MOBIUS-- a procedure for identifying mobility barriers based on dynamic strategic grouping analysis

UNIVERSITY OF

ILLINOIS L.tiRARY

AT URBANA-CHAMPAIGN

BOOKSTACKS

Digitized by the Internet Archive

in 2011 with funding from

University of Illinois Urbana-Champaign

http://www.archive.org/details/mobiusproceduref1226sudh

:-., ■■«*

BEBR

FACULTY WORKING
PAPER NO. 1226

THE LIBRARY OF THE

APR 1 o 1986
UNIV. OF ILLINOIS

MOBIUS — A Procedure for Identifying
Mobility Barriers Based on Dynamic
Strategic Grouping Analysis

D. Sudharshan
Howard Thomas
Avi Fiegenbaum

College of Commerce and Business Administration
Bureau of Economic and Business Research
University of Illinois, Urbana-Champaign

BEBR

FACULTY VORKING PAPER NO. 1226
College of Commerce and Business Administration
University of Illinois at Ur bana -Champa i gn
February 1986

M0BIUS--A Procedure for Identifying Mobility Barriers Based
on Dynamic Strategic Grouping Analysis

D Sudharshan, Assistant Professor
Department of 3usiness Administration

Howard Thomas, Professor
Department of Business Administration

Avi Fiegenbaum, Assistant
Department of Business Administration

ABSTRACT

The Mobility Barriers Paradigm has strongly influenced much recent
work in strategic management research, particularly that of strategic
group research. In this note we propose a procedure that enables the
identification of those variables that act as mobility barriers in an
industry. We also propose a classification of mobility barriers based
on the degree of observed mobility and the extent to which change is
desired on these variables. Our procedure is demonstrated in the con-
text of the pharmaceutical industry.

MOBIUS — A Procedure for Identifying Mobility Barriers Based on
Dynamic Strategic Grouping Analysis

Int roduct ion :

The concept of strategic groups has become an important unit of
analysis in developing theories of competition in the field of strate-
gic management. The commonly accepted definition of a strategic
group, due to Porter (1980: 129), states that "a strategic group is
the group of firms in an industry following the same or a similar
strategy along the strategic dimensions." McGee and Thomas (1986)
provide a thorough review of both the conceptual framework, and current
research work in this area while Harrigan (1985) proposes the applica-
tion of clustering procedures for strategic group analysis using the
"Mobility Barriers Paradigm" (p. 56). To quote Harrigan (1985, p. 57)
"Mobility Barriers are structural factors that protect successful
firms from invasions by adjacent competitors (Caves and Porter, 1977).
They are internal (to the industry) entry barriers which delineate
boundaries between different strategic groups, and they may be con-
trasted with the external entry barriers discussed in traditional eco-
nomic theory which deter outside firms from entering any part of the
industry (Harrigan, 1981)." Further, according to both Harrigan
(1985) and McGee and Thomas (1986), it is important not only to iden-
tify the presence of inter-group mobility barriers, but also to exam-
ine how, and to what extent, these barriers influence competitive
activity.

Several strategic grouping studies (e.g., Dess and Davis (1984))
have used cluster analysis to form strategic groups. Typically,
cluster analysis leads to the development of strategic groups by

-2-

grouping Che companies in an industry based on their "scores" on a set
of important strategic variables. However, such grouping for any one
time period does not in itself provide sufficient Information to
either identify the strategic variables that act as barriers to inter
group mobility or to infer the heights of such barriers. This is
because no information is provided by such a single period grouping
about movement between groups (which clearly requires at least two
grouping periods).

Therefore, this note presents a procedure for identifying the
strategic variables that act as mobility barriers in an industry. The
discussion of the procedure and its rationale is followed by an appli-
cation of this procedure to the pharmaceutical industry. This proce-
dure is called MOBIUS — an acronym for Mobility Barriers Identif ication
Using Strategic Grouping.

The MOBIUS Procedure

The essential idea behind MOBIUS is very simple. Its objective is
to identify strategic variables that act as barriers to movement of
companies from one strategic group to another. If a mobility barrier
is defined in terms of certain strategic variables, then there should
be very little shift in the structure (group membership and number of
groups) of strategic groups (characterized by these variables) over
time. On the other hand, if another set of strategic variables is not
a mobility barrier, then, considerable shifting in strategic group
structure over time, is a likely occurrence. However, it is also
possible that little or no shift in group structure may be observed
for a certain set of factors (variables) because there is no motivation

-3-

on the part of companies to change group affiliations by framing stra-
tegic policy shifts in terms of these variables. While this may hap-
pen, we still define any such set of variables as a mobility barrier.
Figure 1 provides a simple diagram of the possible combinations of
observed structural change and desire to change. The case just
discussed is identified as Mobility Barrier (Type B) in this figure
(that is, Low Density to Change, Low Observed Structural Changes).
Mobility Barrier (Type A) involves those sets of variables which pro-
vide strong barriers but also strong motivations for a company to
change group affiliations. Clearly, the mobility barriers A and B are
quite different in character. For example, considering Type B, it is
unlikely that significant structural changes in strategic groups will
involve variables which companies perceive to have little or no influ-
ence on their desire to change group affiliations. Thus it is perhaps
quite safe to say that those strategic variables which lead to con-
siderable structural shifts are those in the left hand upper quadrant
of Figure 1 (on which better company-environment fits are obtainable
by such shifts).

The MOBIUS procedure defines a measure of structural change which
uses the output of strategic grouping analyses undertaken for dif-
ferent time periods. An index called a "Match Ratio" (MR) captures
structural change and its rationale is explained below. (Table 1
shows the computational formula for this index.)

Insert Table 1 about here

Consider a comparison between strategic groups for two time
periods, T and T9. Let there be m strategic groups in period T and
n strategic groups in period T_. If the same companies belong to the
same groups in both time periods (and, therefore, m=n) then there is
no mobility between these two time periods (i.e., no change in group
structure or membership). If either m is very different from n, or if
there are considerable differences in group membership between the two
time periods, then there is relatively high mobility between the two
time periods, and the corresponding MR is high. The entry corresponding
to row i and column j, in Table 1, is the number of companies that were
in group i in period T, which moved together to group j in period T„.
Clearly if the oil-diagonal entries of this table all have zeros, then
there is a perfect match between groupings in the periods T and T .
If, on the other hand, all the diagonal entries are zeros, it is
implied that no group retains the same members. The MR corresponding
to this condition is 0. The MR corresponding to perfect match between
T and T? is 1. All other realizations of Table 1 will result in MR
between 0 and 1. Thus the index MR is bounded between 0 and I, and
increasing mismatch between groups gives rise to higher off-diagonal
entries which in turn leads to higher MR. It is, therefore, an ele-
gant and informative index of mobility. This index would thus allow
the comparison of observed mobility (or immobility) on different sets
of strategic variables. Mobility barrier strategic variables may thus
be identified.

The complete MOBIUS procedure consists of the following steps

1) Identification of relevant sets of strategic variables.

2) Identifying companies constituting the industry under study.

-5-

3) Obtaining appropriate measures for each company for all relevant
strategic variables across time.

4) Forming strategic groupings for each time period for every set of
strategic variables.

5) Computing the MR for each pair of consecutive time periods for
every set of strategic variables.

6) Inferring of Mobility Barriers from the MR results.

7) Obtaining industry expert opinions to throw light upon the
differences between Type A and Type B Mobility Barriers.

This procedure is demonstrated for the pharmaceutical industry.

Empirical Example: Pharmaceutical Industry

The strategic variables considered here encompass both scope and
resource deployment activities. These variables constitute a union of
those typically used to capture strategic behavior in the strategic
management literature. The details of the extensive procedure
involved in choosing these variables is described in Sudharshan,
Thomas, and Fiegenbaum (1985).

A total of 22 companies, representing 90 percent of the sales in
this industry were selected. Data were obtained on Marketing Stra-
tegy, Financial Strategy, Production Strategy, and Scope variables for
the 22 companies for the period 1974-1980 from COMPUSTAT® tapes.
(Descriptions of the specific variables chosen are shown in Table 2.)

Insert Table 2 about here

For each variable type (i.e., Marketing, Financial, Production
Strategy and Scope) MRs were computed between strategic groups

-6-

(obtained by cluster analysis, as in Harrigan (1985)) for every pair
of adjacent years. These MRs are shown in Table 3. From this table

Insert Table 3 about here

it is evident that there is considerable stability in strategic
grouping structure, in terms of Scope (MR=0.97) and Financial Strategy
(MR=0.93) variables. On the other hand, the strategic group struc-
tures based up on Marketing Strategy variables (average MR=0.54, mini-
mum MR=0.36, maximum MR=0.81), and Production Strategy variables
(average MR=0.64, minimum MR=0.31, maximum MR=0.81) change con-
siderably. Thus it appears that Scope and Financial Strategy
variables act as mobility barriers in this industry, whprpa1?,
Marketing and Production strategy variables do not.

Discussion:

The foregoing example illustrated the first six steps of the MOBIUS
procedure. Thus, Scope and Financial strategy were identified as
mobility barriers. It is not clear based solely on this data whether
these are Type A or Type B mobility barriers. Expert opinion should
enrich the identification of the nature of individual mobility
barriers. Further, the same variable set may be a Type A barrier for
some companies and Type B for others. Thus even more detailed analy-
sis would be necessary. However, in spite of the absence of the
seventh step of the MOBIUS procedure, our illustration demonstrates
its usefulness in analytically identifying mobility barriers in an
industry.

-7-

In addition, the fact that considerable inter-group, inter-
temporal mobility is noticed tor the Marketing and Production strategy
variables, is important for a planner in this industry. The time
horizon for monitoring and forecasting competitive strategic changes
in terms of these dimensions should be shorter than for mobility
barrier variables.

It should be noted that the MOBIUS procedure does not formally
provide for a measurement of the "height" of the mobility barriers.
However, we suggest that this height not be measured as the distance
between strategic group boundaries in the metric (or measurement spa-
ce) used for clustering. We recommend that this distance be converted
into an effort or cost measure (at least for the Type A barriers, for
the Type B ones, of course, this measure is of no immediate conse-
quence). This would make (a) mobility barrier heights on different
barriers comparable, and (b) allow serious consideration of the feasi-
bility of the scaling of such barriers by particular companies.

In conclusion, the MOBIUS procedure, by providing an analytical
approach to identifying mobility barriers in an industry is likely to
be useful (a) to practitioners in understanding their environment and
in planning and (b) to academic scholars interested in empirically
validating theories of competitive strategies within the paradigm of
strategic grouping (mobility barriers). We also hope that this proce-
dure will spark some new testing of theoretical insights such as, for
example, the relationship between the number and types of mobility
barriers and industry life cycles.

-8-

Ref erences

Caves, R. E. , and Porter, M. E. , "From Entry Barriers to Mobility
Barriers," Quarterly Journal of Economics, 1977, May, 241-262.

Dess, G. G. , and Davis, P. S. , "Porter's (1980) Generic Strategies as
Determinants of Strategic Group Membership and Organizational
Performance," Academy of Management Journal, 1984, September,
467-488.

Harrigan, K. R. , "Barriers to Entry and Competitive Strategies,"
Strategic Management Journal, 1981, 3, 395-412.

Harrigan, K. R. , "An Application of Clustering for Strategic Group
Analysis," Strategic Management Journal, 1985, 6, 55-73.

McGee, J., and Thomas, H. , "Strategic Groups: Theory, Research and
Taxonomy," forthcoming in Strategic Management Journal (1986).

Porter, M. E. , Competitive Strategy, New York: Free Press, 1980.

Sudharshan, U. , Thomas, H. , and Fiegenbaum, A., "Strategic Time

Periods and Strategic Groups Research: Concepts and an Empirical
Example," Working Paper, University of Illinois, Urbana-Champaign,
1985.

D/321

Table I: Match Ratio Computation

Period T? Groups

Period T,
Groups

I

2

j

n

Total

1

l'n

r

12

— C.
In

Nn

2

c2i

C22

— C-

2n

N12

Nli

•

•

1 i

n

Vj . .

•

•

m

c™i

cm2—

C

mn

Nlm

Total

N21

N22

N2,

2n

N

c. .

'li

'2.1

number of companies that belong to group _i in Period T. and
group 1 in Period T2. —

total number of companies belonging to group i in Period T.,

total number of companies belonging to group j in Period T?,

m

the total

(1) Z N = E N2. = N (number of).
i=l j=l companies

(2) EN,.* E \„ . means that at least one new company has entered
• i J-1 ■ i 2i . ,

1=1 j=i or exited the mdutry.

m n
E E

MR =

1=1 1=1 i.i

l/2( E N. . + E N_ .)

TABLE 2: STRATEGIC VARIABLES AND MEASURES

Variable

Abbreviation

Measurement

Units

A) Scope*

(VI) Asset ASS

(V2) Sales SLS

(V3) Advertising ADV

(V4) R&D RD

(V5) Inventory INV

B) Resource deployment
1) Finance

(V6) Current ratio CR

(V7) Ouick ratio QR

(V8) Dividend payout DP
ratio

(V9) Time interest TIE
earned
(ViO) Debt equity DE
ratio

2) Production

(Vll) Capital

intensity
(V12) R&D intensity
(V13) Inventory

intensity
(V14) Cost

efficiency

3) Marketing

(V15) Receivable RSI

intensity
(V16) Advertising ADI

intensity
(V17) R&D intensity RDI

CI

RDI

INVI

CE

Gross book value of fixed asset $

Firm's total sales $

Firm's total advertising $

expenditure

Firm's total R&D expenditure $

Firm's total inventory level $

Current assets over current
liabilities

(Cash and short term recs) over
current liability

(Preferred and common dividend)
over income before extra-
ordinary items and discontinued
operations

Operating income before depre-
ciation over interest expense

Debt over equity

Invested capital dollars over

sales dollars
R&D dollars over sales dollars
Inventory dollars over sales

dollars
Cost of goods sold over sales

Receivable dollars over sales

dollars
Advertising dollars over sales

dollars
R&D dollars over sales dollars

*The scope variables are deflated for inflation. These deflators
were taken from Business Statistics 1981 for producer prices in drugs
industry.

Table 3: Match Ratio by Strategic VariabLes Type

Strategic

Variables Aver.

Type 74/75 75/76 76/77 77/78 78/79 79/80 80/81 (Min.-Max.)

Finance L.O .95 .95 .95 .81 .86 .90
Production .31 .54 .72 .72 .63 .81 .81

Marketing .81 .59 .40 .40 .36 .54 .50
Scope 1.0 .90 .95 1.0 0.94 1.0 1.0
Overall .95 .72 .95 1.0 1.0 1.0 .90

(.

,81-1
.64

.0)

(.

.31-. 1
.51

3D

(

,36-J

.97

31)

(

.90-1
.93

.0)

(.72-1.0)

MR » 1 •*■ no mobility * perfect stability

MR = 0 * full mobility ■*■ perfect instability

Figure 1: Mobility and Desire to Change

Desire to Change

Observed

Structural

Change

High

Low

High

Not
Barrier

Not

Barrier
(Unlikely
to occur)

Low

Mobility

Barrier

(A)

Mobilit

Mobi lity

Barrier

(B)

y Barriers

FIGURE 2 : MOBILITY BARRIERS

1.2 T

0.8

MATCH RATIO 0.6
0.4 ..
0.2 ..

♦- FINANCE
> PRODUCTION
»• MARKETING
a- SCOPE

7475 7576 7677 7778 7879 7980 808
YEAR PAIRS

HECKMAN

JUN95

B—J.T..PI™/ N MANCHfcSTER
HWANA 46962