Citation
Demand for honey dew melons in the New York City wholesale market, with special reference to the potential market for supplies from Peru

Material Information

Title:
Demand for honey dew melons in the New York City wholesale market, with special reference to the potential market for supplies from Peru
Creator:
Fajarado-Christen, Adrian
Affiliation:
University of Florida -- Department of Agricultural Economics -- Institute of Food and Agricultural Science
Publisher:
Department of Agricultural Economics, Institute of Food and Agricultural Sciences, University of Florida
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Agricultural Economics thesis M.S. ( LCSH )
Melons. ( LCSH )
Agriculture ( LCSH )
Farm life ( LCSH )
Farming ( LCSH )
University of Florida. ( LCSH )
Dissertations, Academic -- UF -- Agricultural Economics ( LCSH )
Agriculture -- Florida ( LCSH )
Farm life -- Florida ( LCSH )
Melons ( jstor )
Dew ( jstor )
Market prices ( jstor )
Spatial Coverage:
North America -- United States of America -- Florida

Notes

General Note:
Thesis (M.S.)--University of Florida, 1969. Typescript.
Funding:
Florida Historical Agriculture and Rural Life
General Note:
Agricultural economics report - University of Florida Dept. of Agricultural Economics ; no. 14

Record Information

Source Institution:
Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location:
Florida Agricultural Experiment Station, Florida Cooperative Extension Service, Florida Department of Agriculture and Consumer Services, and the Engineering and Industrial Experiment Station; Institute for Food and Agricultural Services (IFAS), University of Florida
Rights Management:
All rights reserved, Board of Trustees of the University of Florida
Resource Identifier:
28998092 ( OCLC )

Downloads

This item has the following downloads:


Full Text





HISTORIC NOTE


The publications in this collection do
not reflect current scientific knowledge
or recommendations. These texts
represent the historic publishing
record of the Institute for Food and
Agricultural Sciences and should be
used only to trace the historic work of
the Institute and its staff. Current IFAS
research may be found on the
Electronic Data Information Source
(EDIS)

site maintained by the Florida
Cooperative Extension Service.






Copyright 2005, Board of Trustees, University
of Florida






,ovember, 1970


Ag. Econ. Report 14


Demand for Honey Dew Melons
in the New York City Wholesale Market,
1 Special Reference to the Potential Mark
for Supplies from Peru


HUME LIBRARY


I.F.A.S. Univ. of Florida


Department of Agricultural Economics
Florida Agricultural Experiment Stations
Institute of Food and Agricultural Sciences
University of Florida, Gainesville


Adrian Fajardo Christen


II r* ----------- CCIILI -- a r z















TABLE OF CONTENTS


ACKNOWLEDGMENTS . . . . ..

LIST OF TABLES . . . . .

LIST OF FIGURES . . . . .

INTRODUCTION. . . ... . .. .

The Problem and Objective . . .
Sources and Nature of the Data . .

ANALYTICAL PROCEDURES . . . ..

Model Specification . . . .
Selection of Marketing Periods . .

ANALYSIS OF PRICE-QUANTITY RELATIONSHIPS. . .

One Equation Models. . . . .
Separate Equation for Each Market Period .

SUMMARY . . . . . .

APPENDIX . . . . . .

BIBLIOGRAPHY . . . . ...


Page

ii

iii

iii

1

1
4

6

6
8

10

10
15

19

25

39













ACKNOWLEDGMENTS


The author expresses his special gratitude to W. W. McPherson,

for his guidance during this research. The author also wishes to

thank, B. R. Eddleman for advice and valuable suggestions.

To M. R. Langham, who took the time to read several drafts

of the manuscript and whose comments and suggestions resulted in im-

provement of the analysis of the manuscript, and to L. H. Myers for

analytical assistance, the author is very grateful. The constructive

reviews of Leo Polopolus, Anthony Prato, and R. W. Ward are appreciated.

J. L. Pearson, in the early stages of this research, was help-

ful in the location of the data. E. R. Mead's experience and insight

into the New York City terminal market was of great help. Information

on potential production areas in Peru and technical matters with re-

gard to honey dew melons was furnished by Alfredo Montes. V. M.

Yellen was adviser in matters of computer use. Computer work was per-

formed by the University of Florida Computing Center.









LIST OF TABLES

Number Page

1 Coefficients for Alternative Seasonal Demand Equations
for Honey Dew Melons in the New York City Wholesale
Market ..... .. .. . . . 11

2 Direct and Cross Price Flexibilities for Each Marketing
Period. . . . . ... ..... 14

3 Coefficients for Each Marketing Period Analyzed Inde-
pendently for Honey Dew Melons in the New York City
Wholesale Market. . . . .. 16

4 Direct and Cross Price Flexibilities for Each Marketing
Period Analyzed Independently . . . 17

5 Summary of Estimated Equations, Price of Honey Dew Melons
in the New York City Wholesale Market, by Market
Period and Month, as a Function of the Quantity of
Honey Dew Melons, Quantity of Spanish Melons and Dis-
posable Personal Income .... .. . 21

6 Honey Dew Melons at New York City Wholesale Market, in
Hundredweights. . ... . . . 26

7 Wholesale Prices of Honey Dew Melons at New York City,
Dollars Per Hundredweight . . . . 28

8 Spanish Melons at New York City Wholesale Market, in
Hundredweights. . ... .. . . .... 30

9 Disposable Personal Income in the United States, Bil.
Dollars . . . . ... ..... .32

10 Simple Correlation Matrix, One Equation Approach. ... .34

11 Simple Correlation Matrix, Separate Equation Approach,
Period 1 (Nov., Dec., Jan.) . . .... 36

12 Simple Correlation Matrix, Separate Equation Approach,
Period 2 (Feb., Mar., Apr.) . . . 37

13 Simple Correlation Matrix, Separate Equation Approach,
Period 3 (May-Oct.) . . . .... 38


LIST OF FIGURES
1 The Three Valleys in Peru" Suitable for Growing Honey Dew
Melons. . . . . . ... 2


iii









DEMAND FOR HONEY DEW MELONS IN THE NEW YORK CITY
WHOLESALE MARKET, WITH SPECIAL REFERENCE TO THE
POTENTIAL MARKET FOR SUPPLIES FROM PERU*

Adrian Fajardo-Christen


INTRODUCTION


The Problem and Objective

In the United States cold and warm seasons occur at different

times of the year in comparison with South America. Consequently,

South American fruits and vegetables are shipped to the United States

during the winter season here. Large supplies of cherries, plums,

nectarines, peaches, pears, melons and grapes are imported from Chile

[14].

Peru's coastal region, which is situated north of Chile, is an

area with climate and soil conditions suited to the growing of tropical

and subtropical fruits and vegetables, including honey dew melons.

Three valleys in this area have particularly favorable conditions for

the production of honey dew melons: the Valley of Viru in the Departa-

mento of La Libertad, the Valley of Canete in the Departamento of Lima

and the Valley of Ica in the Departamento of Ica (Fig. 1).1



This report is based on a M. S. thesis prepared by the author [4].
The author is a Graduate Assistant, University of Florida, Gainesville.

1Departamento is a political division of the country somewhat
equivalent to a state in the United States.

























































o 100 W0 o00 Km.

* Location of potential
production areas.


Figure 1. The Three Valleys in Peru Suitable for Growing Honey Dew Melons.










Even though physical conditions are advantageous, the economic

feasibility of the enterprise is still undetermined. As a first step

it is necessary to determine the demand for honey dew melons in the

United States.

The objective of this study is to estimate the wholesale demand,

on a seasonal basis, for honey dew melons in the New York City whole-

sale market. The results, of course, should be useful to suppliers

and potential suppliers in areas of the United States and other coun-

tries, as well as to those located in Peru.

There have been no published measurements of the demand for

honey dew melons [2], [15], [22]. A study by Suits [12] and one by

Seale and Allen [11] provide information on the demand for watermelons.

Suits used the limited information technique to estimate demand and

supply parameters in a simultaneous system model. Seale and Allen

used two approaches--both were single equation models. In the first

approach, demand schedules were derived for each consuming center from

price and quantity data for that center and for all the marketing periods

into which the marketing season was divided. This type of analysis

failed to yield logical demand schedules. Their second approach reduced

the length of the time period, and consequently reduced the effects of

the changes in tastes and in prices of substitutes. In both cases

their analysis covered the most active marketing season in Mississippi,

June 5 to August 31, for the year 1956 and for 20 consuming centers.

The present study considers only one consuming center, New York

City, and the annual marketing season is divided into three marketing

periods.










Sources and Nature of the Data

Monthly time series data for 1951 through 1967 were used. A

month was thought to be long enough to permit changes in prices in

response to changes in quantities, and short enough to insure that a

relatively homogeneous set of factors was operating. For 1951 through

1959, monthly data on quantity were obtained from Unloads of Fresh

Fruits and Vegetables at New York City [20]. For 1960 through 1967

monthly data on quantity were obtained from Fresh Fruits and Vegetables

Unloads in Eastern Cities [18]. All boat receipts at New York City

were shown as unloads in that city although some were reshipped to

other markets. Unloads cover the New York metropolitan area as follows:

the Five Boroughs, Mt. Vernon, Elmsford, Mt. Kisco, White Plains,

Mamaroneck, New Rochelle and Mineola in New York, and Jersey City,

Hoboken, North Hawthorne, Paterson, Passaic, Rutherford, Kearny, Newark,

Irvington, Elizabeth and Linden in New Jersey.

The unit measure for reporting unloads was the rail carlot, and

truck and boat unloads were converted to carlot equivalents. During

1951 through 1959 a rail carlot contained 500 crates, during 1960

through 1965 it was changed to 600 crates and in 1966 through 1967 it

contained 720 crates. The Consumer and Marketing Service has tried in

this manner to reflect the trend toward heavier rail car loadings through-

out the years.

Honey dew melons are generally packed in flat crates that are

called 7s, 8s, 9s, lls, and 12s, according to the number of melons in

the crate. In the present study, quantity was converted to pounds.

The U. S. Department of Agriculture's Table of Carlot Conversion Factors

Fruits and Vegetables [19] shows for various flat crates a range in










weight from 40 to 45 pounds. A conversion factor of 42.5 pounds per

crate was used to transform unloads to pounds.

Monthly data on prices were obtained from Fresh Fruit and Vege-

tables, Market News [16] and Fresh Fruit and Vegetable Prices [17].

Fresh Fruit and Vegetables, Market News is the original source of prices.

Prices quoted covered sales by primary receivers from 5 a.m. to 12 noon

on offerings of available supplies in less than carlot quantities. All

quotations were on stock of generally good merchantable quality and con-

dition, unless otherwise stated [16].

The monthly wholesale prices quoted in Fresh Fruit and Vegetable

Prices are simple averages of representative prices for each Tuesday

in the month taken from Fresh Fruit and Vegetables, Market News. Prices

quoted in dollars per crate were converted to dollars per hundredweight

for the purpose of the present study. For those months in which prices

were not quoted in Fresh Fruit and Vegetable Prices, monthly prices

were derived from the daily quotations reported in Fresh Fruit and

Vegetables, Market News. When more than one price per month was quoted

a weighted average was calculated.

Monthly data on personal income for the United States were obtained

from the Survey of Current Business [21]. Personal taxes were not

reported monthly. Monthly disposable personal income was calculated by

assuming that the quarterly rates of personal taxes were constant for

the three months of the particular quarter in question.

In order to gain better perspective of the marketing process, a

personal visit by the author was made to the New York City Terminal

Market where its operation was observed and marketing firms were contacted.










ANALYTICAL PROCEDURES2


Model Specification

From the standpoint of Peruvian suppliers, demand at the whole-

sale level is the most relevant point in the marketing process. The

estimation of demand at the wholesale market presented both theoretical

and practical problems. In any ordinary retail market the assumption

that the quantity purchased depends on price is quite reasonable. For

the wholesale market, however, the situation is somewhat different.

Here the retailer may be in a position to bargain over price with the

wholesaler. Therefore, in a wholesale market the unilateral dependence

of quantity on price does not necessarily hold, and it may be disturbed

by a tendency toward bilateral interdependence [23, p. 10]. However,

given the particular characteristics of honey dew melons, their perish-

ability and the fact that planting decisions precede sales decisions,

it may be assumed that the quantity sold at the wholesale level in any

given month or season is predetermined.

The general expression for the model is as follows:


P = f(Qh' Qs Qc' I, P, U)

where

p = wholesale price of honey dew melons,

Qh = quantity of honey dew melons,
Qs = quantity of Spanish melons,
Qc = quantity of cantaloupes,
I = disposable personal income in the United States,
P = population,
U = random disturbance.



The theoretical framework for this analysis is based on Wold [23],
Burk [3], Foote [5], Rojko [9], Salzman [10], and Johnston [8].










After a preliminary analysis, population and the quantity of

cantaloupes were dropped from the original function. Population was

found to be highly correlated with disposable personal income. The

quantity of cantaloupes, originally suspected of having an inverse

effect on the wholesale price of honey dew melons, was not statistically

significant at the .10 level. Thus the quantity of cantaloupes was

dropped from the analysis also. The final derived price-quantity re-

lationship, the one used throughout the remainder of this study, is:

the wholesale price of honey dew melons depends upon the quantity of

honey dew melons and the quantity of Spanish melons sold at the wholesale

level, and disposable personal income. Disposable personal income

refers to the United States as a whole, since the area of ultimate dis-

tribution was not known, but it was known to be much greater than the

New York metropolitan area.

Multiple regression was selected as the method used to estimate

the price-quantity relationship for honey dew melons. The quantity of

honey dew melons supplied was believed to be essentially unaffected by

current price. Also, the demand shifters (disposable personal income

and the quantity of Spanish melons) were believed to be determined inde-

pendently of the price of honey dew melons. The quantity of honey dew

melons and quantity of Spanish melons were assumed to be independent

because the production decisions are made prior to the marketing month

and they are highly perishable. The disposable personal income variable

was considered independent because the expenditures for the honey dew

melons are a very small share of total personal expenditures [6, p. 395].

Some other statistical specifications must be met, if the method of

least squares is to yield best linear unbiased estimates of demand parame-

ters. These additional specifications are that the random disturbances










are serially independent, have zero means and finite and constant vari-

ances. The assumption that the disturbances are normally distributed is

not needed for structural estimation but is required for the tests of

hypotheses which were made. A final statistical specification is the

assumption that the variables are measured without error.


Selection of Marketing Periods

The unload data showed a seasonal pattern with regard to sources

of supplies. In February through April, 97 percent of the supply of

honey dew melons was imported from Chile during 1951 through 1967.

In May the unloads from Chile decrease and the last ones overlap

with the beginning of small quantities of unloads from domestic produc-

tion. During May, from 1951 to 1967, there were 687 unloads of honey

dew melons from Chile compared with 579 from domestic production. From

June through October the honey dew market in New York is supplied from

domestic production. In the month of October the Spanish melon from

Spain appears in the market but in small quantities in comparison with

the unloads of honey dew melons from domestic sources. In October,

during 1951 to 1967, a total of 4,341 unloads of honey dew melons were

supplied compared with only 649 unloads of Spanish melons from Spain.

In November and December the supply of Spanish melon increases

sharply in the New York market. During 1951 to 1967, 2,479 unloads of

Spanish melons compared with 620 unloads of honey dew melons came into

the market. The month of January presents a special situation: Spanish

melons appear to have been supplied beginning in 1958 since no data

were found for previous years. During the 1958 to 1967 period, 138

unloads of Spanish melons compared to 323 unloads of honey dew melons

were supplied. If the full period of 1951 to 1967 is considered then










the 138 unloads of Spanish melons must be compared with 491 unloads of

honey dew melons.

The next step in the selection of the marketing periods was to

plot graphically the monthly wholesale price against the quantity of

honey dew melons. Each month for the 17 years was plotted separately.

The monthly scatter diagrams appeared to provide information about the

general nature of the price-quantity relationships. A close examination

of these diagrams indicated a general tendency for price to be negatively

correlated with quantity. This graphic analysis also reinforced the

previous hypothesis of the presence of a seasonal pattern. The relations

for February through April appeared to be similar, as did those for

November and January. Those for May through October appeared to have

similar slopes whereas the intercepts shifted from month to month. In

December both price and quantity of honey dew melons were reported only

for 1953, 1958, 1960, 1966 and 1967. For the other 12 years either

price, quantity or both were missing. To fill in for these missing

data the simple average wholesale price and the simple average quantity

from the available data were used.

As a result of this graphic analysis, three marketing periods were

chosen tentatively: the first period covers November through January,

the second includes the months of February through April, and the third

consists of May through October. However, in the analysis, the inter-

cepts of the monthly curves within each of the three marketing periods,

as well as among the periods, were tested for significant differences.













ANALY'f e OF PR:.C:-QUANTITY FELATiOi!SHiIE~


There are two general alternatives in the use of regression analy-

sis to estimate class differences. One is to estimate a separate equa-

tion for each group of observations. A second alternative is to use

dummy variables in one equation to allow for differences among classes

while ..1,l, the total (pooled) set of observations [1]. Dumay variables

were used first in an attempt to determine whether or not there would

be gains from pooling the data. Dummy variables were introduced to

allow changes in the slope among market periods and to allow changes in

intercepts both an'ig market periods and among months within each period

when one regression equation was used. When a separate equation was

employed for each market period, dummy variables were used to allow

changes in the intercepts among months within each period. November to

January is market period one, February to April is period two, and May

to October is period three.

One Equation Models

Four sets of coefficients were estimated. The observations used

in the analysis are given in the Appendix (Tables 6-9). The results are

summarized in TIbhi. The first equation is a linear regression of

the price of honey dew melons in dollars per L,,idric.._ h on the

quantity of honey dew melons in !undid..:. iiht, quantity of -I...Ii,



Results are given here in summary form. Details of pro;ii-di..,
are given in Fajardo [4]; for further discussion of statistical pro-
cedures, see Ben-David and Tomek [1].










Table 1. Coefficients for Alternative Seasonal Demand Equations for
Honey Dew Melons in the New York City Wholesale Market.a



Item Equation Number
(1) (2) (3) (4)


Intercept for honey dew melon price-quantity relationship


Annual


8.50


9.08


Period one:


January

b
November

b
December

Period two:

April


February


March


Period three:

October

d
May


June


August


11.45**
(.56)

.36
(.65)

- 2.39**
(.50)


7.50**
(.51)

.21
(.48)

- .47
(.50)


9.72**
(.76)

.73
(.61)

.99*
(.54)

.78
(.50)

1.31**
(.58)


d 1.17*
September 1.17
(.62)


10.97**
(.55)

- .89
(.54)

- 2.64**
(.50)


7.53**
(.53)

.15
(.50)

- .29
(.51)


8.63**
(.56)

1.28**
(.56)

1.33*
(.53)

.86*
(.52)

.94*
(.55)

.68
(.56)










Table 1.--Continued.


Item Equation Number
(1) (2) (3) (4)

Slopes of honey dew melon price-quantity relationships

Base or -.000035** -.000024** -.000024** -.000029**
annual (.000008) (.000005) (.000004) (.000003)

f -.000188* -.000023
(.000062) (.000049)

f .000021** -.000025
(.000010) (.000005)

Demand shifters

Quantity -.000033** -.000029** -.000034** -.000029**
Spanish
Spanish (.000013) (.000013) (.000012) (.000013)
melons

.006726** .007737** .009712** .008367**
Income
(.001191) (.001154) (.001343) (.001392)


R2 .600 .570 .390 .311

Sum of Squared
Residuals 379.11 406.90 577.52 652.39


aThe standard
parentheses.


errors of the coefficients are presented in the


Differences from January.

CDifferences from April.

dDifferences from October.

e
Based on annual mean in equations 2 and 4, and on market period
three in equations 1 and 3.

Differences from period three as a base.
*Significantly different from zero at the .10 level.
**Significantly different from zero at the .05 level.










melons in hundredweight, and disposable personal income in billions of

dollars. Intercepts were permitted to vary among the three market periods

and among months within market periods by means of dummy variables in

order to test for differences [1]. Slopes of the honey dew price-quan-

tity relationships were permitted to vary and tested for differences

among the three market periods. In the second equation, intercepts

among months and periods were permitted to vary but the coefficient

for honey dew melon price-quantity slope was not permitted to change

among months or periods.

In the third equation, slopes but not intercepts were permitted

to vary. Finally, in the fourth equation neither intercepts nor slopes

were permitted to vary among periods nor among months within periods.

In each equation the income and Spanish melon coefficients were not

permitted to vary among months nor among market periods.

The coefficient for quantity of Spanish melons was negative

and significant at the .05 level in each of the four equations. The

income coefficient was positive and significant at the .05 level in

each of the four equations. The intercepts of the honey dew price-

quantity relationships were significantly different among some months

within market periods for r;riods one and three. In the first equation

the slopes of the price-quantity relationships were significantly dif-

ferent among all three market periods.

The estimates of the direct and cross price flexibilities, com-

puted at the mean values of the variables, are given in Table 2. Thus

in market period one, equation one indicates that an increase of 1%

in the quantity of honey dew melons would reduce price by .099%.

Elasticities of demand were not computed, but if Houck's [7, p. 789-792]


















Table 2. Direct and Cross Price Flexibilities for Each Marketing Period.


Market Price Flexibilitiesa
Market ----------------------------
Period Direct Cross
Honey Dew Melons Spanish Melons Disposable Personal Income

1. November to
January -.099 -.038 .198

2. February to
April -.069 -.007 .257

3. May to October -.198 -.007 .222


percentage change in price of honey dew melons associated with 1% change in quantity of
honey dew melons, quantity of Spanish melons, and in disposable personal income, respectively.










findings are accepted, then the reciprocal of the direct price flexibility

coefficient may be considered the lower limit of the price elasticity of

demand. The simple correlation matrix, based on results of estimates

with the first equation, is given in the Appendix (Table 10).


Separate Equation for Each Market Period

Procedures used in this analysis were generally like those used

in the preceding section. The difference is that, in this analysis

the data in each period--i.e., the 51 observations in market period

one, the 51 observations in market period two, and the 102 observations

in market period three--were analyzed independently of the other two

periods. The units of measurements and the variables are identical

with those used in the preceding section. The results are given in

Table 3.

The coefficient for quantity of Spanish melons was not statis-

tically significant in periods two and three. The preliminary analysis

of the unload data and the values of the cross price flexibility esti-

mates from the preceding one-equation analysis also indicate the small

influence that this variable exerted during market periods two and three.

In market periods one and two the intercepts for the first and second

months did not differ significantly from the base or third months.

The estimates of the direct and cross price flexibilities for each

of the market periods are presented in Table 4. The correlation matrices

are given in the Appendix (Tables 11-13). The Durbin-Watson statistic for

market period two indicated no serial correlation at the .05 level,

and for market periods one and three the test was inconclusive at the

.05 level. An analysis of variance was performed with each estimated









Table 3. Coefficients for Each Marketing Period Analyzed Independently
for Honey Dew Melons in the New York City Wholesale Market.a



Equation Number and Period
Item (5, period 1) (6, period 2) (7, period 3)


Intercepts of honey


Base, last mo.
of period

1st mo.


2nd o.b


dew price-quantity relationship


11.16

.23
(.73)

- .69
(.78)


6.94

.23
(.34)

- .33
(.37)


3rd mo.


4th mo.


5th mo.b

Coefficients of independent variables


Quantity honey
dew melons

Quantity Span-
ish melons


-.000330**
(.000079)

-.000062**
(.000019)


Income .010276**
(.002734)


-.000019**
(.000007)

-.000031
(.000060)


.008894**
(.002898)


-.000035**
(.000008)

-.000007
(.000028)

.005458**
(.001408)


2
R .552 .305 .532

Sum of Squared
Residuals 138.38 48.03 161.35


aThe standard errors of the estimated coefficients are presented
in the parentheses.

Differences from base month.

*Significantly different from zero at the .10 level.
**Significantly different from zero at the .05 level.


9.71

1.15*
(.62)

1.43**
(.57)

1.21**
(.55)

1.71**
(.61)

1.50**
(.63)


-- --

















Table 4. Direct and Cross Price Flexibilities for Each Marketing Period Analyzed Independently.



Market Price Flexibilitiesa
Period Direct Cross
Honey Dew Melons Spanish Melons Disposable Personal Income

1. November to
January -.148 -.072 .302

2. February to
April -.094 -.007 .340

3. May to October -.198 -.002 .180


percentage change in price of honey dew melons associated with 1% change in quantity of
honey dew melons, quantity of Spanish melons, and in disposable personal income, respectively.








equation. The F statistic was highly significant for each of the

estimated equations.

The slopes of the honey dew melon price-quantity relationships

differ significantly among market periods. It is quite likely that the

effects of both the quantity of Spanish melons and the income on price

of honey dew melons, as well as the intercepts of the honey dew melon

price-quantity relationship, differ among periods, although these dif-

ferences were not tested statistically. Under these conditions, the

separate equations are believed to provide more accurate estimates of

the coefficients than do the one equation models in which the data for

all periods were pooled. In addition, there are only small differences

in the coefficients of determination between the separate equation

approach and the highest one in the one-equation models for periods

one and three. Thus the separate equations probably would give the

most accurate estimates of prices that might be received from melons

from Peru, at least in market periods one and three.













SUMMARY


Thus, we shall still use the fact that 2
dollars plus 2 dollars are 4 dollars, but
not that 2 raindrops added to 2 raindrops
are 4 raindrops. Two raindrops plus 2
raindrops make a puddle.4


The objective of this study was to estimate the wholesale demand,

on a seasonal basis, for honey dew melons in the New York City whole-

sale market. There was a particular interest in estimating prices that

might be expected for potential supplies from Peru during the winter

months. In the analysis, the price of honey dew melons depended on

the quantity of honey dew melons, the quantity of Spanish melons and

disposable personal income. The year was divided into three periods

which coincided with the geographic origin of supplies, and with the

quantitative importance of Spanish melons. In November through January

(period one) there was a decreasing domestic supply of honey dew melons

and small supplies from Chile and Ecuador, and this was the season of

the peak supply of Spanish melons from Spain. February through April

(period two) was supplied almost entirely from Chile; and May through

October (period three) was supplied from domestic production.

The coefficients of the equations for each season were estimated

by means of least squares regression using one equation as well

as separate equations--a separate one for each of the three periods.



4
Morris Kline, "The Meaning of Mathematics," in Thruelsen and
Kobler [13, p. 88].









In the one equation models, dummy variables were used to permit changes

in intercepts of the honey dew melon price-quantity relationships among

months within market periods, and changes in slopes among periods. How-

ever, the one-equation models had no dummy variables to allow for changes

in the Spanish melon and income coefficients among periods. The results

indicate that slopes of the price-quantity relationship for honey dew

melons differ among periods and that the intercepts differ among months

in periods one and three. The numerical results are summarized in Table

5. The magnitude of the differences in the intercepts between period two

and the other two periods, and in the coefficients for Spanish melons and

for income among periods are great enough to suspect that they are signifi-

cant, but they were not tested statistically. In view of these differences

among periods, the separate equations are believed to provide the more

accurate estimates of the coefficients.

What price might be expected for given quantities of honey dew

melons supplied from Peru? The market periods of major interest are one

and two. It is assumed that honey dew melons from Peru would be of a

similar quality to those from Chile. The procedure in projecting

prices that might be expected for given quantities of honey dew melons

from Peru would be to project the disposable income and Spanish melons

variables, update the coefficients in the estimating equations by

including the most recent observations, project the quantity of honey

dew melons expected from outside Peru and add to this the range of

quantities to be considered from Peru, and then compute the price-

quantity relationships. This same procedure would apply to any other

area, Florida for example, as well as to Peru. Special caution should


5 ,
Peruvian melons are known to contain a higher sugar content than
those grown in Chile. It is possible that Peruvian melons would command
a higher price than those from Chile.









Table 5. Summary of Estimated Equations, Price of Honey Dew Melons
in the New York City Wholesale Market, by Market Period and
Month, as a Function of the Quantity of Honey Dew Melons,
Quantity of Spanish Melons and Disposable Personal Income.a



Partial Regression Coefficients
Disposable
Personal
Month Intercept Honey dew melons Spanish melons Income

Period 1

November: 1 11.81 -.000223 -.000033 .006726

5 11.39 -.000330 -.000062 .010276

December 1 9.06 -.000223 -.000033 .006726

5 10.47 -.000330 -.000062 .010276

January 1 11.45 -.000223 -.000033 .006726

5 11.16 -.000330 -.000062 .010276


Period 2

February: 1 7.71 -.000014 -.000033 .006726

6 7.17 -.000019 -.000031 .008894

March 1 7.03 -.000014 -.000033 .006726

6 6.61 -.000019 -.000031 .008894

April 1 7.50 -.000014 -.000033 .006726

6 6.94 -.000019 -.000031 .008894


Period 3

May: 1 10.45 -.000035 -.000033 .006726

7 10.86 -.000035 -.000007 .005458

June 1 10.71 -.000035 -.000033 .006726

7 11.14 -.000035 -.000007 .005458









Table 5.--Continued.


Partial Regression Coefficients

Disposable
b Personal
Month Intercept Honey dew melons Spanish melons Income

July 1 10.50 -.000035 -.000033 .006726

7 10.92 -.000035 -.000007 .005458

August 1 11.03 -.000035 -.000033 .006726

7 11.42 -.000035 -.000007 .005458

September 1 10.89 -.000035 -.000033 .006726

7 11.21 -.000035 -.000007 .005458

October 1 9.72 -.000035 -.000033 .006726

7 9.71 -.000035 -.000007 .005458


rice
in cwt., and


of honey dew melons in dollars per cwt., quantity of melons
disposable personal income in billions of dollars.


1 is based on the first equation, where all observations were
pooled, and 5, 6, and 7 are based on the fifth, sixth and seventh
equations for periods one, two and three, respectively, where a se-
parate equation was used for each period.










be observed, of course, when working beyond the range of observations

used to estimate the regression coefficients.

This procedure may be illustrated by means of an hypothetical

example. In this example all steps except updating the coefficients

are included. The months of November and March are used for the

illustration:

Projections
November March
Cwt. honey dew melons
from outside Peru 10,000 150,000
Cwt. Spanish melons 51,000 2,800
Income, bill. $ 700 726

Now, suppose the quantities of honey dew melons one desires to consider

as supplies from Peru range up to 5,000 cwt. in November and up to 50,000

cwt. in March. The results calculated by means of intercepts and coef-

ficients given in Table 5 would be as follows:


Cwt. of honey dew melons Estimated price per cwt.
using the separate
From equation model
Total Peru From outside Peruequation model

November 10,000 0 10,000 $12.11
15,000 5,000 10,000 10.46

March 150,000 0 150,000 10.13
200,000 50,000 150,000 9.18


These price-quantity points could be plotted and connected by a straight

line in order to read the estimated prices for any quantity between the

two extremes--i.e., between 10,000 cwt. and 15,000 cwt. for November,

and between 150,000 cwt. and 200,000 cwt. in March.

Future work with regard to Peruvian honey dew melons should

deal with the allocation of supplies among periods and months within






24

periods in order to maximize Peruvian suppliers' net income. The

production aspects were not considered in this study but are recognized

as an important part of the economic problem. Any seasonal differences

in production costs must also be taken into account in the final

allocation model.









































APPENDIX














Table 6. Honey Dew Melons at New York City Wholesale Market, in Hundredweights.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.


1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962

1963


3,400

3,400

10,412

8,925

3,187

2,550

3,825

6,587

12,750

12,240

12,750

10,200

5,865


5,950

12,750

9,775

4,675

17,000

15,937

14,662

26,137

27,837

32,130

38,760

53,040

77,010


5,100

14,662

17,212

17,637

36,975

36,125

15,937

26,562

45,262

62,730

39,015

61,455

88,740


8,075

27,625

30,812

24,437

25,925

24,650

28,900

27,837

41,650

42,585

41,310

31,875

55,590


18,062

8,925

17,850

21,037

13,175

21,887

17,425

3,187

7,437

6,120

25,245

19,380

13,515


18,700

8,712

17,850

42,075

44,837

41,650

22,737

34,000

14,237

16,575

41,820

34,425

60,435


62,687

76,287

66,725

98,812

58,225

88,825

33,575

66,512

57,800

66,555

42,840

46,920

36,975


104,125

88,825

94,562

97,537

109,650

107,100

74,162

74,162

89,250

136,425

93,330

112,200

63,750


101,575

123,675

128,775

84,575

102,425

68,637

84,150

93,925

105,400

120,615

142,035

77,520

116,535


80,325

86,487

71,612

76,500

47,600

58,012

47,387

50,575

53,550

69,105

83,640

48,450

78,285


17,000

7,862

10,200

10,200

7,862

3,187

10,412

3,612

1,487

9,945

5,610

3,825

4,845


1,675

1,675

2,337

1,675

1,675

1,675

1,675

425

212

510

1,785

765

510








Table 6.--Continued.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

1964 6,120 62,985 96,135 65,280 18,870 29,325 27,285 87,720 120,105 46,155 3,060 3,060

1965 2,040 49,725 153,765 77,775 31,365 65,535 33,405 91,290 103,020 42,840 10,200 3,315

1966 10,710 71,910 181,458 59,058 9,180 11,322 66,402 98,226 78,642 28,458 10,098 1,836

1967 1,224 57,222 110,160 62,424 24,480 67,320 30,906 88,740 107,712 33,966 9,486 3,672

Mean 6,834 33,971 59,349 39,753 16,302 33,621 56,514 94,768 103,490 58,997 7,582 1,675


Source: Calculated from data in [18], [19] and [20].















Table 7. Wholesale Prices of Honey Dew Melons at New York City, Dollars Per Hundredweight.



Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.


1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962


12.50

11.34

11.34

10.24

10.97

12.94

14.38

13.34

12.78

12.35

9.32

11.81


8.99

8.54

7.77

9.62

9.32

9.84

10.89

9.72

8.26

8.89

8.09

9.65


8.85

7.65

8.99

8.92

8.47

8.42

9.13

8.64

7.95

7.77

7.77

8.38


10.17

9.58

9.04

9.22

9.20

7.81

8.94

9.72

8.82

8.82

10.17

10.87


9.46

10.75

9.93

10.94

12.82

11.55

14.31

12.66

10.52

13.08

10.49

12.35


11.15

16.07

12.47

10.05

10.89

11.06

12.71

8.82

10.82

12.05

10.64

12.14


9.65

9.93

10.94

9.39

10.66

8.02

14.57

8.17

11.11

10.78

11.25

10.26


8.66

9.98

10.38

8.92

7.74

8.09

10.47

9.37

12.12

9.67

11.41

9.41


9.06

8.73

7.62

9.74

7.77

8.80

9.58

10.26

12.00

9.37

9.46

9.91


7.58

8.42

9.44

9.41

8.28

12.00

9.48

10.85

10.07

9.37

10.00

10.00


1963 11.18 9.13 7.95 9.11 13.20 9.04 12.47 12.94 9.93 9.39 11.18 13.21


7.74

10.61

10.33

9.81

10.87

12.35

8.82

13.93

12.94

9.04

10.00

8.82


13.21

13.21

8.24

13.21

13.21

13.21

13.21

15.29

13.21

8.82

13.21

13.21








Table 7.--Continued.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

1964 11.74 9.34 7.88 8.12 15.11 12.42 12.80 9.65 9.65 9.62 9.48 13.21

1965 12.94 8.61 8.61 7.88 13.72 11.74 10.96 9.88 10.07 9.27 8.35 13.21

1966 12.17 10.07 7.65 8.68 15.48 16.47 10.73 11.06 11.77 10.99 12.45 17.25

1967 21.18 14.71 10.24 10.59 10.92 13.41 13.25 11.18 8.92 10.31 8.24 16.47

Mean 12.50 9.50 8.43 9.22 12.19 11.88 10.88 10.05 9.57 9.69 10.29 13.21


Source: Taken directly or calculated from data in [16] and [17].














Table 8. Spanish Melons at New York City Wholesale Market, in Hundredweights.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.


1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962

1963


0

0

0

0

0

0

0

2,250

225

3,120

7,200

2,400

4,500


0

0

0

0

0

0

0

2,700

0

6,210

540

2,700

3,510


0

0

0

0

0

0

0

1,125

1,125

7,830

0

8,640

6,480


0

0

0

0

0

0

0

900

225

1,620

540

2,160

7,560


0

0

0

0

0

0

0

225

0

0

270

2,160

3,780


0

0

0

0

2,250

0

0

0

0

300

2,100

2,400

2,700


1,750

0

0

250

1,250

0

500

1,750

5,750

7,800

7,200

1,800

0


0

0

1,500

7,500

11,000

3,250

6,500

12,750

7,875

42,600

15,900

18,300

12,000


0

750

11,000

7,500

7,500

3,750

10,500

22,500

28,500

32,400

19,800

51,900

51,000


3,000

2,250

10,750

10,750

25,500

20,750

17,500

13,500

25,500

30,000

20,750

30,250

30,500








Table 8.--Continued.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

1964 12,900 1,350 7,830 8,640 3,780 0 0 0 6,300 27,000 61,500 22,250

1965 600 0 270 810 270 0 0 0 0 12,000 1,500 500

1966 684 2,268 3,564 0 0 0 0 0 0 0 360 750

1967 575 1,620 10,692 7,128 324 0 0 0 2,520 19,080 62,250 38,000

Mean 2,027 1,229 2,797 1,740 636 574 2,169 11,603 21,924 17,794


Source: Calculated from data in [18], [19] and [20].













Table 9. Disposable Personal Income in United States, Bil. Dollars.


Year Jan. Feb. Mar. Apr. May


June July Aug. Sept. Oct. Nov. Dec.


1951 216.5

1952 228.2

1953 246.4

1954 252.5

1955 258.8

1956 278.1

1957 300.8

1958 311.6

1959 326.8

1960 345.5

1961 352.3

1962 372.7

1963 393.6


218.4

231.9

247.3

252.4

259.5

279.2

303.3

310.4

329.0

345.5

351.8

375.8


219.5

231.5

249.9

251.7

261.9

281.4

304.8

312.0

333.5

345.9

356.0

379.1


222.5 223.9

232.0 234.2

249.8 250.3

251.8 252.7

265.4 268.4

284.2 285.7

306.0 308.9

312.4 314.0

335.8 338.8

350.9 353.1

357.3 360.8

380.7 383.2


392.5 394.4 396.8 399.6


225.4

236.1

252.7

253.7

269.6

287.3

311.4

316.2

340.8

353.9

363.9

384.1


225.3 227.4 228.7 231.0


235.4

252.4

253.2

272.3

285.8

312.3

321.7

340.7

353.9

366.8

385.4


241.3

251.6

253.6

272.2

289.5

313.2

321.1

337.3

354.4

365.0

386.5


244.2 244.5

251.1 251.6

255.1 255.2

275.0 275.6

291.3 293.6

312.4 311.9

322.8 322.9

338.1 337.6

354.7 355.9

366.4 369.2

387.4 388.2


402.1 403.1 405.2 408.0 411.0


230.1 231.3

242.8 245.4

250.7 250.5

258.7 260.3

278.5 281.6

294.4 294.3

311.8 310.0

326.5 325.8

342.1 347.1

355.5 353.5

373.4 377.6

390.4 392.6

412.1 415.4








Table 9.--Continued.


Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

1964 421.7 423.9 426.1 433.2 436.1 438.1 439.4 443.6 446.3 445.1 448.9 456.7

1965 454.4 454.9 458.4 459.8 465.0 469.4 473.7 476.5 491.9 486.8 491.6 496.6

1966 494.9 500.1 505.1 503.3 504.2 509.3 510.7 515.4 519.9 522.0 526.0 528.0

1967 531.6 53441 537.1 538.5 540.5 545.7 546.2 550.6 553.4 552.4 559.3 567.0


Source: [21].














Table 10.--Extended.


S1 S2
D D D D D D Honey Dew Honey Dew Honey Dew
22 13 23 33 43 53 Quantity Quantity Price
(cwt.) (cwt.) ($/cwt.)


-.021

-.066

0.132

0.000

0.578

0.000

0.000

0.000

0.000

1.000


-.007

-.127

-.210

0.000

0.000

0.408

0.000

0.000

0.000

0.000

1.000


0.000

-.145

-.072

0.000

0.000

0.408

0.000

0.000

0.000

0.000

0.000

1.000


0.003

-.145

0.109

0.000

0.000

0.408

0.000

0.000

0.000

0.000

0.000

0.000

1.000


0.008

-.129

0.413

0.000

0.000

0.408

0.000

0.000

0.000

0.000

0.000

0.000

0.000

1.000


0.016

-.084

0.482

0.000

0.000

0.408

0.000

0.000

0.000

0.000

0.000

0.000

0.000

0.000

1.000


0.072

-.074

-.229

0.514

0.000

0.000

0.767

0.112

0.000

0.000

0.000

0.000

0.000

0.000

0.000

1.000


0.223

-.082

0.328

0.000

0.786

0.000

0.000

0.000

0.347

0.607

0.000

0.000

0.000

0.000

0.000

0.000

1.000


0.265

0.050

-.429

0.371

-.420

0.043

0.264

-.045

-.157

-.306

0.220

0.176

0.037

-.079

-.147

0.088

-.333
1.000

























Table 11. Simple Correlation Matrix, Separate Equation Approach,
Period 1 ( Nov., Dec., Jan.).



Disposable Spanish Melons Honey Dew Melons Honey Dew
Income Quantity Quantity D11 D21 Price
(Bil. Dol.) (cwt.) (cwt.) ($/cwt.)

1.0000 0.3509 -0443 -.0935 0.0380 0.2740

1.0000 -.2481 -.5067 0.3413 -.2038

1.0000 0.2500 0.3771 -.5113

1.0000 -.4999 0.1436

1.0000 -.4911

1.0000


























Table 12. Simple Correlation Matrix, Separate Equation Approach,
Period 2 (Feb., Mar., Apr.).



Disposable Spanish Melons Honey Dew Melons Honey Dew
Income Quantity Quantity D12 D22 Price
(Bil. Dol.) (cwt.) (cwt.) ($/cwt.)

1.0000 0.5440 0.7747 -.0165 -.0548 0.2216

1.0000 0.5078 -.1654 0.1794 -.2702

1.0000 -.2102 0.2542 -.1238

1.0000 -.4352 0.2724

1.0000 0.3737

1.0000














Table 13. Simple Correlation Matrix, Separate Equation Approach, Period 3 (May-Oct.).



Disposable Spanish Melons Honey Dew Melons Honey Dew
Income Quantity Quantity D13 D23 D33 D43 D53 Price
(Bil. Dol.) (cwt.) (cwt.) ($/cwt.)

1.0000 0.1194 -.0639 -.0206 -.0094 -.0043 -.0029 0.0146 0.2345

1.0000 0.0646 -.1290 -.1758 -.1758 -.1336 -.0163 -.1626

1.0000 -.5617 -.3422 -.0520 0.4329 0.5435 -.6105

1.0000 -.1999 -.1999 -.1999 -.1999 0.3612

1.0000 -.1999 -.1999 -.1999 0.2849

1.0000 -.1999 -.1999 0.0415

1.0000 -.1999 -.1589

1.0000 -.2775

1.0000
















BIBLIOGRAPHY


[1] Ben-David, Shaul, and Tomek, William G. "Allowing for Slope
and Intercept Changes in Regression Analysis." A. E. Research
179. Department of Agricultural Economics, Cornell University
Agricultural Experiment Station, November, 1965.

[2] Biological and Agricultural Index. New York: The H. W. Wilson
Company, 1951-1968.

[3] Burk, Marguerite C. "Survey of Interpretations of Consumer
Behavior by Social Scientists in the Postwar Period." Journal
of Farm Economics. 49:1-31, February, 1967.

[4] Fajardo-Christen, Adrian. "Demand for Honey Dew Melons in the
New York City Wholesale Market, with Special Reference to the
Potential Market for Supplies from Peru." M. S. Thesis, Uni-
versity of Florida, Gainesville, December, 1969.

[5] Foote, Richard J. Analytical Tools for Studying Demand and Price
Structures. Washington: U. S. D. A. Agricultural Handbook No.
146, August, 1958.

[6] Fox, Karl A. Intermediate Economic Statistics. New York: John
Wiley & Sons, Inc., 1968.

[7] Houck, James P. "The Relationship of Direct Price Flexibilities
to Direct Price Elasticities." Journal of Farm Economic.
47:789-792, August, 1965.

[8] Johnston, J. Econometric Methods. New York: McGraw-Hill Book
Company, Inc., 1963.

[9] Rojko, Anthony S. "Time Series Analysis in Measurement of Demand."
Agricultural Economics Research. Vol. XIII, No. 2, April, 1961.

[10] Salzman, Lawrence. Computerized Economic Analysis. New York:
John Wiley & Sons, Inc., 1968.

[11] Seale, A. D., Jr. and Allen, M. B. Reactive Programming of Supply
and Demand for Watermelons Produced in Mississippi and Competing
Areas. Agricultural Economics Technical Publication, No. 1,
Mississippi Agricultural Experiment Station, Mississippi State
University, June, 1959.










[12] Suits, D. B. "An Econometric Model of the Watermelon Market."
Journal of Farm Economics. 37:237-251, May, 1955.

[13] Thruelsen, R. and Kobler, J., Editors. Adventures of the Mind.
New York: Vintage Books, Alfred A. Knopf, Inc., 1962.

[14] U. S. Department of Agriculture, Agricultural Marketing. Consumer
and Marketing Service. Vol. 14, No. 1, January, 1969.

[15] U. S. Department of Agriculture, Bibliography of Agriculture.
National Agricultural Library, 1951-1968.

[16] U. S. Department of Agriculture, Fresh Fruit and Vegetables, Market
News. Metropolitan: New York City, Daily Miscellaneous Report,
Various Issues.

[17] U. S. Department of Agriculture, Fresh Fruit and Vegetable Prices.
Wholesale New York City, Statistical Bulletins, Various Issues.

[18] U. S. Department of Agriculture, Fresh Fruits and Vegetables Un-
loads in Eastern Cities. Consumer and Marketing Service, Fruit
and Vegetable Division, Market News Branch, 1960-1967.

[19] U. S. Department of Agriculture, Table of Carlot Conversion Fac-
tors Fruits and Vegetables, Effective. January, 1966. Consumer and
Marketing Service, 1966.

[20] U. S. Department of Agriculture, Unloads of Fresh Fruits and Vege-
tables at New York City. Production and Marketing Administration,
Fruits and Vegetables Branch, 1951-1959.

[21] U. S. Department of Commerce, Survey of Current Business. Office
of Business Economics, Various Issues.

[22] U. S. Superintendent of Documents, Monthly Catalog of U. S. Govern-
ment Publications. Washington, U. S. Government Printing Office,
1951-1968.

[23] Wold, Herman and Jureen, Lars. Demand Analysis, A Study in Econome-
trics. New York: John Wiley & Sons, Inc., 1953.