Foreign direct investment and bilateral investment treaties: An international political perspective

Journal of Comparative Economics 37 (2009) 372–386

Contents lists available at ScienceDirect

Journal of Comparative Economics journal homepage: www.elsevier.com/locate/jce

Foreign direct investment and bilateral investment treaties: An international political perspective Rodolphe Desbordes a, Vincent Vicard b,* a b

University of Strathclyde, United Kingdom University of Paris I Panthéon-Sorbonne and Banque de France, France

a r t i c l e

i n f o

Article history: Received 4 October 2007 Revised 13 May 2009 Available online 9 June 2009

JEL classification: F21 F53 F59 Keywords: Foreign direct investment Interstate political relations Bilateral investment treaties Institutions

a b s t r a c t Desbordes, Rodolphe, and Vicard, Vincent—Foreign direct investment and bilateral investment treaties: An international political perspective This paper investigates the effect of the implementation of bilateral investment treaties (BITs) on the bilateral stocks of foreign direct investment (FDI). We argue that the understanding of how BITs affect FDI requires recognizing that multinational enterprises (MNEs) are not Stateless and that their investment return may well depend on the quality of political relations between the home and host countries. Using bilateral FDI data and event data to measure political interactions between countries, we show that the effect of the entry into force of a BIT crucially depends on the quality of political relations between the signatory countries; it increases FDI more between countries with tense relationships than between friendly countries. We also find evidence that BITs and good domestic institutions are complementary. BITs should therefore be understood as a mechanism for host governments to credibly commit not to expropriate investors in the future. Journal of Comparative Economics 37 (3) (2009) 372–386. University of Strathclyde, United Kingdom; University of Paris I Panthéon-Sorbonne and Banque de France, France. Ó 2009 Association for Comparative Economic Studies. Published by Elsevier Inc. All rights reserved.

1. Introduction ‘‘Ulysses, too, saw the value of binding himself to the mast. Constraints on sovereignty are, therefore, the aim of the exercise. In a world of international transactions and multiple jurisdictions, constraints on sovereignty are also desirable. Otherwise, the potential for conflict and unpredictability seems almost limitless.” (Wolf, 2005, p. 91) The reduction in trade barriers over the second half of the last Century went hand in hand with increasing regulation of international trade flows, via the creation of supranational institutions – preferential trade agreements at the regional level, and the World Trade Organization at the multilateral level – and interaction between domestic and international institutions, such as the New York Convention (Berkowitz et al., 2006). On the contrary, international capital flows, especially foreign direct investment (FDI), do not benefit from global governance mechanisms enforcing common rules: the property rights of foreign investors are instead protected by bilateral investment treaties (BITs). BITs are signed between two countries in order to encourage, promote and protect investments between them (UNCTAD, 2000). They include expropriation clauses defining what is deemed to be expropriatory behavior and specify compensation and dispute-settlement mechanisms, such as the recourse to international arbitration courts. In this respect, BITs are * Corresponding author. Address: 106-112, bd de l’hôpital, 75647 Paris Cedex 13, France. E-mail address: [email protected] (V. Vicard). 0147-5967/$ - see front matter Ó 2009 Association for Comparative Economic Studies. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jce.2009.05.001

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

373

considered to reduce both the risk and the cost of investing abroad. Using bilateral panel data, Egger and Pfaffermayr (2004) find that the implementation of a BIT increases the outward FDI stock by 30%.1 This average effect may nevertheless hide heterogeneity in the effectiveness of BITs across country pairs. As suggested by Martin Wolf, in the absence of constraints on the host country’s sovereignty, the activities of multinational enterprizes (MNEs) are subject to the host governments’ actions. BITs are thus a means for host governments to credibly commit not to expropriate MNEs in the future. In this paper, we argue that, as a commitment device, the effectiveness of a BIT should depend on the risks to which MNEs are exposed in the host country. MNEs face two kinds of political risks when investing abroad: systemic domestic risk, which is common to all investors, related to the quality of domestic institutions, and an idiosyncratic risk specific to each pair of home and host countries, resulting from interstate political relations. As the existing literature on FDI has largely considered that FDI takes place in an international political vacuum, the latter risk has been ignored.2 Anecdotal evidence and survey work however suggest that interstate political relations have a significant effect on MNEs’ decisions to invest abroad. This effect may be positive or negative. Transparency International (2002) emphasizes that, after corruption, diplomatic pressure is an important means for MNEs to gain business advantages. More crucially, foreign firms may suffer from the retaliatory consequences of deteriorating diplomatic relations between their home and host countries, through various types of expropriation (Boehmer et al., 2001). Foreign investors are, therefore, likely to be sensitive to the quality of interstate political relations, as any deterioration may increase the risk of seizure of their investment return in a given host country. In this framework, BITs should increase the volume of bilateral FDI not only directly by reducing costs, but also indirectly through two channels: first, by offsetting political tensions between states and the resulting expropriation risks; second, as a costly signal that the host government will not diminish the protection of property rights granted by domestic institutions. A BIT and good domestic institutions are complements in attracting FDI (Hallward-Driemeier, 2003). The link between FDI and interstate political relations has only rarely been investigated, due to the lack of information on the quality of these relations over recent decades. We overcome this obstacle by the use of a new database compiling a large number of recent interstate political interactions. This indicator allows us to estimate their impact on bilateral FDI stocks between 30 OECD countries and 62 OECD and non-OECD countries over the 1991–2000 period. We control for the self-selection of countries into BITs by including country pair fixed effects, as suggested by Baier and Bergstrand (2007) in the context of regional trade agreements. We first show that the quality of political relations between countries significantly increases bilateral FDI flows. We then find that the positive effect of a BIT on FDI stocks depends crucially on the quality of interstate relations, increasing FDI more between countries with political tensions. We also find that the effectiveness of the host government’s credible commitment increases with the quality of domestic governance. Good domestic institutions and BITs are indeed complements. This paper proceeds as follows: Section 2 reviews the different arguments explaining the links between interstate political relations, bilateral investment treaties and FDI. Section 3 describes the indicators used to evaluate the quality of interstate political relations, and explains the specification and data used in the empirical estimation. The impact of interstate political relations and BITs on bilateral FDI stocks is then presented in Section 4. Section 5 concludes. 2. Related literature The number of BITs dramatically increased starting in the 1990s. By the end of 2005, 2495 treaties had been signed, of which 1891 had entered into force, suggesting that an increasing number of countries see them as a way of attracting FDI and protecting their FDI outflows. The number of BITs by region is shown in Fig. 1. The emergence of BITs as the main way of protecting foreign investors’ property rights3 occurred during an uncertain and changing period regarding the regulation of international investments (Guzman, 1998). The classic formulation of the customary international law on FDI, known as the ‘‘Hull Rule”, requires ‘‘prompt, adequate and effective compensation” in the case of expropriation by the host government. This rule has been challenged by Latin American countries and former colonies, and weakened by resolutions adopted by the UN General Assembly in the 1960s and 1970s. The Charter of Economic Rights and Duties of states adopted by the UN General Assembly in December 1974 emphasized the sovereignty of host countries regarding their treatment of foreign investors (Guzman, 1998, 2007). In the framework of this charter, domestic courts are the only authority which can determine appropriate compensation. Bubb and Rose-Ackerman (2007) point out that the protection provided by the customary international law on FDI is weak: any claim by an expropriated foreign investor has to be supported by its home government in order to proceed. Moreover, claims can only cover egregious expropriation and not simple breaches of contract. These cases of open expropriations

1 Work using aggregate FDI data has found an ambiguous impact of BITs on FDI, ranging from positive (Neumayer and Spess, 2005) to insignificant (RoseAckerman and Tobin, 2005). 2 Nigh (1985) is one of the few papers in the international business literature to have considered this subject. He finds that conflictual and cooperative diplomatic relationships exert, respectively, negative and positive effects on US manufacturing FDI in developing countries. The empirical work is however restricted to US diplomatic relationships, does not account for other FDI determinants, and only covers the particular period of the Cold War (1948–1978). 3 UNCTAD (2008) reports that, in 2007, 78% of the investor-state cases filed under international investment agreements were initiated following a violation of a BIT provision, 13% under NAFTA, and 6% under the Energy Charter Treaty.

374

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386 World Developed countries Latin America and the Caribbean Africa Southeast Europe & CIS Asia and Oceania

1600

1400

1200

1000

800

600

400

200

0 1980-1989

1990-1999

2000-2009*

Fig. 1. Number of bilateral investment treaties signed by region. Source: UNCTAD. *2000–2009 values are linearly extrapolated based on the 2000–2005 data.

peaked in the 1970s, with the nationalizations carried out in developing countries in the natural resource sector, but have since declined. 2.1. Foreign direct investment and interstate political relations Good legal institutions reduce the risk of expropriation and the cost of operating in a host market, and thus the risk-premium requirements for sunk-cost investment by MNEs (Globerman and Shapiro, 2002). Property rights must not only be protected against the actions of private agents (individuals or enterprises) but also against the State. As outlined by Djankov et al. (2003), a state powerful enough to enforce contracts and secure property rights can also use this power for its own benefit. Since MNEs bear a sunk cost when investing abroad, once they invest they are subject to policy changes and attempts to renegotiate contracts by the host government. In this case, foreign investors can only resort to disinvestment or appeal to political influence, via lobbying or diplomatic pressure. Foreign investors, as informal representatives of their country, may suffer from worsening diplomatic relations between their home and host countries, since their expropriation can be used as a retaliatory instrument in an interstate conflict. Boehmer et al. (2001) show how valuable interstate linkages, such as FDI, can serve as a costly signaling mechanism in conflicts. They argue that the ex-ante destruction of mutually valuable interstate economic links can be seen as a means of communication via which parties in disagreement signal their resolve by sending a credible (and costly) signal. By reducing uncertainty about the preferences of State leaders, this signal fosters the emergence of a peaceful negotiated settlement.4 This revelation of preferences through the targeting of foreign firms should make MNEs less inclined to invest in countries engaged in diplomatic disputes with their home countries, due to uncertainty over their future returns. This uncertainty raises the required return on FDI and reduces, at least temporarily, the range of projects deemed attractive by foreign investors.5 On the contrary, better diplomatic relations should foster bilateral FDI, by reducing the risk of expropriation. 2.2. Foreign direct investment and bilateral investment treaties The expropriation risk sustained by MNEs via interstate relations is related to the very structure of the international system, in which court jurisdictions are delimited by political boundaries and where no multinational legal standards for the treatment of FDI have emerged. Attempts to negotiate multilateral investment agreements have failed. The 1965 Washington Convention established a multilateral dispute-arbitration body for investor-state disputes, the International Center for the Settlement of Investment Disputes (ICSID), but this only focused on procedural issues (Bubb and Rose-Ackerman, 2007). Another attempt to initiate a multilateral investment treaty among OECD countries in 1995 was definitively

4

It is worth noting that in this case, expropriation by State rulers fulfills a number of policy objectives, rather than just the private benefit of rulers. See Rodrik (1991) and Servén (1997) for a real-options approach to investment. They show that if an investment is partially irreversible and can be delayed, it will only be undertaken in an uncertain environment if the return to the investment is much higher than the user cost of capital. 5

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

375

abandoned in 1998. Some regional trade agreements, such as those in the European Union and the NAFTA, do provide extended protection to investors from partner countries. In addition, the 1995 Trade-Related Aspects of Intellectual Property Rights (TRIPS) agreement defined minimum standards for intellectual property right laws and their enforcement. As part of the WTO, disputes over TRIPS obligations are subject to the dispute settlement mechanism. Compliance with the TRIPS agreement should improve the domestic institutions which protect intellectual property rights while leaving signatory governments with a great deal of flexibility in their choice of policies which ensure the minimum level of property rights’ protection enshrined in the TRIPS agreement (Maskus, 2000).6 BITs can be understood as a means of reducing the uncertainty from expropriation risks, allowing host governments to credibly commit to not expropriate investors. BITs generally include provisions prohibiting discriminatory treatment against foreign investors; they also include investment performance requirements and ensure the possibility of repatriating profits without delay (UNCTAD, 2000). More remarkably, many BITs grant foreign investors the right to sue the host government through international arbitration if any action undertaken by the former is deemed to be tantamount to expropriation, i.e. nationalization or even regulatory change (Hallward-Driemeier, 2003). Under a BIT, a contract is binding for both the foreign investor and the host government, since any breach of contract falls under international law (Guzman, 1998). This raises the cost to the host government of reneging on its commitments.7 By signing a BIT, host countries accept the possibility of being sued and send a costly signal that they will not renege on their contracts.8 Two types of interactions between BITs and domestic institutions have been advanced in the literature. First, by signing a BIT, countries can indicate their determination to offer foreign investors an institutional framework with better propertyrights protection than that offered by domestic institutions (Neumayer and Spess, 2005). From this perspective, a BIT acts as a substitute for good domestic institutions. Alternatively, the implementation of a BIT could act as a signal to foreign investors that a country will not threaten the property-rights protection granted by domestic institutions in order to achieve its national objectives and security choices: BITs and good domestic institutions are then complementary (Hallward-Driemeier, 2003). In addition, even without a BIT with their home country, MNEs may take into account any agreement signed with third countries. The number of BITs already signed by a host country with other partners may act as an additional signal of the government’s credibility in protecting the property rights of foreign investors. In this respect, countries which implement more BITs should attract more FDI, even from non-signatory countries. The implementation of a BIT is expected to increase the volume of bilateral FDI both directly by reducing the cost of investing abroad and indirectly depending on the quality of interstate relations between the signatory countries and the quality of domestic institutions. The existing literature has largely ignored these indirect effects of BITs on FDI. In a bilateral framework, Hallward-Driemeier (2003) is the only paper to take the latter indirect effect into account, finding weak evidence of complementarity between BITs and good domestic institutions. The risk which is specific to pairs of home and host countries has not been analysed in this literature, which is not surprising since diplomatic relations have largely remained a ‘‘missing” determinant of FDI to date. 3. Empirical model and data Before turning to the estimation of the impact of BITs, we consider the effect of the quality of interstate relations on bilateral FDI flows. This preliminary step is carried out by constructing a new indicator of the quality of interstate political relations using event data. 3.1. The quality of interstate political relations Two types of data are available regarding interstate interactions: qualitative data on armed conflicts and quantitative data on daily events. In the former, the actors, duration, geographical location and intensity of each conflict are recorded and documented. This can only be undertaken for infrequent high-intensity interstate interactions such as armed conflicts. In the latter, daily events are automatically extracted by computers from wire reports or newspapers and are automatically coded according to their actors and type of action. In contrast with the armed-conflict datasets, it is almost impossible to ascertain whether these individual entries point to different stages of the same case of conflict. However, data on daily events presents the substantial advantage of providing information about both conflictual and cooperative relations between states, whatever the intensity of the underlying event. The evaluation of diplomatic relations between countries is based on a new event dataset, developed by the Kansas Events Data System (KEDS) and made available by Gary King on his website.9 Computers are programmed to read the first sentence of news reports from wire services and to code each event according to the actor, the target, the type of event and the date. King and Lowe (2002) describe in detail this process and suggest that computer coding is equivalent to human coding in the 6

The TRIPS agreement allows a transitional period for transition and developing countries (5 years) and less-developed countries (11 years). Elkins et al. (2006) report that the governments of the Czech Republic, Lebanon and Ecuador had to pay 250, 266 et 70 Million US$, respectively, to foreign firms for having expropriated them. 8 Schelling (1960) develops the same rationale regarding the purpose of an enforced legal system for interactions between domestic firms. 9 http://gking.harvard.edu. 7

376

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

short run and more efficient in the long run. The typology of events comes from the Integrated Data for Events Analysis database (IDEA, see Bond et al. (2003) for a complete description of the coding scheme). In order to aggregate these daily events, the level of conflict or cooperation in each case needs to be taken into account. The Goldstein (1992) scale is one way of transforming daily interactions into two annual flows of cooperative and conflictual interstate political relations.10 The values attributed to each category of event, reported in King and Lowe (2002), are shown in Appendix A. This scale produces a score between 0 and +10 (10) for each event category according to the extent of cooperation (conflict) embodied in each event. We use the sum of annual cooperation and conflict flows to construct a net indicator of the quality of interstate political relations (QIR) between two countries in year t.11 The aggregation of cooperation and conflicts into an annual index yields a yearly measure of the political climate between two countries. The greater the value, the better is the political climate between the two countries. These data are available over the 1991–2000 period. 3.2. The gravity model for FDI The workhorse econometric model for bilateral trade flows, the gravity model, is now increasingly used to analyze FDI (Wei, 2000; Razin and Sadka, 2007; Blonigen et al., 2007). Recently, Head and Ries (2008) and Bergstrand and Egger (2008) have provided theoretical micro-foundations for a gravity model of FDI. The basic gravity specification relates the volume of bilateral FDI to the GDPs of both the home and host countries and to the geodesic distance between them. The larger the market sizes in both countries, the greater is FDI. Distance is a proxy for investment costs, such as management costs, and is expected to reduce FDI. In line with the literature, we add variables reflecting the geographical and historical links between the host and home countries (contiguity and common language) and the GDP per capita of each country. The sign of the latter is ambiguous since greater GDP per capita reflects both purchasing power and nominal wages, which exert positive and negative effects on FDI, respectively (Globerman and Shapiro, 2002; Globerman and Shapiro, 2003). Finally, we include our two measures of investment risks (indicators of the quality of domestic legal and political institutions as a proxy for the host-country specific risk, and the quality of interstate relations as a proxy for the risk specific to the pair of home and host countries), and a dummy for the entry into force of a BIT. Larger risks, by increasing investment costs and uncertainty, should reduce FDI, while a BIT is expected to increase FDI. The baseline specification estimated in the following section is:

lnðFDIijt Þ ¼ b0 þ b1 lnðGDPit Þ þ b2 lnðGDPjt Þ þ b3 lnðGDPPC it Þ þ b4 lnðGDPPC jt Þ þ b5 lnðdij Þ þ b6 C ij þ b7 lnðINST jt Þ þ b8 lnðQIRijt Þ þ b9 BIT ijt þ b10 BIT ijt  lnðQIRijt Þ þ b11 BIT ijt  lnðINST ijt Þ þ Ei þ Ij þ T t þ ijt

ð1Þ

where FDIijt stands for the bilateral stock of FDI in country j originating from country i in year t; dij is the bilateral distance, C ij is a vector of gravity-specific dummies (contiguity and common language), INST jt is a measure of the quality of domestic institutions, QIRijt is our proxy for the quality of interstate political relations, BIT ijt is a dummy variable for a bilateral investment treaty between countries i and j; Ij ðEi Þ corresponds to a host (home) country time-invariant fixed effect, T t is a countryinvariant time effect and ijt is the error term. In order to assess the indirect effects of BITs on FDI, we introduce interaction terms between BIT ijt and the quality of both interstate relations and domestic institutions. As a commitment device, a BIT should be more effective when implemented between countries with poor interstate relations: we therefore expect b10 to be negative. As noted above, BITs may be substitutes or complements for good public governance: the sign of b11 is thus ambiguous. Omitted time-invariant determinants of FDI are captured by both home- and host-country fixed effects, and time dummies control for any macro shocks affecting FDI. 3.3. Data and methodology Our dependent variable is the bilateral FDI stock. This comes from the OECD International Direct Investment Statistics database, which reports data for bilateral stocks among 30 OECD countries, and between OECD countries and 32 non-OECD emerging countries, over the 1991–2000 period. FDI stocks are preferred to FDI flows as the former are less volatile, which is particularly important when working with yearly data. All FDI stocks are converted into millions of current US dollars. A well-known problem in the log specification of the gravity model is the difficulty of accounting for zeros in the dependent variable, as dropping them could create selection bias.12 To deal with this problem, we use a Poisson quasi maximum likelihood estimator (QMLE). This strategy was suggested by Santos Silva and Tenreyro (2006) regarding gravity models of trade flows (see Head and Ries (2008) for an application to FDI). They point out that both standard log-linear and Tobit models yield inconsistent estimates in the presence of heteroscedasticity. Their proposed estimation procedure, Poisson QMLE, is not only consistent in the presence of heteroscedasticity, but it also allows us to incorporate zero values for the dependent variable in our regressions.

10 The mapping of IDEA categories onto the Goldstein scale, first developed for the World Event/Interaction Survey (WEIS), is available from the IDEA website (http://vranet.com/idea). 11 As a result of the log specification of the gravity equation, we will add 491.8 to this number to obtain strictly positive values. 12 Out of our 8032 observations, about 8% are zero.

377

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386 Table 1 Descriptive statistics Variable

Obs

Mean

Std. dev.

Min

Max

FDI ln FDI ln GDP origin ln GDP host ln distance Contiguity Common language ln GDP per capita origin ln GDP per capita host Quality of domestic institutions (host) ln Quality of domestic institutions (host) Quality of interstate relations (QIR) ln Quality of interstate relations (QIR) BIT

8032 7389 8032 8032 8032 8032 8032 8032 8032 8032 8032 8032 8032 8032

3014.08 5.39 12.83 12.67 8.24 0.07 0.08 9.48 9.23 7.46 1.99 523.22 6.25 0.28

12387.55 2.75 1.51 1.51 1.10 0.26 0.27 1.08 1.19 1.27 0.18 91.68 0.16 0.45

0 4.25 8.67 8.67 4.09 0 0 5.63 5.63 2.29 0.83 1.00 0.00 0

303591.7 12.62 16.10 16.10 9.88 1 1 10.69 10.69 9.79 2.28 2179.2 7.69 1

Notes: Millions of US dollars.

Data on GDP ½ðGDP and GDP per capita [GDPPC] in current US dollars are taken from the World Bank World Development Indicators database. Time-invariant bilateral characteristics (distance, contiguity and common language) are obtained from the CEPII.13 In addition, an important determinant of FDI flows is the quality of domestic institutions ½INST (Wei, 2000; Globerman and Shapiro, 2002; Benassy-Quere et al., 2007). Our proxy measure here comes from the International Country Risk Guide, which provides ratings of economic, financial and political risks for a large number of countries.14 We construct an annual index of the quality of domestic institutions using a simple average of four components of the political risk index: government stability, investment profile (which measures contract viability/expropriation, profit repatriation and payment delays), law and order (which measures the strength and impartiality of the legal system and popular observance of the Law), and bureaucracy quality.15 The greater is this index, ranging between 0 and 10, the lower is the perceived risk. The indicator of the quality of interstate political relations ½QIR was described in Section 3.1. Finally, the BIT dummy ½BIT equals one starting from the year when the BIT between the two countries entered into force.16 Data on BITs come from the UNCTAD Investment Treaty Database.17 Summary statistics are shown in Table 1. 4. Results In this section, we first present the estimation of the effect of the quality of interstate relations. We then turn to the estimation of Eq. (1) and examine the direct and indirect effects of BITs on FDI. Last, we present our results on related third-country effects. 4.1. Quality of interstate political relations The results are shown in Table 2. From a host-country perspective, greater income per capita, good public governance, and a shared language exert positive impacts on bilateral investment, whereas the opposite is true for bilateral distance. These results are in line with previous work using the same specification, such as Benassy-Quere et al. (2007) and Head and Ries (2008). Although the signs and significance of our control variables change with the exact specification,18 it is reassuring to note that the coefficient on our proxy for the quality of interstate political relations (QIR) is always positive and significant, indicating that countries with good diplomatic relations invest more in each other. The economic effect is substantial since, from the results in Table 2, a one standard deviation increase from the mean of the quality of interstate political relations increases the bilateral FDI stock from 7% (column (2)) to 16% (column (4)). Columns (5) to (8) of Table 2 show the results of a number of robustness tests. First, the multilateral resistance terms, as emphasized by Anderson and van Wincoop (2003) and Feenstra (2004) with respect to bilateral trade flows, will also likely affect investment flows (Head and Ries, 2008). In a panel context, these are captured by country-and-year fixed effects.

13

www.cepii.fr See http://www.prsgroup.com/. 15 Each of these components is recalculated on a 1–10 scale. 16 Egger and Pfaffermayr (2004) show that a BIT significantly increases bilateral FDI only if it is actually implemented, underlining that the international commitment of the host country must appear credible to foreign investors. Hence, we use the date of entry into force of a BIT rather than its signature date. 17 http://www.unctadxi.org/templates/Startpage____718.aspx 18 The puzzling estimates on GDPs may result from the presence of host and home country fixed effects and the multicollinearity between GDP and GDP per capita (as population varies only slowly). 14

378

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

Table 2 The impact of interstate political relations on FDI Model Dependent var.

(1) ln (FDI)

Estimator

OLS

ln GDP origin ln GDP host ln distance Contiguity Common language ln GDP per capita origin ln GDP per capita host ln Quality of domestic institutions host (INST)

(3) FDI

(4) FDI

(5) FDI

Poisson QMLE b

2.21 (1.07) 3.23a (1.08) 1.02a (0.06) 0.63a (0.19) 0.79a (0.19) 2.57b (1.11) 4.34a (1.12) 0.41c (0.21)

ln Quality of interstate relations (QIR)

(2) ln (FDI)

1.65 (1.01) 2.99a (1.06) 0.95a (0.07) 0.63a (0.18) 0.72a (0.19) 1.90c (1.06) 4.19a (1.11) 0.42c (0.23) 0.43c (0.26)

0.22 (2.40) 1.97c (1.17) 0.47a (0.05) 0.12 (0.14) 0.59a (0.10) 0.44 (2.47) 2.64b (1.12) 0.15 (0.17)

1.25 (2.33) 1.73 (1.12) 0.43a (0.05) 0.08 (0.14) 0.40a (0.09) 0.65 (2.41) 2.42b (1.10) 0.15 (0.16) 0.93a (0.13)

0.44a (0.05) 0.06 (0.14) 0.38a (0.09)

(6) ln (FDI)

(7) FDI

OLS

Poisson QMLE

0.83a (0.08) 0.26 (0.19) 0.57a (0.18)

1.36a (0.14)

1.43b (0.73)

Pair ever in a colonial relationship Ever same country Militarized interstate dispute EU NAFTA GATT

1.30 (2.38) 1.65 (1.12) 0.43a (0.06) 0.05 (0.15) 0.29a (0.09) 0.70 (2.46) 2.33b (1.09) 0.14 (0.17) 0.81a (0.13) 0.38a (0.12) 2.58a (0.36) 0.00 (0.12) 0.27c (0.14) 0.03 (0.15) 0.12 (0.13)

(8) FDI

0.86 (2.41) 1.84c (1.09)

0.21 (2.49) 2.53b (1.07) 0.14 (0.16) 0.66a (0.18)

Year fixed effects Country fixed effects Country-pair fixed effects Country-and-year fixed effects

Yes Yes – –

Yes Yes – –

Yes Yes – –

Yes Yes – –

– – – Yes

– Yes – –

– Yes – –

Yes – Yes –

Observations Number of groups Sargan–Hansen statistic Weak identification F-test Partial R2

8714 – – – –

7389 – – – –

10,025 – – – –

8032 – – – –

8149 – – – –

440 – 1.60 135.44 0.44

7971 – – – –

7590 1086 – – –

Notes: a, b and c denote, respectively, significance at the 1%, 5% and 10% levels. Heteroscedasticity- and autocorrelation-robust standard errors are in parentheses.

Column (5) presents the results from the PQML estimator with country-and-year fixed effects, where the results remain qualitatively unchanged. These results may also suffer from endogeneity. The causality between FDI and interstate political relations may be bidirectional since, according to the liberal peace paradigm, greater economic interdependence fosters better interstate political relations (Polachek, 1980; Oneal and Russett, 1997; Oneal and Russett, 1999). In addition, omitted country-pair specific variables correlated with the quality of interstate political relations may be the true factor driving the impact of the quality of diplomatic relations on FDI. We deal with each problem consecutively, as we do not have an exogenous time-varying instrumental variable for the quality of interstate political relations and the inclusion of country-pair fixed effects requires panel data. The first source of endogeneity, simultaneity, can be dealt with by finding a suitable cross-section instrument. However, since even in a cross-section we were not able to identify good external instruments,19 we resort to internal instruments: the lagged values of the quality of interstate political relations eight and nine years earlier. These lags were chosen by the firststage F-test statistic, the partial R2 and the Hansen (1982) J-tests of overidentifying restrictions. The first-stage F-statistics and partial R2 indicate that these instruments can be regarded as ‘‘strong” since the statistic values are, respectively,

19 We tried to instrument the quality of interstate political relations via the similarity of the alliances the two countries have with other countries, UN vote correlation, religious similarity, and conflict history. However, the Sargan–Hansen test rejected their exogeneity in every case.

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

379

135.4 – well above Stock et al. (2002)’s rule of thumb of 10-, and 0.44, and the Sargan–Hansen test does not reject their exogeneity. Note that the econometric methodology in column (6) of Table 2 is two-stage least squares. To overcome the second source of endogeneity, omitted variables, variables which could be correlated with the quality of interstate political relations are included in the main specification. First, in column (7), historical ties, i.e. the existence of a colonial relationship and the fact that the two countries used to be joined (e.g. the Czech Republic and Slovakia), military conflict and common membership in NAFTA, the EU and the WTO/GATT are controlled for. Second, all time-invariant (unobservable) country-pair characteristics which may affect bilateral FDI, such as cultural proximity, are taken into account by country-pair fixed effects in column (8), replacing the geographic and linguistic bilateral variables and country-specific effects. We use the xtpqml Stata package developed by Tim Simcoe, which computes robust standard errors for fixed-effects Poisson models, as suggested by Wooldridge (1999).20 In this specification, the impact of the quality of interstate political relations on FDI is identified only through the effect of its variation over time. The results controlling for endogeneity confirm our previous findings: the coefficient on QIR is always positive and significant at the 1% level. In the instrumental variable regression in column (6), the coefficient on QIR is significant, and the results remain unchanged when controlling for additional determinants of bilateral FDI in column (7). This shows that historical ties influence bilateral FDI but are not behind the effect of the quality of diplomatic relations. In column (8), the coefficient on QIR falls slightly, as the inclusion of country-pair fixed effects implies that only changes in the quality of interstate political relations identify the relationship, controlling for any time invariant country pair heterogeneity. However, the coefficient remains significant at the 1% level, highlighting the importance of interstate political relations as a determinant of FDI. These tests thus underline the robustness of the impact of the quality of interstate political relations on bilateral FDI. 4.2. Bilateral investment treaties We now turn to the estimation of the impact of BITs on FDI. In Table 3, we introduce our two interaction variables to assess the impact of the entry into force of a BIT conditional on the quality of interstate relations and the quality of domestic institutions. The significance of the BIT coefficient depends on whether country-pair fixed effects are included (columns (1) and (2)). The fact that the coefficient is only significant in the specification including country-pair fixed effects suggests that countries ‘‘choose well” when signing a BIT, as shown by Baier and Bergstrand (2007) regarding regional trade agreements. The decision to sign a BIT is not exogenous: country pairs that have more to gain from a BIT are more likely to decide to enter one. When the factors behind this selection are unobserved, the estimated coefficient on BITs will be biased downwards. The inclusion of country-pair fixed effects controls for the endogeneity bias related to any omitted (unobservable) variables affecting both the level of bilateral FDI and the opportunity of entering a BIT.21 When these factors are taken into account, the average effect of the entry into force of a BIT on FDI is positive and significant (column (2)). Without controlling for the interdependence between BITs and the risk sustained by MNEs, we find an average effect of BITs (39%) slightly larger than that in Egger and Pfaffermayr (2004). The effectiveness of BITs is nevertheless conditional on the risk sustained by MNEs when investing abroad. In columns (3) of Table 3, we introduce our two interactions terms. The coefficient on the interaction between BIT and the quality of interstate relations is negative, suggesting that the entry into force of a BIT increases bilateral FDI less for countries with good diplomatic relationships. The positive sign on the coefficient of the interaction between BIT and the quality of domestic institutions confirms that a BIT and good institutions are complements rather than substitutes. In this specification, the coefficient on BIT becomes insignificant. This does not mean that BITs have no direct effect on FDI. We need to measure the total effect and the statistical significance of the entry into force of a BIT, conditional on the quality of both interstate relations and domestic institutions (Brambor et al., 2006). Fig. 2 shows the marginal effect of the entry into force of a BIT and its 95% confidence interval, as a function of the quality of interstate relations between signatory countries, at three different levels of domestic institutional quality (median, and the first and ninth deciles). This quantification of the effect of BIT on FDI is based on specification (3) of Table 3, which we prefer as the country-pair fixed effects control for any self-selection into BITs. Fig. 2 clearly shows that the effect of the entry into force of a BIT depends crucially on the quality of political relations between the signatory countries. A BIT increases bilateral FDI more when signatory countries are experiencing political tension. It has no significant effect between friendly partner countries and between countries with very confrontational political relations.22 The host government’s credible commitment not to expropriate foreign investors is thus more valuable when MNEs face risks related to interstate political tensions. A BIT increases FDI stocks by 42% when the quality of interstate relations and domestic institutions are set at their median values (INST ¼ 2:01, panel (a) of Fig. 2). When the quality of domestic institutions is set at the first decile ðINST ¼ 1:73Þ in

20 http://www.rotman.utoronto.ca/timothy.simcoe/xtpqml.txt. This methodology is equivalent to including bilateral country-pair dummies when using a Poisson Quasi Maximum Likelihood estimator with heteroscedasticity-robust standard errors. 21 Endogeneity related to reverse causality is not likely to be an issue here since Egger and Pfaffermayr (2004) show that signing a BIT without ratifying it does not increase significantly FDI. 22 The first percentile of ln QIR is 6.16, for which a BIT still increases FDI significantly. On the contrary, the top percentile is 6.91, for which the effect of a BIT is insignificant.

380

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

Table 3 The impact of bilateral investment treaties on FDI Model Dependent var.

(1) FDI

Estimator

Poisson QMLE

ln GDP origin

1.23 (2.33) 1.76 (1.11) 0.43a (0.05) 0.08 (0.14) 0.40a (0.09) 0.64 (2.41) 2.44b (1.10) 0.15 (0.16) 0.93a (0.13) 0.05 (0.12)

Year fixed effects Country fixed effects Country-pair fixed effects Host country-and-year fixed effects Observations Number of groups

ln GDP host ln distance Contiguity Common language ln GDP per capita origin ln GDP per capita host ln Quality of domestic institutions host (INST) ln Quality of interstate relations (QIR) BIT

(2) FDI

(3) FDI

(4) FDI

(5) FDI

0.85 (2.41) 1.94c (1.10)

0.79 (2.39) 2.11c (1.11)

0.83 (2.43) 2.05c (1.11)

2.15 (1.99)

0.19 (2.49) 2.64b (1.08) 0.14 (0.16) 0.66a (0.18) 0.33a (0.09)

0.12 (2.47) 2.81a (1.08) 0.01 (0.19) 0.66a (0.18) 0.70 (2.44) 0.25 (0.41) 0.62b (0.26)

1.50 (2.06)

Yes Yes –

Yes – Yes

Yes – Yes

0.15 (2.51) 2.75b (1.08) 0.02 (0.19) 0.66a (0.18) 0.39 (2.55) 0.11 (0.44) 0.73b (0.34) 0.20 (0.16) 0.06 (0.07) 0.15 (0.16) 0.30b (0.14) Yes – Yes

– 8032 –

– 7590 1086

– 7590 1086

– 7590 1086

BIT*ln QIR BIT*ln INST EU NAFTA GATT TRIPS (host)

0.77a (0.20) 4.33 (2.94) 0.75 (0.47) 0.26 (0.39)

– – Yes Yes 7590 1086

Notes: a, b and c denote, respectively, significance at the 1%, 5% and 10% levels. Heteroscedasticity-robust standard errors are in parentheses.

panel (b) and the ninth decile ðINST ¼ 2:22Þ in panel (c), FDI increases by 24% and 53%, respectively. Together the three panels of Fig. 2 point to complementarity between BITs and good domestic governance. When the host country’s institutions are poor (panel (b)) the total effect of a BIT on the bilateral FDI stock is clearly lower than when domestic institutions are at the median level (panel (a)), but this effect is only slightly larger when institutions change from median to good (panel (c)). The government’s commitment not to attack the property rights of foreign investors is more valuable when domestic institutions are good. In a risky market, where domestic governance is poor, the host government may not be powerful enough to secure property rights and enforce contracts. For domestic institutions that are good enough, the complementarity becomes weaker since an additional increase in the quality of institutions only slightly increases the effectiveness of BITs on FDI. Overall, Fig. 2 suggests that implementing a BIT signals the credibility of domestic governance, as the host country is less likely to encroach on the protection granted by its domestic institutions. In column (4) of Table 3, we control for the home and host countries’ membership in international organizations that provide various levels of protection of property rights to investors from signatory countries. We include dummies for common membership of regional trade agreements, the EU and the NAFTA, and the GATT/WTO, and a dummy indicating whether the host country is part of the TRIPS agreement. These agreements are mainly designed to regulate trade flows; their effect on FDI is ambiguous since it depends on their relative impact on trade and investment barriers (Bergstrand and Egger, 2008). In this specification, the results confirm previous evidence regarding the quality of interstate relations and the quality of domestic institutions and their effect on the impact of BITs on FDI. Finally, in column (5) we include host-country-and-year dummies to control for omitted variable bias. The coefficients on BIT and the interaction between BIT and QIR increase in absolute value, and the coefficient on the interaction between BIT and the quality of domestic institutions falls. Our results nevertheless remain qualitatively similar, underlining their robustness.

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

381

(a) Quality of domestic institutions set at median

Marginal Effect of BIT

2

1.5

1

.5

0

5

5.5

6 6.5 ln Quality of Interstate Relations

7

(b) Quality of domestic institutions set at 1st decile

Marginal Effect of BIT

2

1.5

1 .5

0

5

5.5

6 6.5 ln Quality of Interstate Relations

7

(c) Quality of domestic institutions set at 9th decile

Marginal Effect of BIT

2 1.5

1 .5 0 5

5.5

6 6.5 ln Quality of Interstate Relations

7

ˆMarginal Effect of BIT on FDI˜ ˆ95% Confidence Interval˜

Fig. 2. Total effect of the entry into force of a BIT on FDI. Note: based on specification (3) in Table 3.

Overall, these results lend support to our argument that the purpose of signing a BIT for a host country is to send a costly signal stating its credibility as a third party guaranteeing foreign investors’ property rights. Through the signature of a BIT, the two partner countries reciprocally abandon the use of retaliatory actions against foreign firms and part of their sovereignty in order to mark their determination to offer a safe business climate for foreign investors on a long-term basis. 4.3. Third-country effects In this section, we investigate whether maintaining good interstate relations or implementing a BIT with a given partner exerts any side effects on FDI originating from third countries, as an additional signal of the credibility of the host country’s

382

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

Table 4 Bilateral FDI and third country effects Model Dependent var.

(1) FDI

Estimator

Poisson QMLE

GDP origin

0.80 (2.38) 2.12c (1.10) 0.14 (2.46) 2.83a (1.07) 0.01 (0.19) 0.66a (0.18) 0.71 (2.44) 0.25 (0.41) 0.62b (0.26) 0.01 (0.02)

GDP host GDP per capita origin GDP per capita host ln Quality of domestic institutions host (INST) ln Quality of interstate relations (QIR) BIT BIT*ln QIR BIT*ln INST ln QIR multilateral Dummy 30th percentile QIR Dummy 70th percentile QIR ln QIR multi*Dummy 30th percentile QIR ln QIR multi*Dummy 70th percentile QIR

(2) FDI

(3) FDI

(4) FDI

0.85 (2.39) 2.13c (1.10) 0.18 (2.47) 2.84a (1.07) 0.01 (0.19) 0.63a (0.19) 1.50 (2.53) 0.37 (0.42) 0.63b (0.27) 0.02 (0.02) 0.11 (0.08) 0.01 (0.06) 0.05c (0.03) 0.03 (0.02)

0.82 (2.29) 2.70b (1.20) 0.18 (2.38) 3.43a (1.17) 0.01 (0.19) 0.68a (0.18) 1.91 (2.51) 0.29 (0.39) 0.08 (0.32)

0.83 (2.28) 2.72b (1.20) 0.20 (2.37) 3.46a (1.16) 0.03 (0.19) 0.65a (0.18) 1.88 (2.42) 0.30 (0.40) 0.12 (0.33)

0.19c (0.10) 0.37a (0.13)

0.55b (0.24)

ln BIT total ln BIT total*Dummy 30th percentile QIR ln BIT total*Dummy 70th percentile QIR Year fixed effects Country-pair fixed effects Observations Number of groups

Yes Yes 7590 1086

Yes Yes 7590 1086

Yes Yes 7590 1086

0.51b (0.23) 0.05b (0.03) 0.11a (0.03) Yes Yes 7590 1086

Notes: a, b and c denote, respectively, significance at the 1%, 5% and 10% levels. Heteroscedasticity-robust standard errors are in parentheses.

government. First, we add a proxy for the quality of interstate political relations of a host country vis-a-vis all its partners [QIR multilateral]. This is the average of bilateral interstate political relations, weighted by the market size of partner countries (GDP). The results are presented in column (1) of Table 4. Good interstate political relations with countries other than the partner country do not have any significant impact on FDI. Column (2) presents a Chow-type test by including dummies for the 30th and 70th percentile of the quality of interstate political relations and their interactions with ln QIR multilateral. The joint significance tests indicate the insignificance of the multilateral quality of interstate relations. This test confirms that the impact of interstate political relations on FDI is country-pair specific, and that different foreign investors in the same host country will not experience the same level of property rights’ protection at a given time. In column (3), in addition to the BIT dummy, we add the stock of BITs [BIT total] (the cumulative number of BITs that the host country has signed with all its partners).23 If a BIT acts as a costly signal of the quality of property rights’ protection in the host country and the host government’s commitment to MNEs in the signatory partner country, BITs signed with other countries should also signal the good business climate in the host country to international investors, even those from nonsignatory countries. We find that FDI from a non-signatory country rises with the stock of BITs implemented by the host country. This implies that a greater number of BITs signals the host country’s determination to offer a business climate favorable to MNEs to foreign investors. The Chow-type tests presented in column (4) confirm that the stock of BITs significantly increases bilateral FDI.

23

The data come from the UNCTAD FDI database (http://stats.unctad.org/FDI/ReportFolders/ReportFolders.aspx).

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

383

5. Conclusion Most of the literature dealing with the location of FDI has ignored that MNEs are not Stateless and that their activities take place in the international political system. When investing abroad, the business environment faced by MNEs is not only shaped by the quality of domestic institutions: the return on their FDI can also be affected by the quality of interstate political relations between the home and host countries. In this framework, BITs can be understood as commitment devices enabling host country governments to state their credibility in guaranteeing the property rights of foreign investors. Through the signature of a BIT, the two partner countries reciprocally abandon the use of retaliatory actions against foreign firms and accept constraints on their sovereignty in order to attract FDI. This paper fills this gap in the literature by testing the impact of the quality of interstate political relations and BITs on bilateral FDI. Our empirical results indicate that having good interstate political relations increases FDI. The effect of a BIT depends crucially on the risk sustained by MNEs when investing abroad. BITs have a greater effect when implemented between countries with political tensions while they have no significant effect between friendly countries. Our results confirms that BITs work as a commitment device: the host government’s credible commitment not to expropriate foreign investors is more valuable when MNEs face risks related to interstate political tensions. In addition, they suggest that BITs and good domestic institutions are complementary in attracting FDI. Overall, when the direct as well as both indirect effects are considered, the entry into force of a BIT increases bilateral FDI stocks by 42% when the quality of interstate relations and domestic institutions are at their median level. However, the magnitude and significance of this effect differ crucially according to both the political relations between the home and host countries and the quality of domestic institutions. Acknowledgments We thank two anonymous referees, James Anderson, Paola Conconi, Liza Jabbour, Keith Head, Farid Toubal and Ian Wooton for their helpful comments. A previous version of the paper, entitled ‘‘Being nice makes you attractive: the FDI – international political relations nexus”, circulated and was presented at the API conference in Beirut (2005), the AFSE congress in Paris (2006) and the RIEF conference in Roma (2007). We also thank Thierry Mayer and Antoine Berthou for providing data. Vincent Vicard gratefully acknowledges the financial support from ANR. This paper represents the views of the authors and should not be interpreted as reflecting those of Banque de France. Appendix A. Events and corresponding weights in Goldstein scale Definition

Goldstein

Definition

Extend military aid Extend humanitarian aid Rally support Extend economic aid Improve relations Promise material support Promise economic support Promise military support Promise humanitarian support Agree Collaborate Promise Promise policy support Endorse Forgive Praise Empathize Solicit support Ask for material aid Agree or accept Ease sanctions Host a meeting Assure Extend invitation Grant Provide shelter

8.3 7.6 7.6 7.4 5.4 5.2 5.2 5.2 5.2 4.8 4.8 4.7 4.5 3.5 3.5 3.4 3.4 3.4 3.4 3 2.9 2.8 2.8 2.5 2.2 2.2

Comment Decline comment Pessimistic comment Ask for protection Deny Grant asylum Criticize or blame Reduce routine activity Complain Informally complain Formally complain Accuse Warn Alerts Denounce or denigrate Halt negotiations Reject Reject proposal Refuse to allow Defy norms Impose curfew Censor media Veto Political flight Disclose information Break law

Goldstein 0.1 0.1 0.1 0.1 1 1.1 2.2 2.2 2.4 2.4 2.4 2.8 3 3 3.4 3.8 4 4 4 4 4 4 4 4 4 4 (continued on next page)

384

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

Appendix A (continued) Definition

Goldstein

Definition

Goldstein

Evacuate victims Observe truce Relax censorship Relax administrative sanction Demobilize armed forces Relax curfew Apologize Acknowledge responsibility Travel to meet Release or return Request Ask for economic aid Ask for military aid Ask for humanitarian aid Consult Offer peace proposal Call for action Yield Discussions Propose Yield to order Yield position Optimistic comment Ask for information Animal incidents Economic activity Other human action Human illness Human death Economic status Other human condition Natural disaster Accident Other incident Animal attack Animal death Animal illness Other animal incident Arts and entertainment performance Sports contest Transactions Government transactions Private transactions Government default on payments Default on payment Elect representative Administrative adjustment Non-governmental adjustment Judicial actions Infectious human illness Non-infectious human illness Currency reserves Exchange rates Equity prices Debt yields Commodity prices Affective state

2.2 2.2 2.2 2.2 2.2 2.2 2.2 2 1.9 1.9 1.6 1.6 1.6 1.6 1.5 1.5 1.2 1.1 1 0.8 0.6 0.6 0.1 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Non-specific threats Arrest and detention Political arrests and detention Criminal arrests and detention Administrative sanctions Sanction Strikes and boycotts Demand Expel Protest demonstrations Protest obstruction Protest procession Protest defacement Reduce or stop aid Sanctions threat Threaten Non-military force threats Seize Police seizure Other seizure Carjacking Hostage taking and kidnapping Control crowds Demonstrate Give ultimatum Protest altruism Military force threats Break relations Threaten military attack Threaten military blockade Threaten military occupation Threaten military war Military clash Threaten nuclear attack Military alert Military air display Military naval display Military troops display Military demonstration Military mobilization Military border fortification Riot or political turmoil Bombings Seize possession Abduction and hijacking Military seizure Military occupation Military border violation Force Physical assault Beatings Shooting Bodily punishment Sexual assault Torture Assassination Military engagements

4.4 4.4 4.4 4.4 4.5 4.5 4.5 4.9 5 5.2 5.2 5.2 5.2 5.6 5.8 6.4 6.4 6.8 6.8 6.8 6.8 6.8 6.9 6.9 6.9 6.9 7 7 7 7 7 7 7 7 7.6 7.6 7.6 7.6 7.6 7.6 7.6 8.3 8.7 9.2 9.2 9.2 9.2 9.2 9.6 9.6 9.6 9.6 9.6 9.6 9.6 9.6 10

385

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

Appendix A (continued) Definition

Goldstein

Definition

Goldstein

Beliefs and values Drought Earthquake Flood Hurricane Tornado Volcano Tsunami Wildfire Hazardous material spill Private default on payments

0 0 0 0 0 0 0 0 0 0 0

Military raid Coups and mutinies CBR weapons use Grenade/RPG use Suicide bombing Mine explosion Vehicle bombing Chemical weapons use

10 10 10 10 10 10 10 10

Source: King and Lowe (2002).

References Anderson, J.E., van Wincoop, E., 2003. Gravity with gravitas: a solution to the border puzzle. American Economic Review 93 (1), 170–192. Baier, S.L., Bergstrand, J.H., 2007. Do free trade agreements actually increase members’ international trade? Journal of International Economics 71 (1), 72– 95. Benassy-Quere, A., Coupet, M., Mayer, T., 2007. Institutional determinants of foreign direct investment. The World Economy 30 (5), 764–782. Bergstrand, J.H., Egger, P., 2008. A knowledge-and-physical-capital model of international trade flows, foreign direct investment, and multinational enterprises. Journal of International Economics 73 (2), 278–308. Berkowitz, D., Moenius, J., Pistor, K., 2006. Trade, law and product complexity. Review of Economics and Statistics 88 (2), 363–373. Blonigen, B.A., Davies, R.B., Waddell, G.R., Naughton, H.T., 2007. Fdi in space: Spatial autoregressive relationships in foreign direct investment. European Economic Review 51, 1303–1325. Boehmer, C., A, G.E., Li, Q., 2001. Investing in the peace: economic interdependence and international conflict. International Organization 55 (2), 391–438. Bond, D., Bond, J., Oh, C., Jenkins, J.C., Taylor, C.L., 2003. Integrated data for events analysis (idea): an event typology for automated events data development. Journal of Peace Research 40 (6), 733–745. Brambor, T., Clark, W.R., Golder, M., 2006. Understanding interaction models: improving empirical analyses. Political Analysis 14, 63–82. Bubb, R.J., Rose-Ackerman, S., 2007. Bits and bargains: strategic aspects of bilateral and multilateral regulation of foreign investment. International Review of Law and Economics 27 (3), 291–311. Djankov, S., Glaeser, E., LaPorta, R., Lopez-de Silanes, F., Shleifer, A., 2003. The new comparative economics. Journal of Comparative Economics 31 (4), 595– 619. Egger, P., Pfaffermayr, M., 2004. The impact of bilateral investment treaties on foreign direct investment. Journal of Comparative Economics 32 (4), 788–804. Elkins, Z., Guzman, A.T., Simmons, B., 2006. Competing for capital: the diffusion of bilateral investment treaties: an empirical analysis. International Organization 60, 811–846. Feenstra, R.C., 2004. Advanced International Trade: Theory and Evidence. Princeton University Press, Princeton. Globerman, S., Shapiro, D., 2002. Global foreign direct investment flows: the role of governance infrastructure. World Development 30 (11), 1899–1919. Globerman, S., Shapiro, D., 2003. Governance infrastructure and US foreign direct investment. Journal of International Business Studies 34 (1), 19–39. Goldstein, J.S., 1992. A conflict-cooperation scale for WEIS international events data. Journal of Conflict Resolution 36 (2), 369–385. Guzman, A.T., 1998. Explaining the popularity of bilateral investment treaties: why LDCs sign treaties that hurt them. Virginia Journal of International Law 38, 639–688. Hallward-Driemeier, M., 2003. Do bilateral investment treaties attract foreign direct investment? only a bit. . . and they could bite, World Bank Working Paper, No. 3121. Hansen, L., 1982. Large sample properties of generalizes method of moments estimators. Econometrica 50 (3), 1029–1054. Head, K., Ries, J., 2008. FDI as an outcome of the market for corporate control: theory and evidence. Journal of International Economics 74 (1), 2–20. King, G., Lowe, W., 2002. An automated information extraction tool for international conflict data with performance as good as human coders: a rare events evaluation design. International Organization 57 (summer), 617–642. Maskus, K.E., 2000. Intellectual Property Rights in the Global Economy. Institute for International Economics, Washington, DC. Neumayer, E., Spess, L., 2005. Do bilateral investment treaties increase foreign direct investment to developing countries? World Economy 33 (10), 1567– 1585. Nigh, D., 1985. The effect of political events on united states direct foreign investment: a pooled time-series cross-sectional analysis. Journal of International Business Studies 16 (1), 1–17. Oneal, J.R., Russett, B.M., 1997. The classical liberals were right: democracy, interdependence and conflict, 1950–1985. International Studies Quarterly 41 (2), 267–294. Oneal, J.R., Russett, B.M., 1999. Assessing the liberal peace with alternative specifications: trade still reduce conflicts. Journal of Peace Research 36 (4), 423– 442. Polachek, S.W., 1980. Conflict and trade. Journal of Conflict Resolution 24 (1), 57–78. Razin, A., Sadka, E., 2007. Corporate transparency, cream-skimming and FDI. European Economic Review 51 (5), 1263–1276. Rodrik, D., 1991. Policy uncertainty and private investment in developing countries. Journal of Development Economics 36 (2), 229–242. Rose-Ackerman, S., Tobin, J., 2005. Foreign direct investment and the business environment in developing countries: the impact of bilateral investment treaties, yale Law & Economics Research Paper, No. 293. Santos Silva, J., Tenreyro, S., 2006. The log of gravity. Review of Economics and Statistics 88 (4), 641–658. Schelling, T.C., 1960. The Strategy of Conflict. Oxford University Press, Oxford, UK and New York. Servén, L., 1997. Irreversibility, uncertainty and private investment: analytical issues and some lessons for africa. Journal of African Economies 6 (3), 229– 268. Stock, J.H., Wright, J.H., Yogo, M., 2002. A survey of weak instruments and weak identification in generalized method of moments. Journal of Business and Economic Statistics 20 (4), 518–529. Transparency International, 2002. Bribe Payers Index 2002: Explanatory Notes and Comparative Tables. .

386

R. Desbordes, V. Vicard / Journal of Comparative Economics 37 (2009) 372–386

UNCTAD, 2000. Bilateral Investment Treaties 1959–1989. United Nations, New York. UNCTAD, 2008. Latest Developments in Investor-state Dispute Settlement. IIA MONITOR 1. Wei, S.-J., 2000. How taxing is corruption international investors? Review of Economics and Statistics 82 (1), 1–11. Wolf, M., 2005. Why Globalization Works. Yale University Press, New Haven and London. Wooldridge, J., 1999. Distribution-free estimation of some nonlinear panel data models. Journal of Econonmetrics 90 (1), 77–97.