World Development Vol. xx, No. x, pp. xxxxxx, 2011 2011 Elsevier Ltd. All rights reserved.

. 0305-750X/$ - see front matter www.elsevier.com/locate/worlddev

doi:10.1016/j.worlddev.2011.04.036

Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows

DENIS MEDVEDEV * The World Bank, Washington, DC, USA
Summary. This paper investigates the eects of preferential trade agreements (PTAs) on net FDI inows of member countries using a comprehensive database of PTAs in a panel setting. PTA membership is associated with a positive change in net FDI inows and FDI gains increase with the market size of PTA partners and their proximity to the host country. The estimated relationship is driven by the developing countries in the sample and agreements signed in the late 1990searly 2000s, a period when the majority of deep integration PTAs have been advanced. 2011 Elsevier Ltd. All rights reserved. Key words preferential trade agreements, foreign direct investment (FDI), world

1. INTRODUCTION The prospects of increased foreign direct investment (FDI) are generally recognized as some of the more important reasons for entering preferential trade agreements (PTAs). For example, many observers viewed the expected post-NAFTA increase in cross-border investment inows as the main motivation of the agreement for Mexico (see Lederman, Maloney, & Serven, 2005, chap. 4). However, while the trade outcomes of PTAs have received extensive attention in the theoretical and empirical literature, the link between PTAs and FDI is less well understood. This paper aims to add to the emerging literature on the impacts of PTAs on FDI by combining an extensive sample of PTAs with a set of FDI determinants in a panel data framework. The papers contribution is the identication of an empirical relationship between preferential liberalization and FDI inows through the size of the extended common market created by a PTA and the average distance to the preferential trading partners. Although the increase in FDI inows through joining a larger common market has already been identied in the literature, the results have so far been limited to a fairly small sample of countries and PTAs. This paper aims to broaden this result to nearly all countries and agreements in existence, and also establishes an important role of distance to ones preferential trading partners as a determinant of FDI. Moreover, the paper identies dierential eects for developing and developed PTA members, as well as for the 1980searly 1990s and late 1990searly 2000s (the latter being the period when most deep integration PTAs have been signed). The paper is structured as follows. It begins with Section 2, which provides a brief review of the existing theoretical arguments and the empirical evidence on the relationship between PTAs and FDI. Section 3 estimates an empirical link between preferential liberalization and net FDI inows by country in a large panel of high income and developing nations over the last two decades. It also conrms the robustness of this relationship to controlling for additional determinants of FDI and decomposes the FDI eect by time period and income level of the host country. Section 4 provides the concluding remarks.
1

2. THEORETICAL LINKS AND EMPIRICAL EVIDENCE The various ways in which preferential liberalization may affect FDI could be grouped into four distinct categories: the effects of investment and other non-trade provisions (which have become more common in recent deep integration PTAs), the eects of changes in trade ows, creation of an extended market, and long-term growth eects. Table 1 summarizes these eects along with their transmission channels and a selection of papers identifying the impacts (although in most papers, the specic case of preferential liberalization is not considered). With regard to the impact of non-trade provisions the most relevant paper is Adams, Dee, Gali, and McGuire (2003), who studied the impact of PTAs on stocks of outward FDI for a panel of high income and developing countries between 1988 and 1997. The authors found that thedeep integration provisions of PTAs (investment, liberalization of trade in services, setting and harmonization of standards, competition policy, customs cooperation, dispute settlement, and IPR protection) were an important driver of FDI in ve of the nine PTAs examined. Moreover, the authors argue that the deep integration provisions of these agreements have had a greater impact on FDI than the trade provisions. Other studies have also linked aspects of deep integration to increased FDI, although not necessarily in the context of PTAs. For example, Lee and Manseld (1996) and Maskus (1998) identify a positive relationship between US-sourced FDI and IPR, but the results are tentative and, for the latter paper, limited to developing countries. 1 To the extent that PTAs can lower political risk by locking-in reforms and improving the investment climate, the results of Kolstad and Tondel (2002), who nd a positive association between low political risk and larger FDI per capita in a cross-sectional study of 120 developing and industrialized countries, are also relevant.
* I am grateful to Robert Blecker, Jerey Lewis, Kara Reynolds, Hans Timmer, and two anonymous referees for valuable comments and suggestions and to Michiel Paris for excellent research assistance. All remaining errors are my own. Final revision accepted: April 3, 2011.

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

WORLD DEVELOPMENT Table 1. Potential links between preferential liberalization and FDI Impact area Services liberalization, investment, and other nontrade provisions Transition channel PTA provisions related to services and investment can directly stimulate FDI Increased IPR protection in some PTAs may stimulate FDI Evidence Adams et al. (2003) nd a signicant impact of non-trade provisions on investment ows Lee and Manseld (1996) and Maskus (1998) nd a positive relationship between FDI and IPR, but Manseld (1993), Maskus and Eby-Konan (1994), and Primo Braga and Fink (1998) cannot establish any link Kolstad and Tondel (2002) nd a positive association between low political risk and larger FDI per capita High tari rates on imported goods induced FDI inows into the United Kingdom (Dunning, 1958), Canada (Horst, 1972), and Australia (Brash, 1966) Markusen (2002) and Globerman (2002) show a positive association between trade and FDI based on MNE production structure Lim et al. (2001) cites both survey and econometric evidence that conrms market size as the most robust determinant of FDI (see also Kolstad and Tondel, 2002; Chakrabarti, 2001) World Bank (2001), UNCTAD (1998), and others show that, controlling for other factors, FDI ows are positively associated with economic growth, while Berthelon (2004) nds a positive correlation between regional integration and growth

Merchandise trade liberalization

PTAs may lower political risk by improving the investment climate and therefore stimulate FDI Trade and investment are substitutes (tarijumping FDI) Trade and investment are complements

Extended market

PTAs could stimulate FDI by creating a larger host market

Growth

PTAs may accelerate growth, which in turn attracts FDI

Moving to the relationship between investment and trade ows, the early literature on the subject conceptualized the two as dierent modes of reaching foreign markets (Blomstrom & Kokko, 1997). Chapter 3 of World Bank (2002) cites studies conrming that high tari rates on imported goods induced FDI inows into the United Kingdom (Dunning, 1958), Canada (Horst, 1972), and Australia (Brash, 1966). If FDI is indeed tari-jumping, these studies would suggest that preferential liberalization removes incentives to invest. On the other hand, more recent literature (e.g., Caves, 1996; Markusen, 2002; Globerman, 2002) has emphasized the complementarities between trade and investment owing to the growing importance of MNE production networks and intra-industry and intra-rm trade. This view is supported by a review of cross-country regressions on the determinants of FDI by Chakrabarti (2001), who notes that after market size, openness to trade has been the most reliable indicator of the attractiveness of a location for FDI (see also Kolstad & Tondel, 2002). Thus, the net change in FDI ows as a result of preferential trade liberalization could be either positive or negative, depending on the type of FDI and whether PTAs are on average trade-creating or trade-diverting. While the direction of the relationship between trade and FDI may be unclear, the connection between host market size and FDI is well established. In a literature review, Lim et al. (2001) cites both survey and econometric evidence that conrms market size as the most robust determinant of FDI. In fact, virtually all studies of FDI nd a highly signicant positive eect of the size of the host market on FDI inows (see also Kolstad & Tondel, 2002; Chakrabarti, 2001). To the extent that a PTA creates an extended market through closer integration of PTA partners, this channel suggests a positive relationship between PTAs and FDI. Blomstrom and Kokko (1997) argue that this eect works through increased rm sizea larger market may allow some rms to grow beyond what they would have been able to achieve in segmented national markets, or the competitive pressures of a larger mar-

ketplace may force some rms to expand through mergers and acquisitions of former competitors. 2 In addition to the more static channels above, PTAs may also aect FDI through more dynamic means by generating additional growth. The complementarities between FDI and growth have been well documented in the empirical literature. A number of studies, including World Bank (2001) and UNCTAD (1998), have shown that, controlling for other factors, FDI ows are positively associated with economic growth. Unfortunately, the direction of causation is not clear: while a large literature documents the growth-enhancing knowledge and eciency spillovers from FDI (see, e.g., the review in Lim et al., 2001), other studies, most notably Rodrik (1999), have suggested that FDI tends to be located in more productive and faster-growing economies. Although the link between PTAs and growth is much more tentative, Berthelon (2004) does nd a positive correlation between regional integration and growth in a panel of high income and developing countries between 1960 and 1999 with the caveat that the relationship is signicant only for the larger countries in the sample. Overall, the studies cited above and in Table 1 suggest that there are multiple channels through which PTAs could aect FDI. However, they do not answer the question of what could be the average or expected eect of preferential liberalization on FDI as this would depend on which individual channels are present for each country and agreement as well as on the interactions between these channels. Clearly, the sign and strength of the nal impact as well as its persistence in the presence of other determinants of FDI is an empirical question. In this regard, the empirical work on the relationship between PTAs and FDI can be separated into two main categories: case studies/time series analysis of large, wellknown agreements (such as the European Union, NAFTA, or MERCOSUR) and, more recently, analysis using panel data. A selection of most relevant studies and their main ndings are summarized in Table 2.

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS Table 2. Empirical evidence on the impact of preferential liberalization on FDI Methodology Case studies/time series analysis Paper Lim et al. (2001) Findings FDI-to-GDP more than doubled in the four-year post-PTA period for Portugal and Spain (EU accession), Brazil (MERCOSUR), and Mexico (NAFTA), and increased by 70% for Argentina (MERCOSUR) Similar to Lim et al. (2001) for Spain and Portugal, but no change in FDI inows for Greece following EU accession. FDI into Mexico increased in the rst two years following NAFTA, but leveled o soon afterwards, similar to a stock adjustment experienced by the new EU entrants Sharp increase in intra-EU FDI following the implementation of the Internal Market Program in 1986 Similar to Pain (1997) for German investment into the rest of the European Union Intra-PTA FDI into Canada declined following the CanadaUSA FTA (CUFTA), but extra-PTA FDI increased just enough to oset the decrease. In the United States, net FDI rose as a result of CUFTA, but this was achieved through a large increase in FDI from outside CUFTA. Extra-PTA FDI responded more strongly to the macroeconomic stabilization programs than the early stages of MERCOSUR, but subsequent deeper integration with the establishment of the customs union resulted in signicant increases in the US investment position Following CUFTA and subsequently NAFTA, European FDI to Canada increased much more than FDI from the United States US FDI into Canada was encouraged by CUFTA and NAFTA Positive eect of NAFTA on US-sourced FDI in Mexico only during the rst twothree years of the agreement MERCOSUR led to a signicant increase in FDI, but most of it came from outside the PTA Joining a PTA increases bilateral FDI stocks between members by 27%, while a larger common market aects hosts FDI with an elasticity of 0.1 Six of the nine sample PTAs (including the European Union) are investment-creating, one investment-diverting, and two have no observable impact The expectation of joining a PTA can increase FDI ows by more than onethird, while joining a common market twice as large as the host country can raise FDI ows by 20% or more Signicant positive eect of the beginning-of-period extended market size on end-of-period FDI stocks

Lederman et al. (2005)

Pain (1997) Pain and Lansbury (1997) Blomstrom and Kokko (1997)

Globerman (2002) Buckley et al. (2000) Monge-Naranjo (2002) Chudnovsky and Lopez (2001) Panel analysis Yeyati et al. (2003) Adams et al. (2003)

Lederman et al. (2005)

Jaumotte (2004)

Studies of individual agreements generally nd an increase in FDI following preferential liberalization. This has been documented for Spain and Portugal after EU accession (Lim et al., 2001; Lederman et al., 2005), Brazil and Argentina after MERCOSUR (Blomstrom & Kokko, 1997; Lim et al., 2001; Chudnovsky & Lopez, 2001), Mexico after NAFTA (Lim et al., 2001; Lederman et al., 2005; Monge-Naranjo, 2002), and Canada (Blomstrom & Kokko, 1997; Buckley, Clegg, Forsans, & Reilly, 2000; Globerman, 2002) after CUFTA and NAFTA. Similarly, the implementation of the EU Internal Market Program (IMP) in 1986 led to an increase in intra-EU FDI (Pain, 1997; Pain & Lansbury, 1997). However, the same studies caution that the increase is not automatic (e.g., Lederman et al. (2005) nd no accession impact for Greece) and that in a number of instances the increase was essentially a one-time stock adjustment (e.g., Lederman et al. (2005) and Monge-Naranjo (2002) for Mexico). In some cases, these studies also nd that the FDI responded more to the concurrent policy reforms rather than the preferential liberalization per se. For example, Graham and Wada (2000) show that US FDI to Mexico began to grow rapidly in the late 1980s, much earlier than the implementation of NAFTA, and attribute the increase to Mexicos unilateral policy reforms. For MERCOSUR, Blomstrom and Kokko (1997) show that extra-PTA FDI responded more strongly to the

macroeconomic stabilization programs than the early stages of the Southern Cone project, although the creation of the customs union during the later stages was associated with large FDI gains. Results of cross-country analyses support the positive association between preferential liberalization and FDI identied by the individual agreement studies above. Using gravity-type applications, Levy Yeyati, Stein, and Daude (2003) and Adams et al. (2003) nd positive eects of preferential liberalization on bilateral FDI stocks. Turning to net eects, both Lederman et al. (2005) and Jaumotte (2004) nd a positive effect of extended market size on FDI inows and FDI stocks, respectively. In the case of Lederman et al. (2005), the authors nd an insignicant coecient on the PTA membership dummy but a signicant positive eect of the expectation of joining a PTA and the size of common market; the former increases FDI ows by more than one-third, while joining a common market twice as large as the host country can raise FDI ows by 20% or more. Regarding the contributions of dierent transmission channels identied in Table 1, Adams et al. (2003) and Blomstrom and Kokko (1997) provide some evidence to support the investment provisions/non-trade channel. In many cases the additional investment originates from outside the PTA, giving support to the trade and extended market size hypotheses (the

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

WORLD DEVELOPMENT

latter is identied explicitly by Yeyati et al. (2003), Lederman et al. (2005), & Jaumotte (2004)). No studies have attempted to focus on the growth channel, and the empirical identication is likely to be very dicult due to the large number of potential growth determinants. A major shortcoming of the existing studies lies in their small sample sizes, both in terms of country coverage and the number of PTAs considered. The latter is a particular handicap, since the coecient on the extended market variable will be signicantly biased if only some PTAs of which a particular country is a member are included. Furthermore, all existing studies focus on large, well-known agreements, and, therefore, address neither the question of the average or expected impact of PTAs on FDI, nor the FDI gains that a new signatory to a PTA may expect to receive. In contrast, this paper explicitly takes into account all PTAs in existence by drawing on an extensive database of both WTO-notied and un-notied PTAs (Medvedev, 2010). 3 Moreover, this paper uses the largest sample of countries (153) of all existing studies, therefore, making an attempt at covering both the universe of countries and PTAs in order to identify whether and how the explosion in the number of PTAs over the last twenty-some years has aected the evolution of FDI ows. 3. MODEL SETUP AND ESTIMATION RESULTS (a) Baseline model This section estimates an empirical relationship between preferential trade liberalization and net FDI inows using a panel of 153 countries over the 19802004 period. 4 Net FDI inows correspond to the increase in holdings of foreign enterprises inside a country less the decreases in domestic asset holdings by the same enterprises. 5 The choice of the dependent variable is driven by both theory and data considerations. Globerman and Shapiro (2002) point out that the use of FDI ows is preferable to stocks because the calculation of FDI stocks is often not homogeneous across countries and, to the extent that inward and outward FDI have been going on for a long time, recent and relatively large changes in FDI behavior may not be apparent if FDI stock gures are used. The choice of net FDI inows as the dependent variable precludes the estimation of bilateral ows, but these data are quite scarce and usually limited to high income countries as the FDI source; thus, the use of total net FDI inows allows for inclusion of a larger number of countries and also implicitly takes into account South-South FDI ows. The set of country-level control variables is similar to those used in Lederman et al. (2005): to PTA membership, the log of net FDI inows into country i in time period t is a function of that countrys size, openness, rate of growth, and ination. 6 With regard to market size, Chakrabarti (2001) cites over a dozen papers that found a positive eect of market size (dened as absolute GDP, GDP per capita, GNI, or GNI per capita) on FDI and concludes that market size has, by far, been the single most widely accepted . . .signicant determinant of FDI ows. I dene market size as iGDP i;t lnGDP i;t FDI i;t in order to avoid the endogeneity problem of FDI being included in the measurement of GDP. I use the trade-to-GDP ratio as a measure of openness, X calculated as iOPEN i;t GDPi;t M i;t i;t , where X i;t and M i;t are meri;t FDI chandise exports and imports, respectively. While the tradeto-GDP ratio does not necessarily reect the extent of an economys outward orientationamong other things, it could be measuring country size (inversely)data for alternative mea-

sures such as tari levels are more scarce with less time-series variability. Furthermore, according to the ranking of the most robust determinants of FDI by Chakrabarti (2001), the tradeto-GDP ratio places ve positions above tari levels. A potential concern here is the possibility that GDP and openness may be collinear because smaller countries tend to have larger trade-to-GDP ratios. However, the within-sample correlation of these variables is very low: in a regression of openness on GDP (controlling for country and year xed eects) the overall R2 is just under 3%, while the within R2 is around 1%. I calculate growth rates as annual percentage changes in the GNI per capita (GNIGRO) in order to avoid the previously mentioned endogeneity problem with GDP and also to avoid GDP appearing too many times on the right-hand side. The endogeneity issue may still be relevant, however, because the direction of causation is not clearly established: some studies argue that FDI generates growth, while others suggest that FDI tends to be located in faster-growing economies. But this problem is likely to be less of an issue here because the model allows previous growth to aect FDI, but the reverse causality is unlikely unless current levels of FDI aect growth rates in the earlier period. 7 Another reason to be optimistic is that the causal links from FDI to growth are usually envisioned as a longrun phenomenon, and, therefore, are less likely to be problematic in annual data. Finally, ination CPIGRO, used more generally as a measure of macro instability, is calculated as the annual percentage change in the consumer price index. In addition to these control variables, I include two variables to capture potential eects of PTA membership on net FDI inows. First, to capture the potential FDI gains from obtaining access to a larger market by joining a PTA, I  dene a common market size variable PTAGDP i;t  PN ln j GDP i;j;t 8ij, where j ranges over all of the PTA partners of country i in time t for all countries that are members of at least one PTA. In the case of overlapping agreements (i.e., when the same country pair appears in more than one PTA) I add a given partners GDP to the PTAGDP variable in the year in which the earlier of the two (or more) agreements enters into force while for countries that are not members of a single PTA in a given year, PTAGDP is equal to zero. Second, economic and geographic proximity to PTA partners may also be an important determinant of FDI. On the one hand, similar to gravity models of trade, greater distance could be associated with higher transaction costs and therefore lower FDI. On the other hand, to the extent that trade volumes are negatively related to distance and trade and FDI may be substitutes, we could see a positive relationship between distance and net FDI inows. Therefore, while the expected sign of this variable is not clear and likely depends on the purpose of investment (i.e. whether FDI is market-seeking, tari-jumping, etc.), in either case it could be an important determinant of net FDI inows between PTA partners. To capture this eect, I dene the second PTA-related variable,  P  1 DIST ln N N distij , as the log of average great circle disj tance between country i and all of its PTA partners at time t. This variable is equal to zero when a country is not a member of any PTAs, and has a time series dimension since its value is updated every time a country enters into a new preferential agreement. Data on PTA membership by country have been obtained from Medvedev (2010), who provides a comprehensive account of all PTAs in force. 8 Overall, the sample includes 180 preferential agreements (both regional and bilateral), many of which overlap and which encompass the vast majority of sample countries. As pointed out in the literature on prefer-

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS Table 3. Baseline and additional variables estimation results, net FDI inows Variables iGDP iOPEN GNIGRO CPIGRO PTAGDP DIST PSTAB GOVEFF REGQ ROLAW CONCOR WLDGRO WLDFDI RERGRO PTAEXP CU CUGDP Observations Countries 2,778 153 2,778 153 1,803 99 798 146 2,778 153 0.628** (0.252) 0.544*** (0.028) 0.148 (0.131) 0.162 (0.106) 0.932 (1.428) 0.026 (0.049) 2,778 153 (1) 0.139* (0.078) 0.114*** (0.025) 0.762*** (0.154) 0.022*** (0.004) 0.053*** (0.018) 0.176*** (0.058) (2) 0.618*** (0.063) 0.176*** (0.027) 0.446*** (0.147) 0.022*** (0.004) 0.079*** (0.017) 0.244*** (0.057) (3) 0.201** (0.093) 0.100*** (0.025) 0.855*** (0.201) 0.016*** (0.005) 0.042** (0.021) 0.148** (0.067) (4) 0.025 (0.118) 0.083*** (0.016) 0.523** (0.227) 0.039*** (0.010) 0.149*** (0.037) 0.503*** (0.149) 0.205*** (0.054) 0.005 (0.087) 0.505*** (0.065) 0.008 (0.097) 0.118 (0.081) (5) 0.133* (0.078) 0.113*** (0.025) 0.762*** (0.153) 0.022 *** (0.004) 0.057*** (0.018) 0.175*** (0.058) (6) 0.142* (0.078) 0.113*** (0.025) 0.772*** (0.156) 0.022*** (0.004) 0.053*** (0.018) 0.177*** (0.058)

Standard errors in parentheses. The dependent variable is expressed in natural logarithms * p < 0.1. ** p < 0.05. *** p < 0.01.

ential trade, e.g., Baier and Bergstrand (2004) and Magee (2003), PTA membership may be endogenous as the likelihood of entering into a PTA is signicantly aected by variables that also inuence bilateral trade and investment, such as economic size and distance. However, both Baier and Bergstrand (2004) and Magee (2003) have been unable to come up with instruments that, on the one hand, signicantly determine the likelihood of forming a PTA, and, on the other hand, are uncorrelated with the error term in the gravity equation (i.e., do not aect bilateral trade). More recently, Baier and Bergstrand (2007) have pointed out that the use of panel data, either with xed eects or in rst dierence form, adequately addresses the endogeneity problem as long as the unobservables determining the likelihood of forming a PTA are not time-dependent. This is the assumption adopted by this paper, which leads to the following specication: FDI i;t a ci ht b1 iGDP i;t b2 iOPEN i;t b3 GNIGROi;t b4 CPIGROi;t b5 PTAGDP i;t b6 DIST i;t i;t 1

In order to simultaneously account for heteroscedasticity across panels (BreuschPagan v2 239:38; p 0:00) and serial correlation within panels (Wooldridge F 103:032; p 0:00), I use a three-step feasible generalized least squares (FGLS) estimator. The advantage of this approach is that it allows estimation in the presence of AR(1) autocorrelation within panels and cross-sectional correlation and/or heteroscedasticity across panels. 9 The results of estimating equation (1) are shown in column (1) of Table 3. With the exception of GDP, all of the control variable coecients are signicant at the 1% level and carry the expected sign. Openness and growth have a positive eect on net FDI inows, while the impact of ination is negative. The GDP coecient carries the expected sign and is signicant at the 8% level, but the implied elasticity of net FDI inows with respect to market size (GDP) is substantially below estimates found in other studies. 10 For example, Kolstad and Tondel (2002) obtain an elasticity estimate of slightly higher than one using gross FDI per capita as a dependent variable, while Coughlin and Segev (2000) estimate a similar elasticity with gross FDI data for China. 11

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

WORLD DEVELOPMENT

Both of the PTA-related variables are signicant at the 1% level and show that net FDI inows respond positively to the size of the PTA common market and negatively to the distance to preferential trading partners. The PTAGDP coecient suggests that a 1% increase in the size of a countrys extended market tends to expand net FDI inows by an average of 0.05%. This elasticity is below the 0.1 value estimated by Lederman et al. (2005), but the order of magnitude is similar. To quantify this, consider the example of Costa Rica. By 2004, the country was a member of several PTAs, including the Central American Common Market, Association of Caribbean States, and bilateral agreements with Argentina, Chile, Dominican Republic, Mexico, Panama, Venezuela, and Canada. If Costa Rica were to join DR-CAFTA (which includes the United States) in the same year, the model would predict an 11.5% increase in its net FDI inows. 12 On the other hand, were it instead to join a PTA with Ecuador, Costa Rica could only expect its net FDI inows to rise by 0.08%. The negative estimated elasticity of net FDI inows with respect to average distance to PTA partners suggests that increasing this distance by 1% lowers net FDI inows by 0.18%. To continue with the example of Costa Rica, accounting for the fact that joining DR-CAFTA would increase the countrys average distance to PTA partners lowers the expected FDI benet from 11.5% to 11.1%. On the other hand, if the United States were as far from Costa Rica as, say, Germany, the expected increase in net FDI inows would be limited to 9.6%. (b) Additional variables In this section, I test the robustness of the estimated PTA FDI relationship to the inclusion of some additional variables, such as global growth and FDI, real exchange rate, institutional quality, the expectation of joining a PTA, and membership in a currency union, using Eq. (1) as the starting point. I begin by incorporating global eects into the model similar to the approach of Lederman et al. (2005), who showed that country-level FDI inows respond positively to increases in global FDIreecting aggregate FDI trends and globalization eectsand negatively to increases in global growth because faster growth in the rest of the world makes the host country a less appealing FDI recipient. Because of the endogeneity of global FDI, which is determined by a combination of left-hand side values for all sample countries, I measure this variable as WLDFDI i;t P ln i FDI i;t FDI i;t .Similarly, global growth is calculated  P GDP i;t GDP i;t i as WLDGROi;t 100 P GDP GDP 1 .
i i;t1 i;t1

The results of estimating Eq. (1) with global FDI and growth variables are shown in column (2) of Table 3. In this case, I have to estimate Eq. (1) without time xed eects to ensure that the global variables are identied: because both global FDI and growth have very low within-year variation, they are highly collinear with time eects. The estimated coecients on the global variablesworld growth and world FDIare consistent with expectations and show that faster growth in the rest of the world (controlling for home country growth) makes a particular nation a less attractive location for FDI, while rising total world FDI tends to increase net FDI inows for an average host country. Although the estimated coecients on other model variables (except ination) are substantially dierent from the baseline estimates in column (1), all of them (including the PTArelated variables) retain their expected signs and statistical signicance.

Next is the exchange rate, which many studies cite as an important determinant of FDI (see, e.g., Chakrabarti, 2001). The most basic theories of the relationship between exchange rate movements and FDI argue that a real depreciation of home currency reduces production costs for foreign investors and therefore attracts FDI inows (see, for instance, Cushman, 1985). In addition, a real depreciation lowers foreigners costs of acquiring domestic assets, which should also stimulate FDI. In other theoretical models, the link between exchange rate movements and FDI is assumed to work through capital market imperfections, since the stream of income in a weak currency is subject to greater risk and is, therefore, capitalized at a higher rate. For instance, Froot and Stein (1991) suggest that under information asymmetries, monitoring costs cause external nancing to be more expensive than internal nancing. In this situation, depreciation of the domestic currency not only increases the relative wealth of foreigners, but also increases the relative rate of return for foreign rms that can invest in domestic assets without incurring the monitoring penalty, therefore encouraging additional FDI. Blonigen (1997) also proposes a positive link between exchange rate depreciation and FDI, although his reasoning diers from that of the previous authors: even if domestic and foreign rms may have the same opportunities to purchase domestic assets, these assets can generate returns in currencies other than those used for purchase, which tends to favor foreign investors. In order to test the sensitivity of the PTA-FDI relationship to the inclusion of the exchange rate variable, I dene RERGRO as the annual percentage change in the trade-weighted real eective exchange rate (REER) of country i, with a positive number implying an appreciation. The results of estimating Eq. (1) with the exchange rate variable are shown in column (3) of Table 3. Due to narrower data coverage of the REER variable, the number of observations drops to 1,803 and the number of panels (countries) to 99. Still, all control variables retain their signicance and the coecient estimates are not very far from those shown in column (1). The estimated elasticity of net FDI inows with respect to RERGRO is negative, consistent with theoretical expectations, but the coecient estimate is not signicantly dierent from zero. Finally, the PTAFDI relationship remains signicant and its magnitude is similar to before even after controlling for exchange rate eects. So far, the attempted specications have not included institutional quality variables because they substantially limit the sample size and have two other potential diculties according to Blonigen (2005): rst, since the indices are usually computed using survey responses, cross-country comparability is questionable when the composition of respondents varies across countries and second, since institutions are quite static, the amount of information that can be extracted from time-series variation in these variables is likely to be negligible. Furthermore, using country xed eects controls for a number of country-specic characteristics that display a lot of inertia, such as the quality of institutions. Still, to test whether institutions matter for the estimated PTAFDI relationship, I add the following variables from the World Governance Indicators (WGI) dataset to Eq. (1): political stability and absence of violence, government eectiveness, regulatory quality, rule of law, and control of corruption. The results of estimating the model with institutional variables are shown in column (4) of Table 3. The number of observations falls to 798 but this is mainly due to the loss in time periods rather than countries, as the WGI data are only available from 1996 onwards. In the presence of institutional

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS

variablesof which political stability and regulatory quality signicantly and positively aect FDI while the others are not signicantthe GDP coecient loses its signicance and coecients on the PTA common market size and distance to PTA partners nearly triple in magnitude. However, they remain signicant at the 1% level, and the increase in magnitudes (as well as the loss of signicance on the GDP coecient) is due to restricting the sample to later years rather than the inclusion of institutional controls, a point I will return to when discussing PTAs that have been signed since the mid-1990s in Section (c). The fourth experiment, shown in column (5) of Table 3 investigates the potential signicance of the expectation of joining a PTA. I create an expected PTA dummy, PTAEXP, which is equal to one for two years prior to a PTA coming into force and zero before and after those dates. Contrary to the ndings of Lederman et al. (2005), I cannot identify any significant eects of the expectation of joining a PTA, although the rest of the estimated coecients are very similar to the baseline specication. I also attempted alternative denitions of this variablefor example., extending its duration to three years prior to the PTAs entry into force as well as limiting it to one yearbut the expected PTA dummy was never signicantly dierent from zero. There are several reasons why the expected PTA dummy may not be insignicant. First, since many countries have more than one PTA and I do not rank PTAs in terms of their potential eect on FDI, it may be the case that the expectation of some agreements raises FDI but the expectation of others does not. 13 Since the expected PTA variable is only dened for the rst agreement signed by a given country, identifying a robust anticipation eect could be dicult. Furthermore, if FDI responds to the investment provisions of an agreement or complements the increased trade ows, neither of these can increase before the agreement lifts the relevant barriers. Therefore, the PTA anticipation eect is likely to work

through only a subset of channels that link preferential liberalization to FDI and its empirical contribution to the PTA FDI relationship is more dicult to capture. Finally, I check whether the membership in a currency union makes a dierence to the estimated relationship. I add two variables to Eq. (1): a currency union dummy (CUd) and the combined market size of all members of the currency union (CUGDP). The results of estimating this specication are shown in column (6) of Table 3; although the coecient estimates on all variables are very similar to those shown in column (3) of Table 3, neither the currency union dummy nor the market size of the currency union appear to be significant determinants of FDI in the presence of PTA-related variables. 14 (c) Decomposition by income level and time period In this section, I check whether the income level of destination countries or the time period when the PTAs were signed makes a dierence for the estimated PTAFDI relationship. I begin by testing if PTA membership has a larger impact on net FDI inows of a smaller, rather than a larger country by interacting the PTA dummy variable with the market size (iGDP) of the destination country in Eq. (1). The results of estimating this specication are shown in column (1) of Table 4. Neither the PTA dummy nor the new interaction variable are signicantly dierent from zero while the other coecient estimates are close to those shown in column (1) of Table 3. The only exception is the iGDP variable, which becomes insignicantly dierent from zero once its interaction with PTA membership is added to the specication. The results of the second experimentwhich limit the sample to low- and middle-income countriesare shown in column (2) of Table 4. 15 The sample size falls to 2,244 observations, representing an unbalanced panel of 130 coun-

Table 4. Estimation results by income level and time period, net FDI inows Variables iGDP iOPEN GNIGRO CPIGRO PTAGDP DISTA PTAd PTAd*iGDP Observations Countries (1) 0.079 (0.088) 0.112*** (0.025) 0.796*** (0.155) 0.022*** (0.004) 0.048** (0.020) 0.238*** (0.070) 0.574 (0.973) 0.050 (0.038) 2,778 153 (2) 0.273 (0.084) 0.951*** (0.098) 0.795*** (0.158) 0.020*** (0.004) 0.056*** (0.020) 0.200*** (0.066)
***

(3) 0.181 (0.310) 0.015 (0.036) 0.605 (0.786) 3.051** (1.356) 0.005 (0.053) 0.015 (0.154)

(4) 0.274 (0.110) 0.891*** (0.140) 0.765*** (0.189) 0.011** (0.005) 0.016 (0.027) 0.049 (0.088)
**

(5) 0.167 (0.117) 1.354*** (0.139) 0.551*** (0.193) 0.031*** (0.011) 0.112*** (0.040) 0.386** (0.165)

2,244 130

534 23

1,084 107

1,149 126

Standard errors in parentheses. The dependent variable is expressed in natural logarithms. Column (2) limits the sample to low and middle-income countries and column (3) shows only high income countries. Column (4) limits the sample to 198094, and column (5) shows the estimates for 19952004. * p < 0.1. ** p < 0.05. *** p < 0.01.

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

WORLD DEVELOPMENT

tries between 1980 and 2004. The signs and signicance of all estimated coecients remain unchanged from the baseline specication. Moreover, the coecient magnitudes are quite close to the baseline estimates, with openness and GDP being the two exceptions. In contrast, the results for high-income countriesshown in column (3) of Table 4 are very dierent. Only the ination coecient is signicantly dierent from zero, and it is of the wrong sign. Furthermore, neither the size of the extended common market nor the average distance from PTA partners is a signicant determinant of net FDI inows for high income countries. Therefore, these results suggest that the model of net FDI inows specied in Eq. (1)and, most importantly, the estimated link between preferential liberalization and net FDI inowsapplies only to low- and middle-income FDI hosts and is therefore driven exclusively by North-South and/or South-South FDI. The next experiment is to divide the developing country sample (since the estimated relationship only holds for these countries) into separate time periods: 198094 and 19952004. The 1994 breakpoint was chosen for two reasons: rst, the mid-1990s witnessed an explosive growth in the number of PTAs and many of the agreements signed in that time period and after have been of the third wave kind, putting much more emphasis on deep integration than earlier PTAs. Second, this year splits the sample almost evenly in two halves (1,084 observations in this period versus 1,149 observations in the 19952004 time period). Column (4) of Table 4 shows the estimated coecients for the 198094 sub-sample. The coecients on all the control variables are signicant and reasonably close to the estimates shown in column (2) for the full sample, but neither the extended common market nor the average distance to PTA partners is a signicant determinant of FDI during this period. On the other hand, model estimates for the 1995 2004 sub-sampleshown in column (5) of Table 4-provide strong support for the PTAFDI relationship, with both PTA-related variables signicant at the 1% and 2% level, respectively. Therefore, columns (2)(5) of Table 4 suggest that the estimated link between preferential liberalization and FDI is driven by FDI inows into developing countries owing to the eects of PTAs formed between 1995 and 2004. There are several potential reasons for this result. First, as mentioned earlier, PTAs signed over the course of the last decadeparticularly the North-South typehave generally focused on deep integration provisions much more heavily than the earlier agreements. A particularly important link between the new PTAs and FDI is the vast expansion in trade in services, to which FDI is a natural complement. As an example, consider the recent USSingapore FTA, which is unlikely to have a signicant eect on merchandise trade ows between the two partners (especially taking into account Singapores very liberal MFN tari schedule), but is expected to have a much larger impact on the already substantial trade in services between the two economies. Finally, studies such as World Bank (2002) have documented the tremendous expansion in FDI inows into developing countries during the late 1990s and early 2000s, and at least part of this expansion is potentially linked to the increased global integration brought about by the new deep integration PTAs. Despite these arguments, there also several reasons to be cautious when interpreting the above results. First, causality is much more dicult to establish than correlation. Therefore, even though both the number and scope of PTAs were increasing dramatically at the same time as average country

net FDI inows were surging, it is impossible to know for certain that the latter were directly caused by the former. It may even be the case that, as FDI increased globally in the late 1990s, many countries actively pursued PTAs as a strategy to attract additional net FDI inows. Second, the results of Table 4 do not necessarily imply that the deep integration provisions of later PTAs are responsible for the empirical PTAFDI link, since the deep integration provisions are not the only feature dierentiating the later PTAs and some of the earlier ones also contained chapters dealing with the liberalization of investment ows (e.g., EFTA, MERCOSUR, and USIsrael FTA). 4. CONCLUSIONS A large and growing body of literature has been devoted to understanding the trade creation and diversion eects of preferential trade agreements. On the other hand, the investment consequences of preferential liberalization have received relatively little theoretical and empirical attention. This paper adds to the small but growing literature on the links between PTAs and FDI by establishing a positive relationship between preferential trade liberalization and net FDI inows for a large panel of countries and preferential agreements. There are three main messages emerging from the empirical results. First, the FDI benets of preferential liberalization are increasing in the size of PTA partners and their proximity to the host country. Second, this relationship is driven by the developing countries; and third, the link between preferential liberalization and FDI is only found in the late 1990s and early 2000s, a period when most deep integration agreements have been signed but also a time of a global boom in FDI ows. Overall, the results support the hypothesis that PTAs (and deep integration PTAs in particular) are associated with signicant increases in the net FDI inows of their participants. The estimates of the FDI elasticity of a PTA extended common market are roughly similar to those of Lederman et al. (2005) and Yeyati et al. (2003), who obtain elasticities of approximately 0.1. I also identify signicant proximity eects, although my estimates of the FDI-distance elasticity are much lower than those in the bilateral gravity specications of Yeyati et al. (2003) and Adams et al. (2003). Estimates in this paper are particularly noteworthy because I consider a very wide sample of PTAs, many of which are small and/or poorly implemented. The heterogeneity of PTAs is one possible reason why the estimates of the elasticity of FDI with respect to the size of common market and distance to PTA partners are lower, although it is probably not the only reason. There are a number of questions that the analysis here does not address. For example, I do not consider the source of net FDI inows, and therefore cannot distinguish PTAinduced investment creation and diversion eects. I also do not distinguish between types of PTAsfor example, NorthSouth versus SouthSouth, bilateral versus regionalalthough dierent kinds of agreements may have varying investment eects. Finally, I do not dierentiate between PTAs that include explicit investment provisions and those that do not due to the lack of such data for the majority of PTAs in existence. Answers to these questions are likely to shed additional light on the empirical relationship between preferential trade liberalization and FDI and therefore represent promising directions for future research on the subject.

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS

NOTES
1. On the other hand, Maskus (2000) cites Manseld (1993), Maskus and Eby-Konan (1994), and Primo Braga and Fink (1998), as examples of studies which could not establish a link between measures of IPR protection and the distribution of FDI. 2. It should be noted that this version of the extended common market hypothesis relies on a certain geographic and economic proximity between PTA members. 3. The list of PTAs used in the analysis is provided in the Annex (Table 7). 4. The full list of countries is provided in the Appendix (Table 5). 5. This is in contrast to net FDI, which is the dierence between net FDI inows and net FDI outows (the net increase in home country MNE assets abroad). 6. Variable denitions and sources are provided in the Appendix (Table 6). 7. This is consistent with the arguments of Rodrik (1999). The problem may still arise if both series are highly autocorrelated, although an estimation approach that explicitly controls for serial correlation should minimize any adverse eects. 8. This dataset improves upon previous sources by including agreements that have not been notied to the WTO as well as PTAs found in the WTO database. This addition is important since the PTAs that have not been notied to the WTO account for more than 45% of the total number of PTAs in force. 12. This is equivalent to joining a bilateral PTA with the United States, since Costa Rica already has preferential access to all the other countries in DR-CAFTA through its existing PTAs. 13. The same argument applies to attempts to estimate the model with a variable that captures a number of PTAs that a country has. Since the impacts of PTAs are likely to be very heterogenousfor example, an agreement with a large partner is clearly more valuable than an agreement with a small partnerit was not possible to estimate a robust relationship between the number of PTAs and FDI. 14. The same results also hold if just the currency union dummy, or just the market size of currency union members, are included in the specication. 15. I dene high income countries as 24 developed members of the OECD plus Liechtenstein. 9. I do not iterate the third step to convergence because Greene (2000) notes that no asymptotic gains can be expected from iteration since the estimator is ecient at every step and, while the iterated FGLS estimator converges to the maximum likelihood estimator (MLE) in models without serial correlation, the same does not hold when the disturbances are assumed to follow an AR(1) process. 10. The estimated elasticity is close to unity when time xed eects are omitted from the model specication. 11. Other studies, such as Asiedu (2002) and Addison and Heshmati (2003), impose a unitary elasticity by moving GDP to the left-hand side.

REFERENCES
Adams, R., Dee, P., Gali, J., & McGuire, G. (2003). The trade and investment eects of preferential trading arrangementsold and new evidence. Working paper. Productivity Commission, Canberra. Addison, T., & Heshmati, A. (2003). The new global determinants of FDI ows to developing countries. Discussion paper 2003/45. UNU/WIDER (World Institute for Development Economics Research). Asiedu, E. (2002). On the determinants of foreign direct investment to developing countries: Is Africa dierent?. World Development, 30(1), 107119. Baier, S. L., & Bergstrand, J. H. (2004). Economic determinants of free trade agreements. Journal of International Economics, 64(1), 2963. Baier, S. L., & Bergstrand, J. H. (2007). Do free trade agreements actually increase members international trade?. Journal of International Economics, 71(1), 7295. Berthelon, M. (2004). Growth eects of regional integration agreements. Working paper 278, Central Bank of Chile. Blomstrom, M., & Kokko, A. (1997). Regional integration and foreign direct investment. NBER working papers 6019. National Bureau of Economic Research, Inc. Blonigen, B. (1997). Firm-specic assets and the link between exchange rates and foreign direct investment. American Economic Review, 87(3), 447465. Blonigen, B. A. (2005). A review of the empirical literature on FDI determinants. NBER working papers 11299. National Bureau of Economic Research, Inc. Brash, D. T. (1966). American investment in Australian industry. Canberra: Australian National University Press. Buckley, P., Clegg, J., Forsans, N., & Reilly, K. (2000). United States foreign direct investment into Canada: An empirical analysis with emphasis on the free trade hypothesis. Mimeo, University of Leeds. Caves, R. (1996). Multinational enterprise and economic analysis. New York: Cambridge University Press. Chakrabarti, A. (2001). The determinants of foreign direct investment: Sensitivity analyses of cross-country regressions. Kyklos, 54(1), 89114. Chudnovsky, D., & Lopez, A. (2001). La inversion extranjera directa en el Mercosur: Un analisis comparativo. In D. Chudnovsky, & A. Lopez (Eds.), El Boom de Inversion Extranjera Directa en el Mercosur. Buenos Aires: Siglo Veintiuno de Argentina Editores. Coughlin, C., & Segev, E. (2000). Foreign direct investment in China: A spatial econometric study. World Economy, 23(1), 123. Cushman, D. O. (1985). Real exchange rate risk, expectations, and the level of direct investment. Review of Economics and Statistics, 67(2), 297308. Dunning, J. H. (1958). American investment in British manufacturing industry. London: George Allen and Unwin. Froot, K., & Stein, J. (1991). Exchange rates and foreign direct investment: An imperfect capital markets approach. Quarterly Journal of Economics, 106(4), 11911217. Globerman, S. (2002). Trade, FDI and regional economic integration: Cases of North America and Europe. Paper presented at Enhancing Investment Cooperation in Northeast Asia, Honolulu. Globerman, S., & Shapiro, D. (2002). National political infrastructure and foreign direct investment. Working paper 37. Industry Canada Research Publications Program. Graham, E., & Wada, E. (2000). Domestic reform, trade and investment liberalization, nancial crisis, and foreign direct investment into Mexico. World Economy, 23(6), 777797. Greene, W. H. (2000). Econometric analysis (4th ed.). Upper Saddle River, NJ: Prentice Hall.

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

10

WORLD DEVELOPMENT Maskus, K., & Eby-Konan, D. (1994). Trade-related intellectual property rights: Issues and exploratory results. In R. Stern, & A. Deardo (Eds.), Analytical and negotiating issues in the global trading system. Ann Arbor: University of Michigan Press. Medvedev, D. (2010). Preferential trade agreements and their role in world trade. Review of World Economics (Weltwirtschaftliches Archiv), 146(2), 199222. Monge-Naranjo, A. (2002). The impact of NAFTA on foreign direct investment ows in Mexico and the excluded countries. Department of Economics, Northwestern Unweversity. Pain, N. (1997). Continental drift: European integration and the location of UK foreign direct investment. The Manchester School Supplement, 65, 94117. Pain, N., & Lansbury, M. (1997). Regional economic integration and foreign direct investment: The case of German investment in Europe. National Institute Economic Review, 160, 8799. Primo Braga, C., & Fink, C. (1998). The relationship between intellectual property rights and foreign direct investment. Duke Journal of Comparative and International Law, 9, 163188. Rodrik, D. (1999). The new global economy and developing countries: Making openness work. Policy Essay 24, Overseas Development Council, Washington, DC. Rose, A. (2004). Do we really know that the WTO increases trade?. American Economic Review, 94(1), 98114. UNCTAD. (1998). World investment report: Trends and determinants. New York: United Nations Conference on Trade and Development. World Bank. (2001). Global development nance 2001: Coalition building for eective development nance. Washington, DC: World Bank. World Bank. (2002). Global economic prospects 2003: Investing to unlock global opportunities. Washington, DC: World Bank.

Horst, T. (1972). Firm and industry determinants of the decision to invest abroad. Review of Economics and Statistics, 54, 3745. Jaumotte, F. (2004). Foreign direct investment and regional trade agreements: The market size eect revisited. Working paper 04/206. International Monetary Fund. Kolstad, I., & Tondel, L. (2002). Social development and foreign direct investments in developing countries. Report 11. Chr. Michelsen Institute, Bergen, Norway. Lederman, D., Maloney, W., & Serven, L. (2005). Lessons from NAFTA for Latin America and Caribbean countries: A summary of research ndings. Washington, DC: World Bank. Lee, J. Y., & Manseld, E. (1996). Intellectual property protection and US foreign direct investment. Review of Economics and Statistics, 78, 181186. Levy Yeyati, E., Stein, E., & Daude, C. (2003). Regional integration and the location of FDI. Working paper 492. Inter-American Development Bank. Lim, E. G. (2001). Determinants of, and the relation between, foreign direct investment and growth: A summary of the recent literature. Working paper 01/175. International Monetary Fund. Magee, C. (2003). Endogenous preferential trade agreements: An empirical analysis. Contributions to Economic Analysis & Policy, 2(1) (Article No 15). Manseld, E. (1993). Unauthorized use of intellectual property: Eects on investment, technology transfer, and innovation. In M. Mogee, M. Wallerstein, & R. Schoen (Eds.), Global dimensions of Intellectual Property Rights in Science and Technology. Washington, DC: National Academy Press. Markusen, J. R. (2002). Foreign direct investment and trade. In B. Bora (Ed.), Research issues in foreign direct investment. London: Routledge. Maskus, K. (1998). The international regulation of intellectual property. Weltwirtschaftliches Archiv, 134, 186208. Maskus, K. (2000). Intellectual property rights in the global economy. Washington, DC: Institute for International Economics.

APPENDIX A. DATA DETAILS .

Angola Albania Netherlands Antilles Argentina Armenia Australia Austria Azerbaijan Burundi Belgium Benin Burkina Faso Bangladesh Bulgaria Bahamas Belarus Belize Bolivia Brazil Barbados Bhutan Botswana Central African Republic Canada Switzerland Chile China, P.R. C^te dIvoire o Cameroon Congo, Republic Colombia Cape Verde Costa Rica Cyprus Czech Republic Germany Dominica Denmark Dominican Republic Algeria Ecuador Egypt Spain Estonia Ethiopia European Union Finland Fiji France United Kingdom Gabon Georgia Table 5. Included countries Ghana Gambia, The Guinea-Bissau Equatorial Guinea Greece Grenada Guatemala Guyana Hong Kong Honduras Croatia Hungary Indonesia India Ireland Iran Iceland Israel Italy Jamaica Jordan Japan Kazakhstan Kenya Kyrgyz Republic Cambodia St. Kitts and Nevis Korea, Rep. Kuwait Lao Peoples Dem.Rep Lebanon Liberia St. Lucia Sri Lanka Lesotho Lithuania Latvia Morocco Moldova Madagascar Maldives Mexico Macedonia Mali Mongolia Mozambique Mauritania Mauritius Malawi Malaysia Niger Nigeria Nicaragua Netherlands Norway Nepal New Zealand Oman Pakistan Panama Peru Philippines Papua New Guinea Poland Portugal Paraguay Romania Russia Rwanda Sudan Senegal Singapore Solomon Islands Sierra Leone El Salvador Slovak Republic Slovenia Sweden Swaziland Seychelles Syrian Arab Republic Chad Togo Thailand Tonga Trinidad and Tobago Tunisia Turkey Tanzania Uganda Ukraine Uruguay United States St. Vincent & Grenadines Venezuela, Rep. Bol. Vietnam Vanuatu Samoa Yemen, Republic of South Africa Congo, Dem. Rep. Zambia Zimbabwe

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS Table 6. Variable names and sources Variable FDI GDP Exports Imports GNI CPI PTA DIST REER PSTAB GOVEFF REGQ ROLAW CONCOR CU Name Foreign direct investment, net inows (BoP, current US$) Gross Domestic Product (current US$) Merchandise exports (current US$) Merchandise imports (current US$) Gross National Income (current US$) Consumer price index (2000 = 100) PTA membership Great circle distance between largest or capital cities Real eective exchange rate index (2000 = 100) Political stability and absence of violence Government eectiveness Regulatory quality Rule of law Control of corruption Strict currency union Source WDI WDI WDI WDI WDI WDI Medvedev (2010) CEPII WDI WGI WGI WGI WGI WGI Rose (2004)

11

WDI: World Development Indicators, World Bank, http://data.worldbank.org. WGI: Worldwide Governance Indicators, World Bank, http://info.worldbank.org/governance/wgi/index.asp. CEPII: Centre dEtudes Prospectives et dInformations Internationales, http://www.cepii.fr.

Table 7. List of PTAs considered in the analysis Agreement AEC AFTA Aghadir Agreement (Med-Arab FTA) AlbaniaMacedonia Andean CommunityArgentina Andean CommunityBrazil ACC AMU ArgentinaChile ArgentinaCosta Rica ArgentinaCuba ArgentinaEcuador ArmeniaRussian Federation ACS BAFTA Bangkok Agreement BelarusUkraine BelarusUzbekistan BoliviaChile BoliviaCuba BoliviaMexico Bosnia-HerzegovinaMacedonia Bosnia-HerzegovinaYugoslavia BotswanaZimbabwe BrazilCuba BulgariaCroatia BulgariaEstonia BulgariaIsrael BulgariaLatvia BulgariaLithuania BulgariaMacedonia BulgariaTurkey Burkina FasoCuba Burkina FasoKorea Burkina FasoTunisia Burkina FasoIndia CACM CACMDominican Republic CACMVenezuela Year Notied to the WTO 1994 1992 2003 2002 2000 1999 1989 1989 2000 1983 1984 1993 1992 1994 1994 1976 1992 1993 1993 1995 1995 2002 2002 1988 1987 2001 2002 2002 2003 2002 2000 1999 1987 1998 1993 1995 1961 2001 1993 No Yes No No No No No No No No No No No No Yes Yes No No No No No No No No No No Yes Yes Yes Yes Yes Yes No No No No Yes No No Agreement

Table 7 continued Year Notied to the WTO 1988 1960 1997 2002 1997 1973 1995 1999 1993 1993 1999 1983 2002 2002 1994 1995 2004 2004 1996 1999 1998 1993 1994 2004 2004 1985 1988 1995 1994 1998 2001 2002 1997 1998 1998 1995 1987 1997 1998 Yes No Yes Yes Yes Yes No No No Yes Yes Yes Yes No No No No No No Yes No No Yes No No No No No Yes No Yes No No No No No No No Yes

CAN CanadaAustralia (CANATA) CanadaChile CanadaCosta Rica CanadaIsrael CARICOM CARICOMColombia CARICOMDominican Republic CARICOMVenezuela CEFTA CEMAC CER ChileCosta Rica ChileCACM ChileColombia ChileEcuador ChileEFTA ChileKorea ChileMERCOSUR ChileMexico ChilePeru ChileVenezuela CIS ChinaMacao ChinaHong Kong ColombiaCACM ColombiaCuba ColombiaPanama COMESA CEN-SAD CroatiaBosnia and Herzegovina CroatiaHungary CroatiaMacedonia CBI CubaChile CubaEcuador CubaUruguay CzechMacedonia Czech RepublicEstonia

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

12 Table 7 continued Agreement Czech RepublicIsrael Czech Republic - Latvia Czech RepublicLithuania Czech RepublicSlovak Republic Czech RepublicTurkey EAC EAEC ECLebanon ECO ECCAS CEPGL ECOWAS EEA EFTACroatia EFTAJordan EFTABulgaria EFTACzech Republic EFTAEstonia EFTAHungary EFTAIsrael EFTALatvia EFTALithuania EFTAMacedonia EFTAMexico EFTAMorocco EFTAPalestinian Authority EFTAPoland EFTARomania EFTASingapore EFTASlovak Republic EFTASlovenia EFTATurkey EFTA (Stockholm Convention) El SalvadorGuatemala EstoniaFaroe Islands EstoniaTurkey EstoniaUkraine European UnionCroatia European UnionJordan European UnionMacedonia European UnionAlgeria European UnionAndorra European UnionAzerbaijan European UnionBulgaria European UnionChile European UnionCyprus European UnionCzech Republic European UnionEgypt European UnionEstonia European UnionFaroe Islands European UnionHungary European UnionIceland European UnionIsrael European UnionKyrgyz Republic European UnionLatvia European UnionLithuania European UnionMalta European UnionMexico European UnionMorocco European UnionNorway European UnionOCTs European UnionPalestinian Authority European UnionPoland

WORLD DEVELOPMENT Table 7 continued Year Notied to the WTO 1997 1997 1997 1993 1998 2000 1997 2003 1992 1995 1977 1975 1994 2002 2002 1993 1992 1996 1993 1993 1996 1996 2001 2001 1999 1999 1993 1993 2003 1992 1995 1992 1960 1991 1998 1998 1996 2002 2002 2001 1976 1991 1997 1993 2003 1973 1992 1977 1995 1997 1992 1973 2000 1998 1995 1995 1971 2000 2000 1973 1971 1997 1992 Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Agreement European UnionRepublic of San Marino European UnionRomania European UnionSlovak Republic European UnionSlovenia European UnionSouth Africa European UnionSwitzerland and Liechtenstein European UnionSyria European UnionTunisia European UnionTurkey European UnionUzbekistan European Union (Treaty of Rome) Faroe IslandsIceland Faroe IslandsNorway Faroe IslandsSwitzerland FijiPapua New Guinea GCC GeorgiaArmenia GeorgiaAzerbaijan GeorgiaKazakhstan GeorgiaRussian Federation GeorgiaTurkmenistan GeorgiaUkraine Greater Arab Free Trade Area Group of Three GSTP GuineaMorocco HungaryEstonia HungaryIsrael HungaryLatvia HungaryLithuania HungaryTurkey IndiaBangladesh IndiaSri Lanka IndiaBhutan IndiaNepal IGAD IranSwitzerland IraqEgypt IsraelJordan IsraelTurkey JapanSingapore JordanMorocco JordanSyria Kyrgyz RepublicArmenia Kyrgyz RepublicKazakhstan Kyrgyz RepublicMoldova Kyrgyz RepublicRussia Kyrgyz RepublicUkraine Kyrgyz RepublicUzbekistan LAIA LaosThailand LatviaTurkey LebanonKuwait LebanonSyria LebanonUAE LithuaniaTurkey MacedoniaYugoslavia MRU MERCOSUR MERCOSURBolivia MexicoBrazil MexicoCosta Rica MexicoCuba Year Notied to the WTO 1992 1993 1992 1997 2000 1973 1977 1998 1996 1999 1958 1993 1993 1995 1996 1982 1998 1996 1999 1994 2000 1996 1998 1995 1989 1997 2001 1998 2000 2000 1998 1980 2001 1995 1991 1986 2001 2001 1996 1997 2002 1994 2001 1995 1995 1996 1993 1998 1998 1981 1991 2000 1996 1998 2001 1998 1996 1973 1991 1997 2002 1995 1985 No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes No No Yes No Yes Yes Yes Yes Yes No Yes No No No No No No Yes Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No Yes No No Yes No No No No

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036

BEYOND TRADE: THE IMPACT OF PREFERENTIAL TRADE AGREEMENTS ON FDI INFLOWS Table 7 continued Agreement MexicoIsrael MexicoNicaragua MexicoNTR MexicoUruguay MoldovaArmenia MoldovaAzerbaijan MoldovaBelarus MoldovaKazakhstan MoldovaRussian Federation MoldovaTurkmenistan MoldovaUkraine MoldovaUzbekistan MoroccoTunisia MSG NAFTA NamibiaZimbabwe New ZealandSingapore NicaraguaColombia Palestinian AuthorityUnited States Palestinian AuthorityEgypt Palestinian AuthorityJordan PanamaCosta Rica PanamaDominican Republic PanamaEl Salvador PanamaGuatemala PanamaHonduras PanamaMexico PanamaNicaragua PATCRA PeruCuba PolandFaroe Islands PolandIsrael PolandLatvia PolandLithuania PolandTurkey PTAES PTN RomaniaMoldova RomaniaTurkey RussiaTajikistan Year Notied to the WTO 2000 1998 2001 2001 1995 1996 1994 1996 1993 1996 1996 1995 1999 1993 1994 1993 2001 1985 1996 1998 1995 1973 1987 1974 1975 1974 1986 1974 1977 1994 1999 1998 1999 1997 2000 1981 1973 1995 1998 1992 Yes No No No No No No No No No No No No Yes Yes No Yes No No No No No No No No No No No Yes No Yes Yes Yes Yes Yes No Yes Yes Yes No Agreement RussiaUkraine Russian FederationAzerbaijan Russian FederationBelarus Russian FederationTurkmenistan Russian FederationUzbekistan SAPTA Saudi ArabiaSyria SingaporeAustralia Slovak RepublicEstonia Slovak RepublicIsrael Slovak RepublicLatvia Slovak RepublicLithuania Slovak RepublicTurkey SloveniaBosnia and Herzegovina SloveniaCroatia SloveniaEstonia SloveniaIsrael SloveniaLatvia SloveniaLithuania SloveniaMacedonia South AfricaMalawi South AfricaZimbabwe SACU SADC SPARTECA TaiwanPanama TRIPARTITE TunisiaEgypt TurkeySlovenia TurkeyBosnia and Herzegovina TurkeyCroatia TurkeyMacedonia UkraineAzerbaijan UkraineTurkmenistan United StatesJordan United StatesIsrael United StatesChile United StatesSingapore VenezuelaCuba VenezuelaTrinidad and Tobago WAEMU/UEMOA Table 7 continued

13

Year Notied to the WTO 1994 1992 1995 1992 1992 1995 2003 2003 1998 1997 1997 1997 1998 2002 1998 1997 1998 1996 1997 1996 1990 1964 1969 2000 1981 2004 1968 1998 2000 2003 2003 2000 1996 1996 2001 1985 2004 2004 1989 1989 2000 No No No No No Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No Yes No Yes No Yes Yes Yes Yes No No Yes Yes No No No No Yes

Available online at www.sciencedirect.com

Please cite this article in press as: Medvedev, D. Beyond Trade: The Impact of Preferential Trade Agreements on FDI Inows, World Development (2011), doi:10.1016/j.worlddev.2011.04.036