. The purpose of
simulating a management system is to test
the impact on variables of interest within par-
ticular management policies, before such
policies are implemented, and influence the
real system. Here, the simulation performs the
important function of providing information
about the possible consequences over time of
various alternative management policies. Thus,
it provides answers to the managers' questions
which are of an if-then type. The computer
program is an if-then calculator. Systems
could be simulated using paper and pencil, but
computers can carry out these calculations
more efficiently.
Simulation should be viewed as an iterative
problem-solving technique which involves four
stages: (1) problem definition, (2) mathe-
matical modeling, (3) refinement and testing
of the resulting model, and (4) creative design
and execution of simulation experiments to
provide information relevant to the manage-
ment problem. In Figure 1, arrows indicate
that the general sequence is from problem
definition to application, but the reverse arrows
indicate that the process is iterative, or learn-
ing in nature. A prior stage might have to
be repeated on the basis of information acquir-
ed during a subsequent stage of the modeling
process.
Problem definition is fundamental to build-
ing a simulation model. This study's inter-
disciplinary team, composed of biologists and
agricultural systems analysts, initially met
to determine the types of questions the model
was to answer. The questions fell into three
categories:
1. Biological questions involving the dynam-
ics of the deer population.
2. Economic questions involving the value
or worth of certain events and occurrences
within the system.
3. Management questions which affect the
biological system and have economic and
social consequences.
Problem
Definition
>
Mathematical
Modeling
& Simulation
IP
Model
Refinement
& Testing
1
"
Model
Application
1
¥
Output
Figure 1. — Computer simulation as an iterative
problem-solving process.
In its present form, the model construction
cuts across all three types of questions, and
should be viewed as the first generation model
of a sequence of models which, hopefully, will
be able to answer these questions at more
sophisticated levels. This first generation model
is essentially a population simulator capable
of answering questions mainly of a biological
nature, but provides output for management
questions — in particular, hunting strategies.
Other sections of the output could easily be
given economic interpretation. The second
generation model will include economic vari-
ables such as losses due to deer damage to
agricultural and forest lands, and gains, such
as hunter expenditure and the value of venison.
The proposed third generation model will in-
clude a management component which would
be capable of evaluating management strate-
123
gies in the broader context of their biological,
economic, and social consequences.
DEER HERD SIMULATION MODEL
A comprehensive flow chart of the compo-
nents and interrelationships of a deer herd was
developed. The available data did not permit
all relationships to be quantified and proxy
variables were devised to overcome this diffi-
culty. For other relationships a complete speci-
fication of the biological interactions would
have been possible, but this would have result-
ed in a model of substantial complexity. Model
building is a continual compromise between
abstraction and complexity. Models which
are too abstract are devoid of interest, and the
results will not be easily related to the oper-
ations of the real system. When the models
are large, and incorporate complex mathe-
matical formulations, it can be difficult to
extract meaningful guidelines for management.
Such models may be expensive to run, and
thereby not achieve one objective of the model-
ing process, namely, to simulate the systems
and generate information and knowledge about
the systems at a cost less than alternative
analytical techniques.
The flows and relationships identified for
the Mendocino County deer herd are sum-
marized in Figure 2. In this figure the time
series of events is not obvious. These are dis-
cussed in more detail later in the paper. The
model as depicted in Figure 2 is best viewed
as a summary of the most pertinent inter-
actions which occur each year in the deer
biomanagement system. The basic components
of the system are the birth and death process.
Each year fawns are born into the herd, and
the number of fawns born is a function of the
exponential average of the density in particular
Legend
— Weather
Predators
and
Disease
«r
==dfc
Accidents and
unclaimed
hunter kill
Competition
from wildlife
and domestic
animals
Functional
Relationships
Causal Relationships
Information Flows
Real Flows (Deer)
Other factors,
geophysical,
destruct
- -1
Regulation
and
Management
-»
Feed
Conditions
Figure 2. — Biomanagement system of a deer population.
124
months prior to the time of birth.4 Thus, the
exponential average density is a proxy vari-
able which summarizes all relevant causal
influences of the real system. The casual in-
fluences are indicated in Figure 2, but are not
explicitly programmed into the computer.
In the model, losses are defined as either
natural or due to hunting. Natural losses are
the residual of losses after accounting for the
recorded hunter kill. The natural losses will
include those due to age, the plane of nutrition,
the action of predators, disease, and accidents
on the highways. Both natural and hunting
losses are computed each time period. Natural
losses are computed for each category of deer
by reference to functions relating the density
of deer at the beginning of the period to the
rate of mortality. Here, density is the proxy
variable for an array of causal relationships,
as indicated by Figure 2. These natural mor-
tality functions were based upon biological
theory and the available empirical evidence.
The paucity of data, however, precluded sta-
tistical estimation; hence, use was made of
interpolation techniques between data points
to derive the mortality rates for particular
densities. Natural losses are therefore endog-
enous to the model.
Hunting losses are treated differently. The
hunting loss rates are defined by age category
and the time period in which hunting is allow-
ed, as specified prior to the execution of a com-
puter run. The hunting losses could be made
endogenous, but in the first generation model,
where accent is on formulating a reasonable
biological model, it is advantageous to man-
ipulate these losses to test the model. In the
real world, hunting strategies are fomulated
cognizant of political considerations, regula-
tions, management capability, and the demand
for hunting. They are the consequences of
4 The exponential average density each month is
computed as follows:
EADt = EADt_ , + 1/T (D, - EAD, , )
interactions which are not fully indicated by
Figure 2.
Thus far, the model has been presented as
deterministic. The real world is characterized
by random variability. The response of the
deer biosystem to a particular set of conditions
is variable, due to random, uncontrollable
elements such as the weather conditions. Ran-
domness must be accounted for in any simu-
lation which purports to model reality.
In the deer model, a random number gener-
ator is used to generate variability.5 Vari-
ability is due to weather conditions which are
assumed to result in particular forage quality-
quantity relationships or forage conditions.
The notion of a forage factor is used as an
index of forage conditions. Each year, a random
number is computed which, in turn, implies
a particular forage factor. Only five forage
conditions are identified. A forage factor of
five corresponds to average conditions; and
a forage factor of one corresponds to poor con-
ditions. Forage factors of two and four cor-
respond to below and average conditions, res-
pectively.
The probability distribution of forage factors
can be easily modified, consistent with the
investigation of the impact of changes in the
pattern of forage conditions over time. Once
the forage factor is selected for the year, it is
used to modify the components of the system
which are considered to be subject to vari-
ability due to changes in the forage conditions
— namely, natural mortality rates and birth
rates. The notion of the forage factor has
proved most useful in the development of the
computer model, in addition to its primary role
in carrying out experiments with the model
after development.
Thus, the biomanagement system is present-
ed as a network of flows, rates, and levels. The
system being modeled is complex, but by suit-
able abstraction, a workable dynamic model
which permits examination of the system in
a manner not permitted by the usual compara-
tive statics formulation, can be developed.
where: t = Time period (month)
EAD = Exponential average density (deer/
square mile)
D = Density (deer/square mile)
T = Exponential smoothing time con-
stant (number of months)
5 The computer program generates a sequence of
pseudo-random numbers which provides the facilities
for comparing results of different runs under identical
simulated conditions.
125
Time Sequence of Events
A flow chart of the computer program of
the deer herd is shown in Figure 3. For any
simulation model concerned with the flow of
variables over time, a unit of time must be
defined for purposes of calculation. The com-
puter moves in discrete steps through time,
and calculates the variables at each step. In
the deer herd model, the unit of time is one
month. For each month of a computer run,
the relevant calculations are made, and the
status of the system at the end of that month
is generated. The status of the system is an
array of rates and levels for all variables in
the system. The time counter is advanced one
unit (one month) and the appropriate calcu-
lations for that month are made. Calculations
can be made conditional upon any event or
series of events in the past, but not upon
future events, because they have not occurred.
Starting with the opening inventory shown
at the top of Figure 3, the computer program
selects a forage factor for the year as of Novem-
ber 1, and computes natural losses as a con-
sequence of the forage factor and deer density.
Figure 2 shows the array of interactions which
are summarized in the mortality rate-density
functions. The mortality rate in each age and
sex class is described as an exponentially in-
creasing function of density. Hunting losses
are then computed in accordance with the
hunting strategy specified for the simulation
run, and the closing inventory by age and sex
is calculated. Loss totals are then accumulated,
and can be included in the output as desired
by the analyst. Each month, the above sequence
of events is carried out.
After accumulating losses in May, the num-
ber of new fawns to be introduced into the
herd is computed. The birth rate in each age
class of does is described as a decreasing func-
tion of the exponential average density. The
age categories are then advanced one year.
Bucks and does in their sixteenth year are
removed from the system — represented in
Figure 2 by the sink. Fawns born 12 months
previously are separated into bucks and does,
and redefined as deer in their second year.
Two accounting years are defined in the
computer program. The first is from November
1 to October 31. November 1 is the time when
managers are best able to make population
counts indicative of the age and sex composi-
tion of the herd. The second accounting year
used in the model, July 1 to June 30, facilitates
the summarization of the hunting results for
each year. Selected parameters are printed
at the end of each accounting year. After all
the October operations are performed, the
year counter is advanced and the simulation
proceeds until the specified number of years
has been executed. At the end of each run, sum-
mary statistics are printed.
Input Data
The model is intended to simulate the Men-
docino County deer herd, but the primary
data source was the University of California
Field Station at Hopland, where the deer
population has been under continuous and in-
tensive study since 1951. The investigators
at Hopland compiled these data and integrated
them with the California Fish and Game De-
partment data for the remainder of the county.
Data input for each run is separate from
the computer program. This permits changes
in the data assumptions to be made without
altering the computer program. The program
is designed to be applicable, with minor modi-
fication, to other big game populations.
The data block for each run includes constants
to initialize the run, such as the opening in-
ventory, the area of land available to the herd,
and the length of the run. Other data used in
each year include birth rate and natural mor-
tality functions, hunting loss percentages, and
the distribution of forage factors.
HUNTING STRATEGY RESULTS
While an infinite variety of hunting strate-
gies can be tested in this moi..el, the options
of the wildlife manager are limited because
certain hunting strategies that are biologically
feasible may be socially or politically unaccept-
able. In addition, hunters can usually dis-
tinguish only a few age and sex classes in the
field. Limited hunter access to extensive
areas of private forest and range lands pre-
cludes the achievement of uniform hunting
pressure over the entire county.
126
Figure 3. — Flow chart of the computer program of the deer
herd.
The hunting strategies summarized in Table
1 include the range of options which could be
practically implemented in Mendocino County.
Two kinds of population parameters are shown:
those which can be maximized or minimized
as management goals, and those comparable
with field data to determine whether manage-
ment goals are being achieved. Some
parameters, such as the hunting kill, serve
both purposes. The current program prints
out many other parameters in addition to those
presented in Table 1.
Although it is physically possible to hunt
deer at any time of the year, in California it
is customary to set the deer seasons in late
summer and fall, for numerous biological and
sociological reasons. In the simulation runs
presented in Table 1, all buck hunting was con-
ducted during August and September, in ac-
cordance with existing custom, and potential
doe and fawn hunts were set for November
and December, the months when antlerless
deer are in the best condition. All parameters
other than hunting specifications were held
constant throughout these runs, and the values
shown were selected from the output after
127
Table 1. — Selected parameters of the Mendocino County deer
population as affected by alternative hunting strategies.
Strategy
45% bucks
50% bucks
No
25% adult
30% does
15% does
hunting
bucks
15% fawns
60% fawns
(1)
(2)
(3)
(4)
Total deer
June 1
236,000
251,000
141.000
168,000
November 1
191,000
191,000
117,000
141,000
May 30
150,000
148,000
90,000
93,000
Annual Losses
Natural
85,000
95,000
15,000
22,000
Hunting
-
7,900
36,000
53,000
Natural hunt-loss ratio
-
12:1
0.4:1
0.4:1
Kill as percent of June 1
population
-
3
26
32
Percent composition of kill
Bucks
_
100
42
22
Does
-
0
41
18
Fawns
-
0
17
60
Herd composition data
Fawns/ 100 does
Spring
41
41
90
50
Fall
64
64
83
96
Bucks/ 100 does
Fall
86
43
41
22
stability had been attained. Year-to-year vari-
ability was suppressed to highlight the dif-
ferences among the hunting strategies. The
principal features of each strategy are sum-
marized below:
1. No Hunting: This strategy is presented
mainly for comparison with the other
runs. It is characterized by a high buck: doe
ratio, low productivity, and high natural
mortality.
2. Twenty-Five Percent Adult Bucks: This
is an estimate of the hunting effected in
Mendocino County during the past 10 +
years. Hunting is limited to males with
two or more points per antler. Natural
mortality is higher than in Strategy 1
because the population includes relatively
more does, as indicated by the buck: doe
ratio, and the number of fawns born is,
therefore, higher. Fawns are most sus-
ceptible of all age classes to natural mor-
tality. Overall deer numbers do not differ
markedly between Strategies 1 and 2.
For every deer taken by hunters, about
12 die of starvation and other natural
causes. Although the management goals
are not explicitly denned, current regu-
lations result in the maintenance of maxi-
mum deer numbers and maximum natural
losses. This strategy provides no con-
straint upon overall deer numbers.
Forty-Five Percent Bucks, Thirty Per-
cent Does, and Fifteen Percent Fawns:
Where the hunter is allowed to select
either bucks or does, this strategy repre-
sents the results of the heaviest hunting
128
pressure likely of achievement. Although
hunters generally avoid killing fawns if
possible, data from other areas indicate
that fawns comprise 15% to 20% of the kill
in antlerless hunts. The annual kill of
36,000 Would probably require private
lands to be hunted as heavily as public
lands. Comparison with Strategy 2 indi-
cates that the hunting kill would increase
about 45% , even though the overall popu-
lation decreases about 40% . Natural losses
are also much reduced.
4. Fifty Percent Bucks, Fifteen Percent Does,
and Sixty Percent Fawns: While the previ-
ous strategy would tend to maximize the
hunting kill if hunters were allowed their
free choice of animals, the kill could be fur-
ther increased by selectively hunting fawns.
This strategy is comparable with the usual
sheep management regime in Mendocino
County, where a high proportion of lambs
is marketed annually. Although the kill
would be considerably higher than in the
previous strategy, the total biomass yield
would be slightly lower because of the
relatively small size of fawns. It may be
unrealistic to propose that 50% of the
bucks can be killed annually. However,
if the goal of management is to maximize
the number of animals taken by hunting,
it is necessary to maintain the highest
possible proportion of breeding does in the
herd, and this can be achieved only by
heavy hunting of adult males.
A convenient way of showing hunting yield
and population numbers at equilibrium for
different strategies is by plotting the results
from many computer runs on graphs like these
shown in Figures 4 through 6. These graphs
permit a comparison of the relative effects of
selective hunting pressure directed against
does, fawns, and buck, respectively.
Figure W- This graph depicts population
trends and yields of deer when various per-
centages of does are taken by hunting when
(A) no bucks or fawns are taken, and (B) 50%
of all bucks and 15% of the fawns are taken an-
nually. Several pertinent aspects of population
performance are apparent from this graph:
(1) With no hunting of bucks and fawns, a
slightly higher total population of deer
tends to be maintained when any given
removal of does is carried out.
(2) Maximum productivity or yield of the
population is achieved when approximate-
ly 25% of the does are removed annually.
However, the total yield is approximately
five times higher if bucks and fawns are
taken as specified in Strategy (B).
(3) As hunting pressure on does increases,
overall deer numbers decrease at an in-
creasing rate.
Figure 5: Figure 5 indicates the effect of
increasing fawn removals accompanied by (C)
no buck or doe hunting, or (D) annual hunting
removals of 50% of the bucks and 30% of the does.
It shows that:
(1) The total population will decline only
slightly with the increasing removal of
fawns only, as depicted by (C).
(2) Under the buck-doe strategy in (D),
maximum yield and population size will
diminish rapidly if annual fawn removal
exceeds approximately 30% .
Figure 6: The hunting conditions set forth
on this graph are, (E) no does or fawns are
taken as related to the increasing take of bucks,
and (F) a removal of 30% of the does and 15% of
the fawns in relation to an increasing take of
bucks. The graphs show that:
(1) Buck removal alone has only a slight
effect on yield, and even less effect on
total population.
(2) When does and fawns are taken as speci-
fied in Strategy (F), the total yield of
the population is roughly doubled, as com-
pared to taking bucks only.
GENERAL RELATIONSHIPS
Consideration of the three graphs shows that:
(1) Maximum yield of the Mendocino County
deer population is only achieved through ex-
ploitation.
(2) Reduction of the large, unexploited popu-
lation through hunting produces a more dy-
namic population, with greater turnover. The
basic relationship is to lower stocking rate
on the range, which reduces competition for
available feed, and thereby raises the plane
of nutrition. This, in turn, improves fecundity
and survival.
129
180 _
140
120
100
(A) Stable November 1 Population
Buck Hunting 07.
Fawn Hunting 07.
20 30 40 50
Percent Does Taken Annually
Figure 4. — Yield and population numbers at equilibrium for two buck-fawn
hunting strategies and variable doe hunting percentages.
(3) It appears that maximum population
and yields will probably be achieved with a
hunting removal of about 20-25% of the does,
15-30% of the fawns, and over 50% of the bucks
annually. At this rate of buck removal, there
is no possibility of reducing the breeding suc-
cess of the population, but it is highly unlikely
that such a high rate of buck take can ever be
achieved over the county as a whole. The density
of cover on much of the deer range precludes
it. Under present hunting practices, a buck
removal of possibly 20-25% is being achieved.
At best, this might possibly be doubled.
Likewise, it is highly unlikely that hunters
can be forced to take large numbers of fawns
selectively. Most either-sex hunting efforts
can be expected to produce a take of fawns
of about 10-20% , and it is difficult to increase
this, as hunters try to avoid taking fawns
because of their small size.
(4) Removal of does above the 25% level is
the most powerful means available for total
population control, since it reduces total re-
productive potential. This finding is readily
applicable to the special management problems
in National Parks, where big game numbers
130
160
140
Percent Fawns Taken Annually
Figure 5. — Yield and population numbers at equilibrium for two buck-doe
hunting strategies and variable fawn hunting percentages.
must be controlled, but public hunting is con-
sidered incompatible with other management
goals. In such situations, it is customary for
surplus animals to be shot by park officials.
Our calculations indicate that these removal
programs should be directed solely against
adult females to provide the most effective
population control. This would minimize the
number of animals to be killed, as well as the
manpower requirements, and would additional-
ly maintain a high proportion of the aesthetical-
ly desirable adult males in the population.
CONCLUSIONS
Computer simulation of dynamic biomanage-
ment systems appears to provide a means of
generating information useful to resource
managers and to research administrators. In
building computer simulation models, research-
ers and managers put together their theoretical
and practical knowledge of a system. This
process frequently results in finding existing
gaps in empirical data, and helps to revise
131
160
140
120
100
60
'!
(E) Stable November 1 Population
Doe Hunting 0%
Fawn Hunting 07.
Stable November 1 Population
Doe Hunting 307.
Fawn Hunting 157.
|(F) Yield at Stable Population
) Doe Hunting 30%
30 40 50 60
Percent Bucks Taken Annually
Figure 6. — Yield and population numbers at equilibrium for two doe-
fawn hunting strategies and variable buck hunting percentages.
research plans and data collection procedures
for monitoring the real system. Outside of this
important research administration outcome,
information about consequences of manage-
ment policies which might otherwise not be
obvious can be provided. For example, our
results to date indicate that annual revisions
of the hunting regulations will not, in general,
cause management objectives to be attained
more rapidly than following a fixed hunting
strategy. This is due to the compounding ef-
fects of random variability and the difficulties
in monitoring the system.
The systems analysis approach, and its con-
comitant technique of computer simulation,
can and has been used to study other wildlife
resources such as fish populations. Models de-
veloped for fish populations would necessarily
incorporate the unique features of each system,
and the output would be designed according
to the special needs of the resource manager.
However, further exploration of the usefulness
of computer simulation in studying fish popu-
lations is needed before the optimism shown
for big game management can be expressed
for management of fisheries.
132
Augmentation of Salmon Stocks through Artificial
Propagation: Methods and Implications1
Joe B. Stevens and Bruce W. Mattox2
ABSTRACT
Eighty-one hatcheries on the Pacific Coast now rear significant numbers of salmon
and steelhead for sport and commercial fisheries. Annual operation and maintenance
costs amount to $6.6 million. A production function analysis of 15 Oregon Fish Com-
mission hatcheries produced tentative conclusions that (a) controlled inputs were com-
bined in fixed proportions, (b) constant returns to size were realized, and (c) some
degree of factor substitution existed between the controlled "fixed proportion input"
and water temperature. The latter relationship may allow hatchery managers to im-
prove efficiency at the hatchery level. Uncertainty with respect to downstream en-
vironmental conditions, however, must be considered along with returns to size for
the hatchery production function when new investments are undertaken.
Fixed asset theory was used to conceptualize exit and entry of salmon harvesting
resources between 1947 and 1966. Net entry followed years of good catches, but net
exit did not occur following the bad years. If a major objective of hatchery programs
is to augment fishermen's incomes, consideration must be given to increasing the
opportunity costs of extant resources as well as to limiting entry of new resources.
INTRODUCTION
It is a moot question to ask whether or not
the public sector should involve itself exten-
sively in hatchery rearing of salmon and steel-
head on the Pacific Coast. Eighty-one hatcher-
ies, valued at over $56 million with annual oper-
ation and maintenance costs of $6.6 million,
now rear significant numbers of chinook and
coho salmon and steelhead trout for sport and
commercial fisheries. It is a relevant question,
however, to ask under what conditions con-
tinuing investment of this type should be under-
taken. Although this is a question which can
and should be posed, it is not easily answered;
thus we do not attempt to do so, aside from
exploring some obvious and not-so-obvious
implications. Our major attention herein is
devoted to asking and partially answering
the question: "Given the decision to augment
resource flows by artificial propagation, what
can be gleaned from existing data which will
allow the public sector to increase efficiency
at the hatchery level?" In exploring this ques-
tion, we recognize the dangers of a partial
analysis, i.e., divorcing hatchery objectives
from higher order objectives. Our defense is
pragmatic, i.e., that it is better to start fitting
the pieces of the puzzle together, one by one,
than to not start at all or to theorize how they
might all be fitted simultaneously.
THE CURRENT SIGNIFICANCE OF
SALMON AND STEELHEAD HATCHERIES5
1 Technical Paper No. 3010, Oregon Agricultural
Experiment Station.
2 Associate Professor of Agricultural Economics,
Oregon State University, and Assistant Professor of
Resource Economics, University of Rhode Island, res-
pectively. This publication is supported in part by the
National Oceanic and Atmospheric Administration
(maintained by the U.S. Department of Commerce) In-
stitutional Sea Grant 2-35187. Nothing stated herein is
to be taken as representing the views or policies of the
Oregon Fish Commission.
The first Pacific Coast salmon hatchery was
constructed in Northern California by the
U.S. Fish Commission almost a century ago.
Since that time, artificial propagation of
salmon has alternately been viewed as a
panacea and as no solution at all. Improve-
ments in propagation methods have allowed,
3 Data on the nature and contributions of hatchery
programs were taken freely and gratefully from Wahle,
(1970).
133
and environmental deterioration has forced,
increased reliance on hatchery operations,
especially in the past decade. Eighty-one
hatcheries are now operated by fishery agencies
of Alaska, Canada, California, Oregon, and
Washington, and by the Bureau of Sport Fish-
eries and Wildlife. Extensive evaluation pro-
grams are carried on by the Columbia Fisheries
Program Office of the National Marine Fish-
eries Service and by some of the other agencies.
The evaluative work of the NMFS program
has included extensive fin-clipping, sampling
for marked salmon, and benefit-cost analyses
for brood years by species.
The current status of these resource augmen-
tation programs has recently been summar-
ized by Wahle (1970) of the NMFS, and is
portrayed in Table 1. Survival rates of 4 to 5%
indicate that a multitude of fingerlings must
be released in order to affect resource stocks.
The cost of production for one fingerling, on
the other hand, is relatively low. Our study
revealed that the 15 hatcheries of the Oregon
Fish Commission produced the equivalent of
about 70 million salmon and steelhead finger-
lings between October 1, 1968 and April 30,
1970, at a cost of slightly over two cents per
fingerling.4 Assuming that the survival rates
in Table 1 are appropriate, the cost per fish
caught at some time in the future rises to about
$1.35, disregarding any discounting for time.
The contributions of hatchery-reared fish to
the ocean troll fishery is impressive, ranging
from 30 to 80% of total catch in 1968. Wahle
points out, however, that the proportion of
hatchery fish to wild fish was higher than
usual in that year. The true contribution to
the sport catch of coho, for example, may be
closer to 50% .
It may be useful to this group to have the
hatchery programs put into perspective with
the total salmon catch for the West Coast
States of Washington, Oregon, and California.
Table 1. — Survival rates and contributions to ocean
troll fisheries of hatchery-reared salmon and steelhead
in 1968.
Hatchery-reared fish as
a percentage of total
Species
Survival rate1
ocean troll catch (1968)2
Coho
0.04 (.037)
Commercial: 30%
Sport: 80%
Fall Chinook
.004 (.003)
Commercial: 70%
Spring Chinook
.05
Sport: 65%
Steelhead
.04
—
'Survival rates for coastal streams are shown in parentheses.
2The commercial fishery data for chinook salmon include
landings from the west coast of Vancouver Island, in addition
to landings in Oregon, Washington, and California. The sport
landings include only the latter three States.
SOURCE: Wahle, 1970.
To do so, we have done some quick (and dirty)
calculations for which we assume sole responsi-
bility. The total yearly landings of all salmon
in this region fiucuate widely because of the
odd-year cyclical nature of pink salmon, an
important species for which hatchery propa-
gation work is now in advanced experimental
stages (McNeil, 1969). Averaging one recent
cycle year for pink salmon (1967) with one
non-cycle year (1964), about two-thirds of the
total salmon catch is comprised of coho and
chinook (U.S. Department of the Interior,
1947-1967). Assuming that Wahle's data from
Table 1 are appropriate for coho and chinook,
regardless of method of capture5 (troll, gill net,
purse seine), and using a conservative hatchery-
contribution share of 30% , it would appear
that perhaps 20% of the total West Coast (U.S.)
salmon fishery is supported by hatchery pro-
grams. This share is increasing over time,
and success in rearing pink salmon will pro-
vide further augmentation.
4 This assumes that coho and spring chinook were
released at 15 fingerlings per pound of fish, fall chinook
at 100 per pound, and steelhead at 10 per pound. Costs
include variable operating expenditures plus and im-
puted 5% charge on the $7.5 million replacement value
of fixed facilities (Mattox, 1970 and Wahle, 1970). The
latter sum is no doubt an overestimate of real capital
values.
5 The troll fishery accounted for about 63% of total
coho and chinook capture, averaging 1964 and 1967
data. Ocean troll alone would constitute at least 50%
of total catch.
134
PRODUCTION FUNCTION ANALYSIS OF
HATCHERY PROPAGATION OF
SALMON AND STEELHEAD
The Incentive Framework of
Hatchery "Firms"
As is usually the case, our initial research
objectives were more elegant than could be
accomplished with existing time and data.
Initial plans were to estimate marginal pro-
ductivities for each of several factors of pro-
duction relevant to the 15 major hatcheries of
the Oregon Fish Commission, and to estimate
the total elasticity of production or returns
to size for these hatcheries. If possible, we
wanted to incorporate into the function post-
hatchery phenomena, especially the physical
returns to the fishery of hatchery-reared fish.
In that the NMFS data on the latter were
not yet precise enough to identify differential
returns by hatchery, it was necessary to
restrict the analysis to the hatchery produc-
tion function.
One of the most interesting aspects of the
analysis was the influence on model specifica-
tion of the incentive framework of the hatch-
eries. Federal and State hatcheries receive no
price for their product, have no responsibility
for realizing profit, and are managed by pro-
fessionals trained primarily in terms of bio-
logical relationships. Budget constraints are
imposed by the political rationing process.
Furthermore, the nature of the incentive frame-
work is such that it is only partially con-
ducive to providing data in a form which is
useful for economic analysis.
On the other hand, hatchery managers are
not unaffected by economic forces, since they
face constraints on operating capital and tech-
nology as well as constraints with respect to
factor prices, fixed facilities, and natural phe-
nomena. Among the latter are yearly and
seasonal variations in water quantity, which
often result in the non-use of rearing ponds,
and seasonal variations in water temperatures
which affect metabolic processes of fry and
fingerlings.
The absence of a product price, of course,
does not mean that the conventional econo-
mizing model is not relevant. The influence of
technological, budgetary, and factor price con-
straints seemed sufficiently strong to postulate
that hatcheries attempt to maximize output
subject to these constraints. In one major
respect, however, it was anticipated that the
decision framework of the hatchery managers
would give rise to a type of empirical result
not usually obtained in analyses of private
firms. That is, it was hypothesized that the
particular set of hatcheries we observed were
(a) combining controlled inputs in fixed pro-
portions, and (b) realizing constant returns
to size.
The reasoning behind this hypothesis largely
reflects the institutional nature of the hatch-
eries, although physical attributes of the pro-
ductive factors serve as necessary conditions.
The primary institutional factor is the influ-
ence of centralized supervision on the Fish
Commission hatcheries. Resident managers
appear to operate within guidelines set by the
central office with respect to input combin-
ations, a system which is reinforced by dis-
ciplinary training of both groups. The physical
attributes of factors which would allow them
to be combined in fixed proportion is a rela-
tively high degree of divisibility. The latter
is elaborated below.
The Biological Production Function
The underlying production function for
fingerlings can be viewed as consisting of three
controlled factors — food, labor, and rearing
space — and one non-controlled factor — water
temperature. The food variable is nutritionally
complex, but a convenient one for analytical
purposes since the Oregon Moist Pellet is a
"complete" ration. This food, fed in a variety
of pelletized and mash forms, was specially
formulated to satisfy the nutritional demands
of fingerlings at different ages as well as for
prevention and treatment of disease. Further,
the food is centrally purchased, thus elimina-
ting any price differentials between hatcheries.
Although mechanical feeders have been tried
in some areas, the Fish Commission feeds
entirely by hand application of the pellets.
In that a pool of temporary labor is usually
available to resident managers, both labor
and food variables are quite divisible.
The third major controlled variable, rearing
space, might be described, tongue in cheek,
135
as water surrounded by concrete. Water flows,
as noted earlier, vary in quantity and tem-
perature. Both of these physical dimensions
are largely outside the control of management.
Although some low flow augmentation is ac-
complished, the usual result of low flows during
summer months has been an inability to fully
utilize rearing space. Since the rearing ponds
are fairly small and numerous, low flows are
adjusted for by maintaining water volume
in some ponds and temporarily retiring others.
Thus, the rearing space actually used is also
fairly divisible, although some seasonal excess
capacity may exist.
Although our initial inclination was that
separate marginal factor productivities might
be estimated, discussions with hatchery man-
agers soon revealed the similarity of practices
in combining controlled inputs. Levels of
inputs and outputs at larger hatcheries seem-
ed to be constant multiples of those found at
smaller hatcheries, although opportunities for
variable input proportions seemed to be present
in a physical sense. One could, for example,
stock rearing ponds with fingerlings at dif-
ferent rates, or spread existing water flows
over all rearing ponds. Centralized manage-
ment, of course, may not be conducive to such
experiments. On the other hand, it may well
be that past "experiments", intended or other-
wise, have revealed that other factor combin-
ations involve a greater degree of risk. For
example, disease spreads rapidly in rearing
ponds; overcrowding of fingerlings might be
disastrous. Similarly, lower water levels in
all ponds would increase water temperature
and accelerate the spread of disease.
Our hypotheses of fixed factor proportions
and constant returns to size were equivalent
to expecting that the Fish Commission acts
as if the isoquants for hatchery production of
fingerlings are right-angled, whether they
actually are or not. The hypothesis was strong-
ly dependent, of course, on our prior decision
to analyze Fish Commission hatcheries. A
cross-section analysis over various agencies,
in retrospect, would possibly have yielded more
empirical information.
The non-controlled variable, water tempera-
ture, can be quite important during periods
of either cold or warm weather. Extremes of
either type seem to effect primarily the volun-
tary rate of metabolic activity, rather than
the efficiency of food conversion (Paloheimo
and Dickie, 1966). It was expected, then, that
growth would be retarded in the upper and
lower limits of observed water temperature.
This noncontrolled variable, then, was viewed
as the principal shifter of a constant returns
production function.
Exploratory Estimation
The time period selected for analysis was
October 1, 1968 through April 30, 1970. This
19-month period allowed the propagation pro-
cess to be observed for at least one brood
year for each species of interest (Figure 1).
These included coho, spring chinook, fall
chinook, chum, and steelhead. In the absence
of cost data which were separable by species,
it was necessary to estimate an aggregate
function over all species.6
In view of the fixed proportions hypothesis,
the initial attempt at estimation involved
several of the factors which were thought to
be jointly combined. We were limited in this
analysis by the absence of data on either actual
water flows or rearing space used. As a fairly
unsatisfactory proxy, these variables were re-
placed by a measure of the replacement value
of all fixed facilities. This variable, along with
food, operating expenses (largely labor), and
cumulative water temperature units7 for the
warm weather period and the cold weather
period, constituted the five independent vari-
ables in the initial run.
As anticipated, a high degree of intercorre-
lation resulted between food, operating ex-
penses, and the value of fixed facilities in both
Cobb-Douglas and linear estimations. Correla-
tion coefficients between these three variables
approached or exceeded 0.80, and resulted in
a considerable inflation of standard errors.
Since it appeared that some degree of factor
substitution could be estimated between any
6 An interagency effort is now underway to explore
cost accounting systems by species.
7 A cumulative temperature unit (CTU) is defined
for each day in which the average water temperature
exceeds 32°F by one degree. One month of 40° water
temperature, for example, would constitute 240 CTU's.
136
1968
1969
SPECIES AND BROOD o|n|d|j|f|m| A | M | J J A S ON D J F | M | A
COHO
FALL CHINOOK
^
1968
1969
SPRING CHINOOK
1967
1968
1969
CHUM
^^^^^
1968
1969
^^
Figure 1. - Brood year classification of species propagated from the beginning' of October
1968 through April 1970.
one of these variables and water temperature,
the food variable was retained in further
analyses.
Marginal Factor Productivities
and Returns to Size
Since the underlying functional relation-
ships were unknown, output response func-
tions were estimated in both linear and log-
linear (Cobb-Douglas) forms. Within each func-
tional form, estimates were obtained relating
to two different assumptions about the inter-
cept term.8 The output response functions and
marginal physical productivities are shown
in Table 2.
Several items are worthy of note. First, the
R2 values were uniformly high, regardless of
functional form. Second, the marginal pro-
ductivity estimates appeared reasonable and
were fairly constant over the various functional
forms. The marginal productivity of one pound
of food was about 0.58 pounds of salmon, a
Table 2. — Output response functions and marginal
physical productivities for the 15 Oregon Fish Com-
mission salmon hatcheries.
8 While output would logically be zero if all input
levels were zero, an estimate of the intercept may be
helpful in assessing the "constant returns" argument
for the linear function.
Functional form
Intercept
b
c
b
l
Variables1
b b
2 3
ft-
I. (a) Linear
2-13,998
.572
-15.694
33.324
0.959
3 (.10)
(.01)
(.10)
(.05)
4 .572;
-15.694
33.324
I. (b) Linear
0
.563
-16.735
29.715
.991
-
(.01)
(.05)
(.05)
.563
-16.735
29.715
II. (a) Log-Linear
- 23.74
1.106
- .334
.526
.960
(.01)
(.01)
(20)
(.05)
.620
-11.281
31.618
II. (b) Log-linear
(t
1.047
- .450
.332
.999
-
(.01)
(.10)
(.20)
.588
-15.217
19.958
1 Variables:
y = pounds of output of salmon (released e
fingerlings or swim-up fry)
X = pounds of food fed (Oregon Moist Pellet).
ther as
X^ = aver
ige cumulat
ve tem|
jerature
units (CTU's) of
water from May through October (warm season).
X, = average CTU's of water from November through
April (cold season).
2 Regression coefficients.
3 Significance level.
4 Marginal physical productivities.
137
figure that seems consistent with the literature
in fisheries biology (Paloheimo and Dickie,
1966). Adding one day with water temperature
one degree in excess of 32 °F (i.e., one CTU)
during the cold season would add about 30
pounds to total output; one additional CTU
during the warm season would reduce output
by about 15 pounds. Third, the high R2 values
support the hypothesis of fixed factor propor-
tions, although we recognize that another
analysis, covering several agencies and systems
of management, might well yield different
results. Fourth, the evidence appears to sup-
port the "constant returns" hypothesis, al-
though this is somewhat conjectural. Summing
the coefficients for Cobb-Douglas forms is
hindered by the negative coefficient on warm
season water temperatures. One might, as we
did, view the water temperature variables as
"shifters" of food-input relationship. If so,
the coefficients on the food variable do not
differ significantly from unity.9
Our estimates of marginal productivities
thus enabled us to ask, "What would be the
change in hatchery output if one were to in-
crease (or decrease) water temperatures by a
given amount?" A 10% reduction in CTU's
during the warm season would reduce average
water temperature from 52.97°F to 50.87°F
and cause output to increase by 5,684 pounds,
or about 4.36% of the mean hatchery output.
Raising cold season water temperatures from
43.99°F to 45.19°F would add 6,218 pounds
of output, or about 4.77% of mean hatchery
output.
Factor Substitution
If controlled inputs are combined in fixed
proportions, as evidenced above, the data ob-
viously do not allow estimation of substitution
possibilities. On the other hand, our analysis
does permit us to identify degrees of sub-
stitution between the fixed proportion input,
using food as a proxy variable, and changes
9 The negative intercept on the linear model was
significantly different from zero at the 0.10 level. This
gives some evidence of increasing returns, and is
consistent with the bi estimates of 1.106 and 1.047
for the log-linear models. Acceptance or rejection of
"constant returns" thus, depends partly on one's pref-
erence for significance levels.
in the noncontrolled water temperature vari-
ables. The marginal rates of factor substitu-
tion, as estimated from both linear and log-
linear functions, are shown in Table 3. Al-
though log-linear models no doubt conform
more closely to biological reality, linear rates
of substitution may be appropriate for some
decisions. The degree of isoquant curvature is
largely a matter for the judgment of fisheries
biologists; experimental work in this area
should be useful in checking and refining our
estimates. Our confidence in the linear rates
would be greatest in the neighborhood of mean
CTU values (e.g., Figure 2).
Table 3. — Linear rates of factor substitution between
inputs.1
3 [Food (Xt)]
3 [Food(X,)l
Functional form 3 [Summer CTU's (-X2)
3 [Winter CTU's (X3)\
I. (a) Linear -27.462
I. (b) Linear -29.714
II. (a) Log-linear -18.186
II. (b) Log-linear -25.867
-58.309
-52.762
-50.972
-33.935
1 Estimates are based on mean values. The sign on the X vari-
2
able (warm season water temperatures) is reversed here for
convenience since decision makers would attempt to reduce
summer temperatures and increase winter temperatures.
Increased environmental control, as through
controlling water temperature, is in fact one
means that Pacific Coast fishery agencies are
now considering for output augmentation. Thus
far, the agencies have primarily adapted to,
rather than controlled, this aspect of the en-
vironment. The hatching of fry is concentrated
to some degree in those hatcheries which have
water temperatures most conducive to this
operation; other hatcheries tend to specialize
in the rearing of fingerlings. Control of tem-
peratures would allow both food and transport
costs to be lowered, although empirical data
on factor price ratios were not available. It
was our thinking that the estimates of factor
substitution in Table 3, together with a step-
by-step presentation of "output maximization,
given budget constraints" would aid agencies
in increasing efficiency at the hatchery level.10
10 Specific attention was directed to the problem of
determining factor prices when there is significant
unused capacity at existing hatcheries. As mentioned
earlier, seasonal low water flows often force non-use
of some rearing ponds.
138
2 -
. Actual observa
Q ~ 70,500
Q ~ 130,481
m —
- \^
Q . 2,210, 000
^\Qi
\ ^v
'•'
1 '
1 i Ql
Cumulative
Temperature
(Thousands)
Figure 2. — Observed relationships between food and cumulative
temperature units (November through April).
This information will be made available to
hatchery management through an Oregon
State University Marine Economics publication.
Concluding Comments on the
Hatchery Production Function
Several strengths and qualifications of our
research became clearer as the work progress-
ed. The principal strength is that our conven-
tional cross-sectional analysis of "firms" can
be useful to public decisionmakers in spite
of their "unconventional" incentive frame-
works. Our principal lesson in methodology
has been that differences within frameworks
of the various agencies may be more crucial
than differences between those of private and
public firms if the researcher's objective is to
provide a substantial empirical input. In retro-
spect, had we included a number of agencies
in our study, it may have been possible to esti-
mate additional substitution relationships. If
our limited empirical results are useful to
management agencies, however, we may have
opened the door for a data system reorganiz-
ation which will both allow for improved eco-
nomic analysis and facilitate consideration of
a broader range of production alternatives.
Our policy advice is accordingly limited by
the methodological constraints of this study.
Constant returns from hatchery operations
may exist, ceteris paribus, but the latter may
not be a very legitimate assumption when un-
certainty exists as to downstream environ-
mental conditions. Agencies could, for example,
spread production over many small hatcheries
located on different streams, but it may be
more desirable to construct fewer and larger
hatcheries if environmental protection can be
assured on specific streams.
SOME IMPLICATIONS OF INCREASED
HATCHERY PROPAGATION FOR
COMMERCIAL FISHERIES MANAGEMENT
Associated Harvesting Costs
The principal limitation on policy advice
stemming from our research is, of course,
whether or not increased efficiency at the hatch-
ery level necessarily leads to increased effic-
iency at the fishery level. The problems of
open-access in U.S. commercial fisheries are
well known to this group and will not be re-
peated here (Christy and Scott, 1966 and
139
Crutchfield and Pontecorvo, 1969). Let it be
sufficient to say that there is both theoretical
ambiguity and a lack of empirical information
on the private and public costs associated with
harvesting open-access resources (Bromley,
1969).
Two lines of thought, however, would prob-
ably receive acceptance by this group. The
first is that in the short run, hatchery produc-
tion could increase output in most salmon
fisheries with only minor increases in associ-
ated harvesting costs, since excess capacity
does exist. The Crutchfield-Pontecorvo re-
search supports this for the Pacific salmon
fisheries. The second argument is that the
open-access tradition insures that resource
augmentation through publicly operated hatch-
eries will induce additional effort into the fish-
ery, especially when the additional inputs are
provided without cost to the fishermen. The
resultant equilibrium levels of factor returns,
output prices, and excess capacity may differ
from initial equilibrium levels, but a priori
speculation about empirical magnitudes is just
that. Furthermore, the time pattern of adjust-
ment and the distribution of benefits and costs,
over both time and space, can be discerned
only vaguely.
We would maintain, however, that resource
augmentation efforts should be placed in per-
spective with the total institutional setting.11
Hatchery contributions to fish stocks may per-
petuate the tendency toward excess harvesting
capacity, but it should not have to bear the
entire burden of responsibility for economic
and social ills of the fishery. The tendency
toward excess capacity pervades open-access
fisheries, most of which do not rely on hatchery
propagation. It would be our guess that the
magnitude of inefficiency associated with the
larger issue probably overshadows any un-
desirable effects of hatchery production, if the
latter in fact exist.
Having confessed that we do not have all
the answers, we hasten to add that we do have
some empirical observations on entry and exit,
over time, of salmon harvesting resources. We
view these not as definitive proof of anything,
but as a piece of the empirical jigsaw puzzle
11 We are indebted to Emery Castle for this pers-
pective.
which must eventually be put together if econ-
omists are to be looked to for policy advice.
Entry and Exit of Resources in the
Commercial Salmon Harvest:
Fixed Asset Theory
The rise and fall of the Pacific Coast salmon
harvest has been well documented elsewhere
(Cooley, 1963 and Crutchfield and Pontecorvo,
1969). Peak harvest years were reached in
the 1930's, and catch has trended downward
since that time. The quantity of resources com-
mitted to the fishery, however, has increased
over time. The number of fishermen and the
net tonnage of vessels increased by about 30%
between 1947-1949 and 1964-1966 periods,
total landings declined by about 25%, and the
deflated value of landings per fisherman de-
creased by about 15% (Table 4). It appears,
however, that the deflated average value of
landings per fisherman has remained about
constant since 1950, with year-to-year fluctu-
ations. This can be taken, recognizing the limi-
tations on accuracy of the data, as very super-
ficial evidence of the open-access phenomenon,
i.e., the dissipation of rents through entry of
additional resources.
Even though there has been net entry into
the fishery since 1947, the time path of entry
and exit of harvesting resources has not been
fully explored. In particular, is there a degree
of symmetry between the relationships which
explain entry, on one hand, and exit, on the
other? Miss Peerarat Aungurarat attempted
to answer this question in another portion of
our Sea Grant research at Oregon State Uni-
versity (1970). Her results are especially inter-
esting in light of the increased reliance on
hatchery programs.
Conventional firm theory suggests that a
high degree of symmetry would exist in ex-
plaining entry and exit of resources. Given a
constant factor price, leftward (rightward)
shifts in the marginal value product function
would imply a reduction (increase) in the
utilization of a factor of production. Dissatis-
faction with the state of the arts in explaining
the inelastic supply of agricultural products
led Glenn L. Johnson to formulate a "fixed
asset" theory (1958). Johnson's contribution
140
Table 4. — Salmon fishing effort, quantity of landings (pounds and values) and average values per fishermen in Alaska,
Washington, and Oregon, 1947-1966.
Year
Labor
(number of
fishermen)
Vessels
(net
tonnage)
Landings
(thousands
of pounds)
Value of
i
landings
(thousands
of dollars)
Average landings
per fisherman
(thousands
of pounds)
Average value
1
per fisherman
(thousands
of dollars)
1947
16,249
44,003
486.560
47,541
29.94
2.92
1948
19,334
59,443
395,981
43,222
20.48
2.24
1949
1950
18,451
19.241
59,510
63.156
477,074
321,575
54,441
42,464
25.86
16.71
2.95
2.21
1951
1952
23.589
22.318
70.799
71.842
367,030
344.999
55,840
46,960
15.56
15.46
2.37
2.10
1953
21.889
69.231
304,945
38,500
13.93
1.76
1954
20.321
66,742
315.217
43,925
15.51
2.16
1955
24.608
69.268
277.900
39,389
11.29
1.60
1956
19.522
63.869
312.837
44,651
16.02
2.29
1957
2
2
260.125
38,580
2
2
1958
1959
2
19.990
2
58.099
303.797
194.915
43.976
32.221
2
9.75
2
1.61
1960
21.546
53.285
229.227
40,146
10.64
1.86
1961
23,206
63.060
301,760
45,421
13.00
1.96
1962
21.921
62.767
307,892
49.649
14.04
2.26
1963
23,689
66.553
286.316
41111
12.09
1.78
1964
22,384
66.057
342.765
47.128
15.31
2.11
1965
23,486
65.691
317.068
54.717
13.50
2.33
1966
24,987
67.314
378.066
60.671
15.13
2.43
1 Deflated by Consumer Price Index (1957-59 = 100).
2 Data not available.
SOURCE: Derived from Fishery Statistics of the United States, U.S. Fish and Wildlife Service. Bureau of Commercial Fisheries.
was his recognition of a particular form of
imperfect factor markets, and involved relaxing
the assumption that firms or industries can
at the same price, both buy and sell inputs.
A "fixed asset", by Johnson's definition, is not
fixed because it has a certain physical life ex-
pectancy, but because it is more economical
to keep it in production than to sell it. Two
factor prices are involved, i.e., an acquisition
price and a salvage value. Applied to the fish-
ing industry, the acquisition price is what a
fisherman (or the industry) has paid or must
pay for an additional productive asset, e.g., a
vessel; the salvage value is what the fisherman
(or industry) could derive from the asset if
it were sold rather than used. For individual
fishermen, the difference between the two
prices might be small if the quality of assets
is assumed constant. For the salmon industry
or even a particular segment of the industry,
the margin might be substantial. The more
specialized the gear or vessel, the less one
might expect to derive from selling it to an-
other segment of the industry.
If there is a large difference between ac-
quisition price and salvage value, then, it
would be possible for no change to occur in
the aggregate level of a resource even if there
were significant changes in factor productivity
or product price. Figure 3 illustrates the
variety of adjustments that could conceivably
take place, depending upon (a) the starting
point, (b) the magnitude of the shift in the
MVP function, and (c) the divergence between
acquisition price and salvage value. In the
absence of specific knowledge about these
factors, the notion of symmetry between exit
and entry in the salmon fishery becomes an
empirical question. Fixed asset theory, how-
ever, does provide a conceptual framework for
specifying a statistical model and interpreting
the results.
Empirical Analysis
In that we had access only to secondary
data (U.S. Department of the Interior, 1947-
67), most of the variables in the analysis were
141
Aggregate Input Level
Starting
Point
Parameter
Change
Given
1
MVPQ ^
MVP
PA, PS
0' 0
MVP -»
MVP„
0
pA pS
0 0
2
MVP -»
MVP
„A „S
V po
MVPo-
MVP
PA, Ps
0' 0
3
MVP0-»
MVP
„A S
po- po
MVP -T
MVP
PA, Ps
0' 0
implies :
Entry
1-T 2
3^ 2
3-*-2
2 J»3
2-, 3
(no change)
Figure 3. — Expected factor adjustments, given alternative
assumptions on key parameters.
proxy variables. Additionally, the quality of
Bureau of Commerical Fisheries historical
data on resource levels in specific fisheries is
far from perfect. A major data limitation of
this study was that it was not possible to
separate full-time from part-time commercial
fishermen.
The secondary data precluded any meaning-
ful estimation of marginal factor productivities.
Also, reliable data on factor acquisition prices
or salvage were not available. The first prob-
lem was bypassed by means of three assump-
tions; the second was resolved by the choice
of units of observations. Both require some
explanation.
First, it was assumed that the demand for
salmon is price-elastic at the ex-vessel level.
Some support for this assumption comes from
two studies conducted at Oregon State Uni-
versity under the supervision of Dr. R. S.
Johnston (Charoenkul, 1970 and Wood, 1970).
Second, it was assumed that for the salmon
fishery as a whole, the supply of factors is es-
sentially fixed prior to the fishing season. The
direct implication of these two assumptions
is that increases (decreases) in landings bring
about increases (decreases) in average short-
run rents and/or profits over the industry.
The third assumption was that actual rents
in the year t equal expected rents in the year
t + 1, ceteris paribus. The expectation of cyclical
fish runs should be accounted for empirically,
since ceteris paribus is not a realistic assump-
tion in areas with pink and sockeye salmon.
142
These assumptions, in context with the earlier
discussion of fixed asset theory, imply a statis-
tical model wherein changes in resource quant-
ity (labor or vessels) are regressed on changes
in salmon landings, lagged by one year. Ideally,
the influence on resource use levels of acquisi-
tion prices and salvage values of the produc-
tive factors should also be taken into account.
In that these data were not available, the units
of observation were defined both cross-section-
ally and over time. Specifically, yearly data
between 1957 and 1966 for each of ten NMFS
statistical regions on the Pacific Coast were
used.12 This yielded a total of 80 observations
and allowed us to take into account, in a rough,
implicit fashion, cross-sectional differences
which might give rise to a variety of deviations
between acquisition prices and salvage values.
The statistical model is as follows:
(1) X (, + | ) = f [L{t), C(f+, ), £/(,+ d, D]
where
X' = index of fishing effort (number
of fishermen and net tonnage
of fishing vessels),
L = index of salmon landings
(pounds),
C = cyclical nature of the fishery
(dummy variable: 1 for all ex-
pected good runs, whether or
not they actually materialized,
and 0 for all expected poor runs),
U = unemployment rate of the civil-
ian labor force in the major
labor market,
D = distance from the center of sal-
mon fishing activity in the
region to the nearest major
labor market.
In order to test for symmetry between exit
and entry relationships with this model, the
12 The regions were Southeastern, Central, and West
ern Alaska; Puget Sound and Coastal in Washington;
Columbia River in Washington and Oregon; Coastal
Oregon; and Northern, San Francisco and Monterey
in California (U.S. Department of the Interior, 1947-
1967).
80 observations were divided into two subsets.
One subset, with 42 observations, consisted of
those years in which landings had increased
over the preceding year. Given our assump-
tions of an elastic product demand, fixed factor
supply (in the short run), and rent expecta-
tions, it follows that these observations repre-
sent years in which the MVP schedule of factors
had shifted to the right and was expected to
remain there, ceteris paribus. Similarly, the
35 observations13 in the second subset repre-
sented years in which MVP had shifted left-
ward.
Fixed asset theory would suggest that aggre-
gate factor levels in an industry would either
increase or remain constant following years
of increased landings, and would either de-
crease or remain constant following years of
reduced landings. Table 5 indicates a definite
asymmetry between entry and exit relation-
ships. For example, in years of increased land-
ings, the index of vessel inputs in year t + 1
increased 0.32 per unit increase in the land-
ings index for year t. The coefficient for years
of decreased landings was very slightly nega-
tive, but not significantly different from zero.
Asymmetry is strongly suggested by the fact
that the B\ coefficients for the two subsets are
significantly different from each other in both
the labor and vessel equations.
This ratchet mechanism is illustrated in
Figure 4. Net entry follows years of "good
catches," but net exit does not occur following
the "bad years". This is not hard to imagine
for specialized trolling vessels which may have
low salvage values outside of fishing or even
in other segments of the salmon fishery. It
is somewhat more difficult to rationalize, on
the other hand, for the labor resource, although
the human resource would no doubt be af-
fected by lack of mobility of the capital
resource.
The relationships of resource use levels to
the other variables in the analysis are also
of interest, and are summarized here:
(1) Expectations of cyclical runs in encour-
aging entry were more important follow-
ing years of declining landings than fol-
lowing years of increased landings. This
13 Three observations were not usable due to lack
of a "bench mark" year.
143
Table 5. — Regression analysis of factors affecting resource use.1
Dependent variable
B
B
(LJ
B
Vm>
B
Wt+i> <D>
R2
All years:
Labor
Vessels
Years of Increased Landings:
Labor
Vessels
Years of Decreased Landings:
Labor
Vessels
85.88
89.95
77.33
84.43
102.90
99.39
+0.19
+ 6.48
-1.00
+0.009
(3.40)
(1.15)
(-0.68)
(0.51)
+0.19
+ 6.25
-1.89
+0.02
(3.25)
(1.06)
(-1.22)
(0.93)
+0.31
-2.41
-2.51
+0.04
(3.26)
(-0.26)
(-1.17)
(1.24)
+0.32
-6.41
-4.46
+0.05
(3.26)
(-0.68)
(-1.99)
(1.95)
+0.03
+ 10.77
-1.37
+0.005
(0.43)
(1.62)
(-0.64)
(0.25)
-0.002
+ 1.15
-0.19
-0.005
(-0.003)
(1.89)
(-0.09)
(-0.25)
0.14
0.13
0.24
0.28
0.08
0.13
42
42
35
35
Variables are as defined in text. Parentheses contain "^-values" of the regression coefficients.
t+1 Resources
(E = expected MVP in year t
A = actual MVP in year t_ )
Figure 4. — Asymmetry between entry and exit of
resources.
may be somewhat spurious due to the
2-year cycle of pink salmon.
(2) Increased unemployment rates in major
labor markets reduced entry into the
fishery, especially in years of increased
landings when the incentive to enter
would have been highest.
(3) Increased distance from major labor mar-
kets had a positive relationship to the
index of resource use, and was relatively
more significant in years of increased
landings. In retrospect, both distance and
unemployment rates might contribute
more to an explanation of the B\ coeffi-
cient which related resource levels to
landings r- dXr+ 1 i if these coefficients
LB| ~ bLt
could be estimated for each district, rather
than the overall fishery. Our data did not
permit this to be done.
Policy Implications
Although this analysis was fairly superficial
because of the reliance on secondary data, it
did indicate that entry of resources is systemati-
cally related to profit expectations based on an
increasing level of aggregate landings. The
same may be said for exit from the fishery if
"systematic" is interpreted in terms of consis-
tency with fixed asset theory. The empirical
144
values by which entry and exit are systemati-
cally related to profit expectations, however,
differ markedly.
This policy implication for augmenting sal-
mon stocks through hatchery programs is ap-
parent; it is evidently easier to induce re-
sources into the fishery than to induce them
to leave. If the real social objective of hatchery
programs relates to improving incomes in the
fishery, rather than producing and catching
fish, research and action programs designed
to increase salvage values of labor and capital
resources would seem to be of a high priority.
COOLEY, RICHARD A., 1963. Politics and Conser-
vation, The Decline of the Alaska Salmon. Harper
and Row.
CRUTCHFIELD, J. A. and G. PONTECORVO, 1969.
The Pacific Salmon Fisheries, The Johns Hopkins Press.
JOHNSON, G. L., 1958. Supply Function: Some Facts
and Notions. IN: Agricultural Adjustment Problems
in a Growing Economy. Iowa State College Press.
MATTOX, BRUCE W., 1970. A Partial Economic
Analysis of Hatchery Propagation and Commerical
Harvest of Salmonid Resources in Oregon. Unpublish-
ed Ph.D. Dissertation. Oregon State University.
LITERATURE CITED
AUNGURARAT, PEERARAT, 1970. An Analysis of
Factors Affecting Resource Usage in the Pacific Coast
Salmon Fishery. Unpublished M.S. Thesis. Oregon
State University.
McNEIL, WILLIAM J., 1969. Survival of Pink and
Chum Salmon Eggs and Alevins. IN: T. G. Northcote.
Symposium on Salmon and Trout in Streams: H. R.
MacMillan lectures in Fisheries. The University of
British Columbia.
PALOHEIMO, J. E. and L. M. DICKIE, 1966. Food
and Growth of Fishes II. Effects of Food and Tempera-
ture on the Relation Between Metabolism and Body
Weight. Journal of the Fisheries Research Board of
Canada.
BROMLEY, D. W., 1969. Economic Efficiency in Com-
mon Property Natural Resource Use; A Case Study
of the Ocean Fishery. Working Paper No. 28, Division
of Economic Research, National Marine Fisheries
Service, U.S. Department of Commerce.
CHAROENKUL, VILAILUCK, 1970. Analysis of De-
mand for Canned Pink Salmon. Unpublished M.S.
Thesis. Oregon State University.
CHRISTY, F. M. and A. SCOTT, 1965. The Common
Wealth in Ocean Fisheries. The Johns Hopkins Press.
U.S. Department of the Interior, 1947-1967. Fish and
Wildlife Service. Bureau of Commercial Fisheries.
Fishery Statistics of the United States.
WAHLE, ROY J., 1970. Salmon Hatcheries, An En-
couraging Supplement to a Pacific Coast Fishery.
Paper presented to the Western Division of the Ameri-
can Fisheries Society, Victoria, British Columbia.
WOOD, WILLIAM R., 1970. A Demand Analysis of
Processed Salmon from the West Coast. Unpublished
M.S. Thesis. Oregon State University.
145
Limited Entry : The Case of the Japanese Tuna Fishery
E. A. Keen1
ABSTRACT
Limited entry has been advocated strongly as an important but as yet usused man-
agement tool for U.S. fisheries. Japan has maintained a policy of limiting entry into
its high seas fisheries since 1949 and thus has considerable experience of potential
value to the use of this tool in U.S. fisheries. This paper presents an assessment of the
limited entry system as it has been developed for the Japanese tuna fisheries. At-
tention is given to effects on the acquistion of capital and overall allocation of national
resources, specific effects on the size and nature of the fleet, pressures to permit ad-
ditional entry, and effects on the location of shore-based activities. Special attention
is given to problems that were unforeseen at the time of the initiation of limited entry
that, with experience, could have been avoided. The paper is based largely on field
research conducted in 1963 and 1964.
INTRODUCTION
Limitation of the number of craft in a fishery
has been advocated strongly as a management
tool for American fisheries. The volume of
literature in which its usefulness is analyzed,
primarily by economists, has become substan-
tial and continues to grow. A brief survey of
work by Crutchfield, Scott, Christy and others
readily convinces the reader that economic
benefits to be gained through its use more
than justify its advocates. In the case of the
extremely crowded northeastern Pacific salmon
fishery, limitation of entry appears to be al-
most mandatory if rational management only
for maximum sustained yield from the phys-
ical stocks is to be attained. Whether one is
concerned with maximum sustained yield or
with maximum economic return, limitation
of entry obviously is a powerful tool and one
that deserves greater use.
As with all powerful tools, implementation
and operation of a limited entry system just
as obviously is not an easy matter. Fisheries
cannot be considered apart from the highly
complex human and physical systems with
which they are intertwined. Foreseeing all
effects of a major change in regulatory inputs
is extremely difficult. Decisions once made
and institutionalized are equally difficult to
change. In light of the complexity of fisheries
and of the difficulty with which mistakes can
be corrected, it behooves those who would
design and implement a system of limited entry
to take advantage of actual experience in other
fisheries to the extent possible.
The purpose of this paper is to explore the
experience of the Japanese with limitation of
entry into one of their major fisheries, the
skipjack-tuna fishery.2 Much of this experience
is, of course, specific to this fishery and is
therefore, only indirectly relevant to other
fisheries in Japan or elsewhere. Many of the
problems grew out of the needs of a rapidly
expanding fishery, a condition that is not
likely to occur too frequently in the future.
However, some generalizations can be drawn
from it that can be of use in management of
a number of fisheries. A brief summary of the
initiation and development of the regulatory
system is presented first to show the complex-
ity of its development. This provides back-
ground for a discussion of the major effects,
favorable and unfavorable, that concludes the
paper.
1 Associate Professor of Geography, California State
University, San Diego.
2 The term "skipjack-tuna fishery" is a direct trans-
lation of the Japanese term "Katsuo-maguro gyogyo."
All species of tuna are sought by those in the fishery,
not the skipjack alone as the translation might apply.
146
INITIATION OF THE
REGULATORY SYSTEM
Basic aspects of the system of limited entry
were set by a series of administrative ordi-
nances and laws passed during the Allied
Occupation of Japan. An administrative order
issued in July 1946 required registration of
all skipjack-tuna craft over 20 gross tons in
size as an aid to limit the operation of these
craft to areas designated by the Occupation
Government.3 An ordinance issued by the
Fisheries Agency in July 1947 brought these
craft under a formal licensing system and
forbade the construction of additional craft.
Licenses were issued to all owners of craft
over 20 tons for the gross tonnage of their
existing craft. An ordinance, issued in May
1949, regularized the licensing system, made
provision for building larger craft by combin-
ation of the licensed tonnage of two or more
craft, and limited the activities of craft en-
gaged in the skipjack-tuna fishery on a seasonal
basis. The essence of these ordinances were
all codified into a new basic fisheries law
passed by the National Diet in November 1949.
An important additional measure included in
the new law was that licenses, while issued
for periods of 5 years, had to be reissued to
the original holder or his heirs except in cases
of serious infraction of laws on the part of
the holder. It also created a new category of
fisheries, called Designated Distant Sea Fish-
eries, into which all skipjack-tuna craft of over
100 tons in size were placed. A separate fish-
eries protection law passed by the Diet in 1950
set a limit of 300 skipjack-tuna vessels in the
Designated Deep Sea category.
Conditions were favorable to establishment
of the system during the few years over which
it evolved. The administrative order and the
basic regulatory law were established at a
time when profits from the fishery were low
or nonexistent. In the first years of the Occu-
pation, the Japanese were anything but prone
to resist rules issued in the name of the con-
quering powers. The fleet had been heavily
decimated during the war but recovery, with
encouragement of the Occupation Government,
3 An excellent treatment of the regulatory system
as it developed up to 1962 appears in Masuda (1963).
All tonnage figures used herein refer to metric tons.
was rapid afterward. By the end of 1947, the
fleet had recovered to its approximate prewar
size and was more than adequate to harvest
resources within the area enclosed by the so-
called MacArthur Line.4 Catch per unit of
effort had fallen off rapidly with the increase
in numbers of craft and little opposition was
expressed to institution of the regulatory sys-
tem. Those who already owned craft in the
fishery, of course, stood to profit by limita-
tion of entry and supported it. The low rates
of return of the fishery discouraged outsiders
from protesting because entry was forbidden
to them. The system imposed no onerous re-
strictions on fishing effort, such as closed
seasons or closed areas within the fishing
grounds available to the fleet. It appears to
have been accepted fairly readily by the fishing
community and functioned without change
until near the end of the Occupation in April
1952.
Several factors were put forth to support
imposition of the system during its develop-
ment. However, the main motivations for estab-
lishment of the limited entry system centered
on conditions in the fishery at the time, not
on the condition of the resource. That is to
say, conservation or management of the re-
source was not a real issue. It was an issue
and an important one in controlling entry
into the East China Sea trawl fishery which
was placed under a limited entry system at
the same time as the skipjack-tuna fishery.
Concern growing out of the serious overfishing
by the East China Sea fleet undoubtedly in-
fluenced the lawmakers in their decision to
bring the skipjack-tuna fleet under control
and to limit the number of vessels over 100
tons to 300. However, the skipjack-tuna fleet
exploited species that migrated over great
distances and showed no signs of depletion
from year to year because of overfishing in
waters off Japan used by the fleet. Sufficient
fish might not be available to support the
fleet during that part of their migration that
made them available to the Japanese fleet,
4 The MacArthur Line, as the line bounding the area
open to Japanese fisheries that was established by the
Occupation Government came to be known, originally
included only the waters within 12 miles of Japan.
However, it was gradually expanded eastward and south-
ward and by 1950, included most of the traditional
Japanese skipjack and tuna ground in the northwest
quadrant of the Pacific.
147
but little evidence existed to suggest that re-
duction of the stocks in any one year serious-
ly reduced the runs the following year. Thus,
the main reasons were to prevent overcrowd-
ing and conflict on the fishing grounds and
to maintain economic viability of the individual
fishing enterprise. This latter reason was to
become clearly the overwhelming one in sub-
sequent years.
DEVELOPMENT AFTER THE
OCCUPATION PERIOD
If the system was accepted and proved ade-
quate as it stood during the first years of its
effect, it patently was going to require modi-
fication after Japan regained full sovereignty.
As stated above, the fleet, both in reference
to numbers and size of craft, was more than
adequate to harvest resources in the area to
which it had been restricted by the Occupation
Government. However, Japanese tuna fisher-
men had begun to open up tuna grounds in
the west central Pacific and East Indies waters
prior to World War II. Catch rates had been
high, the resource was known to be large and
many were anxious to return to these grounds
denied them during the Occupation. To do
so, larger vessels were desirable; the resource
could support a larger fleet than existed in
1952. Pressures developed to permit expan-
sion of the fleet — internal pressure from
existing license holders to build larger vessels,
external pressure from nonlicense holders for
permission to enter the fishery.
The following decade was marked by con-
tinual modification of the regulatory system
as the fishery expanded beyond the most san-
guine anticipations of anyone connected with
it in the early 1900's (see Figure 1). The 1949
fishery law was explicit as to the number of
craft that could be licensed, the 1950 law as
to the number that could be larger than 100
tons. The upper limit of 300 craft over 100
tons in size had already been approached. The
only expansion possible without a new law
from the National Diet was of tonnage within
the framework of the existing law. Subsequent
laws and administrative orders based on them
were numerous and increasingly complex. No
attempt will be made to treat all of these in
detail; to do so would become extremely tedious.
However, the first two are covered in some
detail to show the pattern set for expansion
of the fleet.
THOUSAND
TONS
800
700
600
/ Total \
/
500
^ — '
400
/
/
s
300
y'
--"
"* y Long Line
200
/ —
/
/
-""""" /\
^>
,..--'
_^_, . . ^^ '
100
— Pole and Line
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
51 52 53 5U 55 56 57 58 59 60 61 62 63 6U 65 66 67
Figure 1. — Landings of tuna and other species by skipjack pole-and line
craft and by tuna longliners, 1951-67. Data for 1951-1961 from Masuda
(1962, p. 361), and for 1962-1967 from Japanese Tuna Fisheries Federa-
tion (1968 and 1969).
148
The first measure for expansion was con-
tained in an administrative order from the
Fisheries Agency issued in March 1952. This
order permitted enlargement of vessels by a
combination of free, additional, licensed tonnage
and licensed tonnage from decommissioned
existing craft. The owner of a Designated Dis-
tant Seas craft, i.e., one over 100 tons in size,
could build a vessel 40 tons larger than the
existing one without withdrawing additional
tonnage from another license. If the new vessel
were between 40 and 100 tons larger than the
original, a 50-ton vessel had to be withdrawn
from the fleet; if a new vessel 100-200 tons
larger than the original were desired, two
50-ton or one 50- to 100-ton vessel had to be
withdrawn. A similar system was set up for
the "medium-sized" vessels as vessels in the
20- to 100-ton category had come to be called.
The legal requirement that these craft be less
than 100 tons cramped measures to enlarge
them but a graduated system of free and de-
commissioned tonnage was instituted. Any
vessel could be enlarged up to 10 tons with
no restriction but half of any enlargement
over this had to come from vessels withdrawn
from the fleet. Any permitted enlargement as-
sumed, of course, that the new vessel was to
be less than 100 tons in size. This technique
of granting limited free tonnage, to be com-
bined with tonnage withdrawn from other
vessels, became integral to the regulatory sys-
tem during the ensuing decade.
The March 1952 measure was inadequate
to meet pressures for enlargment of vessels
in the existing fleet and did nothing to meet
pressure to permit additional entry. This latter
pressure was especially strong from fishermen
in the offshore trawl fisheries, the resources
for which were judged to be exploited excessive-
ly. The expanding tuna fishery appeared to
offer an opportunity for relief for these fisheries.
The apparent need for additional tuna vessels
could be met by permitting transfer to the
tuna fishery.
These conditions led rather rapidly to modi-
fication of aspects of the 1949 fisheries law
that related to the fishing power of the tuna
fleet. The National Diet passed a law that be-
came effective in July 1953 and that, for two
years, set aside aspects of the 1949 laws that
limited the size and number of vessels in the
fleet. Under the new law, known as the Ex-
ceptional Measures Law, craft already in the
fleet were divided into four size categories
based on their size as of December 1952. Li-
censed craft between 20 to 70 tons were per-
mitted to go to 100 tons, those between 70
and 95 tons to 135 tons, those between 95 and
100 tons to 150 tons, and those over 100 tons
to enlarge with no limitations. Owners of
licenses for the medium-sized craft complained
strongly that the permitted increases were not
adequate. In April 1954, the upper limits for 70-
to 90-ton craft and for 90- to 100-ton craft were
rasied to 160 and 180 tons respectively. The
2-year moratorium, however, was not extended
beyond its original July 1955 termination date.
Pressure for additional entry was also vented
somewhat by the 2-year law. Originally, it
permitted issuance of 100 full-time and 240
part-time skipjack tuna licenses. This aspect,
too, was revised further in April 1954. New
licenses were granted for 120 skipjack-tuna
craft up to 85 tons in size, for 10 craft be-
tween 85 and 100 tons in size, and for 150
part-time licenses of less than 85 tons. These
licenses were granted to craft owners in cer-
tain fisheries deemed to be overcrowded, pri-
marily the offshore trawl and purse-seine
fisheries. Recipients in all cases had to agree
to give up their right to fish in their original
fishery and to withdraw their craft from it.
The Exceptional Measures Law resulted in
a much larger and greatly changed fleet. Be-
tween December 1953 and December 1955,
the number of licensed craft increased from
1,154 to 1,372 or 19% ; gross tonnage increased
from 112,945 tons to 176,026 tons or 57% ;
and craft over 100 tons in size increased from
290 to 621 (Masuda, 1963, p. 354). The 1950
limitation to 300 craft of over 100 tons had
obviously been abandoned.
Fundamental changes had also taken place
in the nature of many of the craft. If defined
by fishing method, the skipjack-tuna fishery
is actually two fisheries, the skipjack live bait
pole-and-line fishery and the tuna longline
fishery. Historically, the pole-and-line fishery
is the older of the two. It developed to exploit
the large runs of skipjack and to a lesser
extent, albacore, that appear off Japan during
the spring and summer months. The longline
149
fishery developed as on offseason activity for
craft in the former and remained subordinate
to it until the end of the Allied Occupation.
Equipment and crew requirements for the
two bear little similarity. The maximum sized
craft that could be used efficiently in the pole-
and-line fishery was about 150 tons at the
time.5 Live bait wells are an absolute essential
for the pole-and-line fishery but are unneces-
sary for the longline fishery. Crew size for
the former is usually a little more than double
that needed for the longline fishery with con-
sequent additional space required for quarters.
The world market for tuna grew rapidly after
World War II and tuna soon provided a higher
return than did skipjack. Larger craft could
operate year round on the new longline grounds
being opened up in the southern Pacific and
Indian Oceans. As a consequence, most of the
craft built when the Exceptional Measures
Law was in effect and afterward were special-
ized vessels for the longline fishery only. Lack
of a live bait well alone effectively denied
their use in the pole-and-line fishery.
Landings of the fishery increased propor-
tionately along with the tonnage of the fleet.
Tuna longliners landed 117,000 tons in 1952;
in 1955 this had increased to 197,000 tons
(Japanese Tuna Fisheries Federation, 1961,
p. 16). The value of the landings fell rapidly;
the average price of yellowfin tuna at Yaezu,
Japan's most important tuna port, dropped
from $289 per ton in 1953 to $192 in 1955
(Yaezu Fishery Cooperative, 1963, p. 25). H
Lingering effects of the Bikini nuclear weapon
incident of 1954 that had greatly reduced de-
mand for fresh tuna in Japan accounts in
part for the lower price. However, the main
reason was excessive supply. The world mar-
ket for tuna, limited at the time largely to
Japan and the United States, was not able to
absorb the added catch at the 1953 price levels.
The Fisheries Agency policy with the end
of the Exceptional Measures Law called for
5 A vessel of about 150 tons is the minimum sized
vessel needed to operate from Japan on the west-central
Pacific grounds to which the pole-and-line fishery ex-
panded in the mid-1960's. In 1967, forty-one vessels in
the 200-500 ton category were used in the newly de-
veloped distant seas pole-and-line fishery (Japanese
Tuna Fisheries Federation, 1969, p. 13).
H Conversions from yen to dollars was made at the
rate of 360 to 1.
absolute restrictions on new entry. However,
it did continue the policy of permitting and
encouraging enlargement of craft. In a few
cases, slight enlargements were permitted
without abolishment of licensed craft. The
heart of the policy, however, was to permit
use of licensed tonnage for medium-sized
vessels for combination with other licenses to
build larger craft. The net effect of this was
to reduce the total number of craft but to in-
crease the number of larger craft for operation
on distant grounds. The rapid increase in
vessels over 200 tons at the expense of those
under that size is shown graphically in Figure
2. The total number of licensed craft decreased
from 1,380 in 1956 to 1,243 in 1957.
Landings continued to grow at about 50,000
tons annually into the early 1960's. The mar-
ket also began to recover after the lows of
1955 and prices began a steady upward trend.
By 1962, the average price of yellowfin at
Yaezu had risen to $328. Small fortunes were
being made by the end of the decade. It be-
came apparent that craft of at least 250 tons
in size were needed to operate efficiently from
Japan on the south Pacific and Indian Ocean
grounds as well as from bases on the newly
opened Atlantic grounds. The value of licenses
for supplementary tonnage increased rapidly.
Supplementary tonnage could be purchased
for about $100 per ton in 1955, rose to about
$500 in 1959, and in 1960 approached $1,000
per ton (Masuda, 1963, p. 556). 7 In 1960, ad-
ditional free tonnage was permitted for craft
of less than 240 tons in size if they were
wooden craft over 6 years old or steel craft
over 12 years. Also, restrictions on the use
of the licenses for the less than 100-ton vessels
issued after 1953 as supplementary tonnage,
licenses that previously could not be used for
this purpose, were relaxed. Another building
boom was underway and the average size of
the vessels in the fleet grew with it (see
Figure 3).
i
Pressure for additional entry into the tuna
fishery, never quiescent, began to rise marked-
ly with the rise in profits from the fishery.
Pressure was especially strong after 1956 from
7 Precise figures on sale value of licenses are difficult
to obtain since profits from their sale is subject to capital
gains tax. Underreporting to avoid taxes appears to have
been the rule.
150
900
800
700
'Number 600
of
Craft
500
400
300
200
100
40-100 Tons
100-200 Tons
""- /
=5^
"7
s
>^ /
over 500 Tons
1951
i i i i i — i — i — i — i — i — i — i — i — r
1955 1960 1965
Figure 2. — Trends in numbers of licensed distant sea skipjack-tuna craft by size
category. Data: (Japanese Tuna Fisheries Federation, 1969, p. 6).
325
Figure 3. — Annual construction of skipjack-tuna craft over fifty gross
tons in size. Data for 1951-52 from (Masuda, 1963, p. 542), for 1963-67
from (Japanese Tuna Fisheries Federation, 1969, p. 9).
151
the salmon fishery as a result of restrictions
on that fishery growing out of the USSR-Japan
agreement concerning it. An attempt to
relieve this pressure was made in June 1957
by raising the lower limit for licensed skipjack-
tuna vessels from 20 to 40 tons. The result
was the almost instantaneous creation of a
39.9-ton tuna vessel fleet.8 A fairly large num-
ber of "39-tonners" were built by owners in
the traditional salmon ports of northern Japan
but a majority of these new "free entry" vessels
appeared in the traditional skipjack-tuna
ports. The measure thus did provide some
relief for the depressed salmon and other fish-
eries but the main effect appeared to be in-
creased investment by those already in the
skipjack-tuna fishery. Pressure from the salmon
fishermen continued and some fifty new
"medium-sized" tuna licenses were given
craft owners in this fishery between 1960 and
1962 in exchange for their abandonment of
the salmon fishery.
A demand to permit increased use of mother-
ships also began to develop in the late 1950's.
Large motherships operating with independent
licensed tuna vessels had been authorized
since 1948. Fairly stringent restrictions had
been placed on the annual catch and on place
of fishing of those "independent vessel mother-
ships" as they came to be called.9 However, in
the late 1950's, the larger tuna longline
vessels began to carry "portable catcher boats"
on board. Once on the fishing ground, these
catcher boats proved almost as efficient in
H Accurate records were not kept on the number of
such craft until a centralized licensing system was estab-
lished in 1964. However, one study by Fisheries Agency
personnel in which an attempt was made to trace the
growth of this fleet showed only three such craft were
launched in 1957, 23 in 1958, 117 in 1959, and 194 in
1960 (Japanese Fisheries Agency, May 8, 1963, p. 6).
No data are available on the number of salmon longline
craft under 40 tons that switched to tuna longlining but
the number probably was substantial.
H Motherships were limited in place of operation to
designated areas in the central and southern parts of
the Pacific and always under a catch quota system. The
maximum number of motherships used in any one year
was six, each with up to 50 independently licensed tuna
long-liners. In the early 1950's, Antarctic whaling mother-
ships were used as tuna longline motherships in the
offseason. However, salmon motherships came to be
used with restrictions on that fishery imposed by the
Japanese-Soviet agreement in 1956. Each mothership
fleet was granted a maximum catch quota before leaving
port. The total quota for all mothership fleets reached
a high of 28,000 tons in 1958.
terms of catch rates per day as the independent
vessels. A new category of licensing was estab-
lished for these craft in April 1961 and re-
vised in September 1962. Two classes of these
"catcher boat carrying motherships," as they
came to be called, were created — less than
2,000 ton craft where the mothership was per-
mitted to fish, and over 2,000 ton craft where
the mothership was not permitted to fish. A
complex system of computing licensed tonnage
was established for the catcher boats. In gen-
eral, it required that regular licensed craft
be decommissioned in considerable larger ton-
nage for the catcher boat than the maximum
size of 20 tons established for each skiff. Re-
strictions were also placed on area of operation
of these two new classes of motherships. Regu-
lations as to place of operation were designed
generally to limit them to the southwestern
Pacific, Indian, and Atlantic Oceans.
The regulatory system had become some-
what outmoded and unwieldy by the early
1960's. The basic fisheries law was inadequate
for proper regulation of the new motherships
and the need for regulation of the new "39-
ton" fleet was becoming apparent. The former
medium-sized vessels that had been allowed
to expand to over 100 tons but held below 200
tons in size, about 150 in number, were proving
to be uneconomical. Not large enough to
operate effectively on grounds south of the
equator, they were too large to compete ef-
fectively with the large number of "39-ton"
"free-entry" craft and less than 100-ton li-
censed craft on grounds adjacent to Japan.
The price of licenses continued to rise to a
peak of about $1,200 per vessel ton in 1962.
Few owners of these "in between" craft could
afford to purchase supplementary tonnage for
craft enlargement at these prices. For these
and other reasons, the realization became
general that a new legal framework for ad-
ministration of the fishery was needed, a con-
dition that was true of other fisheries as well.
A revision of the basic fisheries law by the
National Diet in August 1962 provided a new
framework. In reference to the tuna fleet, the
new law codified the system for motherships
described above, rationalized a number of com-
plexities that had developed in the licensing
system, and lowered the age at which a vessel
could be replaced to 4 for wooden vessels and
8 for steel vessels. The only aspect of the new
152
law that specifically permitted additional ton-
nage to the fleet concerned the "in between"
craft between 100 and 180 tons. These were
granted permission to enlarge to 240 tons,
about the smallest sized vessel that could
operate effectively south of the equator from
Japanese ports.
Landings from the longline fishery peaked
in 1962. Declines in catches from that year,
increased competition in international markets
from the Taiwanese and Korean fisheries, and
sudden rises in labor costs greatly reduced
pressure for further expansion of the fleet. The
"39-ton" fleet was brought back into the
limited entry system in 1964 with a passage
of a law that established a "near seas" skipjack-
tuna industry. The law limited the number of
licenses for 20- to 50-ton craft engaged in the
skipjack-tuna fishery to 1,850 vessels, a number
selected primarily because it was sufficiently
large to cover all craft of this size range al-
ready in the fishery. In 1964, 1,708 craft were
licensed and registered under this law but
the number has declined slightly since.
Changes in the regulatory system since the
near seas fleet was established have been
relatively few in number compared to earlier
years. As longline catches declined, the pole-
and-line live bait fishery received increased
attention. The more substantial changes in
regulations have been designed to permit or
encourage decommissioning of large vessels
to build smaller vessels for this fishery. Strong
pressure has developed since the mid-1960's
for reduction in the size of the fleet. Agreement
appears to be general that this should be done
but as yet an acceptable method to do so has
not been devised.
EFFECTS ON DIFFERENT ASPECTS
OF THE FISHERY
As can been seen from the above overly
simplified description, measures for regula-
tion of the Japanese skipjack-tuna fishery center
strongly on limitation of the size and number
of craft. Only minor use has been made of
catch quotas and restrictions on place of fish-
ing, measures that tend to reduce the efficiency
of use of vessels and equipment. The fleet as
it developed is very much a result of regulation
through use of limited entry and controls on
size of vessels. Discussion will now turn to
the major effects, some obvious and foreseen,
some less obvious and forseen dimly if at all,
that the regulatory system had on the fishery.
Capital Acquisition and Resource Allocation
One of the more striking aspects of the fishery
was the rapidity with which the fleet was ex-
panded after the Allied Occupation ended.
Vessels used in the fishery are not extraordi-
narily large as fishing vessels go nor were
construction costs in Japan high by any stand-
ard. However, they do represent a sizeable
capital investment and requirements for operat-
ing capital are substantial. Owner-operator
enterprises dominated the fishery in the early
days. This meant that most were small enter-
prises headed by individuals with poorly estab-
lished lines to sources of capital. Two- and
three-boat enterprises became common by the
early 1960's but the fishery continues to be
made up largely of small enterprises. The
large fishing corporations of Japan have played
and continue to play a relatively minor role
in the fishery.
The effect the system as applied had on
acquistion of capital is, of course, obvious. Li-
censes from the beginning became, for all
practical purposes, the personal property of
the recipient. As such they were sold, traded,
or used as security for loans. Even at the de-
pressed tuna prices of the mid-1950's, license
values ranged from 10% to 20% of construction
costs for a vessel. At 1962 earning levels, the
value of the license almost equaled that of
the vessel. With security of this nature to
offer, no license holder had any difficulty in
gaining loans for either fixed or operating
capital. Without the limited entry system and
property characteristics of the licenses, the
fishery possibly would have expanded more
slowly, paradoxical though this may sound.
Enlargement of craft also would have been
more dificult had these valuable licenses not
been available to use as security for loans.
One could postulate that the fleet would have
come to consist of a much larger number of
smaller craft without it, although larger craft
constructed and owned by large corporations
may have come to dominate the fishery.
Licenses decreased in value rather precipi-
tously after 1962 to a low of about $330 in
153
1965 (Commercial Fisheries Review, 1966,
p. 73). Rates of indebtedness at the peak of
license values in 1962 had been much higher
than in other Japanese fisheries. Debts on the
fixed capital alone of craft over 200 tons in
1962 averaged 72% , almost an inverse ratio to
the 30% rate in the East China Sea trawl fish-
ery (Masuda, 1963, p. 539). Debts on smaller
licensed vessels averaged over 50% . Improve-
ment in the earning position of tuna vessels
in the late 1960's with the rapid increase in
price of tuna in Japan stabilized the economic
picture for most owners after 1965. However,
many marginal enterprises were forced out
of the fishery during the mid-1960's.
It can also be argued that the licensing sys-
tem as it evolved also led to a misallocation
of resources within the national economy as
a whole. From the standpoint of the national
economy, investment in the tuna fishery ob-
siously was profitable at least through 1962.
However, the high, and at times unrealistic,
value of the licenses in the tuna fishery gave
this fishery an extremely favorable competi-
tive position within financial institutions
specializing in fisheries, and, indeed, in the
national capital market as a whole. The total
investment was substantial and, as proved
later, was larger than needed to harvest the
resource. Where the investment level would
have proved most advantageous is difficult to
determine and no effort to do so is known by
the author. Few would argue, however, that
a better allocation of national resources would
not have been obtained had part of the invest-
ment in the tuna fleet been directed to other
channels.
Size and Nature of the Fleet
That the size and characteristics of craft
in the fleet was shaped strongly by the regu-
latory system is apparent from the earlier dis-
cussion of the development of the system. En-
largement of craft was a basic and continuing
policy throughout the period of expansion.
The most effective measure used to fulfill this
policy was the frequent granting of additional
free licensed tonnage that could only be used
with the licensed tonnage of the old vessel
which was in turn decommissioned. This, and
the practice of allowing only licensed tonnage
from decommissioned "medium-sized" craft
to be used for enlargement under any circum-
stances, hurried the disappearance of these
smaller licensed craft as well as the construc-
tion of larger ones.
The measures used were highly effective as
is shown by the increase in average vessel size
from 91 to 230 gross tons between 1952 and
1962. It also meant that many vessels were
retired well before their useful life was ended.
This wasteful aspect was recognized and an
attempt made to minimize it by placing mini-
mum ages on craft that could be decommis-
sioned. That this time was shortened from 6
to 4 years for wooden vessels and from 12
to 8 years for steel vessels illustrates the pres-
sures applied to take advantage of grants of
tonnage, grants which usually carried a 2-year
maximum for use from the date they were
granted. A recognized shortcoming of the sys-
tem, it was nevertheless one that was never
solved satisfactorily during the period of ex-
pansion.
An unforeseen result, or certainly one that
was predicted poorly, concerned adverse ef-
fects on the structure of individual vessels.
As the fishing grounds became more distant,
a premium was placed on hold capacity for
fuel and fish. Given the absolute limit on gross
tonnage permitted for an individual vessel,
the owners designed around this limit with
emphasis on increased carrying capacity. First
started in the late 1950's, craft with 20% to
30% greater carrying capacity were soon being
built with no increase in computed tonnage
(Masuda, 1963, p. 546). Crew quarters and
below-deck working space became more cramp-
ed in the process and safety equipment was
reduced to the minimum permissible standards
and often stowed in inaccessible places. Sea-
worthiness also often suffered because of re-
arrangement of storage space that decreased
stability, a factor that undoubtedly contributed
to the loss at sea of a number of smaller craft.
Many of these adverse aspects have been cor-
rected subsequently but only through greater
expenditure of administrative time for inspec-
tion, additional tonnage concessions that could
not be used for hold space, and a weakening
of the competitive position of the fishery for
labor because of poor working and living con-
ditions while at sea.
154
Effect on Other Fisheries
One could argue, as was pointed out earlier,
that the superior competitive position of the
tuna fishery possibly had some adverse effects
on other fisheries, primarily in reference to
competition for capital. Comparatively high
returns to labor in the tuna fishery also gave
it a competitive position in this respect. How-
ever, labor was not a major problem for any
fishery prior to the early 1960's and since
labor was generally drawn from families and
acquaintances of vessel owners, the tuna fishery
appears to have had little effect even on the
quality of labor available to other fisheries.
The overall effect on other fisheries, or at
least the administration of them, probably
was positive. Since entry was controlled, re-
lief could selectively be provided fisheries
creating the greatest administrative problems.
Certainly the Minister of Foreign Affairs must
have been happy to see pressure relieved on
the East China Sea and North Pacific Salmon
fisheries in light of the adverse reaction of
mainland China and the Soviet Union to these
fisheries. Had these new licenses for the tuna
fishery been placed on open bid, one could
hardly have expected fishermen from depressed
fisheries to compete for them with any degree
of success.
Effects on other fisheries may be somewhat
nebulous and difficult to define with precision,
but the effect on the live bait pole-and-line
fishery is much clearer. That the two methods,
or fisheries if one wishes, were administered
as a single fishery meant that expansion of
the live bait fishery was neglected for over a
decade. Catches by the live bait method did
not decline during expansion of the longline
fishery, in fact the secular trend was up slight-
ly (see Figure 1). However, resources for this
fishery were underutilized, a fact known at the
time and borne out by the increase in landings
since the mid-1960's. Craft of sufficient size
to properly exploit this resource and permitted
to do so were also the only ones permitted to
fish with longlines for tuna. Given the higher
rate of return on tuna, the choice of a vessel
owner is not difficult to see. That most did
specialize in longlining is shown by the fact
that the number of licensed craft using the
live bait method declined from 737 in 1953
to 231 in 1961; total tonnage of vessels so
used declined from 80,000 tons at the peak to
33,000 tons in 1961 (Masuda, 1963, p. 358 and
546).
That the total catch by the live bait method
continued to be stable throughout expansion
of the tuna longlining can be attributed pri-
marily to unlicensed craft, including the "39-
tonners" after 1957. These craft were sufficient-
ly large to exploit the traditional grounds
adjacent to Japan. However, craft of over 100
tons in size are needed to exploit the large
skipjack resources in more distant southern
waters. By 1960, nearly all craft of this size
had been rebuilt without live bait wells. With
the decline in longline catches, a distant seas
live bait fishery developed fairly rapidly. In
1964, only 138 craft over 100 tons in size
used the live bait method; by 1967, the num-
ber had increased to 224 (Japanese Tuna Fish-
eries Federation, 1969, p. 13). Had craft using
the live bait method been administered sep-
arately, it can be assumed that craft would
have been available to develop these distant
grounds during the 1950's. That this was not
done can be regarded as a loss to the national
economy during the period.
Effects on Location of
Shore-Based Activities in Japan
The regional pattern of economic activities
connected with the fishery changed consider-
ably during the period of rapid expansion.
Fishing ports and the fleet were distributed
fairly evenly between the southern tip of the
island of Kyushu and the northeastern port
of Honshu when the live bait method dominat-
ed the fleet's activities. Most of the fleet would
gather in the south in early spring to pick
up the annual runs of skipjack and to a lesser
extent, albacore, and follow them northward
along the Pacific Coast until they disappeared
in late summer off northeastern Honshu. Land-
ings were made at the nearest port, nearly all
of which had a dried skipjack stick process-
ing industry, the main use for most of the
catch. Craft would then be converted for tuna
longlining on winter tuna grounds adjacent
to Japan. The main market for tuna was in
the Tokyo region and catches from the winter
fishery were landed at ports in that area.
155
As tuna longlining increased in importance
and became a year round activity, one could
easily have predicted that activities would
concentrate in a smaller number of ports.
Grounds for the year round tuna fishery were
so distant from Japan that no port had a
locational advantage of any significance in
reference to the grounds as was the case with
the live bait fishery. The main markets for
tuna were the canneries, export companies,
and the large urban population in the Tokyo
area. As craft became larger, smaller markets
were unable to handle the full load of most
vessels expeditiously, a factor that further
favored concentration. Concentration of eco-
nomic activities of the longliners in a few ports
thus would have been expected quite apart
from the regulatory system.
The regulatory system as applied did, how-
ever, influence the regional pattern signifi-
cantly. Among the more readily apparent in-
fluences perhaps was that it hastened enlarge-
ment of craft and thus increased tendencies
toward concentration in the central ports.
Conversely, in another aspect, it tended to
favor continued dispersion of economic activi-
ties other than landing of the fish. This de-
rived from the fact that ownership of the fleet
was dispersed at the time licenses were issued.
Ties of Japanese fishermen, both economic
and social, to their home port are strong. A
man's boat is his livelihood and sale of the
right to use it is restricted by strong pressures
of tradition. That the value of the license in-
creased steadily during most of the period of
expansion meant that most holders, even in
more remote areas, were able to fund new
craft and expand along with the fishery. With-
out this source of funding, the longliners would
almost certainly have been concentrated in
all respects in the centrally located ports where
capital was more readily available and where
attention to the fishery would have been much
stronger. However, having been given the li-
censes, owners in outlying ports generally
kept pace with the switch to longlining; with-
out the license as security, lack of capital alone
probably would have been a major deterrent
to so doing. Landing and most resupplying
of vessels might be carried out in centrally
located ports such as Yaezu, Misaki, or Tokyo
but the economic stimulation from other activi-
ties such as management, labor recruitment,
and expenditures by management and labor
largely accrued to the ports where the owner
of the license resided. As such, the fishery con-
tinued to contribute to regional economies to
a larger extent than if the regulatory system
had not existed. Thus, the net effect of the
regulatory system appears to have been a
conservative one working against an expected
tendency toward concentration in the major
market ports.
Flow of Capital to Other Countries
A predictable effect of a limited entry system
in a profitable fishery such as the tuna fishery
in which overall control of entry to the fishing
grounds is impossible would be a flow of
capital to other countries. This was recognized
early in the period of expansion and fairly
effective controls were developed to control
it, at least through 1963. The method used
was to restrict export of tuna longliners. The
craft themselves are not particularly complex
nor is the equipment used on them. However,
countries that had the industrial establish-
ment to build them, by and large were not
able to compete with the Japanese in the
fishery because of labor costs. Countries that
desired to enter the fishery and were in a favor-
able competitive position in reference to labor
costs were not able to build the vessels. Given
these conditions, strict controls on export of
longliners were used to prevent Japanese entre-
preneurs from transferring registration to
other countries and using Japanese or foreign
crews and, at the same time, retard the de-
velopment of the fishery by other countries.
Some transfer of registration was permitted
for operation by joint Japanese and foreign
companies from ports in the country of the
latter. However, conditions under which this
could be done were restricted severely; in a
1965 survey by the Fisheries Agency, only 17
vessels were found to be so operated (Com-
mercial Fisheries Review, 1966, p. 85). Pres-
sures to permit export, especially by shipyard
owners in Japan, were great, but were con-
tained until 1964. By this time, other nations,
especially Korea, were developing a capacity
to build longliners and the restrictions were
relaxed.
156
Japanese capital has played an important
role in the development of foreign fleets since
the early 1960's. Large Japanese trading com-
panies handle most of the tuna exported from
overseas bases, bases originally established
to serve Japanese vessels. As other countries,
namely Taiwan and Korea, began to develop
fleets, they also used these bases and sold
their catches to the Japanese companies. In
return, vessels from these countries have re-
ceived financial assistance, largely operating
capital, from these large companies. A new
base opened recently by a large Japanese com-
pany in Mombasa, Kenya reportedly is to be
used almost entirely by Taiwanese vessels
(U.S. Bureau of Commercial Fisheries, Febru-
ary 24, 1969). However, this Japanese invest-
ment must be attributed primarily to the higher
labor costs of Japanese vessels not to restric-
tions on their number. Under conditions in
the Japanese fishery since the mid-1960's, it
is doubtful that any significant increase of
Japanese vessels operating from these bases
could be expected even if the fishery were
opened to unlimited entry.
CONCLUSION
In retrospect, no one in Japan or elsewhere
would consider the regulatory system develop-
ed for the Japanese skipjack-tuna fishery to
be a complete success. However, few would
argue that the fishery and the country were
not served better by limitation of entry than
they would have been had no controls been
imposed on the number of craft. The system
did have a goodly measure of success in refer-
ence to its main goal, that is, to maintain a
high level of economic viability of enter-
prises in the fishery. Without it, a gross
over-investment in small vessels is almost
certain to have taken place in the early 1950's.
Depression of the market, strained financial
condition of enterprises, and a loss of all
economic rent from the fishery likely would
have occurred long before the resource ap-
proached full exploitation. Conflicts on the
fishing grounds, international incidents, and
disasters at sea also would have been more
numerous. Thus, a second major goal, harmony
within the fleet and on the fishing grounds,
was at least partially achieved. If the system
has been less successful since the early 1960's,
the fault can hardly be laid at the feet of the
fishery policy makers and administrators.
Their control over entry of fishermen of other
countries ended with Japanese ability to con-
trol the technology of the fishery. Had fisher-
men from other nations had the wherewithal
to enter the fishery from 1950, acceptance
of the system by the Japanese fishermen
would have been far more difficult to attain.
Mistakes were made, many of them avoid-
able. Perhaps the largest was to raise the
minimum size of licensed vessels to 40 tons.
That it was done appears to have resulted
from an inadequate assessment of technolog-
ical developments. Less than 40-ton craft in
existence at the time were patently too small
to operate on distant grounds but could re-
lieve the need for more vessels to exploit the
annual runs of skipjack and albacore on near
seas grounds. Vessels of 19.99 tons could
never be designed for effective operation on
distant grounds. However, redesign of vessels
of 39.99 tons led to craft with the fishing power
of a 70-ton vessel designed by standards used
in the mid-1950's. At the catch rates and
prices of tuna in the late 1950's, these vessels
could operate profitably on distant grounds
although the large number of disasters sug-
gest they should not have attempted to do so.
The problem of safety was corrected only by
granting permission to increase size of these
vessels to 50 tons with the provision that the
additional tonnage would be used only to in-
crease crew comfort and safety and limiting
their use to waters adjacent to Japan. How-
ever, the number of such vessels far exceeds
needs and the problem of overcapitalization
has been far more intractable.
Some lawmakers and administrators were
troubled also by the tremendous value that
the licenses came to have at no cost to the
holders of the licenses. Had the tremendous
expansion of the fishery and its profitable-
ness been foreseen at the time the fishery was
brought under regulation, some means pos-
sibly could have been devised to siphon off
at least part of the economic rent represented
by the licenses into the public coffers. How-
ever, to have worked out an acceptable scheme
for the fishery after the basic system was al-
157
ready operating would have been extremely
difficult. Certainly it would have added com-
plexities to an already overly complex struc-
ture that possibly would have caused the
entire system to break down. Also, a national
law that singled out one fishery for such treat-
ment probably would not be acceptable to the
lawmaking body. Values of licenses in more
stable Japanese fisheries have never reached
levels considered to be a problem; to impose
controls on these fisheries would create more
administrative problems than could possibly
be justified by gains resulting from the controls.
In short, to have solved this problem, if it was
one, in the political arena of Japan or any other
country with representative government would
have been extremely difficult. Possibly ignoring
it was the wiser route to follow.
The problem of overcapitalization of the
world tuna fleets appears to be approaching
rapidly if it has not already been reached.
The Japanese were able to limit entry to the
fishery and maintain economic viability of
enterprises in it during the period that they
controlled longline technology. Beyond ques-
tion, limited entry could also be used to
control excessive fishing power and the ex-
cessive pressure on world tuna stocks that
it is certain to bring. The Japanese experience
illustrates many of the problems that would
attend the far more complicated problems
foreseeable in establishment of an international
system. It also suggests the benefits, in refer-
ence to stock management as well as eco-
nomic viability of the fishing enterprise,
could be well worth the effort required to
establish the system.
LITERATURE CITED
Commercial Fisheries Review, January 1966. Vol. 28,
No. 1, p. 85.
Commercial Fisheries Review, July 1966. Vol. 28, No.
7, p. 73.
Japanese Fisheries Agency. May 8, 1963. Katsuo-Maguro
Gyogo, No - 40-ton Munar Gyosen ni Kansuro Shirgo.
p. 6.
Japanese Tuna Fisheries Federation. 1968. Statistics
of the Japanese Tuna Fishery.
. 1969. Statistics of
the Japanese Tuna Fishery.
MASUDA, SHOICHI, ed. 1963. Katsuo-Maguro Soran
(Skipjack tuna Overview), Tokyo: Suisanska, p. 758.
Yaezu Fishery Cooperative. 1963. Mizuage-daka Tokei,
no. 11.
U.S. Bureau of Commercial Fisheries. February 24,
1969. Foreign Fishery Information Release 69-7.
158
A Study of the Socioeconomic Impact of Changes
in the Harvesting Labor Force in the
Maine Lobster Industry1
A. M. Huq2
ABSTRACT
The basic question of the mobility of the labor force in the Maine lobster fishery
is investigated with particular emphasis on the productivity of control groups within a
sample and their social, educational, economic, and demographic characteristics. Under
various assumptions which would lead to exit from the fishery of these groups certain
consequences are enumerated, both with regard to those leaving and those remaining
as well as the impact on and role of the local communities involved. A preliminary
assessment of the impact of certain types of management programs upon the labor
component of the harvesting sector is presented.
INTRODUCTION
In any discussion of alternative manage-
ment strategies (e.g., limited entry) that might
affect the labor force in the lobster fishery
in Maine, it is important to examine the socio-
economic repercussions of the contemplated
change. In some circumstances this may in-
volve the dislocation of labor. In this case
one must, for example, investigate whether
alternative employment would be available
to those fishermen who will be excluded be-
cause of limited entry; their employability
(and trainability) relative to the local labor
market, their geographical and occupational
mobility patterns, the adaptability of their
skills, alternative income earning possibili-
ties ("salvage value" of displaced labor), the
potential for upgrading their existing skills
and for the acquisition of new skills, the
barriers to their mobility including sociolog-
ical, psychological, and economic variables
are some of the crucial elements to be care-
fully considered.
Furthermore, the policy maker has to evalu-
1 This paper is based upon a study sponsored by
the National Marine Fisheries Service. In addition to
the author, the research team consisted of Harland
I. Hasey and Anita Wihry, Research Associates.
2 Director, Manpower Research Project, University
of Maine, Orono, Maine.
ate the potential impact on the local and
regional economy in terms of shifts in income
and employment and associated fiscal conse-
quences including welfare expenditures and
changes in tax revenue. Finally, it would be
important to examine how limited entry in
a given fishery such as the lobster fishery
might affect other fisheries such as shrimp
and scallop fisheries. In a comprehensive study,
all these questions need to be investigated
before any definitive conclusions can be reach-
ed. However, the present study is of much
more limited scope and pertains to only some
of these questions bearing on limited entry.
This study focuses on the possible socio-
economic impact of hypothetical reduction in
the harvesting labor force in the Maine lobster
fishery. As to how this reduction is or can be
brought about is outside the scope of the
study. The study utilizes the data obtained
from a sample survey of 131 fishermen from
three selected communities. The problem posed
for investigation was simply this: if a group
of fishermen from this sample is excluded
from lobster fishing based on some specified
criterion, what sort of socioeconomic impact
can be expected: Can certain indicators be
developed to measure such impact in order to
consider alternative management strategies?
For this purpose, it was considered desirable
to (a) introduce the notion of a target group
composed of fishermen regarded as candidates
for limited entry and (b) to develop alternative
159
criteria for the construction of a set of target
groups rather than singling out one specific
target group.
Constrained by time and resources avail-
able for this project, the study addressed it-
self only to selected dimensions of socioeco-
nomic impacts of limited entry into the Maine
lobster fishery. It is to be clearly understood
that some of the findings of this study, be-
cause of its very limited scope, are essentially
for illustrative purposes rather than for use
as supportive materials for or against any
implicit management strategy that may be
suggested by the format of the target groups.
OBJECTIVES
The major objective of the study is to present
an evaluation of the socioeconomic impacts
of limited entry into the Maine lobster fishery.
A complete evaluation may include but not
be limited to the income and employment
effect on the displaced fishermen, income
effect on the surviving fishermen, income and
fiscal effect on the local and regional economy,
effect on other fisheries and so on. However,
for reasons stated above, the limited objectives
of this study are:
1. To make an appraisal of the employ ability
and alternative income earning possibilities
of displaced labor.
2. To derive some measures of social impact
in terms of (a) income effects and (b) income
maintenance burden associated with dis-
placement because of limited entry.
RESEARCH DESIGN
The study was designed as a small-scale
pilot effort, concentrating on three typical
communities rather than encompassing the
entire Maine lobster fishery. These communi-
ties are Phippsburg, Beals, and Corea. The
selection was made in consultation with the
Maine Department of Sea and Shore Fisheries
and the National Marine Fisheries Service.
The existence of some contrasts in the struc-
ture of the local economy and the relative
importance of the lobster fishery in their econ-
omy weighed heavily in the selection process.
Corea represents a highly specialized, isolated
economy where lobstering is the predominant
economic activity. Beals is also highly special-
ized but less isolated than Corea. Phippsburg's
economy is more diversified and in close prox-
imity to sources of alternative job opportuni-
ties. Each of the areas has one feature in
common: the lobster fishery is a major eco-
nomic activity.
It is difficult to say how representative these
three communities are of the entire lobster
fishery. Sufficient information is not readily
available to identify the economic character-
istics of the population of lobster fishermen
in Maine and relate them to those of the
sample fishermen in these communities.
For the purpose of the study the following
hypotheses were formulated for investigation:
1. Limited entry could eventually exclude
a certain fraction of the lobster harvesting
labor force that will be otherwise unemploy-
able. (Alternative hypothesis: a significant
fraction of labor displaced because of limited
entry will be employable, given the conditions
in the local labor market, the type of skill
possessed, the potential for adapting skills
to job market requirements, the availability
of retraining opportunities, motivation for
training, and mobility and so on).
2. Displacement of labor because of limited
entry may adversely affect the local economy
because of loss of income from lobstering not
being compensated for by income from alterna-
tive jobs and from additional lobstering by
surviving fishermen, and because of loss of
income from lobstering on the part of those
who are not in the labor force.
To generate the information needed for this
investigation, a stratified random sample of
131 fishermen was selected. The size of the
sample depended essentially on the estimated
cost per interview and the budgetary con-
straint. The allocation to each stratum was
strictly according to proportion of fishermen
in each community to the total number of
fishermen of all three communities. The survey
data were supplemented by information on
the local labor market obtained through the
cooperation of the regional offices of the
Maine Employment Security Commission.
For the survey, a structured questionnaire
was developed and pretested. Using the modi-
fied questionnaire and personal interviews,
160
the survey was completed in 6 weeks. The
response rate was better than 90% .
The survey resulted in a large volume of
information on the sampled fishermen. The
following broad categories of information
may be identified:
Categories Types of Information
Demographic Age
Family Size and Composition
Mobility
Marital status
Socioeconomic Income
Employment history
Education and training
Monetary return
Parental occupation
Housing
Operational Gear types
Investment in boat and gear
Operating expenses
Maintenance and repair ex-
penditures
Size of operations
Seasonal patterns
Rate of capacity utilization
Behavioral- Reasons for lobstering
Attitudinal Job interests
Attitudes towards leaving the
lobster industry
Job-seeking
Attitudes toward training, views
on excess capacity
ANALYSIS
The Maine Lobster Fishery: Some Basic Facts
The lobster industry in the State of Maine
landed 19.8 million pounds of lobsters worth
$16.1 million in 1969. This accounted for
10.4% of the quantity and 58.3% of the value
of the total fish and shellfish landings for
that year (Maine Landings, 1968-70, p. 3).
There were 5,750 lobster licenses issued in
the State in 1969. These 5,750 lobstermen
fished a total of 805,375 traps or approximately
105.7 million trap-days during the year 1969.
The gross earnings per unit of effort was
$0.18 per trap-day. This value is arrived at
by adjusting Maine landings up by 16% to
include landings not reported. This produced
total landings of 18.7 million which were
divided by total trap-days yielding the re-
turn of $0.18 per trap-day. The average gross
income was approximately $3,000. The total
investment in gear (i.e., boats, traps, buoys,
etc.) is about $10 million.3
There have been fluctuations in the number
of licenses issued over the past 10 years. Table
1 illustrates a seemingly cyclical pattern of
lobster licenses, showing a high of 6,472 in
1961, a low of 5,425 in 1962, and another
high of 6,316 in 1970.
The communities chosen for study — Phipps-
burg, Corea, and Beals — represent 277 fisher-
men or 4.4% of the 6,316 fishermen licensed
in 1970. A sample of 131 of the fishermen was
randomly selected by community as shown
in Table 2. The geographical locations of these
three communities are shown in Figure 1.
Economic Profile of the Sample Communities
Beals is an island community of 658 persons
located across Mossabec Reach from Jones-
port, Maine, population 1,337 (1970 Census —
Preliminary Report, Population Counts for
States). The two communities — Beals and
Jonesport — are integrated as a labor market
but have separate political identities. The only
administrative connection between the towns
is a shared high school.
Employment opportunities are limited to
the fishing industry and service industry oc-
cupations. The Department of Sea and Shore
Fisheries issued 142 lobster licenses to the
residents of Beals in 1969. Other licenses in-
clude worms — 52, and clams — 89. Many
of the fishermen hold more than one license.
No license is needed for shrimping.
Businesses on Beals include seven lobster
pounds, most of which are family owned and
operated. The pounds are used to store lob-
sters until market prices increase and the
3 Information supplied by Robert Dow, Research
Division, Maine Department of Sea and Shore Fisheries.
161
Calais
Ellsworth _ ^""Machiasport
teals
Gouldsboro
(Corea)
Figure 1. — Maine — selected geographic locations.
Table 1. — Number of lobster licenses issued in Maine
1961-1970.
Table 2. — Distribution of the sample fishermen by
Communities.
Year
1961
1962
1963
1964
1965
Number of licenses Year
6,472
5,658
5,695
5,803
5,802
1966
1967
1968
1969
1970
Number of licenses
5,613
5,4 25
5,489
5,750
6,316
Source: Maine Department of Sea and Shore Fisheries.
Communities
Beals
Corea
Phippsburg
TOTAL
Total fishermen
137
73
67
277
Sample
61
27
44
131
162
pound may be filled by the family owning it
or the pound operator may become a dealer
for part of the year, buying from fishermen
until he has the pound stocked. A third use
of the pound- is leasing to a full-time dealer
for his own stocking activities. If the family
does not operate the pound on a part-time
basis, the employment provided rarely ex-
ceeds one job. The two full-time lobster dealers
on Beals employ between two and four labor-
ers each. The 12 boatyards are father and
son operations although occasionally one non-
family employee may be hired. The two clam
shops on the island employ a total of between
25 and 30 persons together — mainly women
who shuck clams for shipment outside the
area. The service industry employment avail-
able on Beals consists of jobs in three general
stores, one garage, one oil company, one
television and radio sales, the local elementary
school, and various part-time jobs available
in the town government (mostly elective posi-
tions) (Table 3).
Table 3. — Occupational distribution of the work force
in Beals, 1960.
Male
Female
Total
Professional
8
8
16
Clerical
15
4
19
Craftsmen
28
28
Operatives
17
17
Service
4
4
Laborers (farm)
11
1 1
Laborers
77
77
Total
156
16
172
Source: 1960 Census Special Report for Maine Employment
Security Commission. Approximately 90% of the
"laborers" may be classified as lobster fishermen.
In Jonesport employment opportunities are
in much the same industries as they are in
Beals. Ninety-nine lobster licenses, 60 worm
licenses, and 81 clam licenses were issued by
the Department of Sea and Shore Fisheries.
Employment opportunities available in Jones-
port include jobs in one restaurant, one bank,
one sardine factory, two grocery stores, one
clothing store, one drug store, four gas stations,
three gas or oil companies (total employment
each is no more than three), one dentist's
office, one doctor's office, two lobster dealers
and a lobster cooperative which has four em-
ployees. Other firms in the area providing sub-
stantial employment are two sardine factories
— one in Milbridge and one in Machiasport.
This employment is part-time and seasonal.
The 1969 value of product given by the
Census of Maine Manufacturers for Beals is
$283,258, the total gross wages are $70,856,
and average gross $2,443. These figures are
for manufactured products only and do not
include income from lobstering, shrimping,
or other fishing unless the catch has been
processed in some manner. Total employment
in these industries is given as 29. For Jones-
port the corresponding figures are value of
product — $681,509, gross wages — $192,495,
and average gross wage — $2,406. Total em-
ployment was 80.
Total assessed value of property on Beals
in 1969 was $237,560. The town budget shows
total receipts of $99,376, and total expendi-
tures of $73,910, of which about $55,000 was
for wages distributed to inhabitants of the
town.
Table 4. — Occupational distribution of the work force
in Gouldsboro, 1960.
Male
Female
Total
Professional
4
4
Managers
21
14
35
Clerical
4
4
Sales
8
9
17
Craftsmen
50
50
Operatives
9
17
26
Private household
8
x
L6
Service
5
5
Laborers
137
137
No information
33
9
42
Total
275
61
336
Source: 1960 Census Special Report for Maine Employment
Security Commission. Approximately 90% of the
"laborers" may be classified as lobster fishermen.
Corea (Gouldsboro): The community in
Corea is part of the township of Gouldsboro.
The 1970 population of Gouldsboro is 1,270,
an increase of 170 people over the 1960 figure
of 1,100. In 1960 there were 363 households.
There were 420 males over 14 years of age
and 406 females.
163
Corea's major industry is lobster fishing,
providing some 70-80 jobs. Other types of
fishing, which are part-time or supplemental,
include seining, clamming, and worming.
There are some nine stores, a boatyard which
employs six-seven people year around, fish
cannery, a naval tracking base, and eight
teachers employed by the town's elementary
school. These activities employ 109 full-time
and part-time workers.
Table 5.
Occupational distribution of the work force
in Phippsburg, 1960.
Male
Fema
Total
Professional
8
4
12
Farmers and farm managers
4
4
Managers
16
1 1
27
Clerical
4
20
24
Crafts
68
68
Operatives
60
73
Private household
20
Services
12
12
Farm labor
12
12
Laborers
71
71
Others
27
8
35
Total
282
70
358
Source: 1960 Census Special Report for Maine Employment
Security Commission. Approximately 80% of the
"laborers" may be classified as lobster fishermen.
Phippsburg: In 1970 the population of
Phippsburg was 1,180, an increase of 59 people.
Of the 1,121 people listed in April of 1960,
397 were in the labor force; 358 were employed,
and 39 were unemployed. Of those over 14
years of age, 394 were men and 403 were
women. There were 335 households.
Phippsburg's major industry is the summer
tourist and summer resident trade. At Phipps-
burg there are several large tenting grounds,
a state park, and many summer residences
located on its several miles of ocean frontage.
Other local industries include fishing, which
consists of a fish factory, several large offshore
fishing boats, and a fleet of lobster boats.
There are also two small construction com-
panies that build and repair summer homes.
The bulk of Phippsburg's employed popula-
tion, however, commute to other towns and
cities for employment. Probably the largest
employer of Phippsburg people is Bath In-
dustries located in the adjacent city of Bath.
Selected Socioeconomic Characteristics
of the Sample Lobstermen
Average age of the lobstermen in the sample
is 42.6 years. There are 15 below the age of
19 and 18 in the age bracket of 65 and over.
The median income for the group is $5,280
and average income in $6,213. There are 13
fishermen with income less than $1,000 and
15 with income over $14,000. Of the 118 fisher-
men who gave reasons for lobstering, 33
(which includes 3 students) responses may
be categorized as "economic" and the rest
"non-economic" including home consumption,
preference for the particular way of life, in-
fluence of family, and so on.
Of the 109 fishermen who supplied informa-
tion on number of traps, slightly over 50%
owned less than 300 traps; 23 fishermen owned
more than 500 traps. Of the 93 fishermen who
gave information on investment in trap gear,
approximately 50% had investment of less than
$2,000; only 3 had investment of $8,000 and
over. The average years of education was 9.8.
Approximately 40% had less than 9 years of
education. Of 131 fishermen, 41 indicated that
they received some type of formal vocational
training in areas including carpentry, metal
working, mechanic, professional and clerical
work. Of 81 fishermen asked about preference
for receiving vocational training, 63 indicated
no preference. Only a small fraction express-
ed preference for training in electrical, pro-
fessional, and carpentry work.
Among the 109 fishermen who supplied in-
formation on income from part-time jobs, 77
indicated that they had little or no income
from this source. Only 7 indicated that they
received more than 50% of their income from
alternative jobs.4
Analysis of Target Groups
In order to analyze the potential socioeco-
nomic impact of limited entry, it is necessary
to identify the possible candidates who might
be considered targets for limited entry or any
4 More detailed information on these and other aspects
of the study may be found in the complete final project
report, available from the Economic Research Labora-
tory, National Marine Fisheries Service.
164
other management strategy that might affect
the harvesting labor force.
For the purpose of this study, four groups
have been constructed, using alternative
criteria. It is not intended that the groups be
mutually exclusive.
The variables chosen for this analysis in-
clude the following: income, investment, effort,
and earnings/effort ratio.5 It should be noted
that with the exception of one target group,
combinations of variables were used to define
the target groups. Admittedly, similar groups
could be constructed using different criteria.
Groups selected appeared to be quite meaning-
ful for the purpose of this study.
Target Group I was chosen on the basis of
a combination of two criteria: (a) low earn-
ings/effort ratio, and (b) low number of trap-
days serving as a proxy for low income. It
was arbitrarily decided that to be eligible
for this group a fisherman had to have an
income/effort ratio of less than 0.3 and had
to fish less than 30,000 trap-days per year.
Those fishing over 30,000 traps were not in-
cluded because they earned sufficient income
for subsistence. Table 6 was especially con-
structed for this purpose.
Forty fishermen met the conditions set for
this group. As it turned out, this group had
an average earnings/effort ratio of 0.182
compared to 0.230 for the entire sample and
they fished an average number of 12,570 trap-
5 The earning/effort ratio was calculated by dividing
the number of trap-days into gross income reported
by the sample fishermen.
days compared to 30,707 trap-days for the
sample as a whole. Their average income was
only $2,061 compared to an average income
of $6,213 for the sample as a whole. The
fishermen in this group fish fewer number of
days and have invested small amounts of
capital in gear and boat.
In any discussion of deliberate or planned
changes in the harvesting labor force in the
lobster fishery, this group with a low earnings/
effort relationship and low absolute level of
income would warrant consideration. Pre-
sumably, the economic status of the remain-
ing fishermen would improve the terms of a
higher ratio of income to effort and higher
absolute level of income, if this group is elimi-
nated. Of course, one has to look at the social
cost of such a change and the political feasi-
bility of such a change. Some measures of
social cost are developed later in this paper.
An alternative approach to the problem
would be to consider only low levels of pro-
ductivity as measured by the low income/
effort ratio, regardless of the absolute size
of income. Here one could argue that shifting
away from lobstering in this case may be
socially gainful, given possibilities for im-
proving the income/effort ratio in alternative
employments. From such a reallocation of
effort as an economic resource, both the dis-
placed fishermen as well as the surviving
fishermen might benefit, as the marginal pro-
ductivity of both groups is likely to increase.
On this premise, Target Group II has been
constructed. Those fishermen who recorded
an income/effort ratio of less than 0.2 were
Table 6. — Distribution of sample lobstermen according to income/effort ratio and trap-days.
Trap-days fished
per year
Earning effort
ratio
5,000
5,001-
10,000
10,001-
20,000
20,001-
30,000
30,001-
40,000
40,001-
50,000
50,001-
60,000
60,000+
N/I
TOTAL
0.100
.100-.199
.200-.299
!
2
5
2
3
1
1
7
8
8
2
5
4
1
4
2
4
6
8
1
-
7
41
27
.300-.399
-
2
2
2
2
1
1
-
-
ID
.400-.499
.500 +
N/I
2
6
2
1
1
1
2
2
1
-
1
1
1
5
19
4
5
37
TOTAL
18
10
20
14
14
7
15
14
19
131
Source: University of Maine Survey Data, 1970.
165
considered eligible for this group (See Table
6). There will be some overlap between this
group and Target Group I.
Different combinations of investment and
effort suggest other possible approaches to
management alternatives. For instance, one
could identify a group that represents rela-
tively high effort and low investment input
combination; another group may represent
relatively higher investment and lower effort
input combination.6 The reasoning for at
least considering these groups as possible
target groups may be explained as follows:
in the absence of any precise knowledge about
the optimum combination of effort and invest-
ment, two contrasting groups — high-effort
low-investment versus low-effort high-invest-
ment — might suggest alternative goals for
management strategies. For instance, one
might consider eliminating excessive capital
versus eliminating excessive effort as possible
goals. As a minimum, the differences in socio-
economic impact of such changes should be
examined.
It is reasonable to assume that excess
capacity exists in the lobster fishery, although
it is difficult to establish whether such excess
capacity is due to excessive effort or excessive
investment or both. Under these conditions,
6 This approach was suggested by Dr. Adam A.
Sokoloski, National Marine Fisheries Service in per-
sonal correspondence dated December 16, 1970.
it seems meaningful to isolate for analytical
purposes, two cases, one showing evidence
of excessive effort and the other of excessive
investment. Admittedly, the state of the art
does not provide absolute measurement of
excess capacity either in terms of effort or in
terms of investment.
Target Group III has been constructed to
reflect excessive effort in the sense that these
fishermen supply a large amount of labor to
their operation relative to their investment.
They fish, on an average, 150 days per year
compared to 109 days for the entire sample;
their average investment amounted to $4,410
compared to $7,575 for the entire sample.
As a practical device, the criteria of those
fishing over 100 days per year with investment
of less than $8,000 in gear were used to select
the candidates for this group of 28 fishermen.
Target Group IV represents excessive capi-
tal in the sense that the fishermen in this
group have substantial investments in gear
relative to the number of days per year fished.
On the average they have invested $12,410
compared to $7,575 for the entire sample and
they fish an average of 78 days per year com-
pared to 109 days per year for the sample.
This group of six fishermen included those
who have invested more than $8,000 and who
fish less than 100 days per year.
Table 7 provides the basic information from
which Target Groups III and IV have been
derived.
Table 7. — Distribution of sample lobstermen by investment and number of days fished.
Investment
in gear
Days fished
2,001-
4,001-
8,001-
12,001-
16,001-
20,001-
per year
2,000
4,000
8,000
12,000
16,000
20,000
24,000
24,000+
N/I
Total
dollars
50
HI
3
_
1
_
_
_
_
3
17
51-100
16
7
8
2
2
1
-
-
2
38
101-150
3
7
8
1
4
3
2
4
-
32
151-200
-
2
5
6
2
2
1
-
-
18
201-250
-
1
2
-
1
-
1
1
-
6
N/I
-
-
-
-
1
-
1
-
18
20
TOTAL
29
20
23
10
10
6
5
5
23
131
Source: University of Maine Survey Data, 1970.
166
Table 8. — Distribution of lobstermen in target groups by trap-days, gross income, and capital invested.
Trap-days
I
502,799
II
Target groups
III
1,753,287
973,198
IV
185,560
Total
Sample
3,470,000
%
*(No.), %
14.5
(40) 32.0
50.5
(48) 38.4
28.0
(28) 22.4
5.3
(6)4.8
(113)
Income
$82,450
$250,233
$161,583
$61,000
$596,500
%
*(No.),%
13.8
(40)41.7
41.8
(48) 50.0
27.0
(26) 27.1
10.2
(5)5.2
(96)
Capital
%
*(No.), %
$97,043
11.6
(40) 36.4
$332,566
39.9
(48)43.6
$123,485
14.8
(23) 25.5
$74,465
8.9
(6)5.5
$833,209
(110)
*The number in parentheses refers to the total number of fishermen relevant to a particular category; the other number is the relevant
number of fishermen expressed as a percentage of the sample.
Source: University of Maine Survey Data, 1970.
Distribution by Trap-days, Income,
and Capital Invested
Table 8 presents a distribution of the lobster-
men in each of the target groups by trap-days,
gross income and capital invested in boat and
gear. Target Group I emerges as a critical
group in that its share in trap-days, income
and capital investment is the lowest relative
to its size in the total sample. Target Group
II contributes more trap-days, more capital,
and more income compared to Group I. How-
ever, relative to its size, its share in income
and capital investment is less than in propor-
tion. Target Group III contributes relatively
more in trap-days and relatively less in capital
and its income share corresponds closely to
its size. Target Group IV accounts for more
capital relative to size and to number of trap-
days and substantially more income relative
to size. For this reason, this group can hardly
be considered as a target group for limited
entry on the basis of income-effort relation-
ship. However, if the income-capital ratio is
considered, this group does not appear to
be equally efficient.
Socioeconomic Characteristics of the
Fishermen in Each of the
Four Target Groups
Beals will be most affected if Target Group
II is eliminated, and Corea the least. If Target
Group I is considered, the impact on the three
communities is comparable. Corea will be af-
fected in the least if one focuses on Target
Group III. The effect on the other two com-
munities is about the same. Target Group IV
does not affect Phippsburg but will affect
the other two communities equally (Table 9).
Table 10 provides average values for certain
socioeconomic characteristics of the lobster-
men in each of the Target Groups.
Table 9. — Geographic distribution
Target groups
1 II
Community No. % No. %
III IV
No. % No. %
Beals1
Corea2
Phippsburg3
18 29.5 31 50.8 16 26.2 4 6.5
7 26.9 3 11.5 3 11.5 2 7.7
15 34.1 14 31.8 9 26.5
Total
41)
4S
'Beals 61.
2Corea 26.
3Phippsburg 44, includes 10 from Bath.
Source: University of Maine Survey Data, 1970.
The average income of Group I is the lowest
attributable both to low labor and low capital
intensity in its operation. In constrast, Group
IV has the highest average income primarily
due to high capital intensity in its operation
in spite of low labor intensity. Group II ranks
second in average income which can be ex-
plained in terms of relatively more effort and
167
Table 10. — Comparative average value for selected socioeconomic variables in the sample of lobstermen and the four
target groups.
Target groups
Socioeconomic variable
Sample
1
II
III
IV .
Family size
3.2(122)
2.9 (38)
3.6(46)
2.9(28)
3.6(5)
Age
42.4(131)
42.5 (40)
44.0 (48)
49.4 (28)
31.7(6)
Education: years
9.8(126)
9.7 (40)
9.7 (48)
10.0(28)
11.0(6)
Investment (gear & boat)
$7,575 (110)
$2,426 (40)
$6,949 (48)
$4,410(28)
$12,410(6)
Gross income
$6,213 (96)
$2,061 (40)
$5,213 (48)
$6,214(26)
$12,200(5)
Months per year fished
7.2(113)
5.7 (40)
8.0 (48)
8.5 (28)
6.6(5)
Trap-days per year
30,707 (113)
12,570(40)
36,526 (48)
34,757 (28)
30,927 (6)
Days per year lobstered
109.2(113)
87.0 (40)
132.2(47)
147.9(28)
78.0(6)
Earning-effort ratio
.230 (96)
.182(40)
.140(48)
.183(26)
.355 (5)
*The number in parentheses refers to the total number of fishermen relevant to a particular category.
Source: University of Maine Survey Data, 1970.
capital used compared to Groups I and III.
Group III ranks third in average income.
Here the high level of labor intensity offset
the effect of low capital intensity. Its income/
effort ratio is almost the same as that of
Group I.
Socioeconomic Impact of Changes
in Harvesting Labor Force
As pointed out earlier, the different target
groups were constructed on the basis of differ-
ent criteria such as low earnings/effort ratio,
low level of both effort and investment, high
labor and capital input combination. The
rationale for this procedure is simply to facili-
tate comparative analysis of alternative man-
agement strategies. For instance, one might
consider limiting entry on the basis of low
earnings/effort ratio combined with low level
of income (Group I); one might also focus
on low earnings/effort ratio regardless of
the level of income (Group II); alternatively,
one might emphasize high labor-low capital
input combination associated with low income
as an indicator of inefficiency (Group III);
finally, high capital-low labor input combina-
tion regardless of a relatively higher level of
income may be construed as an indicator of
excess capacity (Group IV).
It should be noted that it was not the pur-
pose of this study either to advocate or repudi-
ate any particular management strategy and
its implicit goal. The intent here is simply to
analyze the potential socioeconomic impact
of a change in the harvesting labor force in
the Maine lobster fishery if such a change
amounts to reducing inefficient inputs from
given target groups.
For the purpose of this study such impact
is analyzed primarily in terms of employment
effects and income effects relative to the target
group populations and the local economy.
Employment Effects
Taking into consideration the employment-
related variable such as skills either from
currently held part-time jobs or alternative
jobs held in the past, level of education, and
age, a simplified profile of labor market par-
ticipation potential of the target groups is
shown in Table 11.
The category "potentially employable" in-
cludes those individuals who have market-
able skills acquired from formal vocational
training and/or alternative job experience.
This survey information was supplemented
by information on the local labor market
through the cooperation of the regional offices
of the Maine Employment Security Commis-
sion. If there was a match between the kinds
of skills in demand in the local labor market
and the skills possessed, an individual was
considered eligible for the category "potential-
ly employable."
The category "possibly trainable" includes
those who on the basis of age and level of
education would be likely to benefit from and
168
Table 11. — Labor market participation potential of target groups I-IV.
Target
group
Total
number
Potentially
employable1
Possibly
trainable2
Po
core
ential hard-
unemployed3
Not in the
labor force4
I
40
' 100.0%
14
35.0%
4
10.0%
8
20.0%
14
35.0%
IJ
48
100.0%
18
37.5%
4
8.3%
17
35.4%
9
18.7%
III
28
100.0%
11
39.3%
2
7.1%
10
35.7%
5
17.9%
IV
6
100.0%
4
66.7%
1
16.7%
1
16.7%
-
'Those having marketable skills.
2Those having no skill but less than 35 years of age.
3Those having no skill and in the age bracket 35-65 years.
4Students and those over 65 years.
Source: University of Maine Survey Data, 1970.
be capable of participating in a training pro-
gram. Admittedly, this is only a first approxi-
mation.
The category "potential hard-core unemploy-
ed" includes those fishermen who have no
marketable skills other than lobstering and
who fall into the critical age bracket by labor
market criteria, 35-65. In all likelihood, these
individuals, if excluded from lobstering, will
find it extremely hard to make any vocational
readjustment.
The last category, "not in the labor force"
is self-explanatory. This includes those fisher-
men who are either students or over 65 years
of age and are not likely to participate in
the labor market as active job seekers, barring
purely part-time or seasonal jobs.
It should be emphasized that the above
classification is only a preliminary step in
identifying the differences in labor market
participation potential of various subgroups
within each of the target groups. To be sure,
potential employability, trainability, and hard-
core unemployability require considerably
more in depth analysis than was possible in
the present study.
It is apparent from Table 11 that a sub-
stantial proportion of the fishermen in each
of the target groups is potentially employable
(ranging from 35% to 67%). Of those who are
classified under "potentially employable," some
already have full-time jobs and others have
marketable skills. However, Target Groups
II and III are likely to result in more hard-
core unemployment. Paradoxically, the group
that has a high earnings/effort ratio (Target
Group IV) also happens to be the one with a
relatively larger proportion of potential em-
ployability. With the exception of this group,
other groups include several fishermen not
in the labor force, students, and those 65 years
and over. The question of their employability
is, therefore, irrelevant in the present context.
In analyzing the expected socioeconomic
impact of limited entry, the survey data on
each of the fishermen in each of the target
groups were examined in depth by communi-
ties. In this investigation, attention was focus-
ed on such socioeconomic variables as age,
family size, level of education, types of skill,
alternative job experience, alternative source
of income, and so on. On the basis of informa-
tion from survey data combined with informa-
tion on local labor market, Table 12 is recon-
structed to reflect the differences in labor
market participation potential by communities.
Income Effect and Expected
Socioeconomic Impact
To perform the necessary analysis, the
following procedures were adopted:
1. Assume each target group to be a candi-
date for exclusion from lobstering.
2. Estimate private loss of gross income due
to non-participation in lobster fishery.
3. Assume that 50% of the lost gross income
would be subsequently earned by the re-
maining fishermen. The survey date did
169
Table 12. — Labor market participation potential of target groups I-IV by geographic location.
Target aroup
' by
communities
Total
number
Potentially
employable1
Possibly
trainable2
Po
core
tential hard-
unemployed3
Not in the
labor force4
Phippsburg
I Corea
Beals
15
7
18
7
3
4
2
1
1
3
5
3
3
8
40
14
4
8
14
Phippsburg
II Corea
Beals
14
3
31
8
1
9
1
3
4
13
2
1
6
48
18
4
17
9
Phippsburg
III Corea
Beals
9
3
16
5
2
4
1
1
3
7
1
4
28
11
2
10
5
Phippsburg
IV Corea
Beals
2
4
1
3
1
1
-
6
4
1
1
-
'Those having marketable skills.
2Those having no skill but less than 35 years of age.
3Those having no skill and in the age bracket 35-65 years.
4Students and those over 65 years.
Source: University of Maine Survey Data, 1970.
indicate some evidence of excess capacity
in terms of number of traps owned and
number of traps fished and days fished.
It was recognized that the remaining
fishermen may not be willing or able to
capture the entire amount of output at-
tributable to the excluded fishermen, at
least in the short run. Furthermore, the
purpose here is to illustrate what might
happen if this assumption holds. If a
different figure proves to be more realis-
tic, the results will change.
Estimate the savings in effort measured in
trap-days on the basis of (3) and convert
this into monetary values. For this pur-
pose, we first calculated how many trap-
days would be needed by the excluded
fishermen in a given target group to
produce the gross income attributed to
this group. An average earnings/effort
ratio for this group was used to calcu-
late the number of trap-days required.
Next, an average earnings/effort ratio
was computed in the given target group.
This average ratio was applied to 50%
of the total gross income of the group to
come up with the number of trap-days
that would be required to produce this
income by the remaining fishermen. The
difference between the two values for
trap-days is stated as saving in effort.
This quantity multiplied by the average
earnings/effort ratio of the remaining fish-
ermen produced a monetary measure of
saving that can be expected under the
stipulated conditions.
5. Estimate the sum of expected new in-
comes generated by those who are con-
sidered "potentially employable" based
on information of types of jobs available
and skills needed in the local market.
The number of fishermen in each target
group that fits this category was identi-
fied and typical wages for indicated jobs
were applied to the number of employ-
able fishermen to produce a sum of ex-
pected income.
6. Estimate the expected annual income of
170
those that are classified as "possibly train-
able." Assume that training facilities
and programs are made available and
that individuals are willing to paticipate.
Communication from people involved
with Manpower Development and Train-
ing Act (MDTA) programs provided some
information as to typical wages MDTA
trainees can expect post-training. These
figures were used to derive expected in-
comes that the "possibly trainable" fish-
ermen in each target group can expect
if they receive training comparable to
those under MDTA programs.
7. Estimate the training cost of those classi-
fied under "possibly trainable."
8. Estimate the potential income-mainte-
nance burden on society imposed by the
loss of lobstering income of those who
are classified under "potentially hard-
core unemployed" and under "not in
the labor force." Fifty percent of current
gross income from lobstering was used
for estimation purposes. The rationale
for using this percentage is based on the
consideration that the net income from
lobstering is substantially lower than
reported gross income, although exact
figures for net income were not readily
obtainable. During the course of the
interviews, several fishermen indicated
that although they could not provide in-
formation on net income, roughly 50%
of their gross income could be considered
net, after allowing for business expenses.
The assumed percentage is considered
reasonable for illustrative purposes.
The reason why the individuals in these
categories — "potential hard-core unemploy-
ed" and "not in the labor force" — and their
loss of income from lobstering are used as
the basis for measuring the income mainte-
nance burden on society is to indicate the
upper limit of the social burden. This yields
a relative measure of income loss and corres-
ponding welfare loss for a group of people
who are technically outside the labor force.
At least in the short run, the process of ad-
justment will be quite severe for a bulk of
this group. Conceivably, some low level, un-
skilled jobs would be available which would
moderate the impact. However, considering
the high level of current unemployment and
the generally depressed conditions of the local
economies under consideration, it appeared
reasonable to assume that alternative sources
of income would be unavailable in the short
run, thereby imposing a burden on society.
9. The estimated value of investment in
boat and gear by the fishermen in each
of the target groups is included in the
profile of socioeconomic impact of limited
entry because these values have definite
implications for compensation.
Assuming zero salvage value of such capital
equipment, the stated figures provide the upper
limit of the compensation burden imposed on
society. It is reasonable to think actual com-
pensation will differ from the stated figures
because of some positive salvage value. For
illustrative purposes, without making such
allowance, the quoted figures do serve as indi-
cators of upper limits of the cost of compensa-
tion that may be entailed.
Using the above procedure, the following
tabulations were made to present a compara-
tive picture of the socioeconomic implications
of limiting entry of different groups by using
alternative criteria (Table 13).
Group II is likely to cause the largest de-
cline in income from lobstering. It will be
partially offset by additional income from
lobstering by the remaining fishermen, income
from alternative jobs for the displaced fisher-
men, and the savings in effort measured by
the fewer number of trap-days required to
capture at least 50% of the gross income lost.
In absolute terms, this group may present
the severest income maintenance burden on
society. By comparison, Group I is likely to
impose a relatively smaller burden on society.
On a per capita basis, Group III will impose
the severest burden on society.
The proportion of the "potentially employ-
able" and "possibly trainable" among Groups
I-III are quite comparable. The proportion
of the same categories for Group IV is con-
siderably higher. This accounts for the rela-
tively small social burden indicated for this
group. However, it should be noted that this
underestimates the total real burden on society
in that there will be a dissaving in effort and
potential negative difference between their
current income from lobstering and their ex-
171
Table 13. — Profile of socioeconomic impact by target groups.
Impact variables
Target
groups
I
II
III
IV
-82,450
-250,223
-161,583
-61,000
+41,225
+ 125,116
+ 80,791
+ 30,500
+18,574
+ 168,670
+ 31,346
-11,083
+ 19,000
+ 41,500
+ 38,000
+21,000
+24,000
+ 24,000
+ 12,000
+ 12,000
-13,800
- 13,800
- 6,400
- 6,400
-26,775
- 64,225
- 54,200
- 3,500
-97,043
-332,566
-123,485
-74,465
40
48
28
6
1. Loss of income from lobstering (S)
2. Gain of income from lobstering ($)
3. Monetary value of saving in effort ($)
4. Gain of income from alternative jobs (marketable
skills) ($)
5. Gain of income from alternative jobs (post-
training) ($)
6. Training costs ($)
7. Income maintenance burden on society ($)
8. Estimated value of investment in boat and gear ($)
9. Number of fishermen
Source: University of Maine Survey Data, 1970; local Manpower Development Training Act program officials.
pected income from alternative jobs.
It would have been desirable to compute
a ratio of total gains and losses. However,
with the data in hand, it does not appear to
be feasible and meaningful. First, the quanti-
ties calculated are not additive. Second, costs
and benefits have different time dimensions.
For instance, training costs are once-over
cost items whereas the expected income is
a flow over time. Finally, the figures for in-
come maintenance burden on society do not
take into consideration the loss of income
from lobstering of those who are classified as
"potentially employable" but are already
employed. Furthermore, the discrepancy be-
tween current income from lobstering and
expected income from alternative jobs for those
employable but currently full-time fishermen
is also disregarded.
Despite these limitations, the results do
give certain indicator values that should be
considered and comparatively analyzed rela-
tive to alternative management strategies
and implicit goals. Admittedly, these values
involve many simplifying and rather arbi-
trary assumptions, although hard data were
utilized when available. The value of this
type of approach is primarily methodological,
which is to be expected in a pilot study.
CONCLUSIONS
Several qualifications need to be attached
to the foregoing analysis before any general-
ization is made. First, some fishermen who
are considered as candidates for a given target
group may continue to lobster because of non-
economic reasons. Second, expected new in-
comes from alternative jobs for the displaced
fishermen may not materialize because of lack
of motivation and reluctance to move geo-
graphically and/or occupationally. Third,
there is no assurance that the additional new
income earned by the remaining lobstermen
will exactly equal the lost income due to
limited entry. There is, however, a strong
probability that if they were to capture the
same number of lobsters as attributable to
the displaced fishermen, they could do so
more efficiently because of excess capacity
and potential economies of scale. Fourth, there
may be a significant gap between the number
of those considered trainable and those who
will take advantage of training if made avail-
able. Fifth, a fraction of those trained may
still remain unemployed due to labor market
conditions. Sixth, the income maintenance
burden may not be as severe as indicated be-
cause some of the potentially hard-core un-
employed may be absorbed in unskilled jobs
or in the lobster industry as "helpers." Con-
ceivably, jobs may be redesigned to facilitate
the entry of these men into the labor market.
Finally, some of those who are not in the labor
force, e.g., students, will, in course of time, par-
ticipate in the labor market and reduce the
stated social burden.
It is important that in this kind of analysis
one takes cognizance of the time element
relative to the process of adjustment. The
short run impact may appear to be quite
severe because of the imperfections in the
172
labor market. For instance, men who are
unemployed now may not have marketable
skills; men who have marketable skills may
not have information about available jobs or
may have very restricted mobility; job struc-
ture may be such that it precludes entry of
unskilled workers; those who are trainable
may not have access to adequate training
facilities or programs. Given time, however,
some of these market imperfections may be
reduced, partially through deliberate planning
and partially through autonomous changes
in the labor market itself. For instance, the
quality of job information and job counselling
can be improved; training programs may be
initiated; jobs may be restructured; local
economic development may generate new de-
mands for labor; the lobster fishery itself, if
efficiently managed by fewer fishermen, may
need additional helpers.
It is a reasonable expectation that if a
management strategy results in an improved
return to both labor and capital and if de-
liberate efforts are made to aid the process
of adjustment, net social gains are likely to
materialize in the long run. Although the
present study did not consider, nor was in-
tended to consider, any specific management
scheme with respect to its socioeconomic
impact, it did generate data pertinent to such
an evaluation.
A U.S. GOVERNMENT PRINTING OFFICE: 1973-795-774 / 9 REGION 10
173
349. Use of abstracts and summaries as communica-
tion devices in technical articles. By F. Bruce
Sanford. February 1971, iii + 11 pp., 1 fig.
350. Research in fiscal year 1969 at the Bureau of
Commercial Fisheries Biological Laboratory,
Beaufort, N.C. By the Laboratory staff. No-
vember 1970, ii + 49 pp., 21 figs., 17 tables.
351. Bureau of Commercial Fisheries Exploratory
Fishing and Gear Research Base, Pascagoula,
Mississippi, July 1, 1967 to June 30, 1969. By
Harvey R. Bullis, Jr., and John R. Thompson.
November 1970, iv + 29 pp., 29 figs., 1 table.
352. Upstream passage of anadromous fish through
navigation locks and use of the stream for spawn-
ing and nursery habitat, Cape Fear River N C
1962-66. By Paul R. Nichols and Darrell E.'
Louder. October 1970, iv + 12 pp., 9 figs. 4
tables.
356. Floating laboratory for study of aquatic organ-
isms and their environment. By George R.
Snyder, Theodore H. Blahm, and Robert J. Mc-
Connell. May 1971, iii + 16 pp., 11 figs.
361. Regional and other related aspects of shellfish
consumption — some preliminary findings from
the 1969 Consumer Panel Survey. By Morton
M. Miller and Darrel A. Nash. June 1971, iv +
18 pp., 19 figs., 3 tables, 10 apps.
UNITED STATES
DEPARTMENT OF COMMERCE
NATIONAL OCEANIC & ATMOSPHERIC ADMINISTRATION
NATIONAL MARINE FISHERIES SERVICE
SCIENTIFIC PUBLICATIONS STAFF
BLDG. 67, NAVAL SUPPORT ACTIVITY
SEATTLE, WASHINGTON 98115
OFFICIAL BUSINESS
PENN STATE UNIVERSITY LIBRARIES
ADDDD7ED
210