DETERMINING THE EFFECT OF NAFTA ON THE NORTH AMERICAN SWEET POTATO MARKET

Transcription

1 DETERMINING THE EFFECT OF NAFTA ON THE NORTH AMERICAN SWEET POTATO MARKET YOUNGJAE LEE Department of Agrcultural Economcs and Agrbusness Lousana State Unversty AgCenter 242A Martn D. Woodn Hall Baton Rouge, LA Phone: Fax: E-Mal: P.LYNN KENNEDY Department of Agrcultural Economcs and Agrbusness Lousana State Unversty AgCenter 8 Martn D. Woodn Hall Baton Rouge, LA Phone: Fax: E-Mal: LKennedy@agcenter.lsu.edu Selected Paper prepared for presentaton at the 206 Agrcultural & Appled Economcs Assocaton Annual Meetng, Boston, Massachusetts, July 3-August 2 Copyrght 206 by Youngae Lee and P.Lynn Kennedy. All rghts reserved. Readers may make verbatm copes of ths document for non-commercal purposes by any means, provded that ths copyrght notce appears on all such copes.

2 Determnng the Effect of NAFTA on the North Amercan Sweet Potato Market Abstract: In order to analyze the effect of NAFTA on the North Amercan sweet potato market, ths study uses both observable trade data such as trade volumes and home prces and unobservable trade data such as mported prces and nontarff trade costs. The unobservable trade data are estmated by the model used n ths study. Unlke prevous studes, ths study confrms that the degree of ntegraton n the North Amercan sweet potato markets s relatvely low because the three ndvdual sweet potato markets are mostly domnated by ther own product and the dfference between home and mported prces n each NAFTA member country have ncreased from pre-nafta to post-nafta levels. Two counterfactual analyses confrm that low levels of market ntegraton may be due to hgh non-tarff trade costs. Key words: NAFTA, Sweet Potatoes, Non-Tarff Trade Cost, Trade Creaton, and Trade Dverson. The North Amercan Free Trade Agreement (NAFTA) facltated agrcultural trade among three North Amercan countres (Zahnser et. al., 205; and Zahnser and Crag, 2009). Sweet potato trade among NAFTA s member countres has grown remarkably snce NAFTA was mplemented n 994. For example, the trade volume of sweet potatoes among the three North Amercan countres has ncreased by more than 500%, from 44 mllon pounds n 994 to 270 mllon pounds n 20. As prevous studes ndcated (Balassa, 96; Barchello et al., 99; de Janvry, 996; Gould, 998; Grennes, 99; Jayasnghe and Sarker, 2008; and Robbngs, 204), NAFTA has helped to rase the level of ntegraton among North Amerca s sweet potato markets. However, sweet potato prces n the three North Amercan countres are stll sgnfcantly dfferent. For example, average prces of sweet potatoes n post-nafta between 994 and 20 are $0.75 per pound n the Unted States, $0.483 per pound n Canada, and $0.02 per pound n

3 Mexco. Furthermore, trade data suggests that there mght be a much greater dfference between prces of mported and locally grown sweet potatoes. In emprcal analyses, however, researchers face practcal dffcultes because some trade data are unobservable, such as non-tarff trade costs and fnal prces of goods consumed n the mportng country and produced n the exportng country (Anderson and van Wncoop, 2004; Eaton and Kortum, 2002; Khan and Kalraan, 20; and Waugh, 200). Ths s what motvates ths study. Usng the North Amercan sweet potato market, ths study shows how unobservable trade data can be obtaned and n order to determne the effect of NAFTA on the North Amercan sweet potato market. In descrbng the effect of NAFTA on the North Amercan sweet potato market, ths study focuses on ) the effect of NAFTA on trade creaton and dverson, and 2) the effect of non-tarff trade costs on prce dvergence. In order to acheve these goals, ths study adopts assumptons accepted by prevous studes as follows: ) the North Amercan sweet potato consumer has a specfc structure of utlty whch determnes quantty demanded at a specfc prce level; 2) the trade costs, whch represent the dfference between the margnal cost n the exportng country and the prce pad by the consumer n the mportng country, consst of observable costs (tarffs) and unobservable costs (non-tarff trade costs); 3) factors of producton are mmoble between countres, mplyng that each of the member countres produces sweet potatoes usng ther own resources; and 4) there s a unque elastcty of substtuton representng the quantty relatonshp gven home and mported prces. Based on these assumptons, ths study proceeds as follows. In the next secton, we develop a model n whch the mported prce s defned by consumer preference, total factor productvty, producton costs, observable and unobservable trade costs, home prces, factor 2

4 prces, and expendtures. In secton three, we brefly descrbe the North Amercan sweet potato market. Data and estmated parameters are dscussed n the fourth secton. Emprcal results follow n the next secton. Fnally, the paper concludes wth future research recommendatons. Theoretcal Framework Demand Condtons NAFTA provdes ncreased opportuntes to North Amercan consumers n choosng food products orgnatng from dfferent sources. As North Amercan sweet potato consumers recognze dfferentated sweet potatoes sourced from dfferent orgns, consumers choce to maxmze utlty can be formulated by Dxt and Stgltz s CES utlty functon as follows: () n u q, where q represents sweet potatoes consumed n country and produced n country and * * where s elastcty of substtuton (Dxt and Stgltz, 977; Solow, 956; Tomeck and Robnson, 2003; and Varan, 992). Total expendtures by the consumer n country on orgn-dfferentated sweet potatoes can be summarzed as follows: n (2) pq M, where p represents a prce n country comng from country and M s total expendture on sweet potatoes n country. Ths study defnes quantty and prce as follows: q whch s produced n country and consumed n country s the mported quantty of sweet potatoes, q (or q ) whch s produced 3

5 n country (or ) consumpton of domestc sweet potatoes, p (or p ) s the home prce of sweet potatoes sold n country (or ) and produced n country (or ), and p whch s the mported prce of sweet potatoes sold n country and produced n country. Observable trade data nclude q, q (or q ) and p (or p ). However, the prce of mported sweet potatoes from country and consumed n country s not obtanable n the trade data. Therefore, p must be estmated. Gven the utlty and budget structure above, p can be estmated as follows: q (3) p p q. As equaton (3) mples, the dfference between home and mported prces n country wll ncrease f the dfference between market shares of domestc and mported sweet potatoes n country ncreases, whle the dfference between home and mported prces n country wll decrease f the dfference between market shares of domestc and mported sweet potato n country decreases. Ratonal sweet potato consumers n country dsplay an mport demand for sweet potatoes produced n country as follows: (4) q p n M p. When sweet potatoes produced n country move to country, tarff and non-tarff trade costs are ncurred. These trade costs are descrbed as follows: (5) p, t p 4

6 where t represents the value added tarff mposed by country on sweet potatoes sourced from country and t f but t f. The varable,, represents non-tarff trade costs resultng from the movement from country to country and f but f. The prce, p, s the exportng country s home prce. Therefore, p p f but p p f. Whle the causalty of the followng may be reversed, the mplcatons are mportant. One thng ths framework ndcates at ths pont s that f country does not mport from country, non-tarff trade costs wll be nfnte,, so that no consumer n country wll purchase sweet potato mported from country due to the unacceptably hgh prce. Substtutng equaton (5) nto equaton (4), we obtan an mport demand equaton defned by tarffs, non-tarff trade costs, exporter prce, and mporter expendture as follows: (6) q t p n l l l t p ll M. Therefore, mport demand n country for sweet potato produced n country wll be determned by trade costs, exporter s productvty, and mporter s preference. Supply Condtons If the factor market for sweet potato producton s opened, the producton technology of North Amercan sweet potato producers can be formulated by Dxt and Stgltz s CES functons as follows: (7) n q k x, k where represents total factor productvty n country, and n country that come from country k, wth 0. k x represents nput factors used 5

7 The producton costs can be summarzed as follows: K (8) C where k k r k x, k r s factor prce n country for nput factor, k x, produced n country k. If factor market s mmoble, then K =. Ths framework assumes that factors are mmoble between countres. Gven producton costs, the proft of the sweet potato producer n country s summarzed as follows: n (9) p sq k r k x, where K k q s total quantty of sweet potatoes produced n country and s represents the share of q sold n country. Usng Hotellng s Lemma, proft maxmzng sweet potato producers n country lead to a supply equaton defned by total factor productvty, producton costs, and prces of nput factors as follows: (0) C q. K k k r Then, sweet potato supply to country from country s defned as follows: () s C s q. K k k r Equlbrum Home and Imported Prces 6

8 7 In equlbrum, country s demand for sweet potatoes produced n country s equal to country s sweet potato supply to country. Thus, the home prce of exportng country can be derved by settng equaton (6) equal to equaton () as follows: (2) K k k L l ll l l M r t p t C s p. Substtutng equaton (2) nto equaton (5), we obtan the mport prce of country defned n terms of market share, productvty, producton costs, tarffs, non-tarff trade costs, exporter prce, factor prces, and expendtures as follows: (3) K k k L l ll l l M r p t C s p. Gven the elastcty of substtuton, equatons (4) and (3) are used to dentfy the effects of NAFTA on home and mported prces and the trade creaton and dverson effects n the three North Amercan countres. Furthermore, ths study develops two counterfactual scenaros n order to dentfy the role of trade costs n the market ntegraton of the three North Amercan countres. The North Amercan Sweet Potato Market In the North Amercan sweet potato market, the Unted States s a net exporter and Canada and Mexco are net mporters. In 989, U.S. exports to Canada and Mexco were 0.6 mllon pounds and 0. mllon pounds, respectvely. In 20, U.S. exports ncreased to 29 mllon pounds for Canada and 3 mllon pounds for Mexco, representng a 20 fold ncrease for Canada and a 26 fold ncrease for Mexco. In 99, the U.S. home prce was $0.3 per pound. By 20, the U.S.

9 home prce had ncreased to $0.9 per pound. In the pre-nafta perod, U.S. mports of sweet potatoes were 20 mllon pounds n 989, whle the Unted States dd not mport sweet potatoes from Canada and Mexco durng that tme. In 20, U.S. mports remaned relatvely constant at 23 mllon pounds. However, a shft had occurred as U.S. mports from Canada were mllon pounds, although the Unted States stll dd not mport from Mexco. Maor exportng countres to the Unted States were Chna, the Domncan Republc, and Peru. In 99, U.S. sweet potato producton was,9 mllon pounds; by 20, U.S. sweet potato producton had ncreased to 2,696 mllon pounds, representng a 2.4 fold ncrease durng the last twenty-one years. The data mply that the U.S. sweet potato market has been growng under NAFTA. In 988, Canadan mports of sweet potatoes were 9 mllon pounds. Among ther supplers, Canadan sweet potato mports from the Unted States were 5 mllon pounds, ndcatng that 80% of Canadan sweet potato mports came from the Unted States. Canadan sweet potato mports from Mexco was neglgble, wth only thousand pounds n 989. In 20, Canadan mports of sweet potato ncreased to 34 mllon pounds, an 8 fold ncrease snce 988. Durng that tme, Canadan sweet potato mports from the Unted States ncreased to 26 mllon pounds, a 9 fold ncrease. The U.S. share of the Canadan sweet potato market ncreased from 80% n 988 to 95% n 20. Durng the 986 to 988 tme perod, the Canadan home prce of sweet potatoes ranged from $0.3 per pound to $0.4 per pound. From 989 to 20, the Canadan home prce of sweet potatoes ranged from $0.4 per pound to $0.57 per pound. In 988, Canadan sweet potatoes exports were 0.6 mllon pounds, most of whch were exported to the Unted States. In 20, Canadan exports of sweet potatoes were 6 mllon pounds, a 26 fold ncrease n the last twenty-three years. Among these, Canadan sweet potatoes exports to the Unted States were mllon pounds, representng 6.8% of total Canadan sweet potato exports. 8

10 Maor mporters of Canadan sweet potatoes were Australa, New Zealand, Chna, Sant Perre and Mquelon, and the Unted Kngdom. Mexcan mports of sweet potatoes n 989 were 73 thousand pounds, all of whch were mported from the Unted States. In 20, Mexcan mports of sweet potatoes ncreased to 5 mllon pounds, an ncredble 68 fold ncrease snce 989. In 99, the Mexcan home prce of sweet potatoes was $0. per pound; by 20, that prce had ncreased to $0.3 per pound. In 992, Mexcan exports of sweet potatoes were a mere 5 thousand pounds, all of whch were exported to the Unted States. In 20, Mexcan exports of sweet potatoes had ncreased 73 fold, to 3.7 mllon pounds. Between 992 and 20, Mexcan sweet potato producton ncreased from 03 mllon pounds to 6 mllon pounds, an ncrease of only 0 percent durng that twenty year tmeframe. Data and Parameter Estmaton Data Ths study uses data obtaned from secondary sources but also data estmated through the framework developed n ths study. Ths analyss obtans data for domestc sweet potato producton ( q or q ), and mports ( q ) from the Food and Agrculture Organzaton of the Unted Natons [FAO] (205). Gven the start of NAFTA n 994, ths study uses 986 to 20 as the tmeframe for ths research as t ncludes pre- and post-nafta years. The U.S. home prce ( p or p ) data are obtaned from U.S. Department of Agrculture, Economc Research Servce (205). Mexcan home prce data are obtaned from the FAO (205-a). Ths study uses sweet potato prces estmated by Agrculture and Agr-Food Canada as the home prce of Canadan sweet potatoes (Agrculture and Agr-Food Canada, 205). Labor s used as an ndex for nput factors n sweet potato producton. Hourly wages of the three NAFTA member 9

11 countres are obtaned from the U.S. Department of Labor, Bureau of Labor Statstcs (205). Ths study also obtans sweet potato yeld data from FAO (205-b) whch s used as an ndex for total factor productvty ( ). U.S. sweet potato mport tarff data are obtaned from U.S. Internatonal Trade Commsson (205). Import tarff data for Canada and Mexco are obtaned from Duty Calculator (205). Ths study also uses data estmated through our analytcal framework. Ths study estmates the quantty of nput factors used n producton usng equaton (7) at θ = Import prce data are obtaned usng equatons (3) and (3), where ths study fxes θ = 0.75 and σ = -3. Non-tarff trade costs ( ) are estmated usng equatons (5) and (3). Sweet potato producton costs are then calculated usng equaton (8). Market share data ( s ), defned n equaton (), are calculated by q / q where q q. Total expendture on sweet potatoes n country ( M ) s calculated by Parameter Estmaton p q. In order to conduct counterfactual analyss, ths study estmates benchmark values of parameters n equaton (3). The parameter values are estmated usng the methods mentoned above and are presented n Table. For the purpose of ths analyss, the parameter values are estmated for pre- and post-nafta perods. [Table Approxmately Here] Market share parameters, s, show that the sweet potato markets of the three NAFTA member countres are domnated by ther own products; ths trend contnues after the mplementaton of NAFTA, except for the U.S. sweet potato market. For example, the home share of sweet potatoes n the Mexcan market ncreased from 98.3% pre-nafta to 99.6% post- 0

12 NAFTA, the home share of sweet potatoes n the Canadan market ncreased from 65% pre- NAFTA to 87.% post-nafta; and the home share of sweet potatoes n the U.S. market decreased from 98.4% pre-nafta to 95.7% post-nafta. U.S. sweet potato exports to Canada and Mexco ncreased after NAFTA; U.S. sweet potato exports to Canada ncreased from.5% pre-nafta to 2.6% post-nafta whle U.S. sweet potato exports to Mexco ncreased from 0.% pre-nafta to 0.2% post-nafta. However, Canadan and Mexcan sweet potato exports to the Unted States decreased after NAFTA; Canadan sweet potato exports to the U.S. decreased from 34.7% pre-nafta to.6% post-nafta and Mexcan sweet potato exports to the U.S. decreased from.7% pre-nafta to 0.4% post-nafta. Table shows the tarff rate, t. Ths study consders the elmnaton of the tarff between NAFTA member countres n post-nafta years, whle mantanng tarffs between NAFTA member countres and non-member countres n post-nafta years. As Table shows, non-tarff trade costs,, are much greater than that of tarffs. As ndcated prevously, nontarff trade costs are estmated by usng our framework whch reflects quantty based trade costs. Ths study uses equatons (3) and (5) to estmate non-tarff trade costs. In equaton (3), prces of mported sweet potato are derved by dvdng mported volumes by domestc volume, whch s then multpled by the domestc prce. The mport prces s then used n equaton (5) to estmate non-tarff trade costs. Ths estmaton method creates hgh non-tarff trade costs when mported volume s small relatve to domestc volume and low non-tarff trade costs when mported volume s smlar to the domestc volume. Snce the relatve sze of mported sweet potatoes s small relatve to domestc volume n each market, ths study confrms that non-tarff trade costs dd not sgnfcantly decrease from the pre-nafta to the post-nafta perod.

13 Our study uses yeld as an ndex for total productvty,. When s for the Unted States n the pre-nafta perod, s for Canada,.30 for Mexco, and for the rest of the world (ROW). However, after NAFTA s mplemented, s.253 for the Unted States, for Canada, and.32 for Mexco, and.64 for the rest of the world (ROW). When the U.S. wage rate, r, s n the pre-nafta perod, the wage rate s.09 for Canada, 0.07 for Mexco, and for ROW. These wage rates ncrease to.493 for the Unted States,.30 for Canada, 0.68 for Mexco, and for ROW n the post-nafta perod. Emprcal Results Usng benchmark parameter values, ths study estmates home and mported prces, non-tarff trade costs, and trade creaton and dverson generated by NAFTA. Table 2 shows home and mported prces between pre-nafta and post-nafta tme perods. Home prces of sweet potatoes n the three NAFTA member countres ncrease from pre-nafta to post-nafta tme perods. The U.S. home prce of sweet potatoes ncreases from $0.3 per pound pre-nafta to $0.8 per pound, reflectng 37% ncrease durng that perod of tme. The Canadan home prce of sweet potatoes ncreases from $0.36 per pound pre-nafta to $0.48 per pound followng mplementaton of NAFTA, ndcatng a 33% ncrease durng that tme perod. The Mexcan home prce of sweet potatoes ncreases from $0.098 per pound pre-nafta to $0.02 per pound n post-nafta, reflectng a 3.6% ncrease. In the three NAFTA member countres, the ncrease n mported prces from pre-nafta to post-nafta s greater than the ncrease n home prces except for the mported prce of U.S. sweet potatoes n the Mexcan sweet potato market. The mported prces of U.S. sweet potatoes n the Canadan sweet potato market ncrease from $0.4 per pound pre-nafta to $0.3 per pound post-nafta, reflectng a 9% ncrease. Imported prces of Canadan sweet potatoes n 2

14 the U.S. sweet potato market ncrease from $. per pound pre-nafta to $.86 per pound post-nafta, a 67% ncrease. Imported prces of Mexcan sweet potatoes n the U.S. sweet potato market ncrease from $0.68 per pound pre-nafta to $.40 per pound post-nafta, a 07% ncrease. Increasng mport prces relatve to home prces n the three NAFTA member countres are due to hgh non-tarff trade costs. As shown n Table, non-tarff trade costs are much hgher than tarffs mposed by mportng countres; these hgh non-tarff trade costs are not reduced under NAFTA. For example, non-tarff trade costs of U.S. sweet potato exports to the Canadan sweet potato market ncrease from.077 pre-nafta to.75 post-nafta (see Table ). [Table 2 Approxmately Here] Table 3 shows the effect of NAFTA on trade creaton and dverson. Canadan sweet potato mports from NAFTA member countres ncrease from 7 mllon pounds pre-nafta to 43 mllon pound post-nafta, ndcatng a 49% trade creaton effect. Mexcan sweet potato mports from other NAFTA countres ncrease from 0.9 mllon pound pre-nafta to 3.0 mllon pound post-nafta, ndcatng a 25% trade creaton effect. However, U.S. sweet potato mports decrease from.7 mllon pounds pre-naft to 0.5 mllon pound post-nafta. Total sweet potato trade volume among NAFTA countres ncreases from 20 mllon pounds pre- NAFTA to 46 mllon pound post-nafta, for an overall 33% trade creaton effect. Conversely, Mexcan sweet potato mports from non-nafta countres decrease from 6 thousand pounds pre-nafta to 62 thousand pound post-nafta, reflectng a 60% trade dverson effect. U.S. sweet potato mports from non-nafta countres decrease from 8 mllon pound pre-nafta to 7 mllon pound post-nafta, a 5% trade dverson effect. In contrast, Canadan sweet potato mports from non-nafta countres ncrease from 4 mllon pound pre- NAFTA to 6 mllon pound post-nafta. Ths ndcates that total sweet potato mports of 3

15 NAFTA countres from non-nafta countres ncreased from 22 mllon pound pre-nafta to 23 mllon pound post-nafta, reflectng no NAFTA trade dverson effect. As a result, ths study ndcates that, n the case of sweet potatoes, the trade creaton effects of NAFTA overwhelms any trade dverson effects. [Table 3 Approxmately Here] Accordng to the results of ths study, non-tarff trade costs are hgh post-nafta, even though NAFTA elmnates tarff and other trade regulatons. In order to determne the effect of non-tarff trade costs on trade creaton and dverson, ths study develops two counterfactual scenaros. In counterfactual scenaro one, ths study elmnates non-tarff trade costs among NAFTA countres n order to determne the change n the trade volume of the three NAFTA member countres (market ntegraton scenaro). In counterfactual scenaro two, ths study ncreases non-tarff trade costs among NAFTA countres n order to determne the change n the trade volume of the three NAFTA countres (market segregaton scenaro). Accordng to Table 4, the market ntegraton scenaro sgnfcantly ncreases the sweet potato mports of NAFTA countres from ther NAFTA-partner countres whle sharply decreases sweet potato mports of NAFTA countres from non-nafta countres. For example, U.S. sweet potato mports from ther NAFTA-partner countres ncreased by 2,350 mllon pound, reflectng a 38,995% trade creaton effect, whle U.S. sweet potato mports from non- NAFTA countres decreased by 5 mllon pound, reflectng a 526% trade dverson effect. Canadan sweet potato mports from ther NAFTA-partner countres ncreased by 07 mllon pound, reflectng a 622% trade creaton effect, whle Canadan sweet potato mports from non- NAFTA countres decreased by 4 mllon pounds, reflectng a 2,082% trade dverson effect. Mexcan sweet potato mports from ther NAFTA-partner countres ncreased by 0 mllon pound, reflectng a,9% trade creaton effect, whle Mexcan sweet potato mports from non- 4

16 NAFTA countres decreased by 0. mllon pound, reflectng a 200% trade dverson effect. As a result, total sweet potato trade volume among NAFTA countres ncreased by 2,468 mllon pound, reflectng a 2,43% trade creaton effect, whle total sweet potato mports of NAFTA countres from non-nafta countres decreased by 9 mllon pound, reflectng a 86% trade dverson effect. Accordng to the results of the market ntegraton scenaro, the decrease n sweet potato mports of NAFTA countres from non-nafta countres s relatvely small compared to the ncrease n sweet potato mports of NAFTA countres from ther NAFTApartner countres. For example, the decreased volume n U.S. sweet potato mports from non- NAFTA countres s 0.64% of the ncreased volume of U.S. sweet potato mports from NAFTA countres. Decreased volume n Canadan sweet potato mports from non-nafta countres s 3.68% of the ncreased volume of Canadan sweet potato mports from NAFTA countres. Decreased volume n Mexcan sweet potato mports from non-nafta countres s.04% of the ncreased volume of Mexcan sweet potato mports from NAFTA countres. As a result, total decreased volume n the NAFTA countres mports represents 0.77% of the total ncreased volume of the three NAFTA countres mports. [Table 4 Approxmately Here] Table 5 shows that the market segregaton scenaro decreases the sweet potato mports of NAFTA countres from NAFTA-partner countres, whle t ncreases the sweet potato mports of NAFTA member countres from non-nafta countres. For example, Canadan sweet potato mports from ther NAFTA-partner countres decrease by 7 mllon pounds, whle Canadan sweet potato mports from non-nafta countres ncrease by 220 mllon pounds. U.S. and Mexcan sweet potato mports from NAFTA-partner countres decrease by.7 mllon pound and 0.9 mllon pound, respectvely. At the same tme, U.S. and Mexcan sweet potato mports from non-nafta countres decrease by 0.7 mllon pounds and 2 thousand pounds, respectvely. 5

17 The results of two dfferent counterfactual scenaros does not reect the noton that the reason the trade volume of sweet potatoes n the three NAFTA member countres s relatvely small may be due to hgh non-tarff trade costs. [Table 5 Approxmately Here] Concluson Ths study sought to descrbe the effect of NAFTA on the North Amercan sweet potato market. To accomplsh ths, ths study uses both observable (trade volume and home prces) and unobservable trade data (mport prces and trade costs). In order to obtan unobservable trade data, ths study uses the Dxt and Stgltz s CES utlty and producton functons by whch the market equlbrum prce of each NAFTA member country could be derved and allow for the reasonable estmaton of unobservable trade data for the three NAFTA countres. After obtanng both observable and unobservable trade data, ths study estmates home and mported prces at bench market parameter values n both pre-nafta and post-nafta. Unlke prevous studes (Lambert and Mockoy, 2009; Pangarta, 2000; and Zahnser, et al., 205), ths study confrms that the degree of ntegraton of the North Amercan sweet potato markets s relatvely low, gven that the three North Amercan sweet potato markets are domnated by ther home product and the dfference between home and mported prces n each NAFTA country have ncreased wth NAFTA mplementaton. Ths result suggests that there may be hgh trade costs exstng between borders among the three NAFTA countres. Consstent wth our expectatons, the estmated non-tarff trade costs are much hgher than the tarffs mposed by each NAFTA country. Therefore, even though NAFTA elmnated tarffs and other trade barrers between NAFTA-partner countres, market ntegraton wll be slow unless nontarff trade costs are reduced. Non-tarff trade costs come from a very broad range of economc, socal, and poltcal forces n each NAFTA member country. Although t was not purpose of ths 6

18 study to explan from where non-tarff trade costs occur, future research to dentfy the sources of these non-tarff trade costs and feasble optons for ther reducton would be benefcal to the ntegraton of these markets. Ths study develops two counterfactual scenaros to dentfy the effect of non-tarff trade costs on trade creaton and dverson. The market ntegraton scenaro sgnfcantly ncreases sweet potato mports of NAFTA countres from ther NAFTA-partner countres, whle sharply decreasng sweet potato mports of NAFTA countres from non-nafta countres. The market segregaton scenaro decreases sweet potato mports of NAFTA member countres from ther NAFTA-partner countres whle ncreasng sweet potato mports of NAFTA countres from non- NAFTA countres. As a result, these two dfferent counterfactual scenaros suggest that the reason trade volume of sweet potatoes n the three NAFTA member countres s relatvely small may result from hgh non-tarff trade costs. 7

19 Footnote. Snce there s no offcal report on Canadan sweet potatoes, ths study uses the average prce of sweet potatoes as estmated by Agrculture and Agr-food Canada. 8

20 References Agrculture and Agr-Food Canada. (205). Data on the Internet. Retreved at Balassa, B. 96. The Theory of Economc Integraton. Homewood, Illnos; R.D. Irwn. Barchello, R.R., L. Bvngs, C. Carter, T. Joslng, P. Lndsey, and A. McCalla. The Implcatons of a North Amercan Free Trade Area for Agrculture. Internatonal Agrcultural Trade Research Consortum, Commssoned Paper No. 0, November 99. De Janvry, A. NAFTA and Agrculture: An Early Assessment. Paper presented at the Tr-Natonal Research Symposum, NAFTA and Agrculture: Is the Experment Workng? San Antono, Texas, November -2, 996. Dxt, A., and J. Stgltz Monopolstc Competton and Optmum Product Dversty. Amercan Economc Revew 67(3): Duty Calculator (205). Duty Calculator: Import Duty and Taxes Made Easy. Retreved at Food and Agrculture Organzaton of the Unted Natons (FAO). (205-a). Agrcultural Prce Data. FAO statstc webste (FAOStat). Rome: FAO. Retreved at Food and Agrculture Organzaton of the Unted Natons (FAO). (205-b). Crop Producton Data. FAO statstc webste (FAOStat). Rome: FAO. Retreved at Gould, D.M. (998). Has NAFTA Changed North Amercan Trade? Economc revew of the Federal Reserve Bank of Dallas frst Quarter: Grennes, T., J.H. Estrada, B. Krssoff, J.M. Gardea, J. Sharples, and C. Valdes. An Analyss of a Unted States-Canada-Mexco Free Trade Agreement. Internatonal Agrcultural Trade Research Consortum, Commssoned Paper No. 0, November 99. Jayasnghe, S., and R. Sarker Effects of Regonal Trade Agreements on Trade n Agrfood Products: Evdence from Gravty Modelng Usng Dsaggregated Data. Revew of Agrcultural Economcs, 30(): 6-8. Lambert, D., and S. Mckoy Trade Creaton and Dverson Effects of Preferental Trade Assocatons on Agrcultural and Food Trade. Journal of Agrcultural Economcs, 60():

21 Pangarta, A Preferental Trade Lberalzaton: The Tradtonal Theory and New Developments. Journal of Economc Lterature, 30: Robbns, T The Fruts of Free Trade: How NAFTA Revamped the Amercan Det. Natonal Publc Rado, January 9, 204. Solow, R.M A Contrbuton to the Theory of Economc Growth. The Quarterly Journal of Economcs 70: Tomeck, W.G., and K.L. Robnson Agrcultural Product Prces. Fourth Edton by Cornell Unversty Press. Varan, R.H Mcroeconomc Analyss. Thrd Edton by W.W. Norton & Company, Inc. Wllams, G.W., and T. Grennes (eds.) NAFTA and Agrculture: Wll the Experment Work? Proceedngs of the Internatonal Trade Research Consortum, San Dego, Calforna, December, 2-4, 993. Center for North Amercan Studes, Internatonal Trade Research Consortum, and Texas Agrcultural Market Research Center, College Staton, Texas, 994. U.S. Department of Agrculture, Economc Research Servce. (205). Data on the Internet. Retreved at Fresh%20Market/YRBK204_Secton%202_SandU%20Fresh.xlsx. U.S. Department of Labor, Bureau of Labor Statstcs. (205). Hourly Wages. Data on the Internet. Retreved at U.S. Internatonal Trade Commsson. (205). Import tarff data. Retreved at Zahnser, S., S. Angadvand, T. Hertz, L. Kuberka, and A. Santos NAFTA at 20: North Amerca s Free-Trade Area and Its Impact on Agrculture. USDA. Report Number W- 5-0, February 205. Zahnser, S., and Z. Crago. NAFTA at 5: Buldng on Free Trade. Economc Research Servce, USDA. Report Number WRS-09-03, March

22 Table. Bench Mark Parameter Values Pre NAFTA Post NAFTA s t τ US CANADA MEXICO ROW US CANADA MEXICO ROW US CANADA MEXICO ROW US CANADA MEXICO ROW US CANADA MEXICO ROW α r σ 3.00 θ

23 Table 2. Home and Imported Prces n Pre NAFTA and Post NAFTA Pre NAFTA: : $/Ib P us P ca P mx P row US Canada Mexco N/A ROW Post NAFTA: : $/Ib P us P ca P mx P row US Canada Mexco N/A ROW Percentage Change between Pre NAFTA and Post NAFTA P us P ca P mx P row US Canada Mexco 7.5 N/A ROW

24 Table 3. NAFTA's Trade Creaton and Dverson Effects at Bench Mark s =NAFTA Countres Trade Creaton Effect Pre NAFTA Post NAFTA ) Δq Δ% US CANADA MEXICO TOTAL =ROW Trade Dverson Effect Pre NAFTA Post NAFTA 2) Δq Δq % US CANADA MEXICO TOTAL ( Δq Δq ) 00% US CANADA 6.99 MEXICO 5.0 TOTAL 2.90 ) Δq 0 means "trade creaton effect" Δq < 0 means "no trade creaton effect" 2) Δq < 0 means "trade dverson effect" Δq 0 means "no trade dverson effect" 23

25 Table 4. NAFTA's Trade Creaton and Dverson Effects at Market Integraton Scenaro s =NAFTA Countres Trade Creaton Effect Pre NAFTA Post NAFTA ) Δq Δ% US CANADA MEXICO TOTAL =ROW Trade Dverson Effect Pre NAFTA Post NAFTA 2) Δq Δq % US CANADA MEXICO TOTAL ( Δq Δq ) 00 US 0.64 CANADA 3.68 MEXICO.04 TOTAL 0.77 ) Δq 0 means "trade creaton effect" Δq < 0 means "no trade creaton effect" 2) Δq < 0 means "trade dverson effect" Δq 0 means "no trade dverson effect" 24

26 Table 5. NAFTA's Trade Creaton and Dverson Effects at Market Segregaton Scenaro s =NAFTA Countres Trade Creaton Effect Pre NAFTA Post NAFTA ) Δq Δ% US CANADA MEXICO TOTAL =ROW Trade Dverson Effect Pre NAFTA Post NAFTA 2) Δq Δq % US CANADA MEXICO TOTAL ( Δq Δq ) 00 US 44.4 CANADA MEXICO.3 TOTAL 5.60 ) Δq 0 means "trade creaton effect" Δq < 0 means "no trade creaton effect" 2) Δq < 0 means "trade dverson effect" Δq 0 means "no trade dverson effect" 25