Diversity of Sources for Fresh Produce:Implications for Local Markets

Luanne Lohr and Steven D. Hanson*

Abstrac~

Number of suppliers, approximation of equal-shares market condition and market shareheld by in-state sources were calculatedto determinediversityof sourcesfor 10 fresh fruits andvegetables in eight U.S. wholesalemarkets. Specificityof growing conditions is associated withfew supply sources, unequal market shares and limited purchases from in-state suppliers. Forcropswith few sources, lower perishability and greater transportability are correlated with greater balancein market shares. For crops with many supply sources, greater perishability and greatertransportability are consistent with large market share from imports. Diversity across allcommodities can increase market share for local producers.

Key Words: concentration index, fruits and vegetables, source diversity

The U.S. food supply is considered one ofthe most diverse in the world. An amazing varietyof food products is available in the typical grocerystore, On average, 280 different fresh fruit andvegetable items are stocked year-round, with 310different items in summer months (Litwak andMaline), Many of these commodities travel longdistances across country, or even internationally,between production and consumption regions.There is an entire infrastructure of production,transportation and marketing that supports thissystem.

Yet, the apparent abundance may mask alack of diversity in supply sources. Much of thisproduce comes from four or fewer states, with someitems available only from a single source. Relianceon limited numbers of production regions at greatdistance from consumers reduces market share forlocal producers and increases risk of consumer andretailer impacts from supply disruptions.Expenditures by state departments of agricultureon advertising to encourage demand for locally

grown fresh produce demonstrate support for greatersource diversity.

Little research has been directed towardevaluating diversity in fresh produce markets andexplicitly linking it to effects on local producers,consumers and retailers. Previous research hasfocussed on production diversity, ignoring theinteraction of supply and demand factors in themarket as well as the influence of marketing factorsand consumer preferences on measures of diversity(Tauer; Bacon and Gempesaw). Using data onwholesale shipments collected at demand centers isa simple way of representing the equilibria inmarkets. Thus, in our approach, consideration maybe given to changes beyond the farm-gate that canenhance diversity and self-sufficiency goals.

We describe the relationship betweensupply diversity and local markets, develop an indexof diversity and apply it to data on wholesaleshipments to describe eight major U.S, markets for10 fresh fruits and vegetables in terms of relative

*Luanne Lohr is assistant professor, Department of Agricultural and Applied Economics, University of Georgia. StevenD. Hanson is associate professor, Department of Agricultural Economics, Michigan State University. The authors thanktwo anonymous reviewers for their comments.

J Agr. and Applied EcotL 27 (2), December, 1995:510-521Copyright 1995 Southenr Agricultural Economics Association

511 Lohr and Hanson: Diversity of Sources for Fresh Produce

supply source diversity. The index accounts fornumber of sources and evenness in market power,as well as market share for local producers. Fivefruits (apples, grapes, strawberries, oranges, andcantaloupes) and five vegetables (iceberg lettuce,carrots, tomatoes, celery, and cucumbers) wereselected for this study, based on consumerpreference rankings in 1990 (The Packer), Thepreference criterion for selection is consistent withthe importance of evaluating consumer effects ofinadequate diversity.

The objectives of this study are to measuresupply source diversity for selected fresh fruits andvegetables, and to discuss the implications ofexisting diversity conditions for local self-sufficiency. In addition, this research seeks todescribe the association between commoditycharacteristics and differing degrees of sourcediversity, and to make recommendations forenhancing self-sufficiency.

Source Diversity and Local Markets

There are two competing concerns for localmarkets - diversity of sources and share supplied bylocal producers. Diversity can be assessed bybalance in market share among suppliers. Freshfruits and vegetables are packed and shipped fromnear the production site with minimal processing,Thus, production regions may be considered thesource units. The most diverse systems haverelatively many supply regions with nearly equalmarket shares. Greater diversity (number of andbalance among sources) reduces risk of supplyinterruption due to transportation system andgeographically specific production problems.Greater demand for local products improves theeconomic situation for farmers, provides fresherproduce for consumers, and reduces dependence ondistribution systems. Interest in increasing themarket share held by local producers has led statedepartments of agriculture to encourage demand forlocally grown commodities through advertising(McClure).

The conflict between these two objectivesis one of balance. Becoming self-sufficient impliesthe same susceptibility to geographically specificproblems (weather and pest infestations) that canresult from concentrating production in 1imited

numbers of distant regions. Failure to develop localproduction possibilities leads to dependence on longdistance transportation and distribution systems tocoordinate delive~ of fresh produce. Thesesystems have become highly concentrated, with ahandtld of firms controlling global distributionpatterns of fresh fruits and vegetables (Friedland).Disruptions in transportation (fuel cost increases,worker strikes) or distribution (contract failures,quarantines) are more significant when the marketrelies on long distance suppliers. Both total self-sufficiency and concentration of market shareincrease the risk to the local food system. Foodsystem diversity should be evaluated on the basis ofnumber of sources, balance in market power andshare held by local producers.

Market Share for Local Sources

Bacon and Gempesaw found thatproduction of fmits and vegetables is geographicallyconcentrated, with net surpluses in the ratio of farmproduction to retail demand only for the Pacific,Mountain, and South Atlantic regions. The supplyregions that can bring produce to markets at thelowest cost can capture the majority of marketshare. Farm production in these regions isdominated by California, Oregon, Washington,Arizona, Colorado, and Florida. Since most ofthese areas grow a variety of crops, rather thanspecializing in a single commodity, impacts ofgeographically specific problems may affect a largeshare of fresh fruit and vegetable production. Thisrisk is not calculated in the marginal cost ofproduction, but suggests aspects other than price areimportant in market analysis.

While most states cannot achieve self-sufficiency in supply of fresh produce due toclimatic and edaphic conditions, it is possible togrow a variety of crops for immediate sale orstorage. Apples, tomatoes, potatoes, lettuce, onions,celery, carrots, mushrooms, and corn accounted for38 percent of supermarket produce sales in 1991(Litwak and Maline). Weimer and Hallamcalculated that production of I I vegetables and twotypes of melons in Iowa would generate a net gainof $17 million in new revenues to farmers, anddisplace more distant competitors in local markets.Higher marginal costs for production factors,cooling systems and distribution channels reduce

J. Agr and Applied Econ,, Decembe~ 1995 512

competitiveness of local producers, but riskreduction may offset these costs. Supermarkets arewilling to purchase locally grown produce whenquality is comparable to more distant suppliers(Coupe). Currently, foreign sources are increasingmarket share, reducing local competition whileincreasing reliance on long distance transportationand limited distribution channels (Litwak andMaline).

From the farmer’s perspective, failure todevelop local sources of fresh fruits and vegetableslimits choices in diversi~ing enterprise mix toreduce farm-level risk. Sales of fresh fi-uits andvegetables may be relatively lucrative, with averagefarm gate price at $185 per ton in 1983 (Bacon andGempesaw). If fruit and vegetable farming requiresdifferent factors of production, local input suppliersmay also benefit by diversifying. By notmaintaining local sources, consumers may also lose.If proximity equates with freshness, then longertravel distances result in lower product quality.Coupe cited retailer experience that consumersperceive this difference whether or not physicalindicators support this conclusion. To the extentthat information about farm practices is important toconsumers, obtaining details from local suppliers isprobably easier than with distant producers.

Diversity ofSources

Consumer effects from imbalance in marketshare and limited numbers of supply sources dependon the susceptibility of the consumption region tostochastic disruptions in supply. Besidesgeographically specific factors, transportation anddistribution, involuntary cancellations of as many as35 pesticide/commodity registrations under theDelaney Clause could result in abrupt decline inproductivity in important fruit and vegetable regions(Stimmann and Melnicoe). The more concentratedproduction is in a limited number of regions, thegreater the probability of noticeable market effects.Flexibility of consumption regions in adjusting tostochastic effects is Iimited. Since crop decisionsare made prior to the start of the growing season,the total quantity of produce available at harvest ina given year is fixed. Since product will flow to thedemand regions with the greatest ability to absorbprice increases, quantity that remains will bedistributed according to ability and willingness topay. Some consumption regions may be unable to

obtain a share of desired commodities, as supply isshipped to areas where demand and price arehighest.

The potential impacts - price increase,quality deterioration and restricted selection - wouldbe experienced most intensely by marginalconsumers of fresh produce. Those with the leastdisposable income may experience adisproportionate effect through inability to increasespending to compensate for price increases or tosearch out alternative supply outlets to offset qualitydeterioration and restricted selection. The size of aprice increase determines how many people will bepriced out of the market for fresh fruits andvegetables. Those unable to afford fresh producemay be forced to accept substandard quality or tosubstitute other commodity choices. Thedistributional impacts on these consumers may bedrastic, since substitutes for fresh produce - frozenand canned fruits and vegetables and vitamin pills -are even more costly per unit of nutrient. Further,

as Gussow noted, constituent products of foods willnot necessarily provide the same nutritional value asthe food from which they are derived; an exampleis catsup compared with fresh tomatoes,

With the large selection in types of freshfruits and vegetables, some measure of substitutionamong items is expected. However, as opposed toseasonal differences in price, quality andavailability, which are known in advance,adjustments to sudden changes within the seasonalpatterns are more difficult to make. Consumptionpatterns are set, with over 50 percent of sales in just11 categories of fresh produce. The factorspreviously described may limit availability of aclass of produce, such as all citrus from a particularregion of California. Item replacement is notalways possible nor desirable. For example, findingsubstitutes for fresh tomatoes or potatoes may bedifficult. Familiarity with substitutes and householdpreferences play a major role in willingness topurchase substitute fruits and vegetables, so theeffect is not necessarily a dollar-for-dollarreplacement. It is not necessary for an entire cropto be lost or all supply sources to be eliminated inorder for these effects to occur, since distributionsystems reallocate among consumption regions,potentially creating pockets with these conditions.Lack of diversity in both numbers of suppliers andbalance among sources intensifies this risk.

513 .Lohr and Hanson: Diversily of Sources, for Fresh Produce

Retailers also may be affected by lack ofsupply diversity. Fresh produce is important instore profitability. In 1991, more than 10 percent ofsupermarket sales, a weekly average of $21,861,was fresh produce (Litwak and Maline). With anaverage gross margin of 38.2 percent, fresh fruitsand vegetables contributed 18.7 percent ofsupermarket’s total gross profit. Increasedwholesale prices due to random shocks reduce themargins on produce items, Obtaining alternatesupplies sufficient to meet local demand can bedifficult, particularly if suppliers have alreadycontracted their output. Dominance within thesupermarket indust~ is on a regional, rather than anational basis, so there is strong competition at thelocal level for consumer spending. Consumerperception of stores may be negatively affected ifproduce selection is limited, quality is poor, orprices are high. Inability to stock sufficient quantityand quality of preferred items in season encouragesconsumers to shop elsewhere. The greater thedominance by a limited number of sources, themore vulnerable retailers are. A stochastic shockmay not affect all retailers equally, since chains,rather than individual stores, negotiate wholesalepurchases.

Vulnerability to lack of diversity in freshproduce sources is a matter of degree. With thelarge number of items in produce departments, somesubstitution by consumers will occur in the event ofa supply shock. Supply recovery depends on thenature of the shock, its effect on productive capacityand the economic incentives for new entry, Themain factor in determining susceptibility is themeasure of diversity of supply sources.

Measurement of Source Diversity

An appropriate index of diversity shouldinclude measures of both the number of supplysources for a market and the distribution of marketshare among the sources. One such index used inecological studies (Magurran) and in marketconcentration studies (Tauer; Hannah and Kay) is

r 11

[1l(a) = ~ S,a m ,a*l,=1

(1)

r“ 1f(a) = exp l-~ S, JnSil ,a=l

where we define S, as the share of total marketshipments for a given commodity sent from the ithorigin to a given terminal. Separate values of l(a)

may be calculated for terminals representingdifferent markets.

The parameter a may be set over a rangeof nonnegative values to compare the number anddistribution of items in a unit sampled. Hannah andKay defined a as an elasticity parameter, whoserole is to indicate how much weight to attach to theupper portion of the distribution relative to thelower. In this case, the upper portion includes thedominant supply sources, based on quantity, whilethe lower portion is composed of other sources whomake smaller contributions to overall supply,Hannah and Kay noted that in general, high valuesfor a give greater weight to the role of the largestsuppliers in the distribution, and lower values of a

emphasize the presence or absence of smallersuppliers. If u = O,the index counts the number ofsupply origins, n, which can be seen by substitutingO for w For this value of c.t,the index makes nodistinction between relative dominance of smallerand larger supply sources. Higher values of a

accentuate the evenness of the distribution of marketshares among the supply origins. As a approachesinfinity, the index tends toward the reciprocal of theshare of the largest supply source. For ct = 1, theindex is equivalent to the Theil entropy index ofconcentration, so that a ranking based on the indexis the same as that given by entropy (Hannah andKay). For ct = 2, this index becomes the Herfmdahlindex used to measure market concentration(Hannah and Kay).

Both Tauer and Magurran noted thatchanges in the index value are small for cc > 2.Hannah and Kay tested values of a in comparingrankings of industry concentration, and found arange of a from 0.6 to 2.5 sufficient to makedistinctions in concentration. Thus, a can be alteredto reflect the concentration measure deemed mostappropriate for the problem. In our case, theproblem was to evaluate the dominant supplierrelative to the equal shares condition, so we used a= Oand a = 2. We calculated indexes for values upto a = 5, but found these values did notsubstantially alter the results,

L ,=1 J

J. Agr and Applied Econ., December 1995 514

Inere IS no absolute indicator of idealnumber of suppliers in a market, nor of distributionof market shares that eliminates susceptibility tostochastic supply disruption. Hannah and Kaydefined the process of market concentration as anincrease in the extent to which economic activity iscontrolled by large firms, or in the present case,large quantity suppliers. Differing marketconditions for terminal locations and commoditiesmake determination of single threshold valuesimpossible. Instead, relative diversity acrossmarkets may be compared using weighted marketshares and numbers of sources. Since weightedmarket shares depend on the number of supplysources, a common basis of comparison that isindependent of number of supply origins must beestablished.

We frame this index in terms of a perfectlybalanced market in which each source contributesan equal share of the market supply. This is notnecessarily the cost-minimizing market condition,since this situation may be costly to establish andmaintain. However, if the preceding analysis iscorrect, it is more likely to result in risk reductionthan a condition with imbalance of market share.Without reference to the equal-shares marketcondition, the values generated by the index inequation 1 with u = 2 can be used only fordetermining balance within a specific market. Todetermine how closely the actual marketapproximates an equal-shares condition, and tocompare balance across markets, the index mustaccount for equal-shares conditions that are specificto each market.

Markets are defined by commodity ofinterest and terminal location, Depending on thenumber of supply sources, each market has its ownidealized equal-shares condition. Since therequirements for an equal-shares condition are thesame for all markets, a basis of comparison isestablished that shows how well each marketapproximates its idealized condition. Incombination with numbers of supply sources, thisinformation provides a perspective on relativediversity across markets,

The opposite of an equal-shares situation isa monopoly, where one origin supplies the entiremarket. Under this condition Si = I in equation 1

for the ith origin, the monopolist, and l(a) = 1,regardless of the value of a. The farther from themonopoly situation a market is, the closer it is to anequal-shares condition. When this difference ismaximized, market shares are balanced among all nsupply origins. However, to avoid speci~ing thisdifference as a function of the number of origins ina market, so that markets are not perceived to bebetter balanced simply by virtue of having moreorigins, we normalize the monopoly and equal-shares conditions by the number of supply origins.Then we may compare how closely the actualmarket approximates the equal-shares condition byusing the monopoly situation as a point of reference.

We propose the index

where a > 0, and I(a) is the index of marketdiversity from equation 1, calculated for the actualmarket. The total number of supply origins in themarket is n, which is the same as 1(0) in equation 1.The specific value of n for a given market isdetermined by the data, l“(a) is the diversity indexfor a monopoly situation, which is divided by n tonormalize the index. t“(a) reduces to 1, since theone source supplies 100 percent of the market.

The denominator in equation 2 is themarket specific equal-shares condition, given by thedifference between the normalized indices for theperfectly balanced and the monopoly situations.The normalized diversity index for the perfectbalance situation has a value equal to 1. Since S, isthe same for each source, the diversity index fromequation 1 is equal to n, and is divided by thenumber of sources in the market, normalized to anequal-shares index equal to 1. The numeratordescribes how close to a monopoly the actualmarket is, given as the difference between thenormalized indices for the actual and monopolyconditions. Substituting the parameter a = 2 andsimplifying equation 2 gives

515 Lohr and Hanson: Diversity of Sources for Fresh Produce

1 (2)-1L(2)= ~_, ,n>l

(3)

A(2) = o ,n=l

The resulting diversity index, L(2),measures actual market concentration as aproportion of the equal-shares condition for a givenmarket. Since this index is a proportion, L(2) isbounded by zero and one. If L(2) = O, the actualmarket is controlled by a single supply source. [fA(2) = 1, market shares are evenly distributedamong sources. Values of L(2) between zero andone tend toward either single source dominance ifnear zero or the equal-shares condition if near one.

Approximation of the equal-sharescondition, combined with information about numberof supply origins and shares from in-state andforeign sources, provides an indication ofvulnerability to disruption. This index is richer thanconventional measures of concentration such asmarket share of the four most dominant sourcesbecause it simultaneously accounts for all suppliersin the market. Diversity depends as much on thenumber and shares of other suppliers in the marketas it does on the top four sources. The equal-sharesindex quantifies the difference between the actualmarket and a perfectly balanced market,

Data and Results

Arrival data published by the U.S.Department of Agriculture (USDA) quantifyingshipments of fresh produce by truck, rail, air andboat are classified by commodity, state or countryof origin and wholesale terminal destination. Thesedata represent more than 90 percent of the volumeof shipped produce received at U.S. terminals(How). This constitutes on average 94 percent ofproduce sold through retailers (Litwak and Cepeda).

Data selected for this study were 1990cumulative annual arrival quantities of fresh fruitsand vegetables, measured in hundredweights (cwt),from state or country of origin to eight majorwholesale terminals (USDA, 199la; USDA, 1991b).Regional dominance by the Pacific, Mountain, andSouth Atlantic states evident in this data implies that1990 shipments may be taken as representative of

domestic supply sources for at least the last 10years, reflective of similar trends exhibited in self-sufficiency ratios for fruit and vegetables calculatedby Bacon and Gempesaw for 1983, 1970, 1960, and1949 data, Cumulative data account for storage andsales from inventory and the foreign supplies thatalternate seasons with domestic sources. The eightterminals chosen represent four locations on theAtlantic Seaboard (Baltimore-Washington, Boston,New York-Newark, and Philadelphia) and four sitesin the Midwest (Chicago, Cincinnati, Detroit, andSt. Louis).

Eighty-eight separate fmit and vegetablecommodity categories are tracked by the USDA.Selection was based on consumer purchases. Thesecommodities are relevant since consumer reaction toprice increases, quality changes and reducedavailabilityy depends on preferences, which arereflected in purchase patterns. Though rankedsecond among fruits, bananas were not includedbecause no origin data were available, so the sixthranked fi-uitwas added. Fruits selected, by percentof consumers purchasing, were apples (96 percent),grapes (87 percent), strawberries (79 percent),oranges (78 percent), and cantaloupes (77 percent).Vegetables chosen were iceberg lettuce (88 percent),carrots (88 percent), tomatoes (84 percent), celery(8 I percent), and cucumbers (78 percent).

The equal-shares index A(2) was calculatedfor each of the 10 commodities identified and eightwholesale terminaIs. Proportional market shares foreach origin were based on total quantity shipped foreach crop and terminal combination. Market sharesattributable to the in-state and foreign sources werebased on percentage of deliveries from each source.For the New York-Newark market, both New Yorkand New Jersey shipments were considered in-state.

Table 1 shows approximations of the equal-share condition, given by values of L(2), and thenumbers of sources for the five fruit commoditiesand eight terminals. Values of A(2) in tables I and2 may range from 0.000 to 1.000. The closer is)42) to 0,000, the closer is the market to dominanceby a single firm, As L(2) approaches 1.000, themarket share is more uniformly divided among thesources. Table 2 provides the same information forthe five vegetable commodities and eight terminals.TabIe 3 gives the percentage of arrivals from in-state and foreign sources for the 40 fruit-terminal

J. Agz and Applied Econ., DecembeC 1995

combinations, while table 4 gives this informationfor vegetable commodities.

The most important factor in supplydiversity and self-sufficiency is uniqueness ofproduction requirements. Regions that meet uniquegrowing requirements will dominate all markets forthat crop. Among the sources that can produce agiven crop, balance may depend on seasonality,perishability and transportation and distributionsystems. The shorter the domestic season for acrop, the larger share foreign sources may supply.The more perishable a crop, the higher the marginalcost of transporting it, and the more likely that lessdistant sources can supply part of the market. Themore constrained the transportation and distributionsystems for a crop, the less distant the sources tendto be from the demand center. These hypothesesmay be examined using the results on tables 1through 4.

Overall, most markets for fruits andvegetables exhibit relative]y imbalance conditions,

516

regardless of whether there are only a few or manysupply sources. Among individual fruit markets, thegreatest balance exists in the Detroit market fororanges (0.486) and the St. Louis market for grapes(0.429), yet each of these has few supply sources(two and three, respectively). For both consumptionregions, markets for grapes show the greatestbalance, averaging 0.296 in the Atlantic Seaboardand 0.321 in the Midwest. The least balance is inthe Philadelphia and Chicago apple markets (0.033and 0.036, respectively), with the apple marketdisplaying the least balance over both regions, withaverages of 0.069 in the Atlantic Seaboard and0.098 in the Midwest.

Among individual vegetable markets, thegreatest balance exists in the New York-Newark andCincinnati markets for cucumbers (0.360 and 0.370,respectively). The cucumber markets average thegreatest balance, with averages of 0,293 in theAtlantic Seaboard and 0.307 in the Midwest. Forthe five vegetable commodities, the AtlanticSeaboard region overall tends to experience less

Table 1. Equal-shares Approximation and Number of Sources for Five Fruit Commodities and EightU.S. wholesale Terminals in 19!XF

Apples Grapes Strawberries Oranges Cantaloupe Average

ATLANTIC SEABOARD

Bakimore- A(2) 0.070Washington n 12

Boaton A(2) 0.10514

New York- :(2) 0.06sNewark 14

Philadelphia ;(2) 0.033n M

Average X(2) 0.069

MIDWEST

Chicago k(2) 0.036n 14

Cineirmati A(2) 0.0s4n 11

Detroit A(2) 0.1s6n 10

St. Lmria X(2) 0.0s7n 6

Average X(2) 0.098

0.36430.22550.30420.2924

0.296

0.19740.32840.33040.4293

0.321

0.33920.07730.05630.2172

0.172

0S2430.16s30.14330.0582

0.123

0.27630.07340.22s30.1404

0.179

0.0s640.25020.4s620.1632

0.246

0.204u0.170

100S2980.121

13

0.156

0.0S4140.19290.17770.178

11

0.158

0.251

0J30

0S57

o.159

0.105

0204

0.264

0.182

s Equal-shares approximation is X(2). Nmnber of sources is n,

517 Lohr and Hanson: Diversity of Sources jbr Fresh Produce

Table 2. Equal-shares Approximation and Number of Sorrrees for Five Vegetable Commodities andE@t U.S. Wholesale Terminals in 1990’

Lettuce Carrots Tomatoes Celery Cucumbers Average A(2)

ATLANTIC SEABOARD

Baltirnore-Wasbington

Boston

New York-Newark

Philadelphia

Average X(2)

MIDWEST

Cbieago

Cincinnati

Detroit

St. LQuis

Average A(2)

X(2)nX(2)nA(2)nX(2)n

X(2)nX(2)nA(2)nA(2)n

0.03811O.oiss80.20950.0739

0.097

0.06150.21230.13030.2523

0.164

0.23090.062

100.02360.0777

0.098

0.05160.16140.31630.1003

0.157

0.20Q140.075

250.166

150.153

16

0.148

0.167150.213

140.227

120.234

13

0.210

0.13970.15760.09740.1765

0.142

0.03060.10540.11550.1533

0.101

0.292140.234

19o.36il

130.’287

14

0.293

0.229160.37090.336

100.292

11

0.307

0.180

0.119

0.171

0.153

0.108

0.212

0.225

0.206

a Equal-shares approximation is A(2). Number of sources is n.

Table 3. Percentage of Arrivala from In-state and Foreign Sources for F&’eFruit Commoditirx andEight U.S. Wholesale Terminals in 1990

Apples Grapes Strawberries Oranges Carrtaloups

ATLANTIC SEABOARD

Baltimore- In-stateWashington Foreign

Boston In-stateForeign

New York- In-stateNewark Foreign

Pbifadelphia In-stateForeign

MIDWEST

Cbieago In-stateForeign

Cincinnati In-stateForeign

Detroit In-stateForeign

St.IKruia In-stateForeign

0.60.91.13.4

19.01.14.52.3

1.30.92.51.2

44.64:93.30.3

0.029.30.0

30.90.0

13.50.0

34.8

0.022.90.0

38.90.0

40.80.0

33.6

0.00.00.01.80.01.20.00.0

0.01.00.03.71.80.00.00.0

0.00.00.01.20.00.00.00.1

0.0O.oa0.00.00.00.00.00.0

4,625.60.0

30.60.0

26.40.3

23.8

0.414.50.5

28.83.6

29.20.0

19.6

a Thu vafue is less than 0.1 but greater than 0.0.

J. AgK and Applied Econ., Decembe< 1995 518

Table 4. Percentage of Arrivals from In-state and Foreign Sources for Five Vegetable Commodities andEight U.S. Wholesale Terminals in 1990

Lettuce carrots Tomatoes Celery Cucumbers

ATLANTIC SEABOARD

Baltimore- In-stateWashington Foreign

Boston In-stateForeign

New York- Iu-stateNewark Fore@r

Philadelphia In-stateForeign

MIDWEST

Chicago In-stateForeign

Cmciiati In-stateForeign

Detroit In-stateForeign

St.Louis In-stateForeign

0.00.40.00.90,20.00.00.9

0.00.00.00.07.30.00.00.0

0.041.53.011.30.02.9

0.05.8

0.02,20.00.6

24.90.90.00.0

6.511.90.68.40.0

13.84.2

11.3

1.614.23.5

11.84.2

24.31.3

13.5

0.02.00.00.00.00.00.00.6

0.00.82.80.06.10.00.00.0

5.220.4

1.920.111.224.03.2

17.2

7.62s.3

1.420.010.639.40.6

19.6

balance in market share except for celery, Among

individual markets, the carrot market in New York-Newark (0.023) and the celery market in Chicago(0.030) exhibit the least uniformity, The lowest

average regional values are 0.097 for lettuce and0.098 for carrots in the Atlantic Seaboard, andO,101for celery in the Midwest.

There is greater variability in the averagevalues of A(2) for the commodity categories than forthe terminal locations. Across the five fruit crops,Detroit shows the greatest balance (0.264), whileChicago exhibits the least (O.105). For thevegetable commodities, the outcome is the same,with the highest average for Detroit (0.225) and thelowest for Chicago (O.108). These results suggestthat, at least among the top five fruits andvegetables, there is relative dominance by a singlesupply source. This is consistent with the regionalproduction dominance described by Bacon andGempesaw.

Specificity of growing conditions,perishability and transportation and distributionfactors are reflected in the number of supply sourcesfor fruit and vegetable crops. For fruits, markets forgrapes, strawbemies and oranges use from two to

five supply sources, while markets for apples andcantaloupes receive from produce from six to 15sources, For vegetables, there are greater numbersof suppliers overall. This may be due to greaterdemand for vegetable commodities. Litwak andMaline noted that vegetables outsell fruit by about20 percent, Markets for lettuce, carrots, and celeryhave the fewest sources, from three to 11, whilemarkets for cucumbers and tomatoes are supplied by9 to 25 sources. For the vegetable crops, there is apositive relationship between number of sources andmarket balance, so that with more sources, there isgreater balance. However, this result does not holdfor fruits. The apple market has the most sources,but the least balance in supply sources.

For tomatoes and cucumbers, relativelylarge market share is attributable to in-state andforeign sources. Table 4 shows that from up to11,2 percent of supply to cucumber markets is fromin-state sources while up to 39.4 percent isimported. For tomatoes, from up to 6.5 percent ofsupply is produced in-state and up to 24.3 percentis imported. Table 3 shows that grapes andcantaloupes have significant shares from foreignsources, up to 30.6 percent for cantaloupes and up

519

to 40.8 percent for grapes. Among fruits, appleshave the largest shares from in-state production,ranging from 0.6 to 44.6 percent.

The results on tables 1 through 4 supportseveral hypotheses related to seasonality,perishability and transportation and distributionsystems. Greater specificity of growing conditions(oranges, strawberries, grapes, celery, and lettuce)results in fewer sources of supply, with almost noin-state supply outside of the dominant productionregions. Within this group, greater seasonality, lowperishability, good transportability and welldeveloped international transportation systemspromote a large market share for foreign sources(grapes). Market share is most balanced for cropswith lower perishability and greater transportability,regardless of share supplied by in-state and foreignsources (grapes and oranges).

For crops with relatively many supply

sources, there is less specificity in growingconditions (apples, cantaloupes, carrots, tomatoes,and cucumbers). Large foreign market share ispositively related to greater perishability but goodtransportability (cantaloupes, tomatoes, andcucumbers). Mexico is the primary foreign sourcefor these three crops in all markets. Relatively

lower share from imports is correlated with lessperishability (apples and carrots). Market balanceis greater for crops with high perishability and largeforeign market share (cantaloupes, tomatoes, andcucumbers).

For the eight terminals, self-sufficiency isgreatest for cucumbers, tomatoes, cantaloupes, andapples, based on market share for in-state sources.The Detroit market has the greatest in-state sharesacross all crop categories, ranging from 1,8 percentfor strawberries to 44.6 percent for apples. Grapesand oranges were the only crops for which theDetroit market did not obtain in-state supplies.Detroit also exhibited the closest approximation ofequal-shares across the fruit and vegetablecategories. Tauer found that Michigan ranked fifthin the U.S. in 1988 in terms of production diversity,with 44 agricultural commodities of all typesproduced and a diversification index (1(2) fromequation I) equal to 9.72, By contrast, the Chicagomarket had the lowest equal-shares indexes forfruits and vegetables, and ranked 34th in production

Lohr and Hanson: Diversity of Sources ,for Fresh Produce

diversity, with 22 commodities produced and adiversification index of 4.62. These correlationsbetween equal-shares index values for theconsuming regions and production diversificationindexes tend to hold for other markets as well. Thissuggests that greater diversity in a state’s crop mixsupports greater diversity and balance in theassociated market terminal, beyond the effects of asingle additional supply source.

Certain countries dominate the importshares of domestic markets. Mexico is the primarysupplier of cantaloupes, tomatoes, and cucumbersfrom outside the U.S. Canada is the main sourcefor carrots. Chile supplies most of the importedgrapes, which constitute mostly out-of-seasonsupply. Increased imports are expected in the wakeof the North American Free Trade Agreement(NAFTA) (Litwak and Maline). This mightimprove market balance as Mexico and Canadaincrease their share of the U.S. market, but willlikely reduce market share for in-state producers,

Conclusions

Diversity of sources for fresh produce maybe quantified by approximation of equal-sharescondition, number of suppliers, and market shareattributable to local (in-state) growers. The equal-shares index values for five fruits and fivevegetables in eight market terminals suggest thatregional dominance in production creates substantialimbalance in market shares. However, diversityacross all commodity groups in a state’s crop mixcan increase market share for local producers andmove markets toward the equal-shares condition.

Balance in market shares is primarilyaffected by specificity of growing conditions.Certain crops, such as citrus, cannot be grown inmost areas of the U.S. Such specificity isassociated with few supply sources, highly unequalmarket shares and almost no purchases from in-statesuppliers outside of the dominant production region.For crops with few supply sources, lowerperishability and greater transportability arecorrelated with greater balance in the market. Forcrops with relatively many supply sources, lessspecificity is evident in growing conditions. Greaterperishability and greater transportability areconsistent with large market share from imports.

J. Agr and Applied Econ., Decembec 1995

Transportation and distribution systems for importedcrops are critical in increasing foreign market share.

Self-sufficiency may be enhanced byexploiting consumer preferences for local produce,which is generally perceived as fresher. SourceIabelling requirements and advertising to encouragepurchase of locally grown crops would encouragethis preference. Use of technologies such as post-harvest cooling systems would reduce producercosts and improve local product quality. Gainsfrom such policies would flow to both producers

and consumers.

References

520

The relationship between vulnerability to

stochastic supply disruption and the equal-shares

index has not been determined. Data ontransportation and distribution systems are lacking.Future research may determine to what degree lessdiversity implies greater susceptibility to temporarysupply disruptions. While unlikely to threaten thetotal supply of fresh produce, localized consumerand retailer impacts are likely. Information ondemand and supply characteristics is needed topredict the response of producers, consumers andretailers. Diversity studies of other commodity-terminal combinations are needed to assess thevalidity of these results.

Bacon, J. R., and C. M. Gempesaw 11.“Surplus and Deficit Regions in Fruit and Vegetable Production.”Produce Marketing Association Yearbook 1988. Newark, NJ: Produce Marketing Association,

1988. Pp. 137-150.

Coupe, K. “Consumers Get Fresh With Local Produce.” Supermarket Bus. 43(March, 1988):23-28, 108.

Friedland, W. H. “The Global Fresh Fruit and Vegetable System: An Industrial Organization Analysis,”in P. McMichael, cd., The Global Restructuring ofAgro-Food Syslems. Ithaca, NY: CornellUniversity Press, 1994. pp. 173-189.

Gussow, J. D., “Food Security in the United States: A Nutritionist’s Viewpoint,” in L. Busch and W. B.Lacy, eds., Food Security in the United Slates. Boulder, CO: Westview Press, 1984. pp. 207-230.

Hannah, L., and J. A. Kay. Concentration in Modern industry - Theoty, Measurement and the U.K.Experience. London: MacMillan Press, Ltd, 1977.

How, R. B. Marketing Fresh Fruits and Vegetables. New York: Van Nostrand Reinhold, 1991.

Litwak, D,, and J. T. Cepeda. “199I Produce Operations Review.” Supermarket Bus. 46(October,1991):41-49.

Litwak, D., and N. Maline. “Fifth Annual Produce Department Operations Review.” Supermarket Bus.

47(October, 1992):35-42.

Magurran, A. E. Ecological Diversity and ILS Measurement. Princeton, NJ: Princeton Univ. Press, 1988.

McClure, B. S. “A New Marketing Force for Produce: State Department of Agriculture PromotionPrograms.” Supermarket Bus. 43(August, 1988):50-51.

Stimmann, M. W., and R. Melnicoe. “Delaney Clause Ruling May Trigger Pesticide Cancellations.” Cal.Agr. 48(January-February, 1994):30-35.

521 Lohr and Hanson: Diversity qfSources for Fresh Produce

Tauer, L. W. “Diversification of Production Agriculture Across Individual States.” J. Prod. Agr.

5(1992):210-214.

‘he Packer, Fresh Trends 1992- A Projile of Fresh Produce Consumers. 1992. pp. 20, 24.

U.S. Department of Agriculture. Fresh Fruit and Vegetable Arrivals in Eastern Cities - By Commodities,

States and Months, FVAS- 1 Calendar Year 1990. Market News Branch, Fruit and VegetableDivision, Agricultural Marketing Service, Washington, DC, 1991a.

Fresh Fruit and Vegetable Arrivals in Western Cities - By Commodities, States and Months,

FVAS-2 Calendar Year 1990. Market News Branch, Fruit and Vegetable Division, AgriculturalMarketing Service, Washington, DC, 199lb.

Weimar, M, R., and A. Hallam. “Risk, Diversification, and Vegetables as an Alternative Crop forMidwestern Agriculture. ” N. Cent. J. Agr. Econ. 10(1 988):75-90.