CAN IMMIGRANTS HURT TRADE? Tomáš Konečný CERGE-EI Charles University Center for Economic Research and Graduate Education Academy of Sciences of the Czech Republic Economics Institute WORKING PAPER SERIES

Please download to get full document.

View again

of 30
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Spiritual/ Inspirational

Publish on:

Views: 22 | Pages: 30

Extension: PDF | Download: 0

CAN IMMIGRANTS HURT TRADE? Tomáš Konečný CERGE-EI Charles University Center for Economic Research and Graduate Education Academy of Sciences of the Czech Republic Economics Institute WORKING PAPER SERIES (ISSN ) Electronic Version 329 Working Paper Series 329 (ISSN ) Can Immigrants Hurt Trade? Tomáš Konečný CERGE-EI Prague, May 2007 ISBN (Univerzita Karlova. Centrum pro ekonomický výzkum a doktorské studium) ISBN (Národohospodářský ústav AV ČR, v.v.i.) Can Immigrants Hurt Trade? Tomáš Konečný * CERGE-EI May, 2007 Abstract This paper estimates the impact of immigrant network spillovers on international trade. Contrary to previous studies focusing mostly on the trade enhancing role of immigrant networks, the present framework allows for potential trade diverting effects. A simple matching model that incorporates both trade creation and diversion channels furthermore points at the importance of relative as opposed to absolute measures of immigrant networks. Using a new dataset of 19 OECD countries, I find that while immigrant networks indeed facilitate exports from host to source country, they simultaneously hurt trade with the host country s other trading partners. In addition, I find that the impact of information-related trade barriers might be negatively related to the economic size of the trading partner. In particular, the larger the trading partner, the smaller the trade benefits of lower information costs due to a shared common language. Abstrakt V článku odhaduji vliv vazeb imigrantů na domovskou zemi v mezinárodním obchodě. Narozdíl od předchozích studií, které zkoumaly pouze jejich potenciální přínos, se zaměřuji i na možné náklady ve formě odklonu obchodních toků. Součástí práce je jednoduchý model zachycující jak přínosy, tak i náklady imigrantských vazeb. Model dále poukazuje na důležitost relativních ukazatelů množství imigrantů, které byly v dosavadní literatuře opomíjeny. V empirické části používám nová data zahrnující 19 hostitelských zemí v rámci sdružení OECD. Výsledky naznačují, že ačkoliv imigranti rozvíjejí obchod s domovskou zemí, mají zároveň negativní vliv na obchod hostitelské země s ostatními obchodními partnery. V práci také ukazuji, že dopad informačních bariér na mezinárodní obchod je inverzní funkcí velikosti obchodního partnera. Keywords: international trade, immigration, informal trade barriers JEL Classification: F22, O24 I would like to thank CERGE-EI and the Czech Academy of Sciences for generous financial support during my stay at the University of California, San Diego. Special acknowledgment goes to James Rauch, UCSD, who helped substantially to formulate the ideas in the paper. I would also like to thank Libor Dušek and Štěpán Jurajda for their invaluable comments. During the course of my work on this paper, I benefited from the feedback provided by Julian Betts, Peter Katuščák, Andreas Ortmann and Krešimir Žigić. The usual disclaimer applies. * CERGE-EI, Politických vězňů 7, Prague, , Czech Republic. CERGE-EI is a joint workplace of the Center for Economic Research and Graduate Education, Charles University, and the Economics Institute of the Academy of Sciences of the Czech Republic 1 1 Introduction Informal trade barriers have become one of the central points in the debate launched by McCallum s mystery of the missing trade (McCallum, 1995), i.e., the nding that nations tend to trade too much intranationally and too little internationally. Particular attention has been directed towards the lack of information on available trading opportunities and imperfect contract enforcement. In the former case, the insu cient information about foreign partners seems to be pronounced especially in more di erentiated industries where product characteristics vary along multiple dimensions and price happens to be only one of several decision criteria. The resulting higher search costs can then make otherwise e cient crossborder matches unpro table (Rauch and Trindade, 2003; Casella and Rauch, 2003). In the second case, given the existence of insu cient contract enforcement institutions when trade parties originate from di erent jurisdictions, the potential renegation from the contract and the following loss accrued by the a icted party decrease the incentives to engage in trade and, again, might prevent otherwise successful international matches (Greif, 1994). Some social networks seem to be well equipped to deal with both kinds of informal trade barriers. These networks, often de ned by common ethnicity or religion, can provide useful information and trade contacts to their members and/or dispose of some sort of collective punishment mechanism that could substitute for inadequate enforcement institutions. In particular, numerous studies on informal barriers examine the impact on the trade of immigrant networks (e.g., Head and Ries, 1998; Gould, 1993; Girma and Yu, 2002). The results of these studies consistently support the notion that immigrant links indeed facilitate bilateral trade between the host and source countries. 2 However, the combination of pervasive informal trade barriers and country-speci c knowledge possessed by immigrants can also lead to the reduction of trade. Consider a German tractor producer who wishes to export its engines to either Vietnam or Thailand. Finding a reliable trade partner with the capacity to provide local distribution, marketing or maintenance services can be costly both in terms of money and time. Other things equal, if the informal trade barriers are uniform across both countries and trade is still pro table, the producer will be indi erent as to where to export. If, on the other hand, the producer is of Vietnamese ancestry, or perhaps employs Vietnamese o cers in its trade department, the contacts and knowledge of local conditions might bias the export choice in favor of Vietnam. 1 While from the perspective of Germany the total exports do not change (or they even increase if immigrants are more e cient in nding suitable matches), its bilateral trade with Thailand becomes lower than it would have been in the absence of immigrant networks. In this case, trade diversion from Thailand occurs due to a lost fraction of transactions that would have been realized by otherwise indi erent exporters. Gere s study on o shoring in the apparel industry provides yet another illustration (Gere, 1999). The author mentions the case of Taiwanese rms channeling large portions of their o shore investment into Malaysia and Thailand, despite markedly lower wages in other parts of the region. A large part of both economies is, however, controlled by ethnic Chinese who maintain extensive social networks. Gere argues that it is these networks that seem to shape many investment decisions. 2 In his example, the trade diversion would 1 Heerander and Saavedra (2006) cite Peng s (1998) survey on the characteristics of trade intermediaries located in the U.S. According to this survey, 40 percent of U.S. intermediaries o cers or managers are foreign-born. 2 Rauch and Trindade (2002) nd that for trade between Southeast Asian countries with high population shares of ethnic Chinese, the smallest average portion of trade in di erentiated products attributable to 3 take the form of unrealized o shoring projects in countries such as Bangladesh or Sri Lanka, i.e., in destinations with very low wages but lacking links to Chinese networks. Since the previous studies did not address the potential trade-diverting mechanism driven by immigrant links, the present paper aims to ll the existing gap in the literature. It develops a simple framework for the study of network e ects on trade and derives theorybased expressions for both trade creation and trade diversion by networks. Furthermore, the paper evaluates the empirical relevance of the trade-diversion channel using a new dataset on foreign-born populations located in 20 OECD member countries. The following section reviews the spare theoretical literature on the trade-diversion e ects of network ties. Section 3 covers the existing empirical research on the role of immigrant links in international trade and presents the model, while Section 4 discusses the data employed. The subsequent section covers econometric issues, results and sensitivity analysis. Section 6 concludes. 2 Theoretical literature Several theoretical studies formalize the trade consequences of various aspects of asymmetric information. Casella and Rauch (2003) focus on the relationship between network ties and international prices. Rauch and Trindade (2003) study the impact of decreasing information costs on trade volumes, prices and welfare. Rauch and Watson (2003) analyze the mode of match formation in an environment with positive search costs and uncertainty relating to the trade partner s type. 3 Of these studies, Casella and Rauch (2003) is the only one that ethnic Chinese networks reaches nearly 60%. 3 A rm can either engage in a small joint project in order to learn more about its trade partner s quality, or it can place a large scale order right away and risk a higher probability of failure, or it can reject the matched partner and resume searching. 4 mentions potential trade-diversion e ects driven by network ties. The basic setup employed by Casella and Rauch (2003) is a 2 country x 1 good x 2 factor model in which the match quality represents the second factor of production besides the internationally immobile labor. The authors assume producers consisting of di erent types, each type summing to mass 1 located on a unit circle. Each producer can form a joint venture of the best possible quality with either a domestic partner or a foreign party. The quality of the joint venture is measured by the distance between any two producers and it directly impacts gains from a successful match. A crucial ingredient in the model is network ties. While producers in the Home (laborscarce) country possess complete information regarding potential matches in the domestic market, they do not know anything about their foreign counterparts and hence have to search randomly. Only a subgroup of domestic producers tied in a network bene t from full information on the available foreign matches. The model thus de facto introduces a matching technology that reduces existing frictions, yet this technology is reserved for a subgroup of producers only. The equilibrium solution depends on the di erence between the countries labor endowment ratios. 4 In the case that the labor endowments happen to be su ciently close, matches between domestic and foreign producers do not occur. The implied wage di erential is simply too small to outweigh the uncertainty of matching with a foreign partner. As the labor endowment ratio and hence relative wages fall, tied home producers become active and with additional decreases a fraction of untied producers engage in further trade. This breakdown due to information ties leads to a partial insulation of one country from another, meaning 4 Remember that the number of producers located in each country is set to one. 5 that the country s wages tend to be more sensitive to changes in the domestic rather than foreign labor supply and wage convergence becomes weaker. The introduction of ties also raises income and hence aggregate welfare relative to the benchmark case with pervasive uncertainty. The positive welfare contribution of network ties in a two-country model could, however, become reduced or even reversed once a third country is added. Casella and Rauch (2003) argue that in the case when the strongest ties do not persist between the countries with the largest wage di erentials in the absence of ties, trade creation by network ties will be mitigated by the previously nonexistent trade diversion channel. The strength of the trade diversion will depend on the size of the wage di erentials between the two foreign countries. Compared to the baseline two-country solution, the closer the wages of a foreign country with network ties in comparison to the foreign country with the lowest relative wage, the smaller the trade diversion e ect and the larger the trade creation e ect (arising from transfer of labor to a cheaper country) would be. Large wage di erentials between foreign countries will, on the contrary, magnify the trade diversion e ect. Finally, if the wage di erentials become wide enough, ties-induced trade diversion will be removed as producers will avoid the foreign country with network ties altogether and all international matches will be formed with the now substantially cheaper trading partner. Unfortunately, Casella and Rauch (2003) do not present analytical expressions for the trade diversion channel. While for the three-country case such expressions could be derived manually, it remains unclear as to how the model would extend to a multicountry setting. In particular, the extension to three countries would not provide any information on the proper aggregation of wage di erentials and ties to other trading partners (i.e. of the third 6 country). In order to avoid the complexities inherent in Casella and Rauch (2003), the simple multicountry framework developed in the next section allows for the trade diversion channel without any reference to relative wages. 3 Empirical studies and model 3.1 Previous empirical research on immigrant ties The scarcity of theoretical studies on trade-diversion e ects of immigrant networks is further supplemented by the absence of any empirical work in this area. The earlier pioneering studies focused on whether immigrant ties in uence trade patterns at all and if so, their sole focus has been on trade creation. Nonetheless, the existing output still provides a useful consistency check, at least for the trade creation estimates discussed in Section Gould (1993) analyzes migration in ows in the U.S. using panel data from 1970 to 1986 and nds a signi cantly positive impact of immigrants on bilateral trade with their source countries. The implied long-run elasticities suggest a 10-percent increase in immigrant stock to increase U.S. exports to a concerned source country by 4.7 percent and U.S. imports by 8.3 percent. Gould (1993) explains the di erence in estimates by the presence of both a taste e ect and immigrant network e ect in the imports equation, whereas for exports the taste e ect should be absent. A similar exercise for the Canadian economy has been produced by Head and Ries (1998). The authors use two di erent measures of immigrant ties, namely the cumulative sum of immigrant in ows after 1970 and the imputed immigrant populations using census data, and report a 10-percent increase in the immigrant stock to increase Canadian bilateral exports by percent and imports by percent. Rauch and Trindade (2003) s study on ethnic Chinese networks delivers both an economically and 7 statistically signi cant role of Chinese minorities in bilateral trade promotion. Moreover, the authors conclude that while for Southeast Asian countries with larger population shares trade facilitation occurs mainly through the alleviation of information asymmetry (related mostly to trade partner characteristics), for other countries where ethnic Chinese form a rather small fraction of the population, the substitution for weak legal institutions becomes a key factor. Interesting research on the relevance of immigrant ties for the United Kingdom s trade has been conducted by Girma and Yu (2002). Girma and Yu (2002) work with the terms universal and non-universal immigrant ties, the former term being a label for personal contacts broadly consistent with Casella and Rauch (2003) s matching frictions argument, the latter serving as a proxy for the general knowledge of source country s social institutions and market speci cs. The operating hypothesis states that while universal links should matter equally for all countries, non-universal links should matter more in source countries with a more dissimilar institutional (legal, social, cultural) setup. Exploiting the di erent historical experience of Commonwealth and non-commonwealth countries, the paper nds robust evidence that the impact of immigrant ties from Commonwealth countries on the UK s exports is insigni cant. The results thus favor the non-universal against the universal dimension of immigrant ties. More recent studies include Combes et al. (2005) and Herander and Saavedra (2005). The careful econometric study by Combes et al. (2005) analyzes the impact of internal migration in France on inter-regional trade. Migrants are found to double the volume of inter-regional trade ows in comparison to the situation without migrant networks. The estimated impact of business ties proves to be even stronger, magnifying trade ows by ve times in some speci cations. Furthermore, the authors quantify the share of informal 8 trade barriers that can be explained by migrant and business networks. According to their results, the negative impact of transport costs falls by as much as 60 percent and the traderestricting border e ect drops by more than three times. Finally, Herander and Saavedra (2005) explore the spatial dimension of immigrant networks. Focusing on trade-creation e ects of immigrant networks operating within and between the U.S. states, the results show a consistently stronger impact on U.S. state export volumes to a source country for local as compared to out-of-state populations. In particular, their results qualitatively conform to previous estimates in that a 10-percent increase in the local state immigration should on average increase the state s exports by 1.6 percent. The estimated impact of the out-of-state population, i.e. of the immigrant network geographic spillovers, then raises the states export volumes by 0.7 percent only. The research by Herander and Saavedra (2005) is to my knowledge the only one to consider trade spillovers by immigrant networks. However, the present study aims to estimate a rather di erent dimension of network spillovers. While Herander and Saavedra (2005) deal with trade facilitating spillovers generated by immigrant networks of the same nationality located in di erent U.S. states, I instead focus on the relevance of potential trade diverting spillovers by immigrant networks of another ethnicity within a given country. The following subsection presents the estimation framework. 3.2 The empirical model For the empirical evaluation of the trade diversion hypothesis I use a simple gravity framework that explicitly allows for matching in trade. The gravity relationship proportionally linking trade ows to the incomes of trading economies can be derived from a wide range 9 of international trade models. 5 The present model, however, does not build on any of them. Instead, it shows that the gravity might be consistent even with a very simple world economy with matching and no di erences in productivity, endowments, or preferences. Index a source country from which the immigrants come i = 1 ; ::;I and their host country j = 1; ::; J. Assuming no di erences in productivity, a country j s share in world output equals its population share in total population, i.e. N j N = GDP j GDP ; (1) where N j stands for the total population in country j and N stands for the total world population. Further assume that the total N j population in economy j consists of natives and immigrants from other countries i, where native agents are poorly informed with regard to available matches and immigrants from country i (denoted by m ij ) are fully informed. That is, assume native agents j could trade only with agents who were not matched by fully connected immigrants. Similarly, assume the immigrants possess full information on potential matches in their source country i, yet no information at all on matches relating to any other trading partners, so th
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks