Kiwifruit and the Lemon Problem: Do Minimum. Quality Standards Work?

Transcription

1 Christpher Fergusn and Hy Carman Kiwifruit and the Lemn Prblem: D Minimum Quality Standards Wrk? ABSTRACT Mandatry minimum quality standards fr size, grade, r maturity are a cntrversial prvisin included in many U. S. fruit and vegetable marketing rder prgrams. These mandatry rules, which apply t all prducers f the regulated fruit r vegetable, prevent the sale f any prduct that des nt meet minimum specified standards. Califrnia kiwifruit grwers apprved a federal marketing rder in 1985 with a prvisin establishing minimum quality standards in grade, size, and maturity that were first used fr kiwifruit during the marketing year. This paper examines the effects f minimum maturity standards (MMS) n the average price differentials between kiwifruit frm Califrnia and frm New Zealand. The analysis f terminal market data fund evidence f a lemn type prblem fr eight ut f 20 cases fr which data sets were available. Results are cnsistent with the prpsitin that minimum maturity standards crrected a lemn prblem f asymmetric infrmatin n sweetness f Califrnia kiwifruit in the Bstn, Ls Angeles and Philadelphia terminal markets. The early seasn price premium that New Zealand fruit enjyed ver Califrnia fruit decreased significantly after the impsitin f the minimum maturity standards. Intrductin Christpher Fergusn is Ecnmist, American Express in Phenix, Arizna and Hy Carman is Prfessr,

2 Minimum quality standards fr size, grade, r maturity are a cntrversial prvisin included in many U. S. fruit and vegetable marketing rder prgrams. These mandatry rules, which apply t all prducers f the regulated fruit r vegetable, prevent the sale f any prduct that des nt meet minimum specified standards. A pisning-the-well argument fr minimum maturity regulatins is ften used - early shipments f immature fruit that lk gd but taste sur will depress demand fr days r weeks t the detriment f bth prducers and cnsumers. This market failure, widely knwn as the lemns prblem, arises because sellers have better infrmatin n quality than d the buyers. Akerlf (1970) demnstrated that the lemns prblem is characterized by relatively lw quality and prices as bad prducts drive ut gd prducts. Leland (1979) illustrated that the infrmatin prvided by minimum quality standards can crrect the lemns prblem and increase ttal scial benefits. Adpting Leland s reasning, supprters f minimum maturity standards fr fresh fruit argue that bth prducers and cnsumers benefit frm the quality assurance ffered by such standards. Sme ppnents f minimum maturity standards bject t any interference with the peratins f a free market. Others believe that such standards are unnecessary r that they may interfere with individual effrts t establish a high quality r brand reputatin. Califrnia kiwifruit grwers apprved a federal marketing rder in 1985 that has a prvisin establishing minimum quality standards in grade, size, and maturity that were first used fr kiwifruit during the marketing year. Kiwifruit maturity, which is nt bvius frm visual inspectin, is measured with the sluble slids test (Brix level) fr sweetness. The minimum maturity standard requires a Brix level f 6.5 percent at the time f inspectin, which is dne sn after picking. Since kiwifruit cntinue t cnvert starch int sugar after they are picked, they shuld reach a Brix level f at Department f Agricultural & Resurce Ecnmics, University f Califrnia, Davis.

3 least 14 percent by the time they are cnsumed. If picked t early (at Brix levels less than 6.5 percent), kiwifruit may nly reach a Brix level f 12 r 13 percent and never taste sweet. Objectives This paper examines the effects f minimum maturity standards (MMS) n the average price differentials between kiwifruit frm Califrnia and frm New Zealand. Mre specifically, it uses a series f statistical tests t determine if the minimum maturity standard crrected a lemn prblem fr Califrnia prduced kiwifruit, and if s, the nature f the crrectin. Prcedures The test fr market crrectin f a lemn prblem is based n price differences f fruit frm different rigins within a single market. New seasn Califrnia prduced kiwifruit typically cmpetes with imprts frm New Zealand and Chile during the Octber - December perid. New Zealand, which develped the market fr kiwifruit with high quality standards, typically enjyed a price premium fr their fruit. The existence f the New Zealand price premium, in fact, helped stimulate Califrnia grwers interest in minimum quality standards. If a lemn prblem existed prir t the use f minimum quality standards, we hypthesized that this premium wuld be reduced by minimum maturity standards fr Califrnia fruit f the same size and grade. The analytical mdel used is similar t the mdel that Sextn, Kling and Carman (1991) develped and used t test fr perids f inefficient allcatin, spatial price discriminatin and prduct substitutability in markets linked by arbitrage in the U.S. celery industry. Daily data fr majr U.S. whlesale markets, that reprted kiwifruit prices fr cmparable sizes, grades and packs, were analyzed t determine the effect f Califrnia minimum maturity standards n the price premiums enjyed by New Zealand kiwifruit.

4 The initial assumptin is that fruits f different rigin are perfect substitutes and prices are differentiated nly by a stable premium, G, which may accunt fr grade classificatin and differences in bservable characteristics that yield infrmatin n taste. Nte that G may vary by size, terminal market, time, and cuntry f rigin. The arbitrage cnditin is dented by Equatin 1. The fruit s rigin is indexed by the superscript; d indicates the whlesale price f fruit prduced dmestically while is the whlesale price f fruit prduced in ther cuntries. Size f fruit is differentiated by z, i is the terminal market, and t is the time perid. If the arbitrage cnditin did nt hld, prfit pprtunities wuld exist and prices f discunted fruit wuld be bid up s that the equality is restred. d P z, i, t = P z, i, t - G z, i, t =... = P z, i, t - G z, i, t (1) By altering Equatin 1, the impact f the minimum maturity standards is measurable if it leads t the crrectin f a lemn prblem. Fr instance, prir t the marketing rder s impsitin f the minimum maturity standards at the cmmencement f the marketing year, dmestic fruit may have featured premiums r discunts t imprted fruit, depending n the degree f the lemn prblem. The impsitin f the minimum maturity standard stensibly eliminated the asymmetric infrmatin prblem by guaranteeing cnsumers that dmestically prduced fruit was f an adequate expected level f sweetness. Thus, if the minimum maturity standards were t crrect prblems due t asymmetric infrmatin, fruit f dmestic rigin wuld feature a rise in price relative t imprted fruit. Equatin 2 captures this effect. D is a binary variable equal t 1 after the impsitin f the minimum maturity standard. α represents the (psitive) impact n the price f dmestically prduced fruit caused by the minimum maturity standards. α may vary depending n the terminal and the fruit size since the degree f the lemns prblem may vary. Nte that the assumptin f perfect substitutability enables the relative vlumes f fruit frm different rigins t be ignred; the premium will remain

5 cnstant ver time even if imprted fruit is in shrt supply relative t dmestic fruit. Similarly, an imprtant assumptin is that bservable quality differences between fruit f different rigins are stable ver time. d P z, i, t = P z, i, t + G z, i, t d + α z, i D =... = P z, i, t + G z, i, t d +α z, i D. (2) The minimum maturity standards nly affect the premium n early seasn fruit. Fruit picked later in the seasn will ripen t meet the minimum maturity standards regardless f the 6.5% standard. A cmparisn f imprted and dmestic fruit is nly pssible when the tw are sld in direct cmpetitin, which ccurs nly near the beginning (Fall) and end (Spring) f the marketing years. Hence the binary variable is set t 1 nly during the Fall after the impsitin f the minimum maturity standards; the Spring premium (r discunt) is assumed t be unchanged by the minimum maturity standards. Estimatin The Califrnia Fruit Reprt has daily kiwifruit price data which are differentiated by gegraphic rigin and size. Prices are available frm a number f whlesale terminals including Bstn, Chicag, Cincinnati, Dallas, Detrit, Ls Angeles, Miami, Mntreal, New Yrk, Philadelphia, Pittsburgh, San Francisc, and Seattle. Unfrtunately, several factrs limit testing f the mdel t selected terminals, rigins, and sizes. First, Dallas, Miami, Mntreal, and Seattle were either characterized by having fruit f different (and nn-cmparable) packaging, such as ne layer flats frm New Zealand and vlume filled bags frm Califrnia, r fruit f different rigins were gruped tgether and priced identically. Hence these terminals culd nt be further examined. Secnd, the number f bservatins fr imprts is limited. While New Zealand sld fruit thrughut the 1980 s and early 1990 s, Chile did nt becme an imprtant factr until the 1990 s. Italy, which is nw the Wrld s

6 largest kiwifruit prducer, had nly a few bservatins in the 1990 s. Third, terminals may have featured bth imprted fruit and Califrnia fruit, but f different sizes (fr example, trays with 36 pieces f fruit versus trays with 42 pieces f fruit). All else equal, larger fruit cmmands higher prices, s different sizes are nt truly cmparable. A furth severely limiting factr is that cmparable sizes may change ver time. Fr example, in 1987, New Zealand and Califrnia may have bth sld size 36 fruit in a given terminal, and in 1988, they may have bth sld size 33 fruit in the same terminal. Unfrtunately, the tw sizes are nt cmparable, s these data cannt be used in estimating the mdel. Finally, while fruit f different rigins had a large number f cmparable prices in the 1990 s, very few fruit were cmparable in the early and mid-1980 s, prir t the minimum maturity standards impsitin, again restricting pre- and pst-1987 premium cmparisns. Since bth the lw and high prices are reprted fr the week, the lw premium and high premium were bth examined. Fllwing Spiller and Huang (1986, p. 137), Sextn, Kling and Carman (1991) chse t cmpare lw prices, citing that lw price seems t be that mst affected by cmpetitin, that it avids the issue f transactin versus list price, and that lw price is usually quted by independents wh rarely discunt. In the case f kiwifruit, hwever, high prices may be mre accurate since they are the prices fr the highest quality fruit. During the end f the seasn, when fruit f a given rigin becmes sft and heavily discunted, high prices may be mre reflective f cmparable quality. Occasinally prices were reprted fr a given size spread, say the high and lw weekly price f sizes fruit. In this case, it was assumed that the high and lw prices were representative f size 36 and size 42, but that intermediate sizes, such as 39 culd nt be fairly assumed t be priced, and hence wuld nt be cmparable t a size 39 fruit frm a different rigin.

7 Given the data limitatins described abve, we were able t estimate 20 separate regressin equatins fr six different terminal markets using rdinary least squares methds. The sample perid extends frm the 1985 t the 1992 marketing years and estimatin was perfrmed using the Limdep sftware package. Only eight f the 20 regressins exhibited significant explanatry pwer and/r cefficients. Durbin-Watsn statistics indicated the presence f autcrrelatin s AR(1) regressins were als estimated. While we reprt the AR(1) regressin results, the magnitude and significance f the cefficients is virtually identical fr bth mdels. The nnsignificant regressins that were drpped frm further cnsideratin included Chicag 36 lw and high, Chicag 39 lw and high, Cincinnati 39 lw and high, Ls Angeles 30 lw, New Yrk 33 lw and high, New Yrk 36 lw and high, and Philadelphia 39 lw. We cnclude that there was n statistical evidence f a lemn prblem fr kiwifruit f the specified sizes in these markets. The eight cases that have evidence f a lemns prblem were in the Bstn, Ls Angeles, and Philadelphia markets. Table 1 presents the regressin results fr the Bstn terminal market. The estimated cefficients are in the lwer part f the table with t-statistics in parentheses; bld-fnt cefficients are statistically significant at the 95 percent cnfidence level. Fr the sizes and prices (high r lw) listed, the results strngly supprt the hypthesis that the impsitin f the maturity standard in 1987 led t a decrease in the premium f New Zealand fruit ver that f Califrnia fruit. New Zealand kiwifruit had a significant and psitive estimated premium ver fruit frm Califrnia prir t minimum maturity standards. The estimated α cefficient in each case is significantly negative, suggesting that the minimum maturity standards led t a decrease in the premium. In the Bstn market, the initial premiums fr the lw price bservatins were quite similar fr bth the 39 and

8 Table 1. Minimum Maturity Standards Regressin Results Fr The Bstn Terminal Market by Fruit Size and Price Bund, Terminal Bstn Bstn Bstn Bstn Size Price Lw High Lw High Observatins pre-1987 pre-1987 & Spring Parameter Parameter Estimate and t-statistic Initial N. Z..87 Premium (2.01) MMS Dummy (-3.69) Rh.77 (16.58) 1.06 (2.67) (-4.57).79 (17.87).89 (2.90) (-8.48).81 (15.83) 1.09 (4.32) (-9.47).75 (13.03) 36 sizes, as were the initial premiums fr the high price bservatins fr bth sizes, with high price bservatins having a larger premium. The decrease in premiums, as measured by the minimum maturity standards dummy was larger fr the larger size 36 kiwifruit. Regressin results fr the Philadelphia and Ls Angeles markets are shwn in Table 2. The results fr the Philadelphia, high price, size 39 kiwifruit are cnsistent with the Bstn results in Table 1, but the cefficients are nt as large nr as significant. This can be at least partially explained by the small number f bservatins fr the Philadelphia market, especially prir t The signs n the estimated cefficients fr Ls Angeles size 30 kiwifruit are the ppsite expected, althugh the cefficient fr initial New Zealand premium is nt significantly different than zer. The number f bservatins fr Ls Angeles, while mre than Philadelphia, is still cmparatively small. In additin, the large size f cmparable fruit in Ls Angeles may be a factr. If size is crrelated with maturity, size 30

9 fruit may nt have exhibited a lemn prblem prir t 1987; after the minimum maturity standards it is cnceivable that the supply f larger fruit frm Ls Angeles rse relative t that frm New Zealand as fruit was harvested later, thus Table 2. Minimum Maturity Standards Regressin Results Fr The Philadelphia and Ls Angeles Terminal Markets by Fruit Size and Price Bund, Terminal Phil. L. A. L.A. L.A. Size mixed mixed Price High High Lw High Observatins pre-1987 pre-1987 & Spring Parameter Parameter Estimate and t-statistic Initial N. Z..58 Premium (2.74) MMS Dummy -.75 (-2.62) Rh.19 (1.63) -.18 (-.99) 1.13 (4.07).68 (7.60) 2.23 (4.12) (-3.59).75 (11.94) 1.33 (4.85) (-3.42).53 (6.58) diminishing the premium during the latter time frame. Nte that this explanatin requires fruit f different rigins t be imperfect substitutes. The cmparable size requirement was relaxed fr Ls Angeles in rder t examine the effects f a larger number f bservatins prir t Weekly high and lw prices are available frm the Ls Angeles whlesale market frm in the Ls Angeles Whlesale and Vegetable Reprt. The nly prblem is that prices are reprted fr a spread f fruit sizes which changes peridically.

10 Hence the assumptin must be made that the premium is relatively stable acrss sizes. As shwn in the last tw clumns f Table 2, the results are similar t thse fund in Bstn and Philadelphia. Again, the minimum maturity standards appears t have reduced the premium between New Zealand and Califrnia fruit, which is cnsistent with minimum maturity standards crrecting a lemn prblem. The abve results are based n the assumptin that kiwifruit frm different rigins are perfect substitutes. We relaxed this assumptin and estimated a switching regressin mdel that allwed fr imperfect substitutin with perids during which the premium fr dmestic fruit culd rise r fall. The results were very similar. The initial premium fr New Zealand fruit ver Califrnia fruit was significantly psitive except fr Bstn 36 Lw and the minimum maturity standards binary cefficient was significantly negative except fr L.A. 30 Lw. Analytical results are cnsistent with the prpsitin that minimum maturity standards crrected a lemn prblem f asymmetric infrmatin n sweetness f Califrnia kiwifruit in the Bstn, Ls Angeles and Philadelphia terminal markets. The early seasn price premium fr New Zealand fruit ver Califrnia fruit decreased significantly after the impsitin f the minimum maturity standards. This result was cnsistent fr each f the mdels emplyed. The interpretatin is that the minimum maturity standards guaranteed a higher quality fruit t buyers, enabling them t pay a high quality price withut the fear f paying high prices fr lw quality fruit. As expected, the benefits vary acrss gegraphic regins and fruit size as the size f the lemn prblem varies, thugh a significantly reduced premium was cnsistently evident. Managerial Implicatins Market imperfectins caused by asymmetric infrmatin abut quality between cnsumers and the upstream sectrs f the marketing chain can reduce prices and prfitability. Results f this study

11 demnstrate the psitive price impact f mandatry minimum maturity standards in the Califrnia kiwifruit industry. While similar price advantages are theretically available fr firm level quality prgrams, realizatin f higher prices is dependent n clear cmmunicatin f prduct characteristics with custmers. While brands can prvide infrmatin that custmers trust, brands recgnized by fruit and vegetable cnsumers have been difficult t establish. Price premiums are available t innvative firms that find a way t cmmunicate their higher quality standards t all f their custmers. Quality standards based n gegraphic rigin have been successful marketing tls fr bth New Zealand and Califrnia kiwifruit. This strategy has bvius shrtcmings, hwever, fr multinatinal firms with glbal markets and surces. The New Zealand Kiwifruit Marketing Bard has recently adpted the Zespri brand name fr their prduct. If successful, the Zespri brand will permit such things as internatinal surcing f the new glden kiwifruit develped by the Bard as part f a strategy t license prductin f the new prduct and establish year-arund supplies f the brand t the market. References Akerlf, Gerge A., "The Market fr 'Lemns': Quality Uncertainty and the Market Mechanisms," Quarterly Jurnal f Ecnmics 84(1970): Leland, Hayne E., "Quacks, Lemns, and Licensing: A Thery f Minimum Quality Standards," Jurnal f Plitical Ecnmy 87(1979): Sextn, R. J., C. L. Kling, and H. F. Carman, Market Integratin, Efficiency f Arbitrage, and Imperfect Cmpetitin: Methdlgy and Applicatin t U. S. Celery, Amer. J. Agr. Ecn., 73(1991): Spiller, P. T., and C. J. Huang, On the Extent f the Market: Whlesale Gasline in the Nrtheastern United States, J. Indust. Ecn. 34 (1986):

12 USDA, Agricultural Prices, varius issues. USDA, Fruit and Vegetable Shipments, varius issues. USDA/AMS, Federal-State Market News Service, Califrnia Fruit Reprt, daily issues. USDA/AMS, Federal-State Market News Service, Ls Angeles Whlesale and Vegetable Reprt, daily issues.