Historic, archived document
Do not assume content reflects current
scientific knowledge, policies, or practices.
UNITED STATES DEPARTMENT OF AGRICULTURE
In Cooperation with
The Port of New York Authority, New York City
DEPARTMENT BULLETIN No. 1411
Washington, D. C. August, 1926
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES !
By Cuartes E. ArtmMan, formerly Research Agent in Marketing, Bureau of
Agricultural Economics 2
CONTENTS
Page Page
City distribution exemplified by New York Analysis of types of store operation__________ 19
MELLOPOMEAW Area = tos ee ee 1 Classification of store types________________ 20
The metropolitan distribution system _____ 2 Pxbent ol dataasa. esas bs Oe TS 20
Basis of analysis of distribution expense_-_-_-_ 4 Manner of making comparisons___________ 21
Collection of original price data_________ 5 PCHTUS IMC TGR Ole alae ee oe re 21
Adjustment for shrinkage in retailing____ 6 Differentials showing contrasts in distribu-
Method of data analysis________.________ it OMe pense; peli care = a oo eee 22
Analysis of commodity differences_________- 7 Differentials showing contrasts in prices to
Contrasts in service requirements_________ 8 COUS TIM CTS ease ee seeee oe Pe ee 24
Size of sale as criterion of service require- Summary of store operation contrasts_-____ 26
THC SH: SETTLES. is ewe PPS ee ee Othe Generaliconclusionst:322. 22221 it 342 Sei 28
Size of sale as an adequate explanation of Implications in commodity differences_______ 30
percentare mMAareinss = 22 ee 11 Influence of common monetary and physi-
Deductions from size-of-sale amalysis______- 12 Cals (Sie re ee ee 30
Apportionment of distribution expense be- Influence of a new unit of distribution_____ 33
tween jobber and retailer_-_____________- 14 How retail prices are set_________-________- 33
Reasons for variability of jobbers’ portion- 16 Summary of application of theory_________ 34
Influence of wholesale price changes on job- Limitations of percentage differentials for
DCESHS DrGad es =e eee Se ee 17 COMMMALINE SECS sara soa ee ae ae eee 36
Conclusions regarding variability of job-
DELS PRICE Spreads es es ese ee se 19
Marketing fresh fruits and vegetables presents certain difficulties
that are intensified by the perishable nature of these commodities.
This makes it especially important that efficient methods be employed
in distributing them to consumers. An attempt is made here to
explain some factors which account for the large proportion of the
expenditures of consumers which under present methods is absorbed
im the expenses of city distribution.
CITY DISTRIBUTION EXEMPLIFIED BY NEW YORK METROPOLITAN
AREA?
The analysis deals with distribution of perishables in the metro-
politan area in and about New York City, as exemplifying conditions
that prevail generally in urban centers throughout the United States.
The wide geographical extent of producing regions which supply the
1 This is one of a series of marketing analyses made through cooperation of the U. S. Department of Agri-
culture with the Port of New York Authority. ; ;
2 Acknowledgment is made to H. D. Comer, formerly Research Agent in Marketing, for assistance in
the statistical analyses and interpretations included in this bulletin. ee
3 For description of the New York marketing system and statistical tables, see ARTMAN, C. E. FOOD
COSTS AND CITY CONSUMERS, New York. 1926,
86186°—26——1
2 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
perishable food requirements of this area gives the New York City
market national importance from the standpoint of producers.
Similarity of distribution methods for serving this metropolitan popu-
lation to those employed in other large cities makes analysis of these
methods likewise a matter of general interest to consumers.
More than 180,000 carloads of fresh fruits and vegetables, having
an estimated wholesale value exceeding $200,000,000, were shipped
or hauled in the calendar year 1923 for consumption in the New
York market. Ninety per cent of this food supply came from pro-
ducing sections ranging from 30 to 3,000 miles distant. Over one-
half of the total was transported 500 miles or more. Neighboring
States produced only about 30 per cent of the total. Over one-
fourth came from the Pacific coast, and one-seventh from Florida.
The average length of haul for perishables consumed in the New
York market area in 1923 was 1,500 miles.
THE METROPOLITAN DISTRIBUTION SYSTEM
The system by means of which perishables are distributed from the
New York wholesale market to consumers in the metropolitan area is
illustrated by Figure 1. Most of the fresh fruits and vegetables for
the entire metropolitan area pass through a highly centralized whole-
sale district on the lower west side of Manhattan Island. From this
wholesale market the produce is hauled by motor truck or team to five
jobbing markets in different parts of the area. The car-lot receipts of
the wholesale markets are thus broken down into jobbing lots con
venient for handling by jobbing firms. The jobbers in turn split up
their purchases into small-sized lots required by individual retailers.
The principal function of the wholesaler is to receive certain com-
modities in large quantities from the producing points, whereas the
function of the jobber is to assemble from various wholesalers a con-
siderable variety of different commodities in relatively small quanti-
ties. The retailer carries the process of breaking up the shipping
units one step further, and expands greatly the variety of articles
which he distributes.
In distributing perishables in the New York metropolitan area, the
sreater part—from 75 to 80 per cent of the total receipts—passes
through retail stores. The remainder is disposed of by pushcarts,
hucksters, hotels, restaurants, and other agencies. Consideration is
given here to this major group of retail stores. These are subdivided
into three general types: (1) Independent grocery stores, carrying
perishables as an adjunct to their grocery business, which comprise
nearly three-fifths of the total number of food stores; (2) specialized
fruit-and-vegetable stores which handle no other commodities, and
which comprise about one-fourth of the total number; and, (3) chain
grocery stores, whose number in the metropolitan area is estimated
to be approximately one-fifth the total number of retail food stores.
The trade of the unit and chain grocery stores is estimated to consist
of about 20 per cent fruit anc vegetables, whereas these commodi-
ties constitute practically the entire business of the specialized fruit-
and-vegetable stores. Giving consideration to the numerical impor-
tance of these three store types and the proportion of perishables sold
by each, it is estimated that each type handles approximately the
following proportion of fruits and vegetables retailed through
3
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES
poejeijueou0s ATYSIY B 0% ‘so[lUL QOe‘T SuIsVIoAV ‘soduLysrp Buoy, poddrys oe12
x
N
SS
SANS OZNOO ZW DS WO
\
S010JS 9TGe49S0A-JIN.Y poztperoods pue $9108 4109013 [edo, YSNOIY Ss1ouINsuod 04 Pe NqiIystp ore sydeooer
[8407 94} JO SYIANOJ-0o1y} UeYy} o10py ‘solouese surprejot JO sodAy [B19Aes AY UOTYNIAYSIp JO} SJoyIeuL ZuIqqol peuorse4 ysno1g} eoueyq} ssed Aoqy, (
SINVYUNVISAY
“3 SWuI4 SEI
S13.LOH SNOANW1139SIW
SWul4 os
MUAUVMAN
Gi, | SidIaDau
SWuI4 O8 T1V dO
LYOOASSNV9
%O06
syadnNdOud 1V9071
LAYXYVW
SAIVSATOHM
SWul4 SII HMYOA MAN
LnNoavItVWM
SHL
SAYOLS
Sa1EsVLa93A - LINYS
(AN 4940249 ) SWUId OSI
S3YOLS AYRDONO *A’N NMOL-NMOG
0OoO'O!
SYS TIVLAY sYsaegor oog
X1ddNS AIGVLa9OsSA ANV LINYS HSSY4 SLI
SALNGIYLSIG GNV SSAISDSa LOIWLSIG MYOA MAN SHL MOH
SYAIVSSIOHM OSZ2
“JOYIVUL o[esojoyM
Bore ULvp]odoIjouL FOX MON oy} Ul pournsuod sopqeysiied oy} JO yuoo Jed AyoUIN—'T ‘or 7
4
L
y
y
4
Y
y
CLI TAT Ls
2
°
JZ
Nj
\
Ne
N
\
N
N
N)
N
NJ
X
N
N
N
N
N
s
Ny
N
AN)
N
N)
\.
N
N
N)
N)
N)
N)
N)
N
N
N)
N)
N)
N)
N
N)
N
N)
N)
N
N
N
N
N
N
N
N
NJ
N
N
N
N
N
NJ
N)
NJ
N
N
N)
N
\
N)
x
N
N
\
Ny
N
yj
VY)
=x
2)
—]
4
eK
xe
nn
ed
ul
<
3
4 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
metropolitan stores: Fruit-and-vegetable specialists, two-thirds of the
total; independent grocery stores, one-fourth of the total; and chain
erocery stores, one-tenth of the total.
DISTRIBUTION OF FRESH FRUITS
AND VEGETABLES
New York Metropolitan Area
Passaic : bt OP pa
»
——f
) ,
Hy i. @
¥ Bayonne)
PUES
STATEN
<3 oS LAND \\
Primary Wholesale Market
©) Jobhing Markets: (W, Wallabout, Brooklyn; G, Gansevoort, Manhattan;
H. Harlem, Manhattan, N. Newark, N.J-)
@ Retail stores included in this study
Fic. 2,—From the primary New York wholesale market, P, channels of distribution lead through
regional jobbing markets—W, G, H, N, (Wallabout in Brooklyn, Gansevoort and Harlem in
Manhattan, Newark Market in Newark, N. J.) to local retailers. Location of stores supplying
retail price data is shown by dark circles.
BASIS OF ANALYSIS OF DISTRIBUTION EXPENSE
_n attempting to determine the influence of various factors on the
expense of distributing perishables, the plan was to collect a large
number of original price records for several commodities from stores
~
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 5
representing different kinds of management, varying degrees of
service, and diverse types of trade, in a number of representative
localities in different parts of the New York metropolitan area.
COLLECTION OF ORIGINAL PRICE DATA
Nearly 14,000 sets of individual price records were made, covering
a continuous period of 16 months, from February, 1923, to May,
1924, inclusive, records being taken for an identical day in each
week of this period. Fifty retail stores cooperated in supplying
retail prices. Most of the quotations were obtained from 30 re-
tailers located in Manhattan, Brooklyn, Bronx, Newark, Passaic,
and Elizabeth.* The geographical location of these stores and their
relation to the primary wholesale market and the intermediate job-
bing markets, are indicated in Figure 2.
The retail price quotations were collected on Friday of each week
by regularly accredited reporters who were actively interested in the
project. They were principally housewives and representatives of
women’s clubs or home organizations in the respective localities, who
obtained the prices from shops in their immediate neighborhoods
where the family trading was done. Wholesale prices for corre-
sponding dates were taken from the daily market reports of the
United States Department of Agriculture, with an allowance of one
day’s lag for distribution from the wholesale market to the retailers.
Representative dealers in four jobbing markets of the area supplied
jobbing prices for the days corresponding to the retail quotations.
The commodities chosen included the fruits and vegetables which
are consumed in largest quantities as shown by records of annual
car-lot receipts in the New York market. The selection was made
to embrace only articles which are generally sold in retail stores,
and of a sort that admitted ready identification and comparison.
With subdivisions due to differences in varieties or sources and in
methods of marketing, the commodities were ciassified in the 14
groups shown in Table 1
TaBLeE 1.—Commodity classes and number of price quotations, New York metro-
politan area, February, 1923, to May, 1924
I
| Number | Number
Commodity | of quota- Commodity of quota-
| tions tions
PRiiOrnia OfNgEs = ee | ZOt7a\, Southern cabbage 222 sh 688
picllow: ONIONS ts 952s Jeet ros 21 | kore) Southernypotatoes._- 2 )55.) se 538
Nertherh potatoess=— = a GOGH WOSLCEMIGLEGCO. > 484
Ce ADCS 7 te oe ee ee kee 4S 5e | Cantaloupessss i on 2 eres Le EY) os eee 353
LEO GGA GS a POY IE CAGNCS mrp rear tis eet Bie Wa ae a 270
INECEINC LEH CO tet en ee if 1G ANY DIte ONION Sah ae ene ee te aT 237
MICHEL DOLALOES 2) Set i Pee St 1, 252 —
mimeanorn Capnare 94-5 2-2 = 2 = 774 Total, 14 commodities___________---- 13, 971
Retail prices were stated in terms of quantities prevailingly quoted
by retail stores—by the pound, quart, peck, head, dozen, or single
unit, or for an advertised value such as 25 cents worth. These
4 Aid in collecting price data was given by: Women’s City Club of New York; New York League of
Women Voters; Teachers College of Columbia University; Henry Street Settlement, New York;
Pratt Institute of Brooklyn; Brooklyn Society of Ethical Culture; Contemporary Club of Newark,
N. J.; Housewives’ Economic League of Passaic N. J.; Women’s Club of Elizabeth, N. J.
6 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
prices were converted to the original package or hundredweight basis,
to correspond with units quoted in the wholesale and jobbing markets.
Reporters gave description of sizes and qualities in each case, so that
prices for the various marketing stages would represent corresponding
grades of goods. All possible care was taken to procure typical sell-
ing prices that were truly representative of the situations studied.
By restricting the reports to standard varieties and grades, the error
from comparing unlike conditions was reduced to a minimum.
ADJUSTMENT FOR SHRINKAGE IN RETAILING
To make the retail prices exactly comparable with wholesale
prices, a correction was made for shrinkage of contents of packages
in retail selling. This was done by multiplying the mean retail
package price by the percentage of the shipping package which is
prevailingely sold at retail. This shrinkage factor does not allow for
the indefinite and irregular losses that occasionally arise because of
deterioration or decay, but only for actual physical shrinkage in
contents of original packages. The adjustments were made as
noted in Table 2.
TABLE 2.—Adjustment for shrinkage in retail selling, New York metropolitan area,
February, 1923, to May, 1924
Percent-
Gross | Quantity} age of
Commodity Shipping unit weight | usually | shipping
per unit!} retailed unit
retailed
Pounds | Pounds | Per cent
Northern: potatoes:22 2 se ee ee een ee ee 100 pounds________}. 100 95 95
Southern potatoes: =<. 2 Sh ee ee eae 5 (pt ea ese 100 92 92
CaliforniacorangeS:— t_ oe Pes Pe eee IBROX. oes 75 75 100
Reaches =. 28:2 2 oe ee Cratée af eens 35 33 94
Sweet. potatoes 22: five. Tes BSNS Se see eee 100 pounds____--_- 100 90 90
Cantalowpes = 22 See ie ee ee ee Cratet ons 2 Sere 60 60 100
Boxed “apples: Se ae Se eee ear eee BLA BOWS. 2 Sat coe 40 40 100
Southemn: cabbage. 22 Sscess se eee Sea ee 100 pounds_-___---- 100 90 90
Barreled apples =) wees ee 2 ee ane eee ee Barrels ee 150 135 90
MD ASLO TITAS E BUT CC eee a a ep eee ne iEFampersee 22s = 34 34 100
Western: lettuce 22.5 a3 re Aa eee ae eee Crates =>. Seenes 48 48 100
Yellow Onions 2s 222 2220 sate Pas as ae Eee ee 100 pounds_______: 100 95 95
INorthern: Gabbagess 222 = eee ee ane he teak ewan Cc (oe Se Nery 100 90 90
Wihite oniOnS23<5 =) oe ee ee ee ee ae 5 (0 OR Se ee 100 95 95
1 As used in converting original quotations.
Against each retail quotation was matched the wholesale figure
for the corresponding date. Jobbing prices also were added for inde-
pendent unit stores. No jobbing prices were entered for chain stores,
since these buy their goods mainly in the wholesale market, whereas
unit stores deal through jobbers. Especial care was taken to have
each of the stages in the price series represent identical goods and
conditions. In the absence of bias in obtaining the price data, errors
in particular quotations should compensate one another so that the
records for the succession of weeks which are included in the analysis
period should be an accurate representation of prevailing price con-
ditions. The care taken in selecting data, the variety of conditions
included, the length of time covered, the volume of data obtained,
and the carefulness in the method of analysis, unite therefore to make
this analysis an accurate and representative presentation of actual
conditions.
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES t
METHOD OF DATA ANALYSIS
The individual price records, received on printed cards, were tran-
scribed to working sheets, and thence to punched cards for machine
tabulation. The information was thus classified according to the
_ different conditions to be analyzed. In making comparisons of the
5
different groups of data, the arithmetic mean of prices was used
because this was found to be the most representative and the most
workable kind of average.
Since the purpose of the analysis is to compare spreads between
wholesale and retail prices, under various market conditions, it is
necessary to relate these prices to uniform bases. The method com-
monly employed for this purpose is to express price spread in terms
of a uniform outlay of $1 by the consumer. In this case the spread
represents the number of cents of the consumer’s dollar that are
absorbed by the expense of city distribution. In other words, it is
the difference between the value of the dollar’s worth of goods at
retail and the value of the same goods at wholesale.
With such a base the spread is conveniently expressed also as a
percentage margin. This may be derived by dividing the price differ-
ence for any quantity of goods by the retail price of the same quan-
tity. Thus, if R represents retail price, and W represents wholesale
price for a given quantity of goods, the price difference, or spread, is
mencnted by R-W. The percentage margin is then represented by
R
A preliminary inspection of the average percentage margins in the
various classifications of data was made to find out which of the
different market factors considered were accompanied by the greatest
margin differences. The widest divergence was found to exist be-
tween the different individual commodities in the series of 14 articles.
Next to the contrasts in commodities, the divergence of margins was
greatest in different types of retail stores, as indicated by differences
in management and in selling policy. These two sources of con-
trast overshadowed all other distribution factors that were consid-
ered. The analysis was therefore focused upon the factors which
appeared to account for the greater amount of difference in distribu-
tion expense: (1) The nature of the commodity and (2) the type of
store operation.
ANALYSIS OF COMMODITY DIFFERENCES
The composite means of retail and wholesale prices of each com-
modity as a whole for the entire period embraced in this analysis are
shown in Table 3. Jetail store prices are weighted according to the
importance of different store types, and are adjusted for shrinkage in
retailing. The average distribution expense per package is the differ-
ence between mean retail price and mean wholesale punee. It is
expressed as a percentage of the retail price in the last column.
These average prices and percentage margins are the basic figures
used in making the commodity analyses.
5 The work of tabulating and classifying the large number of compilations was done by the machine
Bo ulation section of the Bureau of Agricultural Economics in Washington, under the direction of E. J.
ay.
8 BULLETIN 141i, U. S. DEPARTMENT OF AGRICULTURE
TABLE 3.—Mean prices and percentage margins, New York metropolitan area,
February, 1923, to May, 1924
Mean Mean Per-
Commodity Physical unit retail | wholesale} centage
price price margin
Per cent
INorthern:potatoes22= 2a" 5 Sek ae eae 100 pounds___-___-_- $3. 87 $2. 43 37
Southernipotatoes =. ene ee GOs. Ae eee 6. 76 4.18 38
California OFan gees ee ee eee ee ‘BOxtS Sa See 8. 21 4. 86 41
RGA ches us ee: ee Sao en As Sah pes ae Crates eee 3. 97 2. 20 45
SWeet POtCAtOes as eS ae eae ee eae 100 pounds_______- 8. 00 4, 44 44
Cantaloupesesi2 5 ee Ss Be Oe ee Cratetc sere be 4.57 2. 45 46
Boxed apples: 2222 8 ea er eee BOK Sete ie ee 4.35 DSR 46
Southernicabbagen ws eS ee eee 100 pounds________ 8. 42 4.42 48
Barreledvappless ies 5 2 oe ee eae Barrel.a2. su? 2. 10. 76 5. 53 49
WMastermn lethuCe se os ee ek eee ee Em perasaea ee 4. 43 2. 16 51
Wiesternilet tice ts ss Se ee ees Peers Grate: esse 7. 05 3. 37 52
Mel OWw*OMI OMS. hae ee Se ele es | GOe Sas eee 6. 70 ealea 53
Northern: cabbage aks be oa pyc ees See ee ler 100 pounds_______- 4. 68 1. 96 58
White .onions.c 2" 5. eee ae ee ea lee dO: See 8. 59 3.22 63
14 commodity. weighted mean 255s. See Re | SEES Se ee |e | 45
In the series of articles, there is a range in the margins from a
minimum of 37 per cent for northern potatoes to a maximum of 63
per cent for white onions. The question at hand is to determine
what differences exist in the marketing conditions of these commodi-
ties which are adequate to explain so wide a divergence in the portion
of coe outlays required for retail distribution of such similar
articles.
Some of the conditions which might reasonably be expected so to
influence the manner of handling as to establish these commodity
differences, are: (1) Total volume of commodity marketed through-
out the season; (2) regularity of supply; (8) perishability; (4) vari-
ability in price; (5) value of total quantity inatieoted® (6) comparative
value per unit of commodity. Various detailed tests were made to
ascertain the association of each of these conditions with the variations
in the percentage margins. A great many tabulations were compiled
and numerous diagrams and curves were applied to test the relation-
ships. In no case was the relation sufficiently regular to account
for the margin contrasts. It was therefore necessary to seek further
for an explanation of the differences in the methods of marketing
the different articles.
CONTRASTS IN SERVICE REQUIREMENTS
It was suggested that variations might exist in the extent of ser-
vices required for distributing different commodities, which would
suffice to explain the margin differences. Services rendered by the
retail storekeeper may be considered as of several distinct kinds: (1)
Assembling a considerable variety of different articles in a single place
Peet Banly accessible to consumers; (2) maintaining at all times a
fresh daily supply of articles available in the market; (3) selecting,
grading, and arranging goods for retail sale; (4) displaying his stock of
goods for inspection and selection by customers; (5) breaking up the
original packages or jobbers’ units into smaller quantities required
by the retail trade: (6) waiting upon individual customers.
A typical display of fruits and vegetables of a Manhattan grocery
store is shown in Figure 3. The different services maintained for
the convenience of consumers, require on the part of the retailer
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 8)
the expenditure of a great deal of time and effort. In addition to
the general services enumerated in the first five items, the filling of
customers’ orders alone involves a number of separate acts. These
include (1) taking the goods desired by the customer from the dis-
play space or from the reserve stock of goods, (2) weighing or measur-
ing or counting the desired quantity, (3) putting this up in a satis-
factory package for delivery to customers, (4) computing the amount
of the sale, (5) making a record of the transaction, and (6) obtaining
payment. These several acts have to be repeated with each sale,
pene the personal attention and time of the retailer or his em-
oyee.
fe Retail sales in the metropolitan area are prevailingly made in
small quantitites of a few pounds or quarts, a dozen, or a single unit,
according to commodity. Individual sales range in volume from
Fig. 3.—A typical retail store display of fruits and vegetables
1 pound to 7 or 8 pounds of articles sold by weight, from 1 to 5 quarts
Shen sold by measure, half a dozen to a dozen fruits, 1 or 2 heads of
lettuce, 1 or 2 melons, and corresponding small quantities of other
perishables. From 15 to 50 separate retail transactions are usually
required to dispose of a single shipping package or hundredweight.
Although sales to customers are occasionally made in larger volumes
than these, this small-sized retailing is the prevailing practice in the
New York metropolitan area, and exists to a great extent in other
large cities. Accompanying the multiplication of services required
mn such small-unit retailing, there is also a considerable element of
shrinkage, arising from division of original packages into numerous
small-sale units. The smaller these units, the more numerous are
the opportunities for loss from this item of shrinkage.
The quantity of goods bought at one time by an individual cus-
tomer bears little relation to the amount of service and attention
86180 —26 —_2
10 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
required from the retailer. Practically as much of the storekeeper’s
or clerk’s time is needed to wait upon a customer who purchases a
small quantity of goods as for the person who makes a large purchase.
It is a reasonable supposition, therefore, that the selling expense
should be fairly uniform for each retail! sale, irrespective of its size.
SIZE OF SALE AS CRITERION OF SERVICE REQUIREMENTS
Careful inquiries in the retail trade revealed the fact that the size
of the average retail sale varies distinctly with different commodities.
These specific inquiries embraced the extensive experience of two
large metropolitan chain-store systems, several independent retailers,
and a considerable number of individual families whose size of pur-
chase were included with the original price data.
Although retail prices are advertised in terms of*uniform physical
units, such as the number, quart, head, dozen, or a similar magnitude,
actual sales are made in various multiples of these individual units.
To be sure, there is a great deal of variation in the size of individual
sales on account of differences in buying habits of individual cus-
tomers. Yet sufficient regularity exists throughout the retail trade
to establish a typical or prevailing size of sale for a given commodity.
In retail selling the significance of the variation in quantities of
goods sold per sale lies in their relation to the money value of the
goods so sold. The value of the retail sale is determined by two
variable elements. One of these is the physical quantity of goods
disposed of; the other is the price per physical unit of goods.
From the representative sources referred to, which may be con-
sidered typical of retailing practices in the metropolitan area, it was
possible to ascertain with considerable definiteness the prevailing
range in size of sale for each commodity. The size of the typical
or standard retail sale for each commodity may reasonably be re-
garded as approximating the mid-point of the prevailing range. Use
of the mid-point 1s justified as an approximate indicator of prevailing
size of sale, in view of the fact that extremes, such as unusually small
sales and unusually large sales, were excluded from the ranges given.
This method involves some degree of approximation, but in the ab-
sence of more definite data for arriving at specific accuracy, which
could be obtained only by recording exact details of a large number
of individual sales under representative conditions, the method
here employed is justified as the best that was available.
Retail price per unit of goods, which is the second variable in the
value of the retail sale, was computed uniformly per pound, on the
same basis as that used for expressing the size of sale. Retail price
per pound was calculated from the mean retail price of each com-
modity as derived from the original quotations. These were first
converted to the package or hundredweight basis and thence to the
mean price per pound.
The value of the standard retail sale was calculated as the product
of the mean retail price per pound and the number of pounds in the
standard retail sale. The prevailing range in size of sale for each of
the 14 commodities, the mid-point of this range, and the value of the
standard retail sale, are shown in the first three columns of Table 4.
Spread between the wholesale cost of goods and their retail value
may be computed either by subtracting the wholesale price of a given
quantity from its retail price. or indirectly by multiplying a given
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 11
retail value by the computed percentage margin for the commodity.
The percentage margins and the resulting price spread in terms of the
standard retail sale are shown in the last two columns of this table.
TaBLE 4.—Size and value of standard retail sale, percentage margins and price
spread per sale for 14 commodities in all store types, New York metropolitan area,
February, 1928, to May, 1924
Prevail-
7 Price
ing Tid. Mean | Value of 5 s
Goumedie= range in ee of | _ Tetail | standard Pet ee spies
ae at range nee Der retail margin standard
eae pound SHE retail sale
Pounds | Pounds Cents Cents Per cent Cents
Nori neriepotatoeS== 5 =. ae eee 5-8 6. 50 4.1 26. 7 37 959
SOUGHT Msp OLALOCS saan ae ee eee 3-4. 5 3h (5s 7.4 27.8 38 10. 6
Californiaiorancesa se ee ee 2-3 2. 50 11.0 Qiao 4} ik 3
SWeCl DOULLOES: = s=e 22a se tee a ee ae 2. 5-3 Fe US) 8.9 24. 5 45 HO
RCAC CS eee ee re WS a! 1. 5-3 2.20 11.9 26. 8 44 Ipsn|
IBOXCAE ap LES see er a en ee ee ee 1. 5-3 2a20 10.9 24.7 46 11.4
@antalowpes= = -2- sec fee eee 3-3. 5 O40 7.6 ZA.7 46 11.4
Southernicabbare ss. 52-522. Sareea 2-3. 5 Qo 9.4 25. 9 48 12. 4
BarreledsapDles se sees ae a ee Se 25-3. 9 3. 00 8.0 24.0 49 11.8
iastern-letiices=.~ 5 = 2 Se eee | 1. 5-2 1.75 13.0 22. 8 51 TRS
Western lettuce <.- 2s 32. Sees 1-2 1. 50 14.7 22.1 52 11.6
Mellow, OniOnSae==5s- anes See eee 2. 5-4 525 eal 23a 53 WY
Northern cabbagec 2 = = 3. saeeeerete es _ 3-5 4.00 5.2 20.8 58 12.1
WWihtte: ONIONS SY 2.232 So. 2S ee 1. 5-3 Pas PASS 9.0 20. 3 63 12.8
Wieighted=m Gan! sae» sawetemametnetomet fo 3. 28 Text 2588 45 11.3
1 Adjusted for shrinkage in retailing.
SIZE OF SALE AS AN ADEQUATE EXPLANATION OF PERCENTAGE MARGINS
When price spread is expressed in cents per standard retail sale,
instead of being shown as a percentage margin, it is seen to have a
remarkable degree of uniformity among the 14 commodities. This
spread varies only from a minimum of 9.9 to a maximum of 12.8, a
difference of less than 3 cents per sale. Moreover, there is a high
concentration about the mean of 11.3 cents for the series.
The mean size of the standard retail sale for the 14 commodities
weighted for each commodity according to its total volume marketed
is approximately 314 pounds. There is a variation in size of sale
among the different articles of 5 pounds, nearly twice the mean
size of the standard sale for the series.
The mean value of the standard retail sale for the series is 25.3
cents. There is a total variation in its value among the 14 articles
of 7.5 cents—from a minimum of 20.3 to a maximum of 27.8. This
is less than one-third the mean for the series. The value of the
standard retail sale is thus decidedly less variable than is the size of
sale. This results from the fact, as shown graphically in Figure 4,
that variations in size of sale within the series are generally offset
by reciprocal variations in price per pound, in such a manner that
variation in their products is diminished materially.
Furthermore, despite high variation in size of sale and a con-
siderable variability in its vaiue, price-spread per sale is found to
be very nearly constant for each article of the series. The relations
are shown graphically in Figure 5.
Thus it appears that commodities with low value per retail sale
require practically the same monetary amount per sale for distribu-
tion as do articles with high value per sale. Variations in percentage
12 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
margins result from a combination of uniformity in price-spread per
sale and variability in the value of the sale. It is the variable
amount of money prevailingly spent for the consumer’s single pur-
chase, accompanied by a constant spread per sale, which fixes the
variable proportion of the consumer’s expenditure which is absorbed
in city distribution. Variations in size of the standard retail sale,
accompanied by constancy in distribution expense per sale, thus
yield an adequate explanation of differences in percentage margins
within the series of commodities here analyzed.
SIZE OF RETAIL SALE AND
RETAIL PRICE PER POUND
FOURTEEN LEADING FRUITS AND VEGETABLES
NEW YORK METROPOLITAN DISTRICT, 1923-1924
MEAN RETAIL PRICE PER POUND - MEAN NUMBER OF POUNDS
IN CENTS PER SALE
iS 10 5 18) 18) i Vee os 4-.-5 6 7
W. LETTUCE
E. LETTUCE
PEACHES
BOXED APPLES rare : Weighted Mean
WHITE ONIONS
CALIF. ORANGES
S. CABBAGE
SWEET POTATOES
BBLD. APPLES
CANTALOUPES
YELLOW ONIONS
Weighted Mean
S.POTATOES
N.CABBAGE
N. POTATOES
Fic. 4—Commodities such as lettuce, sold in small retail quantities, have high retail price per
pound, whereas staples like potatoes, which are retailed in larger-sized lots, have low retail
price per pound
DEDUCTIONS FROM SIZE-OF-SALE ANALYSIS
This analysis of variations among different commodities in the
margin or distribution expense per dollar’s worth, demonstrates that
the dominating factor in the variability of percentage margins is
the size of the prevailing retail unit of sale. The quantity of goods
prevailingly taken at a time by the individual customer is definitely
and regularly associated with the proportion of the consumer’s
outlay which is required in the services of city distribution.
The percentage margin signifies the amount of money im a retail
dollar’s worth of goods which is absorbed in these services. A
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 13
dollar’s worth of goods, expressed in number of retail sales, is the
reciprocal of the value of the retail sale. It requires only four
individual sales to dispose of a dollar’s worth of goods if the value
of each sale is 25 cents, whereas five sales would be required if the
value of the sale were reduced to 20 cents. Thus, because of the
prevailing uniformity in distribution expense for each sale, regardless
of its size, the amount of money required to sell a dollar’s worth of
goods varies directly with the number of sales which the retailer
must make to receive $1. As the number increases the distribution
SPREAD PER RETAIL SALE AND SIZE
OF RETAIL SALE
FOURTEEN LEADING FRUITS AND VEGETABLES
NEW YORK METROPOLITAN DISTRICT, 1923-1924
SPREAD PER RETAIL SALE IN MEAN NUMBER OF POUNDS
CENTS PER SALE
(e) | 2 FAM, OM CEU Gy? .o7
W. LETTUCE
E.LETTUCE
BOXED APPLES
WHITE ONIONS
PEACHES
CALIF. ORANGES
S SWEET POTATOES
S.CABBAGE
BBLD. APPLES ighted Mean
7 YELLOW ONIONS
CANTALOU PES
S.POTATOES
N.CABBAGE
N.POTATOES
Fic. 5—Commodities differ widely in the prevailing size of retail sale, but the difference between
wholesale cost and retail value per sale is nearly uniform
expense per dollar increases likewise. Hence the greater the number
of sales per dollar the greater is the proportion of the dollar which
is required in distributing the goods; that is to say, the greater is
the percentage margin. Differences in margins are thus found to be
directly due to differences in size of the retail sale.
The practice of marking retail prices so that they will yield for
different articles a fairly constant money spread per sale, regardless
of the size of sale, thus explains the differences in percentage margins
so adequately that the influence of other factors which might appear
to be effective is obscured by this one. Such characteristics as total
14 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
annual volume, total value, regularity of supply, perishability, or
variability in wholesale price, in so far as they influence the percent-
age margins, are of secondary significance, because they operate
indirectly through prices. The dominating factor is the size and
value of the standard retail sale.
APPORTIONMENT OF DISTRIBUTION EXPENSE BETWEEN JOBBER
AND RETAILER
Two kinds of distribution services are included in the spread be-
tween wholesale and retail prices. One set of services is rendered
by the distinctive retailing agents, whereas the other is performed
by intermediate jobbers who break up the wholesale shipments into
lots of convenient size for handling in retail stores. :
To ascertain the relation which these two portions of the distribu-
tion expense bear to each other, the spread per retail sale was split
into its two component parts. The portion attributable to the
jobber was measured separately from that of the retailer. The
retailers’ portion of the price-spread is the difference between retail
price and jobbing price; the jobbers’ portion is the difference between
the jobbers’ selling price and the cost of goods in the New York
wholesale market. ae of these two portions is restricted to
independent unit stores, since in chain stores the functions of jobber
and retailer are performed by a single agency.
The retailers’ portion and the jobbers’ portion of the total price-
spread per standard retail sale in unit stores are shown for each
eo ramiolite in Table 5. Comparison of the two portions which make
up the total spread per sale shows that the retailers’ part is more
nearly constant. throughout the series than is the jobbers’ portion.
The coefficient of deviation from the mean jobbers’ s saan ofs24
cents is 18 per cent of the mean spread. For the retailers’ portion,
on the other hand, the coefficient of deviation from the mean re-
tailers’ spread, 9.7 cents, is only 8 per cent. Deviation of the
jobbers’ price-spread from the mean for the 14 articles is over twice
as great as the deviation of the retailers’ price-spread from its mean.
TaBLE 5.—Retailers’ and jobbers’ portion in standard retail sale for unit stores only,
New York metropolitan area, February, 1923, to May, 1924
Fi
|
Value of ’ .
: Total | Jobbers’ |Retailers’
Commodity see’ spread | spread | spread
——__—_— sss aaa SQ _° _cc“l—S————————— =
ms Ss
@alifornmiaioran cess. — 22.2 ae ae a ee a 28. 0 11.8 2.0 9.8
Southernypotatoes =: 3S 2 Se ee ee ee 27.8 10.8 1.6 9.2
INorthern-potatoes..+- 3. = ee a eee 27.3 10. 4 Ie 8.7
IRC ACHES ata. Saat: ae eek sere ee a eee ee 27.0 12.2 2.5 9.7
Southernvcabbagens'- = -18s. 2 ys ieee ee Se EE ea ee ee 26. 4 12% 2.9 9.8
Sweetrpotatoes.—- a 55 Sirs ae Oe Pe as Ny ae 25.9 11.9 1.8 10.1
IBOXCGap PleSHeea es Ke eS ee ee ae eee 25. 2 11.8 F457 9.1
Cantaloupeswites. © 22403 0a. =. setae fe eee ee eee 24. 7 11.6 2.0 9.6
IBALTCLE CADDIES Sa sees os oe ee ee ae ee were 24. 6 12.3 2.0 10.3
i asternviettwcest oes. 2) eh See Bebe Pee AISA Ae eee 73h > iba 3.0 9.2
WellowOnlons See es Bee ak ne i 23.4 12.4 1.6 10.8
SWWESTErI Le CEICO me een a ee ee 2 ee eee ee ee ees 22.4 11.9 1.8 10.1
INortbernicabbave saves) Soe . Sar at es A eee Lee ee ee 21.6 12.5 1.9 10.6
WI tevOnl ONS te ener eee a oak Rc ie ae eee 20.9 13. 4 20 10.9
14-commodityawelghtedimeane 252-22 bee he 25. 9 11.8 2.1 whl
;
15
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES
rar Ouse
SS ae ea
aPea Oe
Be bord
meus
HOO 8 oO
a5" be a
Fo @ bee =
D
ran
io)
fc}
”
os
retailer accounts chiefl
shout the commod
icle the split up of the
e component parts.
ereater regularity in
®
a
aD)
a
ere
~~
fas}
ac
aS)
i) =
B Se
a Gag
nN o> a)
PEO eae ce
gdh aa B
colle yi ges
a D> joule)
Oem PRG
ogo po
fo) S Sy
GCTnADo Dn
Bg of OD
PrPHYy od
Oe Oe aa
SOS SD aes
SEOs thd
~~ &
Bao 5 et
He)
Qa 8.8
ROHS a ©
SS n gos
Se ere none
O°8.5 6 Bra
OO |
CS Es bos
eal ol
Sn Sey &
SPLIT-UP OF CONSUMER'S OUTLAY PER STANDARD RETAIL SALE
( UNIT STORES )
FOURTEEN LEADING FRUITS AND VEGETABLES
NEW YORK METROPOLITAN DISTRICT.
wig2s=1924
CENTS
30
25
20
15
10
>
o,
?
XX
5
2
5%
2
2
om
eaeeeacrecirars
SKEREEGG
KS
WHITE ONIONS --
N. CABBAGE ------
WWE THE UK eS SS Soc
YELLOW ONIONS [eee
Eo EEmnuiC Ea —-<-
BARRELED APPLES
4
oe
S24
% SERN )
,
2
OO
SKS
OOOO
SEEK
weterens
OR
KK
OOS
SZ
ras
ees
ars
xx
°,
°,
\
t
vp)
LJ
ao
=)
Oo
a
rss
re,
<{
0
BOXED APPLES --- @
ROS SRO
o,
es
00,00,
0,9,
n
uJ
=~
cs
ees
=2
23
Ow
(BV,
+
OQ?
O07
SSR
Nerereceten
\
°,
L>
RRR RRR
SWEET POTATOES
°
S. CABBAGE ------
SIREN
5252
OO
Ox
ost
os
~
SS
oS
SAY
oes
2
$5
2,
>
kK
week
WL
PEACHES ---------§
N. POTATOES -----
S. POTATOES ----—
XOO
BERKS
0.0.0.4
Go
O
K>
RS
CALIF. ORANGES-
ea Retailer
y exists in wholesale value of goods per retail
, but the jobber’s portion varies considerably
J/A Jobber
ae Wholesa/e cost of goods
Fig. 6.—Among different commodities wide variabilit
The retailer’s return per sale is fairly constant
sale.
©
ro
=
°
~~
=)
D
rf
© es
ie D
~ fas]
© 5
ro) 4
Pa sesh
™ < oo
bo go
or Sano
aSnod
SRS g
as 0 ori
Oodow
ee tees
roe 4
qaspq6
tS ORR Gey
OO wm o
g.d_m6b
slis)
Cn ago
oO 82.4
5 Ao 4
ab
oo mS}
Aagduq
3 OO
Senta.
RM aes
SHS as
W~ ose
So So oe
O82 845 8
=) beg e
oak mhes
Ore pe piney
goods sold,
required for jobber’s
he amount required
al uniformity of the
|
16 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
REASONS FOR VARIABILITY OF JOBBERS’ PORTION
Variability in jobbers’ spreads might be explained by demonstrating
its association with variations in size of the jobber’s sale, in a manner
similar to the explanation of variablity in percentage margins. To
test the existence of such associaticn, the prevailing size of the
jobber’s sale was ascertained for each commodity of the series and
this was compared with the size of the retail sale. The figures for
size of jobber’s sale were obtained from the books of representative
jobbers in Brooklyn and Newark, covering transactions extending
over several weeks.
SIZE OF RETAILER'S SALE AND
SIZE OF JOBBER'S SALE
FOURTEEN LEADING FRUITS AND VEGETABLES
NEW YORK METROPOLITAN DISTRICT, 1923-1924
MEAN SIZE OF JOBBER’S SALE MEAN SIZE OF RETAILER'S SALE
POUNDS POUNDS
250) 200. 150 N00) 250 0) 0 I 2c St ID et tO peed
W LETTUCE
| |
E. LETTUCE
PEACHES
Weighted Mean Weighted Mean
BOXED APPLES
WHITE ONIONS
CALIF. ORANGES
# SWEET POTATOES
S. CABBAGE
BBLD. APPLES
CANTALOUPES
YELLOW ONIONS
S. POTATOES
N. CABBAGE
N. POTATOES
Fic. 7.—General symmetry in these two series of bars shows that the quantity of a commodity pur-
chased by the retailer is proportional to the quantity prevailingly sold to the individual customer
The relationship of size of jobber’s sale to size of the standard
retail sale was found to be substantially regular throughout the
series, aS Shown in Table 6. There was a range in the number of
retail sales per jobber’s sale among the 14 articles from a minimum
of 24 to a maximum of 45 sales, but the grouping around the weighted
mean number, 32.5, was fairly close. With 8 of the commodities,
moreover, the ratio was between 30 and 35; with 3 of the remaining
6 it was below 30; and with the remaining 3 it was above 35. A
general tendency is thus apparent for size of the jobber’s sale to
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 17
vary directly with the size of the retailer’s sale, as shown graphically
in Figure 7. Since general regularity exists in the association of
size of jobber’s sale with size of retail sale, and since there is also no
association between jobbers’ spread and retailers’ spread, it is not
possible to explain variations in jobbers’ spreads by the variability
in size of the jobber’s sale.
TaBLE 6.—Relation between size of jobber’s sale and size of retailer’s sale, New York
metropolitan area, February, 1923, to May, 1924
Mean Mean | Number
number | number | of retail
Commodity of pounds} of pounds} sales per
per job- |per retail-| jobber’s
ber’s sale | er’s sale sale 1
Masternlettuce = 2-- 2. 222 et ee ge A ee 54 1.75 31
Wiestermulettces=— 4-05.25) es ee st. . eee 58 1. 50 39
TR GACHIOS roe eee ee ee ee Pee a ie ete ee NR SET Se oe 65 225 28
Sweet;potatoes 4s 222 2-2 et es ae oak 2 se eee eee 75 2. 75 24
Boxed iapplessas* sss aha ase noes oe sees ta So. lp Sk eae 80 2. 25 35
@aliformianOrean G eS 2s peat ees eae Seer re eae dS 2 Sek ehh ee ee 84 2. 50 34
Southernicabbages: A.) sc2bs 228k. tL = eee eee 95 2. 75 31
(ORVATTA OTTO E es ets See Se ee eo ee: Eee eE EE 96 3. 25 30
Barrelediapples-<—. = /=2 22 2-222 == 4... eee eee 100 3. 00 30
Wihite onids + 22te 2352 ee =e 2d _ eyesore 100 94, 5 43
BYZE OSV OTNNO TNS Ress are eee ge ged ee cre fe ae eer a ee eee pene eee ee 110 375) 32
INOnE erie Cab Dao Ce = ae ere eh eee eS 125 4. 00 28
Southern Potatoes. —— = =o i: wee esa cy ep 183 Bb Ui 45
Northern: potatoes.= =... _-_ -- SoS ea ae ee eee 225 6. 50 33
AY TAGY HA FS Las 0 YS g Reta ee tie a an ot a Raa me ea pare 112 3. 28 32. &
1 Allowing for shrinkage in retail selling according to Table 2.
INFLUENCE OF WHOLESALE PRICE CHANGES ON JOBBERS’ SPREAD
ei gaurantee risks due to price variations in the wholesale market
might, with some reason, be expected to influence the selling policies
of dealers in the jobbing market and thus to account for variability
in jobbers’ price spreads for different kinds of commodities. Tests
were therefore made to ascertain if any regular association existed
between the jobbers’ portion in the standard retail sale and varia-
bility in wholesale price.
A measure of variation is required that will avoid the effects of
pronounced seasonal price trends. ‘The usual measures of dispersion,
such as average deviation and standard deviation, are unsatisfactory
for.this purpose. An adequate quantitative measure is required to
indicate comparable price changes in the variable seasons when
different articles are in the market. For this purpose, the wholesale
price of each commodity was taken for an identical day in each week
of the season during which it was officially reported in the New York
wholesale market. The average week-to-week change in price,
either up or down, was determined and expressed as a percentage of
the season’s mean wholesale price for the given commodity.
This percentage is an accurate measure of tendency-to-change in
esc price, which may be used as an index for comparing the
different commodities. The price variability of eastern lettuce, for
example, is shown by its index of 26.4 per cent to be very much
6 Prices for Thursday of each week were taken from the Daily Market Report of the United States
Department of Agriculture. Orange prices were secured from the Daily Fruit Reporter, published by a
private company.
86186°—26——_3
18 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
ereater than that of northern potatoes, whose index is only 3.2 per
cent. The meaning of these figures is that the average week-to-
week change in wholesale price throughout the market season for
lettuce is 26.4 per cent of the mean price for the lettuce season, whereas
for potatoes the mean week-to-week change in wholesale price is
> cent of the average price for the potato season.
¢)
INDEX OF VARIABILITY OF WHOLESALE PRICES
THIRTEEN LEADING FRUITS AND VEGETABLES
PER CENT
O 5 10 15 20 25 30
LOW VARIABILITY
N. POTATOES
BOXED APPLES -
SWEET POTATOES
BARRELED APPLES Be
CALIF. ORANGES -- §
MEDIUM VARIABILITY
S. POTATOES ose
N. CABBAGE
YELLOW ONIONS --&
HIGH VARIABILITY
W. LETTUCE --- =
CANTALOUPES ----
PEACHES a pas
S. CABBAGE - Ss
E.LETTUCE ---
Fic. 8.—Highly perishable and seasonal commodities have high variability in wholesale price, while
staples have low variability. Wholesale price changes, however, do not appear to influence the jobbers’
portion of the retail sale
Naturally a wide range exists in the indices, as shown in Table 7.
‘The commodities are seen to fall rather definitely into three dis-
tinct groups. Five of the articles have a distinctly low index—less
than 10 per cent. Three of them are in a medium group, with in-
dices between 10 and 15 per cent. With the remaining five articles
there is high variability, near or above 20 per cent. Graphic com-
parison of the three groups is afforded in Figure 8. 3
TaBLe 7.—Index of variability in wholesale price, New York metropolitan area,
February, 1923, to May, 1924}
Low | Per cent Medium Per cent High Per cent
|
Northern potatoes -______ | 3.2 | Southern potatoes______ 11.8 | Western lettuce_-_-_____ 19. 6
Boxed apples___________ 4.6 | Northern cabbage_-_-___ 12.8 | Cantaloupes_--_-_______ 22.8
Sweet potatoes__________ 5.4 | Yellow onions____-____- 13"5:|“iPedehess2 4 te 2S 24.1
Barreled apples________- 8.0 Southern cabbage__-_-_- 25.9
California oranges_____- 9, 2 Eastern lettuce_______- 26. 4
1 White onions omitted because of lack of continuous price quotations.
This grouping of commodities according to variability in wholesale
prices agrees in general with their relative perishability. The low-
variability group includes the more staple articles which keep fairly
well. These may be supplied to the market or withheld from it
according to its demands, so that variation between supply and
a
-
,
a
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 19
demand is kept fairly evenly balanced throughout the season. In
the high-variability group, on the other hand, are the distinctly
seasonal articles with limited keeping qualities. They must be
shipped from producing areas as soon as ready for market, and they
are thrown upon the market immediately after arrival. Conse-
quently, because the erratic seasonal changes in supply are not
balanced by corresponding adjustment in consumers’ demand, they
suffer wide price fluctuation in the New York wholesale market.
Does any regular association exist between indices of wholesale-
price variability in these three groups and jobbers’ spread in the
retail sale? Apparently not, judging from comparison of the two
columns in Table 8. Two of the articles whose spread per sale is
lowest are in the middle-variability group; boxed apples, which have
a bigh price spread, are in the low-variability group; and three
other articles which also have high price spread per sale are in the
high-variability group. From the data on which these comparisons
are based there appears to be no regularity of association between
variability of wholesale prices and the portion of the consumer’s
outlay required for jobber’s service. The results are generally
)
|
negative.
TABLE 8.—Relation of jobbers’ price spread to wholesale price variability, New York
metropolitan area, February, 1923, to May, 1924
Jobbers’ | Whole- | Jobbers’ | Whole-
spread | sale price | spread_| sale price
per retail varia- per retail} varia-
sale bility sale bility
_ Cents Per cent |j Cents Per cent
Mellow iqnlons:-s2i!22 22:2 1.6 13.5 }| California oranges___--_______ 2.0 9.2
Southern potatoes__-__-_------ 1.6 11.8 || Barreled apples____ __________ 2.0 8.0
Northern potatoes___.________- sere Sie dl ECC ACHCSHee: Coke ae a mie 8 2.5 24.1
BIKE POLALOES cr ues Tes oe epee, 1.8 5.4 || Boxed AD DIESSEs. Bee ee La Qe 4.6
Brestert letiiice-- > es Se 1.8 19.6 || Southern cabbage------_---_- 2.9 25.9
Northern cabbage___________-_- 1.9 12.8 || Eastern lettuce_-____2__-____-- 3.0 26. 4
Grantaloupes 2-22 =o ice tee 2.0 22.8 ||
|
CONCLUSIONS REGARDING VARIABILITY OF JOBBERS’ PRICE SPREAD
When the proportional expense of distribution is assigned to jobber
and retailer, there is found to be considerably greater constancy in
the retailers’ portion thanin that of the jobbers. The distribution
services rendered by the retailer are thus the dominant factor in de-
termining the regularity of distribution expense. Variability in job-
bers’ spread is not accounted for by variations im size of jobber’s sale.
Neither does variability in the jobbers’ portion of the price spread ap-
pear to be associated with variations of price in the wholesale market.
t must be accounted for by factors outside of those considered here.
ANALYSIS OF TYPES OF STORE OPERATION
The present analysis is concerned with measurement of the influence
_ of some typical forms of store operation in the expense of distributing
erishable commodities. A two-fold classification of store types is
ere considered for the purpose of determining the extent to which
the form of operation affects the expense of distribution and cost
of goods to consumers.
Ee
20 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
CLASSIFICATION OF STORE TYPES
The first classification divides all retail stores into two groups,
based on the kind of management: (1) Unit stores under independent
operation by individual proprietors, and (2) chain stores operated as
parts of centrally organized systems. The other classification dis-
tinguishes stores on the basis of special services extended. The
service distinctions, which apply to unit stores only, divide these into
three general groups. In the first service group are unit stores whose
regular policy is to extend credit and to deliver orders to a majority
of their customers. ‘The second group includes stores which operate:
prevailingly on a cash basis, but grant a limited amount of delivery
service. The third group comprises unit stores which do a strict
cash-and-carry business, extending neither credit nor delivery. No
stores of a credit-and-carry type were represented in the data. Sub-
divisions of chain-store data were not attempted on the basis of
service, since the price averages in all reporting chain stores were
nearly identical with those of the prevailing cash-and-carry type.
Distinctions between retail stores as to class of trade or clientele
coincide generally with distinctions on the basis of service rendered.
The more discriminating high-class trade of well-to-do neighbor-
hoods is generally served by stores which operate on a full credit-
and-delivery basis. Many independent retailers in thrifty middle
class neighborhoods conduct their business on a cash basis, but render
a limited amount of delivery service to regular customers. The
poorer and middle-class neighborhoods, where low prices are the main
consideration, are served mainly by cash-and-carry stores, which
dispense entirely with credit and delivery services. Chain stores do
an extensive business in the low-price sections also.
Although the original data for unit stores were carefully tabulated
in two additional groups on the basis of specialization, as grocery
stores and fruit-vegetable stores, the slight contrasts in their dis-
tribution expense indicated that, in comparison with management
and service factors, the factor of specialization is of minor signifi-
cance. further study of the influence attributable to specialization
was therefore discontinued, in order that attention might be concen-
trated upon the factors of major significance.
The present analysis thus takes into consideration distinctions
between five types of retailing agencies, classified as to form of man-
agement and extent of service in the following manner:
Management: Unit stores,’ chain stores.
Service policy (unit stores only): Credit-delivery, cash-delivery,
cash-carry.
EXTENT OF DATA
Only 7 of the 14 commodities previously considered were used for
these store-type analyses. The insufficient number of quotations
for the other articles in some store groups did not permit representa-
tive comparisons. Approximately three-fourths of the original num-
ber of quotations are included here, however. The 7 commodities
retained comprise 68 per cent of the total annual volume of the larger
series, and 70 per cent of their total annual retail value. Moreover
7 In the subsequent comparisons, the all-unit or typical-unit-store figures are regarded as representing
the degree of service generally prevailing in metropolitan unit stores. The typical unit store is thus to be
considered as a composite, rather than an actual type, since the figures are based upon averages of original
quotations, which were obtained from stores with all three types of service policy.
a
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES PA
the weighted mean percentage margins for these 7 articles are prac-
tically the same as those for the complete series. The smaller num-
ber of articles may therefore be regarded as representative of fruits
and vegetables in general. Table 9 shows the number of price quota-
tions for each commodity and their total in each of the five groups
and the proportion of the 14-commodity series in each group.
TaBLE 9.—Number of quotations, by store types, New York metropolitan area,
February, 1923, to May, 1924
F Unit Unit Unit F
miletores stares ees a peers
Commodity
delivery | delivery carry
OGENEEN POLALOS: - 22 he eee ee 1, 650 1, 399 833 457 109 251
alifonnia Oranges: sees ya Pee 2, 064 1, 843 1, 109 631 103 221
DSCC Taf Ua LORS eee ae is 1, 281 1, 148 698 387 133
Boxed:ap ples oe 1, 453 1, 285 811 367 107 168
Barreled ap plesi- ee eee iE Bil7/ 1, 184 706 389 89 133
Wastern-lettuce = ea ee ee 1, 331 1, 219 775 391 53 112
BYellO Ws ODN OMS =o sa sere eS 1, 806 1, 524 903 515 106 282
Total 7 commodities____..__-_-____-- 10, 902 9, 602 5, 835 3, 137 630 1, 300
Proportion of 14-commodity _ series
EA re, ae te a 3 ee eee per cent__ 76 73 74 72 75 77
MANNER OF MAKING COMPARISONS
The relative advantage of each of the five forms of store operation
in the distribution of these typical commodities is indicated by the
contrasts or differentials in their respective prices and in their price
spreads. These differentials are presented in two forms. The first
form shows contrasts in the expense of distribution, as represented by
the spread between wholesale price and retail price, in each of the five
types of store. These differentials in cost of retailing are of primary
interest to dealers and other food-handling agencies, which deal in
large quantities of goods. The contrasts are therefore presented on
a per-car basis, in terms of dollars per car. The second form shows
contrasts between different store types in retail selling prices. Since
it is the final retail prices, rather than the intermediate handling
costs, which are of primary interest to the individual consumer, the
price differentials of the various store types are expressed in cents per
standard retail sale, the unit which is of direct interest to the consum-
ing public.
ADJUSTMENT OF DATA
To make entirely valid comparisons of results from the various
groups of data, it was necessary to adjust retail prices to allow for
certain discrepancies occuring in the original wholesale prices. These
irregularities arose from lack of identity in dates of quotations, or from
variability in grades of goods reported by different store types. To
accomplish the adjustment, a common weighted average wholesale
price per car was computed as a base for the five store types, giving
chain stores and unit stores the respective weights of 1 and 9, accord-
ing to their relative importance as metropolitan distributors. An
adjusted retail price for each store group was then constructed, with
this weighted average figure as a base, by adding to this the same
spread as existed between the original wholesale and retail figures.
oe
a
22 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
Spreads between the weighted average wholesale and adjusted re-
tail prices per car remain the same as before adjustment, but retail
prices now reflect only the contrasts due to variations in type of store.
The weighted average wholesale price for the seven commodity series
is $1,180 per car. In Table 10 are given the figures for the original
wholesale and retail prices, with the spread in each of the five store
groups. The same figures after adjustment, based upon the uniform
wholesale figure of $1,180 per car, are given in Table 11. The spread
in Table 11 for each store type is identical with that derived from the
original prices. The price spread for each of the seven commodities
in the five store groups is shown in Table 12.
TaBLE 10.—Original (unadjusted) wholesale and retail prices, and price spread per
car in five store types, seven commodity weighted averages, New York metropolitan
area, February, 1923, to May, 1924
Store type Wholesale| Retail Spread
Shain oo 4. Ones 220 5 ep at hie Ee Be eS ee A Ne ae $1, 130 $1, 700 $570
STAM a Ges ae he ne rea ag rc eng eet Senne 1, 185 2, 180 995
Cash=Carryisi 22 on St ae ee ee ee = ee eee oe 1, 135 1, 960 825
Wash-=deliveny.2 22 22 Oe Sone ae ee ee ee ee ees 1, 190 2, 095 905
Credit-deliveny - 22-222 222 ee Ee a ae eee ee 1, 200 2, 275 1, 075
‘TaBLE 11.—Adjusted wholesale and retail prices and price spread per car in five
store types, seven commodity weighted averages, New York metropolitan area, Feb-
ruary, 1923, to May, 1924
Store type Wholesale; Retail Spread
Gol OG dias Seed 2 es ee eee en ae aE Aes a ees ee aa Se ea CooL $1, 180 $1, 750 $570
IAN unittest S26 a SSE Se ae See a oede Eae e ee eee eee een 1, 180 2, 175 995
(CaSD=CANY + 4 34. eos sae ako bE BES eee: eee Re Eee ree ee ae 1, 180 2, 005 825
Wash=Gelivieny seek oats ec eS Petes Stes ra ea es ee 1, 180 2, 085 905
CTedit-delivery css Hs 2 iN Se 2 ed ee eae eee 1, 180 2, 255 1, 075
TaBLe 12.—Price spread per car for each commodity in five store types, New York
metropolitan area, February, 1923, to May, 1924
Chain | All unit Cash- Cash- Credit-
Commodity carry delivery | delivery
stores stores stores stores stores
Northern potatoes. 22 aes ee ee eee 210 615 600 580 645
California vOLan geese S22: BI ee en ee ae 870 1, 465 985 1, 260 1, 635
SWCCLIDOLALOCS cere). eee ee Le a a ee 330 880 470 815 990
Boxcdappless ses = perro. Ce eee. ON. psa 1, 010 1, 575 1, 340 1, 445 1, 685
Barmeledvapplests esos ea ee eee 570 960 830 880 1, 045
mastern lethiice ses: ws ae ea ae ee 695 940 885 845 990
E]IO WIOMIONS es soo Fee ee a 8 Sa yl ee 675 905 745 870 970
Wrershtedimean #2: .2i\ ty). Gch by ee i a ep 570 995 825 905 1, 075
DIFFERENTIALS SHOWING CONTRASTS IN DISTRIBUTION EXPENSE PER CAR
Contrasts of store types in distribution expense are shown by
differences in their respective price spreads. These indicate the
relative advantage of each store type as a retailing agency. Differ-
entials in distribution expense per car are shown in Figure 9. A
summary of these price-spread differentials is given in Table 13.
In the last column of this table they are shown as percentages of the
)
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES Pies
respective price spreads of the indicated store types. Table 14
shows the actual differentials per car for the individual commodities.
SPREAD PER CAR IN VARIOUS TYPES OF STORES
SEVEN LEADING FRUITS AND VEGETABLES, (WEIGHTED AVERAGES)
NEW YORK METROPOLITAN DISTRICT, 1923-1924
DOLLARS
PER CAR
1200
Credit -Delivery
rN
>
!
'
~ t aa
Ali Unit Stores Cash
Differential
($170)
>
All Stores
'
Cash-Delivery
a s aa
Carr
Differential
($80)
¥
Cash-Carry”Ditferential
“
Cash-Carry
($255)
~
wy |
Ny
+
4!
~~!
3
~
c!
o |
Si
o
it
<
Q|
a
LI
oO}
~
a!
S
i
<!
oT]
Hl
aa
Sy
XK
o,
Net Chain Store Differentia/
Cash-Delivery
<=
All Chain Stores Cash-Carry
Fig; 9.—Differentials among store types are here shown in distribution expense per cur. ‘The ‘‘net”’
chain store differential of $255 indicates the advantage of the chain type of management over the
independent type of store. The ‘‘cash-carry’’ differential of $250 shows the saving to independent
stores from eliminating credit and delivery services
‘TaBLE 13.—Dzifferentials in price spread per car in five store types, seven commodity
weighted averages, New York metropolitan area, February, 1923, to May, 1924
Percent-
: Dollars age of
Types of store operation compared per car price
spread 1
MANAGEMENT: CHAIN STORES WITH UNIT STORES
pGross chain decrease below typical unit store___....----_-----.-..----_---------=---- 425 43
Net chain decrease below cash-carry unit store______________________________________ 255 31
SERVICE: DIFFERENT UNIT STORE TYPES
sCash-carry decrease below credit-delivery -___.._.....-..--.-.---___-.--..----------- 250 23
*Cash-delivery decrease Yelow credit-delivery_-_____________________________________- 170 16
meash-carry: decrease below cash-delivery__....__222 22-9 80 9
1 These percentages are rounded, hence they do not harmonize exactly.
24 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
TaBLE 14.—Differentials in five store types in price spread per car for each com-
modity, New York metropolitan area, February, 1923, to May, 1924
Differential per car
7 Weighted
Types of store operation compared | North- Cali- | Sweet | poxeq | Bar- | East- |-yenow|. mean
ern F
fornia | pota- reled ern Z for seven
Hes oranges} toes apples apples | lettuce | P25 | ¢om-
modities
MANAGEMENT: CHAIN STORES WITH
UNIT STORES
Gross chain-store decrease below
typical unit-stere=—- $405 $595 $550 $565 $390 $245 $230 $425
Net chain-store decrease below cash-
Cannysunit st Oreeaee eee 390 115 140 330 260 190 70 255
SERVICE: DIFFERENT UNIT-STORE
TYPES
Cash-carry decrease below credit-
delivery: 23332253 45 650 520 345 215 105 225 250
Cash-delivery decrease below credit-
GeliV CGY 52-2 es eee 65 375 175 240 165 145 100 170
Cash-carry decrease below cash-
delivery-2- 222i = ae =20 275 345 105 50 —40 125 80
DIFFERENTIALS SHOWING CONTRASTS IN PRICES TO CONSUMERS
Contrasts of the various store types in prices to the individual
consumer are shown by differences in their selling prices. These
contrasts are expressed as differentials in value of the standard
retail sale. The value for each of the seven commodities in each
of the five store types is given in Table 15, with the weighted mean
for the series, as computed from the classified retail price data and
from the size-of-sale data. The values are thus uncompensated
for adjustments in wholesale price differences.
TaBLE 15.—Value of standard retail sale,| New York metropolitan area, February,
1928, to May, 1924
Chain | Allunit | Cash- | .Cash- | Credit-
Commodity carry delivery | delivery
stores stores stores stores stores
Cents Cents Cents Cents Cents
Northern potatoes e223 fe a ee eee 19.9 2. 26. 27.4 28. 2
Califormiavoranges 3.22 = ee ee 23.8 28. 1 24.2 26. 3 29. 7
Sweetpotatoes= =. 2.35 ee ee a ee ee 16. 5 25. 9 18.2 23. 8 27.6
Boxed.ap ples. 26225. - oe ee Gets cnchig Aeet ee eeea 19.5 25. 4 22.9 24. 3 26. 4
iBarreled!ap ples-f- <= bak eee 8 = Aas eel 19.1 24. 6 22.1 23. 7 25.8
ASTER eth Cee ee ee ee re een 18. 4 23. 4 22.8 2250) 24. 6
SVellO We ONIONS pss aa re a 20. 4 2355 21.2 23.4 24.4
Viereh ted mean 22255 8 ie ee eens 20. 2 25.9 23.3 24.9 27.0
1 Unadjusted for wholesale price differences.
Differentials in the value of the retail sale for the various store
types, when the retail prices are compensated for the wholesale
differences noted above, are synonymous with the original differ-
entials in price spread per retail sale. For convenience in computa-
tion, therefore, the original price spreads per car were employed here.
By dividing the per car figure by the computed average number of
erry
5 bee!
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 25
retail sales per car for each commodity, the real spread in value of
the retail sale is derived for the various store groups. The average
price spread for the series was obtained by dividing the per car figure
by 8,405, the weighted mean number of retail sales per car for the
seven-commodity series.
VALUE’ OF STANDARD RETAIL SALE IN VARIOUS
LY PES GF STORIES
SEVEN LEADING FRUITS AND VEGETABLES, (WEIGHTED AVERAGES)
NEW YORK METROPOLITAN DISTRICT, 1923-1924
CENTS PER
STANDARD
RETAIL SALE
T
rap Credit-Delivery
>
!
All Unit Stores “Cash”
Differential
(2.0¢)
!
>
All Stores \
!
Cash-Delivery
A
'
“Carry”
Differential
1 0¢)
v=
Cash-Carry
!
ae]
<i
9°
wy |
Racal
31
a
c
3 |
u
o !
>
!
SS
Q!
o /
Sl
9
a {|
ai
oS
iS
<<!
91
41
%
o!
<1
aii
I
1
Cash-Delivery
<
All Chain Stores Cash-Carry
* Adjusted for differences in wholesale prices
Fic. 10.—Differentials among store types in cost to consumers of a uniform quantity of goods are
shown in cents per retail sale. The chain store form of management accounts for a difference of
3 cents per sale, in contrast to the independent store; while in unit stores the elimination of
credit and delivery service accounts for a further difference of 3 cents per sale.
The summary of differentials in value of the standard retail sale
among the five-store groups for the commodity series as a whole
is presented in Table 16. In the last column these differentials are
given also as percentages of the retail price for each store-type.
Figure 10 shows these contrasts in value of the retail sale. Diuffer-
entials for each of the seven commodities are given in Table 17.
26 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
TaBLe 16.—Differentials between store types in value of standard retail sale,}
weighted for seven commodities, New York metropolitan area, February, 1923, to
May, 1924
Gents Percent-
Store types compared per retail ae of
sale Ect
price 2
Gross chain-store decrease below typical unit store____--__-_______ paige Ge See 2e 5
Net chain decrease below, junit cash-carry/store! === 25-2 ee ee 2 13
@ash-carry: decrease below (eredit-deliver yea e ee ee 3 11
@ash-delivery decrease below; credit=deliveiy = ee 2 8
Cash-carmy Gecreaseybel wa Gas lel e Miiy ers ye aes eee ee a 1 4
i Adjusted for differences in wholesale prices.
2 These percentages are rounded, hence the last three do not agree exactly.
TaBLE 17.—Differentials in five store types in value of the retail sale for each com=
modity, New York metropolitan area, February, 1923, to May, 1924
Differentials per retail sale 1
| | Weighted
Types of store operation compared | North- : ;
ern Cali- | Sweet | Boxed | Bar- Eastern| Yellow | ™¢an for
pota- fornia | pota-_ apples Teled lettuce | onions | S@Ve?
toes | Oranges toes apples com-
modities
MANAGEMENT: CHAIN STORES AND
UNIT STORES
Gross chain store decrease below | Cents | Cents | Cents | Cents | Cents | Cents | @ents Cents
GY DLCAl SUE STORG seem 7.0 4.8 7.5 4,3 5.0 3. 2 3. 2
Net chain-store decrease below
cash-carry unit store____________- 6.7 .9 1.9 745.8 3.3 2.4 1.0 3
SERVICE: DIFFERENT UNIT STORE
TYPES
Cash-carry decrease below credit-
GeliVieny se ee ee ke AS ee .8 5.2 (bil 2.6 Pat 1.4 ah il 3
Cash-delivery decrease below cred-
It-deliveny ss oes a een ee ial 3.0 2.4 1.8 2.1 1.9 1.4 2
Cash-carry decrease below cash-
1
delivery] 2222S Soe ee —.3 Fe 4.7 .8 6 —.5 ed,
1 Adjusted for wholesale price differences.
SUMMARY OF STORE OPERATION CONTRASTS
Consideration has been given to differences resulting from con-
trasted kinds of store management and service policy, the contrasts
being expressed both by differences in their respective distributing
expense, and by differences in price paid by consumers.
Comparison of different store types as to their respective spreads
between wholesale and retail prices indicates the relative efficiency
of various forms of store operation. Between the typical unit store
and the chain-store type, there was an average eross difference in
selling expense, in favor of the chain-store form of management and
service, of $425 per car. The average spread in wholesale and retail
prices for chain stores was 43 per cent below the spread prevailing
in the typical form of unit store. In other words, cash-and-carry
chain stores required for distributing expenses 43 per cent less than
the amount required for these expenses be typical unit stores which
gave their customers the prevailing amount of special service. When
the cash-and-carry chain store is compared with the cash-and-carry
al mein ~ sits
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 27
unit store, both types being on the same nonservice basis, the con-
trast in favor of the chain store is reduced to $255 per car. Chain
stores distributed fruits and vegetables at an expense 31 per cent
below that required in unit cash-and-carry stores. The form of
management in the chain store thus accounted for a saving of nearly
one-third of the total distribution expense.
In the comparison of independent unit stores offering various
degrees of special service the cash-and-carry type required in the
process of distribution $250 per car less than the amount required
for distribution by the credit-delivery type. The spread between
wholesale and retail prices was thus 23 per cent less than that in the
unit stores giving credit and delivery service. In other words,
nearly one-fourth of the credit-delivery store’s distribution expense
was accounted for by the expense of delivery service and the granting
of credit.
To the individual consumer the meaning of these differentials in
selling expense is more clear when they are expressed as differences
in retail prices. Between the chain store and the typical unit store
there was found a gross price difference of 5 cents per standard
retail sale. Retail prices in chain stores averaged 17 per cent below
those prevailing in typical unit stores which gave the prevailing
amount of service. When the cash-and-carry chain store is compared
with the cash-and-carry unit store, both types being on the same non-
service basis, there is a net difference in retail prices of 3 cents per
sale, in favor of the chain store. This is 13 per cent below the cash-
and-carry unit store price. This figure is the truer measure of the
advantage of the chain store form of management, as the former
larger percentage difference includes some difference in service.
Comparing unit stores by themselves according to service policies,
credit-and-delivery service together are found to have cost the con-
sumer 3 cents per retail sale, in contrast to selling prices in cash-and-
carry unit stores. ‘This service differential is 11 per cent of the retail
price in the full-service type of store. From the data from which
these differences were derived, the credit element accounts for a
greater proportion than does the delivery element, but this may be
due in a measure to the limited nature of delivery service maintained
by reporting stores in the cash-and-delivery group.
A presentation of the differences found in this analysis of manage-
ment and service factors, showing the split-up of the consumer’s
outlay under various forms of store operation, is given in Figure 11.
In the bars for different store types the dark portion represents the
wholesale cost of the goods disposed of in the standard retail sale.
This wholesale portion is made uniform for each type—14 cents.
The remaining part of each bar shows the distribution expense for
services. In the chain store all distribution services were rendered
at a total expense of 6.8 cents per sale. In the unit-store types there
is added a charge for jobber’s service of 2 cents per sale. The general
retailer’s service in the cash-and-carry unit store costs 7.8 cents.
Total distribution expense for retailer and jobber in the cash-and-
carry unit store was thus 3 cents per sale greater than in the chain
store. In the unit store which operated on a cash basis with limited
delivery service the expense of delivery adds 1 cent more. In the
type which grants credit in addition to delivery, 2 cents are added
again to cover the added credit service.
28 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
GENERAL CONCLUSIONS
This study brings to light three features of metropolitan distribu-
tion which are of outstanding significance in determining the expense
of distributing perishable foods to city consumers.
The first is the fact that the expense of city distribution is influ-
enced to a remarkable extent by the purchasing habits of consumers.
The prevailing size of the individual retail sale has great influence
in determining the proportion of the consumer’s expenditure which
is absorbed in the distribution process. The price spread necessary
to cover the services involved in bringing supplies from the city
wholesale market to metropolitan consumers is found to be fairly
SPLIT-UP OF CONSUMER'S OUTLAY’ IN VARIOUS TYPES OF STORES.
( PER STANDARD RETAIL SALE )
SEVEN LEADING FRUITS AND VEGETABLES,(WEIGHTED AVERAGES)
NEW YORK METROPOLITAN DISTRICT, 1923-1924
CENTS
TOTAL O IS 20
CONSUMER'S
sites —<—— SERVICES >
CENTS i
CHAIN STORES 208 ee : - ae
SROs
>
UNITCASH-CARRY 255 an eas
. <: wT
GOODS
Wholesale cost of Goods
/4 Cents
SERVICES
Jobbers Service P77 General Retailers Delivery Credit
2 Cents Z L4 Service 7.8 Cents Eg Cent [Teen
“Adjusted tor differences in wholesale prices
Fic. 11.—The consumer’s outlay per standard sale in different types of stores is apportioned thus:
Cost of goods in the wholesale market, 14 cents; total services in chain stores, 6.8 cents; retailers’
service in unit stores, 7.8 cents, with 3 cents additional for jobbers’ service. Delivery service
adds 1 cent and credit 2 cents more to the consumer’s outlay
rae per individual retail sale, irrespective of the physical size
of sale
Selling prices seem to be fixed by retailers at such a point above
cost of goods in the wholesale market as will yield a fairly uniform
money return per sale, to cover the expense of service which is ren-
dered. The extent of service involved in distributing a given quan-
tity of goods thus fixes the proportion which the retailer must charge
above cost to cover his operating expenses. The larger the con-
sumer’s purchase the smaller is the proportion of the outlay which is
absorbed by distribution charges and the greater is the proportion
left to pay for merchandise.
All services involved in city distribution have to be paid for out
of the price charged by the retailer for the individual sale. Since
every sale is a profit-making opportunity, the retailer must so appor-
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 29
tion his expense among the individual sales as to yield him a living
above the cost of the goods he sells and the distribution expense which
he undergoes. The great number of conveniences given to the con-
sumer under present methods of city distribution must be paid for
by the spread between cost of the retailer’s goods and their selling
price. Maintenance of a continuous, well-selected stock, readily
accessible in wide variety at all times, and the splitting up of this
stock into small portions to meet the day-to-day needs of consumers
in the immediate neighborhood require a large outlay by the retailer
for the services involved. The expense occasioned in rendering these
services is influenced by the number of separate transactions required
to dispose of a given quantity of goods rather than by the gross
volume of goods sold. Prices to consumers are therefore established
at a point that will assure a fairly uniform money surplus on each
individual transaction. The retailer has to adjust his price policy
to the prevailing buying habits of his customers. Retail prices are
thus scaled to accord with consumers’ predominating practice of
making many oft-repeated small purchases.
This analysis shows further that the special services involved in
delivery of goods and in extension of credit require a material addition
to the consumer’s food outlay. Stores which operate on a cash-and-
carry basis are able to sell goods at considerably lower prices than
those which operate with a credit-and-delivery policy. Although
many consumers doubtless find the convenience of the credit-and-
delivery store well worth the added expense, those to whom economy
is the first consideration may enjoy a material saving by buying from
stores whose prices are based upon a cash-and-carry policy.
' A third significant factor in the expense of distributing fruits and
vegetables is the form of organization or management of the retail
store. The standardized operation of chain stores, centralized pur-
chase of supplies in large quantities, and sale of goods on a cash-and-
carry basis give this form of management distinct advantage in
economy of distribution. Demonstration of the saving in distribution
expense by the chain-store method points to this form of organization
as a fe means for reducing prices of goods to city consumers.
Ofisetting the economy of the chain form of operation, distinct
advantages are offered by the independently operated neighborhood
unit store. The personal atmosphere of the independent store, its
readiness to serve the preferences of individual customers, and the gen-
erally greater variety and wider choice of qualities give the neighbor-
hood unit store a strong hold upon its local clientele. As with the
option regarding special credit-and-delivery service, the preference
between chain store and unit store is a matter of relative emphasis
upon economy or convenience. One portion of the consuming public
os the fuller advantages of the neighborhood unit store with its
igher prices, whereas another portion prefers to dispense with these
advantages for the economy of lower prices prevailing in chain stores.
the consumer’s outlay for fruits and vegetables, these studies
show that from one-third to more than three-fifths is absorbed by
distribution expense after arrival of goods in the city wholesale market.
Such a portion is required to cover the expense of various services that
inhere in the prevailing methods of city distribution. With all refine-
ments that are possible for reducing service requirements, the task of
meeting the daily food demands of large city populations remains
30 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
primarily a matter of splitting up the centralized stocks of goods that
arrive annually in the wholesale markets into millions of separate
small retail parcels for daily consumption. Any generally effective pro-
gram for reducing the expense of distribution to city consumers must
therefore take into consideration the basic proposition of reducing the
expense on the individual retail sale.
Establishment of the size of the individual retail sale as a deter-
minant of distribution expense, and application of this principle to
the distribution of perishable commodities supported by extensive
statistical information, are results of far-reaching importance in
marketing science.
PERCENTAGE MARGINS OF COMMODITIES. AND
THEIR TOTAL RETAIL VALUES
PERCENTAGE MARGIN TOTAL ANNUAL RETAIL VALUE, 1923
CENTS PER CONSUMER'S DOLLAR MILLIONS OF DOLLARS
100 80 60 40 20 0 COMMODITY 0 4 8 12 165= 205 (24
WHITE ONIONS
N. CABBAGE
S. CABBAGE
W. LETTUCE
SWEET POTATOES
CANTALOUPES
E. LETTUCE
YELLOW ONIONS
PEACHES
S. POTATOES
BBLD. APPLES
BOXED APPLES
CALIF ORANGES
N. POTATOES
Fic. 12—When commodities are arrayed according to their total 'annual retail values, the corre-
sponding percentage margins lie generally in inverse order. The article with least annual value
has the highest margin, and the commodity with greatest annual value has the lowest margin
IMPLICATIONS IN COMMODITY DIFFERENCES §
Certain theoretical implications are revealed in the analysis of commodity
differences, whose interpretation in mathematical form will explain how the
theory of a constant price spread per retail sale accounts for the contrasts in dis-
tribution expense of different articles.
INFLUENCE OF COMMON MONETARY AND PHYSICAL UNITS
In the early part of the analysis of commodity differences, an inverse curvilinear
relationship was discovered between percentage margins and total retail values
in 1928, when these two magnitudes for the 14 commodities were plotted in a
scatter diagram. The articles with the greater total retail values had generally
lower percentages than those with lesser total retail values. The proportion
absorbed by distribution expense varied somewhat inversely with the respective
total commodity values, as is shown in Table 18 and Figure 12. The chart is a
graphic presentation of the fact that when the series of articles is arrayed in
ascending order according to total retail values in 1923 the corresponding per-
centage margins lie in generally descending order.
8 The analysis here presented was developed with the assistance of H. D. Comer.
i Vee se
—
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES Sl
TABLE 18.—Relation of margins to total annual retail sale, New York Metropolitan
area, February, 1923—May, 1924
Total © Percent- Total Percent-
Commodity 1923 retail age Commodity 1923 retail age
value margin value margin
e
Thousands ‘ Thousands
of dollars Per cent of dollars | Per cent
Northern potatoes_________- 7p Wl 30 |) Hasternelepiices sas 8, 944 51
California oranges___-_______ 22, 419 Al=||eCantaloupess= a. -=*= = 7, 469 46
iBoxed:apples- 22-2 = 21, 040 46 || Sweet potatoes_____--_-___- 4 601 44
Barreled apples_. ________-=- 17, 936 49 || Western lettuce___________- 4, 404 52
Southern potatoes__--_--_-_-- 15, 855 38 || Southern cabbage_________- 4, 211 48
IREAcheS Setar a ae he 9, 288 45 || Northern cabbage_________-_ 2, 373 58
Niellow OnlONSe =e eee ee 9, 037 ps (h \WMouiive: @youoyat. ee 1, 290 63
To illustrate the general relationship, let special consideration be given to
three of the articles whose total values differ widely—northern cabbage, barreled
apples, California oranges. Whereas the items in the total retail value series
are in ascending order, the items in the percentage margin series are in descending
order, thus:
Total
1923 retail | Margin
value (per cent)
BNOLinermicabhacom .os tetieel: cle ete tat eke Ses 1 LOY mah ADT ST eel eT oe $2, 370, 000 58
BaTreledsap plese seyret ae a Spy teen we ee epee NO we ae ve eh seek 17, 940, 000 49
alifOTmIAvO RAN CES ee eens Pee we Re eae eee ee NG NS I er oe 22, 420, 000 41
The question arises: Do not the differences in total retail value explain ade-
quately the differences in percentage margins? Is not the lower margin on
barreled apples, in comparison with that of northern cabbage, due simply to
the fact that the metropolitan area as a whole spends more money per year for
_ apples than it does for cabbage? Such a supposition might arise from the view
that merchants could handle articles which yield their principal income at less
expense per dollar’s worth of goods than they could distribute commodities
which bring a minor return.
The meaning of the inverse association between total values and percentage
margins is made clear by a little mathematical analysis. Let the total annual
retail value of a commodity be indicated by R. and its total annual wholesale
value be represented by W. The difference between R and W will then repre-
sent the total annual expense of distributing the given commodity in the metro-
politan area. The ratio of this total distribution expense to the total retail
value, * will then indicate the percentage margin for the given commodity.
The statement that margins vary inversely with total retail values is expressed
in mathematical terms by the formula Bet varies as = Now in the series, the
percentage margin saa may be diminished concurrently with an increase in the
total retail value R, by any one of three conditions affecting the relation between
R and W, namely: (1) if R-W remains constant, (2) if R-W declines, (3) if R-W
advances ‘less rapidly than R increases. The last one of these conditions really
embraces all three, for with constancy in R-W or with a decline in R—-W, the
total retail value R shows greater increase than that of total distribution expense
R-W. The statement that the percentage margin varies inversely with the
total retail value of an article therefore means merely that within the commodity
series an increase in total retail value is accompanied by a proportionally smaller
increase in total distribution expense. A satisfactory explanation of this pecu-
liar relationship is needed to interpret the general inverse association between
margins and total retail values.
32 BULLETIN 1411, U. 8. DEPARTMENT OF AGRICULTURE
Total retail value, total distribution expense, and percentage margin for each
of the three illustrative commodities are
R R-W R-W
Total Total
retail distribution] Margin
value expense
Per cent
Northern: cabbage 22 5252 a ae ade eee $2, 370, 000 | $1, 380, 000 58
Barreled appleSjc- 252-22 oss. 2 oe eee Se Se eee 17, 940,000 | 8, 710, 000 49 @
California Oranges 2232.2 52s 22 tee ea Se eee a ee ees 22, 420,000 | 9, 150, 000 41
In each case a change in R is accompanied by a relatively smaller change in
R-W;; thus
While R for apples is 7.5 times R for cabbage,
yet R-W ” 7? Only. 0.3). ee to ee Leos }
and R for oranges is 1.3 times R for apples, F
but R-W ” a only. Jot 22 RW,
Now the percentage margin for each commodity is a quotient derived by divid-
ing the figure in the second column by that in the first column. The decline in
the margin for apples (49) from that for cabbage (58), results from the fact that
R-W for apples is only 6.3 times R—W for cabbage, whereas R for apples is 7.5
times R for cabbage. The margin for apples is therefore = times 58, which is
49 per cent. Similarly, the decline in the margin for oranges (41) from that for
apples (49), results from the fact that R—W for oranges is only 1.1 times R—W for
apples, whereas R for oranges is 1.3 times R for apples. The margin for oranges
is therefore i times 49, which is 41 per cent. Throughout the 14-commodity
series, it may be demonstrated similarly that the inverse association between
percentage margins and total retail value results from the fact that total retail
value increases from one article to another more rapidly than does the total
distribution expense.
It remains still to explain why increases in total retail values are accompanied ~
by relatively smaller increases in distribution expense. The total retail value of
any commodity may be conceived of as the product of (1) retail price per pound
and (2) total number of pounds sold annually. The total distribution expense
may be regarded either as the product of price spread per pound and total
number of pounds, or as the difference between total wholesale value and total
retail value. If for any commodity, r represents the retail price per pound, w the
wholesale price per pound, and P the total number of pounds sold in the metro- ~
politan area in 1923, then the total retail value R is equivalent to r times P; the
total wholesale value is equivalent to w times P; and the total distribution expense
R-W is rP minus wP. The percentage margin is therefore a Es By can-
celling out the P’s in this fraction, the percentage margin becomes = This —
means that for any commodity in the series the percentage margin is the
same, whether based on total values or on values per pound. The conclusion is,
therefore, that percentage margins are independent of physical volume as a
separate factor. Any influence exerted by physical volume is expressed already
in the prices themselves.
The analysis shows that a difference in physical volumes of two given com-
modities affects R-W and R identically. The reason why R—W fails to increase
as rapidly as R, within the series of articles, is that r-w does not increase as
rapidly as r. In other words, price-spread per pound does not increase pro-
portionally with retail price per pound. Calculations from the per pound figures
for the three illustrative commodities shows it to be true that an increase in
retail price is accompanied by a relatively smaller gain in price-spread, thus:
Gases
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 33
T w I-w ; I—w
Mean retail |Mean whole-| Mean price aire
‘ price per sale price | spread per Margin
pound! | per pound ! pound
Cents Cents Cents Per cent
IN ORUHERINCAD Dae Be eae = ee ee a eee Seo es 5. 20 2.18 3. 02 58
Barreled ap plessetar ase. Sees eee eee See | 7. 97 4.10 3. 87 49
@aliforniaOrau eeSs= see = a ee a a 10. 95 6. 48 4, 47 41
1 Adjusted for shrinkage in retailing.
The percentage margin for cabbage is the quotient of 3.02 divided by 5.20.
The decline in the margin for apples from 58 per cent, the margin for cabbage,
results from the fact that while r—w for apples is only 1.28 times r—w for cabbage,
yet r for apples is 1.53 times r for cabbage. The percentage margin for apples
2
1.53 :
times r—w for apples, while r for oranges is 1.37 times r for apples. Hence the
margin for oranges shows a decline from the margin for apples, because of the
is therefore dee times 58, which is 49. Similarly for oranges, r—w is only 1.16
difference in these two ratios. It is as: times 49, which is 41 per cent.
Thus is demonstrated the simple but important fact that margin variations in
the series of commodities are synonymous with the varying relation of distribution
expense per pound to the respective retail price per pound. The statement that
one article has a lower percentage margin than another is synonymous with the
statement that the ratio of price spreads per pound is less than the ratio of the
respective retail prices per pound. The margin is the same for a given commodity
regardless of the physical quantity of goods considered in computing it.
INFLUENCE OF A NEW UNIT OF DISTRIBUTION
In the comparisons and analyses of commodities thus far, the consumer’s
dollar’s worth and the pound were the assumed common units of measurement.
Differences were noted in the expense of distribution per dollar’s worth and per
pound of the various commodities, but no satisfactory explanation of these
differences has been found.
Upon reflection, no sound reason exists for expecting either the margin per
dollar’s worth, or the price spread per pound to be uniform. ‘The expense in-
curred in distribution arises from various services rendered to consumers by
distribution agencies. If more service is required to retail a dollar’s worth or a
pound of one article than to sell a dollar’s worth or a pound of another article, the
cost of the additional service is logically reflected in a higher retail price. The
amount of service given by the dealer with each dollar’s worth of commodity
depends to a great extent upon the number of separate sales he must make to
receive a dollar from his customers. This depends in turn upon the average size
of sale to the individual purchaser.
Although retail sales of some articles are prevailingly made in larger quantities
and larger monetary amounts than are sales of other commodities, yet each sale,
irrespective of size or value, entails an approximately uniform expense for re-
tailer’s service. Apportionment of the service expense on the basis of the dollar’s
worth or the pound ignores the contrasts of different commodities in their service
requirements. A logical means of comparison should place all commodities on a
comparable service basis. ‘This is done when the comparisons are made on the
‘basis of the individual retail sale.
HOW RETAIL PRICES ARE SET
Determination of how retail prices are set is the vital part of this theoretical
‘discussion. Here is where the size of sale to the consumer enters into the
analysis.
The size of the retail sale, among the series of commodities, has been shown to
vary inversely with the retail price per pound, as illustrated in Figure 4. Com-
modities having low retail price per pound are sold to consumers in lots of several
pounds at a time, whereas articles with high price per pound are sold in smaller
lots. The average retail price of northern potatoes was about 4 cents a pound,
and the prevailing size of sale was 64% pounds. Western lettuce, whose retail
34 BULLETEN 1411, U. S. DEPARTMENT OF AGRICULTURE
price averaged about 15 cents per pound, had a prevailing size of sale of but 1144
pounds. For the three illustrative commodities, the retail price of each per pound
and the number of pounds per retail sale were found to be:
: .
Mean Size of
retail price | mean retail
| per pound sale
Cents Pounds
INortherm cab bages= 222 area ee eee ee ee a 5. 20 4.0
Barreled: appleS = =. 25. 3 a ee 7. 97 3.0
@alifornla oranges. = 2S) o= o e e es 10. 95 225;
When these commodities are arranged in ascending order of retail price per
pound, it is observed that size of retail sale is in descending order. This is rep-
resentative of the general tendency for all commodities, as shown in Figure 4.
A means is now at hand for explaining what determines retail price per pound,
and the price spread per pound. ‘The difficulties arising from the unsuitableness —
of the pound as a unit for comparison of distribution factors are removed by
using the standard retail sale as the unit of distribution. With the size-of-sale
data, price spread may be computed per mean retail sale. The spread per sale
is the product of the price spread per pound and the number of pounds per sale.
For the three illustrative articles the mean spread per pound, the mean size of
retail sale, and the mean spread per sale, are:
Mean | Mean size Mean
spread per of retail spread per
pound sale retail sale
Cents Pounds Cents
INOrheEmMcab bag e tes 5s se ee ee eee 3. 02 4.0 1p A!
IB ArreleG sai ples ae ee ae ee ee re eee 3. 87 3.0 11.6
Californiatorancest (22 2" Bele ere es ee ee eee 4, 47 255 11.2
When the commodities are arranged in ascending order of spread per pound,
the size-of-sale series is seen to be in descending order, as was true in the preceding
instance with the price-per-pound series. In consequence of the inverse relation-
ship of the spread-per-pound series and the size-of-sale series, the mean spread
per retail sale, which is the product of these two, is nearly constant (fig. 13).
This is illustrative of the general tendency for all 14 commodities, as shown in
Figure 5. The general conclusion is therefore that retail prices are set at such
levels above wholesale prices as will tend to make the spread between wholesale
and retail values of the standard retail sale the same for all commodities.
This theory of a constant spread per sale throws light upon several perplexing
problems. It explains why the portion of the consumer’s dollar which is ab-
sorbed in the expenses of city distribution should be greater for some articles than
for others. It indicates the existence of a peculiar price-setting practice which is
based on the prevailing size of the consumer’s individual purchase. Further-
more, it shows that any significant relationship between physical volumes and
margins may be traced to associated differences in size of sale,
SUMMARY OF APPLICATION OF THEORY
Steps in application of the theory of a constant spread per retail sale as an
explanation of contrasts in percentage margins among the 14 commodities are
shown graphically in Figure 13. The interpretation of these steps is set op-
posite the respective diagrams.
Differences in percentage margins within the commodity series, ranging from
37 for northern potatoes to 63 for white onions, come about from use of the
dollar’s worth of goods as the unit of measurement. These differences are merely
a reflection of the fact that to distribute a dollar’s worth costs more for some
commodities than for others. The assumption that distribution expense per
dollar’s worth should be uniform for different commodities is illogical, because it
ignores the fact that more service is absorbed with a dollar’s worth of some articles
than with a dollar’s worth of others, in consequence of differences in their pre-
; es
EXPENSE FACTORS IN CITY DISTRIBUTION OF PERISHABLES 3D
GENERAL RELATIONSHIPS wuicu RESULT IN
TENDENCY TOWARDS CONSTANCY of SPREAD
per RETAIL SALE ror 14 COMMODITIES
PER CENT
MARGIN
/
TOTAL RETAIL VALUE
TOTAL DIST.
EXPENSE
me
TOTAL RETAIL VALUE
SPREAD PER
POUND
ms
RETAIL PRICE PER POUND
SIZE OF RETAIL
SALE
‘a
RETAIL PRICE PER POUND
SPREAD PER RETAIL
SALE,
is
RETAIL PRICE PER POUND
ALL COMMODITIES
Fic. 13.—Inverse relationship between percent-
age margins and total retail value of different
commodities is explained by constancy ofspread
per retail sale
A. Among the 14 commodities there
is an inverse curvilinear. relationship
between percentage margins and total
retail values.
B. The above association is the result
of a direct curvilinear relationship be-
tween total distribution expense and total
retail value, with the curve bending
toward the total-retail-value axis.
C. Analysis of the relationship be-
tween total distribution expense and total
retail value reveals a direct curvilinear
association between spread per pound and
retail price per pound, with the curve
bending toward the retail-price axis.
D. There is an inverse curvilinear re-
lationship between size of the retail sale
and retail price per pound.
E. Spread per retail sale is the product
of spread per pound and number of
pounds per retail sale. Combination of
the direct relationship in ‘“C’’ and the
inverse relationship in ‘‘D”’ results in a
tendency toward constant spread per
retail sale for all commodities,
36 BULLETIN 1411, U. S. DEPARTMENT OF AGRICULTURE
vailing size of sale. Variation in percentage margins is therefore the result of —
wrongly using the dollar’s worth, with its variable service requirements, as the unit
of distribution. When the individual retail sale is taken as a new distribution —
unit, the actual margin per sale is nearly uniform for all commodities, and no ~
appreciable differences remain to be explained.
LIMITATIONS OF Se CG ee FOR COMPARING
ICES
Use of percentages for analyzing price differences incurs certain mathematical
difficulties which may vitiate their meaning. This made it necessary to abandon
the original plan of making detailed comparisons of margins as percentages of
retail prices and to analyze the actual prices and price differences in their stead.
The margin concept assumes a constant money expenditure by the consumer, ~
representing a dollar’s worth of goods under given price conditions. Any varia-
tion in retail price involves, therefore, a change in the quantity of goods secured
for $1. The margin represents in cents of the consumer’s dollar the spread
between wholesale and retail prices for a variable quantity of goods, whose volume ©
changes with any change in the selling price. Use of differentials between per- —
centage margins to measure the relative efficiency of a given money outlay is
therefore logically unsound. :
The difficulty in interpreting percentage differentials may be illustrated by
making a comparison of percentage margins in two types of retail stores. The
general margin in independent credit-delivery stores for the whole commodity ~
series is 47 per cent of the mean retail price, while in cash-and-carry stores it is
only 42 per cent of the retail price. This is equivalent to saying that of a dollar’s
outlay by the consumer in a credit-delivery store, 47 cents are required to cover
handling expenses, leaving 53 cents to pay for the goods in the wholesale market;
whereas of the dollar spent in a cash-and-carry store only 42 cents is required for |
handling and 58 cents is left to pay for goods in the wholesale market.
Thus the apparent difference in handling cost of 5 cents on a dollar’s worth of
goods, actually turns out to be quite otherwise than a difference of 5 per cent of |
the retail price. Out of the dollar expended by the consumer, 58 cents of the |
cash-and-carry customer’s money is used to buy goods in the wholesale market,
whereas only 53 cents of the credit-delivery customer’s money may be so used.
Hence the cash-and-carry customer will obtain 23 times the quantity of goods
received by the eredit-delivery customer. For an identical quantity of goods,
therefore, the cash-and-carry customer would pay only 23 of the other’s outlay. |
The latter fraction, for the cash-and-carry store, is 91.4 per cent of the amount |
required for the same quantity of goods in the credit-delivery store. The actual —
differential between prices in the two store types is therefore 100—91.4, or 8.6 |
per cent, of the credit-delivery price, instead of the apparent 5 per cent. If it |
were desired to express the differential in terms of the cash-and-carry price, |
then the corresponding inverted fraction, 2%, would be used, which is 109.4 _
per cent. This indicates that the selling price for a given quantity of goods in a
credit-delivery store is 9.4 per cent higher than that for the same quantity in a
cash-and-carry store.
The difficulty here illustrated exists wherever comparisons of percentages |
derived from varying or noncomparable bases are attempted. Any effort to
make accurate price comparisons by comparing percentage margins is therefore |
likely to be misleading and to confuse the real differences. |
ADDITIONAL COPIES
OF THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.
AT
10 CENTS PER COPY
V
“the
os.
ra