OUP user menu

Legal System Pathways to Foreign Direct Investment in the Developing World

Hoon Lee, Glen Biglaiser, Joseph L. Staats
DOI: http://dx.doi.org/10.1111/fpa.12026 393-411 First published online: 1 October 2014

Abstract

Building on recent works showing the role that legal institutions can play in attracting foreign capital (Jensen 2003, 2006; Li and Resnick 2003; Li 2006; Biglaiser and Staats 2010; Staats and Biglaiser 2012), and drawing on insights obtained from Powell and Rickard (2010), we use panel data for 114 developing countries from 1970 to 2007 to demonstrate that developing countries with common law legal systems attract greater foreign direct investment (FDI) than countries that have civil law or Islamic legal systems because common law systems are more inclined to promote the rule of law and protect property rights and can be understood to provide more efficiency in the law, better contract enforcement, more judicial autonomy, and more market-oriented regulations.

The surge of foreign direct investment (FDI) coming into developing countries, from below $11 billion in 1970 to nearly $515 billion in 2006 (World Bank 2008), has led political economy scholars to explore the determinants of FDI, often focusing on political risk. Many studies investigate broad institutional factors, most specifically whether it matters in terms of risk that the governing regime is democratic or authoritarian (Olson 1993; Feng 2001; Jensen 2003, 2006, 2008; Li 2006, 2009a). Still other studies consider the particular political institutions, processes, and alignments that attract investors by either lowering risk (Li and Resnick 2003; Biglaiser, DeRouen 2006; Biglaiser and Staats 2010; Jensen and Johnston 2011; Staats and Biglaiser 2012); providing a climate of receptivity to foreign investment, including financial inducements (Li 2006; Pinto and Pinto 2008; Jensen 2012); or causing some combination of risk lowering and receptivity (Jensen et al. 2012).

We follow the lead of prior works by seeking to demonstrate that the type of legal system a country has, whether common law, civil law, Islamic, or some mixture, is an important factor in attracting FDI. Our research builds on prior works suggesting a link between international economic transactions and legal institutions, including Jensen (2003, 2006), Li and Resnick (2003), Li (2006), Biglaiser and Staats (2010), Staats and Biglaiser (2012), Globerman and Shapiro (2003), and Powell and Rickard (2010). Our aim is to provide a deeper appreciation of the pathways in the law that have helped or hindered in drawing in foreign investment. Based on regression analysis of data over a period of 1970–2007 covering 114 developing countries, 1 we find that developing countries with common law systems provide greater adherence to the rule of law and a better risk environment for investors, thereby attracting more FDI than countries with civil law or Islamic systems.

Our results hold important implications for both theory and policy. First, the approach presented here delineates how specific legal institutional arrangements affect FDI. Second, our project provides a platform for further investigation on the role that legal systems have in attracting foreign capital. Third, our study has implications for countries that seek to draw in FDI. Our findings suggest policymakers need to pay closer attention to building strong and independent judicial systems and greater adherence to the rule of law if they hope to attract interest from foreign investors.

In the first section, we address the importance of legal factors in attracting FDI. In the second section, we discuss the pathway by which a common law system has positive effects on FDI, namely by being more successful at adhering to the rule of law and thereby lessening investor risk. We also advance three testable hypotheses based on our proposition that common law systems are better able to attract FDI. Section three presents the data and methods used in our analysis. The fourth section reports the findings and the fifth section concludes the work.

Legal Factors and FDI in Developing Countries

Although the effects of differences in legal systems on FDI have received limited attention from scholars, there are ample reasons to expect a link. First, there are FDI studies suggesting that property rights protection is of importance to foreign investors (Feng 2001; Li and Resnick 2003; Biglaiser et al. 2006; Jensen 2008; Li 2009a), and strong and effective courts can assure investors that a host country government will respect their property rights. In addition, robust and independent courts tend to constrain arbitrary and abrupt government decisions that cause private property to diminish in value (Landes and Posner 1975, 882; North and Weingast 1989, 819; Brunetti and Weder 1999, 2). The presence of strong and effective courts will not always dissuade governments from actions detrimental to business interests, but it will require them to follow established procedures, which slows the process down and may in many cases result in government not doing what it might otherwise do in the absence of constraints. This gives investors greater certainty when making investment decisions.

Effective courts provide a convenient and relatively low-cost venue for investors to protect property titles and enforce contracts in commercial transactions (North 1990:34, 58; Olson 1993:572). While foreign investors often insist on arbitration clauses in their contracts, arbitration is not always an effective substitute for reliable courts. Arbitration adds a layer of costs to commercial transactions (Dezalay and Garth 1995:61), and an aggrieved party often has to ask for court order to have a matter referred to arbitration and the prevailing party after arbitration must go to court to have an award enforced.

Foreign investors also want to earn profits on their investments and thus are expected to take into account whether the economy is on a path for growth and prosperity. The extent to which an economy expands depends in part on whether the government can give credible commitments to all economic actors that it will uphold the rule of law and protect property rights (North and Weingast 1989:803). Unless restrained in some way, current government policies facilitating economic growth may be short-lived, as when existing or future politicians change course for personal or political gain that harms economic vitality (Brunetti, Kisunko and Weder 1997:10–11). Added to concerns about corruption is the possibility that government will impose high business taxes to finance social welfare initiatives or strictly regulate business operations to satisfy populist demands. Government brings credibility to its commitments by delegating to courts the authority to force current and future governments to conduct themselves in accordance with the law (Landes and Posner 1975:882; North and Weingast 1989:819; Brunetti and Weder 1999:2). As we said before, there is no guarantee that current policies will persist, but courts can require government to follow established legal procedures when it wants to do something with an economic effect (Zywicki 2002:4), which might require advance notice, public hearings, formal findings, legislation, or even amendment of the constitution. This court-enforced “constitutionalization” of rights (see, e.g., Schneiderman 2001:521; Hirschl 2004:146–147) tends to slow things down and supports investor confidence in the stability of policies (Jensen 2003, 2006, 2008; Li 2006, 2009a,b).

Strong courts and adherence to the rule of law also contribute to political stability, a concern for foreign investors. Two theoretical strands are relevant here: insurance policy theory and a variation of the credible commitment theory previously discussed. Insurance policy theory suggests that under competitive conditions, a governing party must contemplate the possibility of someday being out of power and therefore is motivated to give up prerogatives while in office by delegating powers to the courts as the price to pay to avoid having seriously diminished rights in the future when out of power (Ginsburg 2003:73, 78; Chavez 2004). Credible commitment theory, of the sort discussed here, suggests an agreement between all political participants, backed by the courts, to continue playing the democratic game rather than defect to extra-legal regime change and/or potentially unstable authoritarian rule (Ginsburg 2003:73, 78). Thus, strong courts contribute to political stability, which, in turn, provides a climate of predictability for investors.

Common Law Advantages over Other Types of Legal Systems

In the previous section, we argued that countries with better-performing legal systems had an advantage for attracting FDI. In this section, we turn our attention to the theoretical reasons why common law legal systems are better positioned than civil law or Islamic systems to provide foreign investors greater confidence about the safety of their investments.

We begin by looking at the historical differences between common law and its main competitor in democratic regimes, civil law. The Roman code system, from which the civil law system traces its origins, spread throughout Europe during the heyday of the Roman Empire but fell into disuse for a time after the Empire's demise. In the eleventh and thirteenth centuries, European legal scholars revived and revised the Roman codes (Glenn 2010:139–140) and ultimately revised forms of the codes became incorporated into French, German, and Italian law, a process that profoundly affected the legal systems of Europe and many of its colonies (David and Brierley 1985; Powell and Mitchell 2007:398). Noteworthy about the role of judges in the civil law system is that they are expected to be passive receivers of law expressed in the codes, not instruments for making or changing law (Powell and Mitchell 2007:399).

The common law had an accidental birth coinciding with the Norman conquest of England. Surrounded by strange languages and animosities of conquered peoples, the Normans needed a new legal system to be simultaneously responsive to Norman interests and local needs (Glenn 2010:238). Under this new legal order, judges were expected to take on roles well beyond the passive application of given law. As the new rulers of England, the Norman conquerors “needed a corps of loyal adjudicators, able to bring a newer, more efficient and modern king's peace to the different parts of the realm” (Glenn 2010:239). There arose in these demanding circumstances a doctrine favoring judge-made legal rules and reliance on case precedent (that is, stare decisis).

We contend that common law legal rules and reliance on case precedent have the potential to lessen risk in economic transactions, thereby providing a climate favorable for FDI. Hayek (1960, 1973) is a good place to start in understanding why the common law is most able to lessen economic risk and increase capital flows. Hayek maintains that common law systems have less government restrictions on economic liberties and give stronger protection to property rights, as opposed to civil law systems that expand the power of the state and limit the jurisdiction of judges to protect property rights and assure enforcement of contracts. The tendency of the civil law to have lesser concern for property rights traces most notably to the period during and immediately after the French revolution. The situation in France was quite different from that in England. Landed aristocrats and merchants in England wanted a system of law that would provide strong protection for property and contract rights and limit the Crown's ability to interfere in markets. Revolutionary France toppled the landed aristocracy of the ancien régime and the revolutionary generation, and Napoleon after it, wanted to alter property rights and ensure that judges could not interfere (Mahoney 2001:504–505). This set the stage for civil law systems, whether in France or elsewhere, to be less solicitous to property rights than was the case under the common law.

An understanding of why even today common law judges are better positioned than judges in civil law systems to constrain state power can be obtained by comparing fundamental differences between the two types of systems and the jurists who work in them. Appointment or election of common law judges “comes as a form of crowning achievement relatively late in life…[,] a form of recognition that brings respect and prestige[, and their] opinions will be discussed in the newspapers and dissected and criticized in the legal periodicals[, as they] are very important people” (Merryman and Pérez-Perdomo 2007:34–35). Civil law judges assume their judicial positions shortly after graduating from law school and passing a competitive exam, and work their way up the judiciary ladder by some combination of ability and time in service. Intelligence and hard work pay off for civil law judges. In Japan, for example, the better the work of an individual judge, the more attractive will be the job that the system gives him (Ramseyer and Rasmusen 1997:261, 277). Civil law judges, as bureaucrats of the law, are respected as government officials but are not accorded any high order of prestige. 2 They are representatives of the state apparatus rather than being independent sources of power positioned to protect individuals and property from state-instigated encroachments.

Civil law, by making distinctions between individual rights in matters of public law as opposed to private law, imposes fewer restraints on public officials (David and Brierley 1985) and provides greater power and opportunity for the state to use the legal system for its own political ends (La Porta, Lopez-de-Silanes, Shleifer and Vishny 1999; Levine 2005). Acting as independent agents, they are able to check extralegal and excessive use of power by the political branches of government. This independent exercise of judicial power leads to greater protection of economic rights generally and especially protection from government predation (La Porta, Lopez-de-Silanes, Pop-Eleches and Shleifer 2004:447, 449; La Porta, Lopez-de-Silanes and Shleifer 2008:286, 293). What we are talking about here is the relative lack of judicial autonomy imbedded in the civil law, not ultra vires manipulation of the judicial sector. Politicians need not intrude on the judicial sector to get what they want, because the judicial sector more often than not can be counted on to give them what they want (Ramseyer and Rasmusen 1997:260). Politicians get what they want because of control over individual judges coming from within the judiciary itself. To judges in civil law Japan, judicial independence means that judges decide cases independently of direct control from the executive and legislative branches, but not that they decide cases independently of other judges within the judiciary (Ramseyer and Rosenbluth 1997:163). The judicial apparatus is structured so that individual judges who stray from the norm expected by political leaders can expect to incur penalties in terms of promotions and assignments (Ramseyer and Rosenbluth 1997:161–179).

Another reason to suppose that common law systems are better at attracting FDI is that they provide more certainty to investors than do civil law systems. They do so through the promulgation of legal rules that become embedded in the law through the doctrine of stare decisis. Except when first promulgated through court decision, common law legal rules are set ex ante rather than ex post. Common law rules, arising from case law and published as part of the decision of the case, are widely known and understood and followed. This provides a level of certainty. Although legislatures enact codes ex ante, there is ambiguity in what the codes mean without benefit of judicial interpretation. No legislation is exact, no matter how hard the legislature may try to make it so. Courts step in to make the law more exact through interpretation. However, in civil law systems, judicial interpretation of the laws does not get regularized because these systems do not generally recognize the doctrine of stare decisis because of a general distrust of judges (Mahoney 2001:504–505; Merryman and Pérez-Perdomo 2007:480). While certainty in the law is the norm for common law judges (Merryman and Pérez-Perdomo 2007:480), each judge in a civil law system is free to interpret codes in the manner that his or her understanding and conscience dictates, unguided by any higher judicial authority. The result is uncertainty (Zywicki 2002:15).

Of course, no legal system enables parties to know for certain what the outcome of litigation will be. Stare decisis constrains judges as they conduct a trial or make a decision in a case, but judges have flexibility nonetheless. Moreover, we know from realist scholarship that judges, especially those on high courts, will often abandon legal accuracy as a goal and search after case precedent to justify a decision that conforms to what they believe the law ought to be (Segal and Spaeth 1993:2002; Epstein and Knight 1998; Baum 1999). Nevertheless, we argue that application of stare decisis provides relative certitude in the law, especially as compared to case-by-case interpretation of law as engaged in by civil law courts. We are careful, however, not to overstate the differences between the two systems. High courts in civil law systems these days tend to publish their opinions on cases, and constitutional courts, in countries that have them, certainly do so. But in civil law systems, because of the general distrust of judge-made law, there is no systematic use of case law to decide cases. Some judges, at some times, may use prior cases informally to decide a current case “because they are impressed by the authority of the prior court, because they are persuaded by its reasoning, because they are too lazy to think the problem through themselves, because they do not want to risk reversal on appeal, or for a variety of other reasons” (Merryman and Pérez-Perdomo 2007:47). Such informal resort to prior cases is hardly the equivalent to that which goes on in common law systems.

A notable exception to the proposition that civil law judges do not apply prior case law involves rulings by constitutional courts. Ordinary civil law courts may be asked to apply a law that was previously found unconstitutional by the constitutional court. In such a circumstance, the ruling by the constitutional court would be supreme and “given erga omnes effect, meaning they are binding for all future cases” (Ginsburg 2003:40). However, this doctrine has limited consequences for investors, since they are most affected by the application of ordinary law in cases where constitutionality is not questioned.

A final reason to suppose that common law systems more effectively attract FDI involves what some argue is the greater efficiency of common law systems (Zywicki and Stringham 2010). Common law judges have leeway to apply pragmatic and justice-oriented legal rules to fit the facts in individual cases to existing law. While this may seem inconsistent with the argument that the common law brings certainty to the law, we are not suggesting that common law judges have unlimited flexibility in this regard. Rather, common law judges have an arsenal of established equity rules that they can apply to insure that a case outcome is fair and just. These equity rules arose originally in the so-called English equity courts to counteract harsh outcomes brought on by strict application of common law rules and are now firmly embedded in the common law. Under rules of equity, a judge can mitigate what would otherwise be an unjust result from strict application of a statute, or can allocate property or responsibility according to the facts of an individual case in order to bring fairness. Rules of equity recognize that some problems are so complex as to make it impossible for the legislature to dictate consequences of all possible combinations of the facts in any given case (Merryman and Pérez-Perdomo 2007:45).

In contrast, civil law systems require judges to cram the facts of cases into existing law whether or not there is a good fit between facts and law, and this is so whatever the economic consequences and no matter whether this leads to fair and efficient administration of justice (Merryman and Pérez-Perdomo 2007:52). Efficiency in administering law finds a home in the common law by virtue of a decentralized process of “many judges and litigants resolving many cases in concrete factual disputes that arise from particular conflicts, rather than a judge essentially articulating a rule for the economy” (Zywicki and Peter Stringham 2010:32). The difference between the two types of systems in this regard is similar to that between a command and market economy (Zywicki and Peter Stringham 2010:32). Common law judges operate in what can best be described as a “legal market,” wherein individuals and entities need and want clearing and closure of disputes and society wants regularity, stability, and safety in social relations. In the area of contract law, for example, the goal of the common law is to help parties accomplish their joint purposes. The common law achieves this goal by providing certainty of the promised performance (or just compensation on failure of performance) and does so by adopting conventions that most parties themselves would adopt, most of the time especially with respect to unforeseen contingencies that may have led to a dispute in the first place (Aronson 1992:307).

Common law systems provide mechanisms for better enforcement of contracts, a source of concern for investors in foreign venues (La Porta et al. 1999, 2004; Djankov, La Porta, Lopez-de-Silanes and Shleifer 2003; Nunn 2007). Common law systems also provide opportunities for markets to flourish and for lower levels of government regulation, also matters that foreign investors should find attractive. “Essentially, the toolkit of civil law features more prominently such policies as nationalization and direct state control; the toolkit of common law features more litigation and market-supporting regulation” (La Porta et al. 2008:308).

We turn now to a comparison of the common law system to that of Islamic law. The position of courts and of legal procedure in Islamic law is closely connected to the historical development of Islam in the seventh century. Islamic law at its inception did not provide for separation between judicial and executive powers (David and Brierley 1985; Glenn 2007). The Prophet Mohammed exercised judicial, legislative, and executive powers, and after the Prophet's death, ultimate political power, including judicial, resided in the ruling Caliphs. Although some powers to administer justice in accordance with Islamic law were delegated to specialized judges, the Caliph, as the supreme ruler of Islam, had the final right of review (Sherif and Brown 2002:4–5). As discussed earlier, independent courts provide a check on governmental authority that might otherwise be wielded unfairly and detrimentally to investor interests. Any legal system that blurs the line between executive and judicial authority cannot be relied upon to constrain government actions, and future government actions and policies are less predictable. There is no “credible commitment” from government that investors can (more or less) rely upon. Islamic law countries are regarded to have weaker protection of property rights, principally due to the “vertical bonds of authority” in Islamic countries (Landes 1998) and the continuation of weak separation of powers (Sherif and Brown 2002; Lau 2004; Glenn 2007). Islamic law systems also distinguish between public and private, which is to say that public authority is given special recognition not accorded to matters of private concern (Sherif and Brown 2002; Lau 2004). Islamic legal systems tend to downplay the importance of individual property rights inasmuch as “[a]bsolute ownership of property is seen as vested ultimately in God” (Glenn 2007:182).

The effect of Islamic law traditions on modern economic transactions is investigated by Kuran (2011), who puzzles over the fact that Western-style reforms have been instituted in Islamic law countries, yet they remain, for the most part, mired in economic backwardness. He answers this puzzle by suggesting that a course of path dependence, a persistence of historical influences, in three parts. First, “modern organizational forms have been transplanted into societies with social norms inimical to their efficient use: relatively high corruption and nepotism, and low trust in organizations.” These norms, he says, are among the legacies of traditional Islamic law. Second, the “weaknesses of private sectors and civil societies, which are rooted in the region's institutional history, breed complacency toward autocratic rule.” Third, “[e]conomic failures have created fertile ground for (i) inward-looking ideologies that limit adaptation to changing global realities; and (ii) Islamism, which foster political uncertainty and limits experimentation in certain areas” (Kuran 2011:294). The persistence of institutions that foster economic backwardness is hardly what foreign investors want to see when contemplating where to put their resources.

Also of concern for investors is the possibility for political instability in countries that adhere to the Islamic law tradition. A system where courts are subordinate to the political branches, as is true in Islamic law systems, may not secure the delicate balance necessary to insure political stability.

The foregoing discussion of the theoretical literature suggests two pathways through which common law systems better attract FDI than any of the other legal systems. 3 The first pathway is direct, meaning that investors are expected to know that common law countries provide better enforcement of contracts, provide a climate for more market-based regulations and policies, and impose more systematic restraint on the political branches and government bureaucracy. Common law systems also decide cases in ways that are more economically efficient and effective for investment. The other pathway is indirect, and here we say that common law systems are more effective at assuring adherence to the rule of law and protection of property rights, things of importance to investors. Common law courts have the autonomy and political clout to force government to follow established constitutional provisions, relevant laws, and established norms of conduct. We therefore expect to find that common law systems have more impact than the other systems on FDI, but also that they have more of an impact on higher levels of rule of law and protection of property rights, and that higher rule of law and protection of property rights in turn lead to increased levels of FDI. We argue that common law systems provide more effective and efficient ways for the rule of law to flourish and that this attracts investors who want to lessen risk and enhance profits.

We should mention, however, that there are contrarian views on the supposed superiority of common law systems. Hathaway (2003:164, fn. 234), for one, suggests that the doctrine of stare decisis may cause common law systems to become locked into suboptimal legal rules. 4 Another area of concern is that the worldwide constitutionalization of rights in recent years has increased the power and prestige of civil law judges, thereby arguably lessening some distinctions between common law and civil law judges previously mentioned (Stone Sweet 2000:114–115; Merryman and Pérez-Perdomo 2007:158). But it is mainly judges on constitutional courts who have benefitted from this trend, not ordinary law judges where most legal decision making takes place (Merryman and Pérez-Perdomo 2007:158). Our answer to those who question whether common law systems are superior to other types of systems is to put the issue to empirical test, which we will do presently.

Based on the previous theoretical discussion, we propose three hypotheses suggesting that developing countries operating common law systems attract more FDI than developing countries that use other legal systems, and that this is so largely because they provide better adherence to the rule of law, which foreign investors favor. Hypothesis 1: Countries with common law legal systems will have higher levels of FDI inflows than countries with civil law, Islamic, or mixed legal systems. Hypothesis 2: Countries with common law legal systems will have greater adherence to the rule of law than countries with civil law, Islamic, or mixed legal systems. Hypothesis 3: The greater the adherence to the rule of law in a country, the higher will be the levels of FDI inflows.

Research Design and Methods

Utilizing country-level time-series panel data, we use three stages to test our hypotheses on the effects of legal systems on FDI inflows. The first stage investigates the impact of type of legal systems on FDI. Our second stage tests the factors that support respect for the rule of law, helping to understand the connection between type of legal systems and the rule of law. The third stage investigates the determinants of FDI, but this time we substitute the rule of law variable for legal system measures, completing the circle linking legal systems, rule of law, and FDI inflows.

We start by testing Hypothesis 1, whether common law systems are better at attracting FDI than other legal systems. For our dependent variable, we use net FDI inflows as a percentage of GDP 5 for 114 developing countries from 1970–2007. FDI data come from the World Bank (2008). For testing Hypothesis 1, as well as Hypothesis 3, we also lag the dependent variable and place it on the right-hand side because previous FDI may affect future FDI inflows. For the independent variables of main interest – the different types of legal systems – we employ 0–1 dummy variables labeled Civil, Islamic, or Mixed to indicate whether a country has a legal system other than common law. If a country does, then we assign a 1 for the appropriate dummy variable and 0 for the remaining dummy variables (that is, a common law country receives a zero on all dummy variables). Thus, negative coefficients on the dummy variables indicate that common law has a greater effect on the dependent variable than the type of legal system with which it is being compared. Legal systems data are from Powell and Mitchell (2007). 6

We also consider other variables that are commonly used in FDI studies, grouped into the three categories: macroeconomic conditions and natural resource abundance, economic reforms, and political factors. Among macroeconomic conditions and resource endowments, we employ GDP per capita for our Economic Development variable, GDP growth for Economic Growth, and natural resource abundance for Natural Resource Endowment. 7 The expectation is that strong macroeconomic conditions and natural resource abundance offer opportunities for investors to receive a high return on their monies. We use data from the World Bank (2008) for GDP growth and GDP per capita (both in constant 2000 dollars). For natural resource endowment, we follow work by Epstein et al. (2006) and Biglaiser and Staats (2010) who identify countries that have more than 75% of national income derived from sales of minerals or petroleum as resource rich, and countries with less than 75% are designated as less resource abundant.

Among the economic reforms, we include variables labeled Trade and Financial Openness and one representing the signing of bilateral investment treaties, which we label BITs. Investors presumably see higher trade levels as facilitating the outsourcing of products for re-export to the host country. Similarly, investors are likely to prefer investing in countries that have fewer capital flow restrictions and that have signed bilateral investment agreements with the United States because these factors enhance capital safety. 8 For the trade variable, we use OPENK ((exports + imports)/Real GDP) data (in constant 2005 dollars) from the Penn World Tables (2010). To assess financial openness, we adopt a four-point measure from Chinn and Ito (2008), where a higher score indicates fewer controls on capital by the host country. For BITs, we consider whether the country has signed a BIT (or not) with the US, the largest country with foreign holdings globally. 9 The data come from UNCTAD (2011).

Lastly, we include five factors to control for the effects that political institutions bring to the equation: Democracy, Political Stability, Executive Ideology, and Conflict. A large debate in the FDI literature identifies the role of regime type for lessening investor risk (Jensen 2003, 2006; Li and Resnick 2003; Tuman and Emmert 2004). We use Polity IV to measure Democracy, rescaled from a 10-point plus and minus scale to 0–20, after which we created dummy variable scores of 0 for countries scoring less than 16, representing nondemocracies, and 1 for scores of 16 or more, representing democracies (Marshall and Jaggers 2006). 10 We assess political stability by following Cheibub et al. (2009), measuring the number of years the current regime has operated under the same political institutions. Executive ideology is predicted to affect FDI inflows, with leftist governments presumed to be less supportive of foreign investors because FDI may cause job losses for leftist constituents (Geddes 1994). Executive ideology comes from Beck et al. (2002) and is coded 1 for leftist executives and 0 otherwise. We expect investors to stay clear of countries beset by conflict because of instability concerns. For this, we use data from Marshall (2009), which employs a dummy measure (1 if there is conflict; 0 if there is no conflict). Finally, we created a dummy variable labeled Post-Cold War (1 if after 1989, 0 if before 1990) to account for the expansion in capital markets following the collapse of the Soviet Union and its satellite countries. We lead the dependent variable a year in relation to all other variables as changes in the economic and political climate take time to have consequences to investors.

Because the literature on legal systems suggests that common law and civil law systems are becoming more similar to each other over recent years (see a discussion in Mattei and Pes 2008 , 11 ), we include a Distance Years variable that indicates for each year the number of years it is from the end point of 2007. Following Powell and Rickard (2010), we also include interaction terms created by multiplying each of the legal system dummy variables by Distance Years. This allows us to account for any convergence over time between common law and other legal systems and for the possibility that increased resort to private arbitration in international transactions may make distinctions between types of legal systems and foreign investment less relevant over time (see Dezalay and Garth 1995:30; Powell and Rickard 2010:351–352). Both of these trends, if actual and substantial, could mitigate the effects of the type of legal system on FDI over time.

Hypothesis 2 concerns the effect of common law systems on rule of law as an intermediary pathway by which common law affects FDI. For our rule of law-dependent variable, we call upon the widely utilized Worldwide Governance Indicators (WGI; Kaufmann, Kraay, and Mastruzzi 2009). The WGI rule of law measure, which is drawn from 35 separate data sources, is defined as the “extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence” (Kaufmann, Kraay, and Mastruzzi 2009:2). 12 Yearly coverage of this measure begins in 1996, so we start here and continue through 2007. 13 The rule of law measure is scaled from −2.5 to 2.5, with higher values indicating better rule of law outcomes. Many of the same factors seen as determinants of FDI also are included to explain the rule of law except for a few economic and political variables that do not appear relevant whether a country adheres to the rule of law. We also exclude the Post-Cold War variable since the rule of law measure does not provide coverage for the Cold War years.

Hypothesis 3 proposes that respect for the rule of law impacts FDI, squaring the circle on the effects of different types of legal system on capital inflows. We test this hypothesis in stage three of our models and use the rule of law measure as an independent variable in place of the legal system dummies used in stage one. We include all other variables used in stage one only excluding the Post-Cold War time variable for coverage reasons. We ran correlations of all variables and conducted a variable inflation factor (VIF) test on all data in all models to determine collinearity, the results of which indicated no cause for concern.

Because of factors unique to our hypotheses, we use different estimating procedures to test each hypothesis. For Hypotheses 1 and 3, we include models based on time-series panel data. Results from Hausman tests indicate the need to control for country fixed effects in these models, which we do by including dummy country variables. Performing a Wald test (testparm in Stata), however, we found that we need not control for temporal fixed effects. To check the stationarity of our data, we ran a Fisher-type unit-root test, which is based on augmented Dickey–Fuller tests. The results of the tests strongly reject the null hypothesis that the panel contains unit roots 14 (inverse χ2, inverse normal, inverse logit t, and modified inverse χ2: p < .0000, respectively; Choi 2001). Because our independent variables of interest in Hypothesis 1 (types of legal system) do not vary over the period covered by our study, we cannot employ the types of estimating procedures normally called for when dealing with time-series panel data (that is, Driscoll–Kraay standard errors (DKSEs); OLS with panel-corrected standard errors (PCSEs)). We utilize instead two alternatives. The first is a three-stage estimation procedure introduced by Plümper and Troeger (2007) called fixed effects with vector decomposition (FEVD). 15 Because of questions raised by some about the efficacy and scope of FEVD (Beck 2011; Breusch, Ward, Nguyen and Kompas 2011; Greene 2011; Plümper and Troeger 2011), we also run in a separate model an OLS pooled regression on cross-sectional data averaged for all variables in 5-year increments across the period covered by our study. 16 This gives us breadth of temporal coverage in a way that avoids the fixed-effects problems of time-series data. For testing Hypothesis 3, our variable of interest is not invariant as is the case for Hypothesis 1, so we do our estimates using DKSEs. DKSEs not only adjust for disturbances in terms of heteroskedasticity and cross-sectional correlation, but also produce more precise statistical results than an OLS regression with PCSEs (Hoechle 2007). We might otherwise be inclined also to estimate with PCSEs for robustness purposes, but PCSEs tend to perform suboptimally when, as is the case with our data, the panel's cross-section dimension N is greater than the time dimension T (Beck and Katz 1995; Hoechle 2007). For a robustness check of our results, we employ in separate models Newey–West and OLS with fixed-effects estimations.

For Hypothesis 2, the results from a Hausman test indicate that we need not model for fixed effects. Therefore, FEVD is not appropriate. For testing this hypothesis, we employ the 5-year average OLS method used as an alternative in testing Hypothesis 1.

Findings

We report the results of our models testing the first hypothesis, type of legal system on FDI, in Table 1. Common law system is the excluded legal system variable in the models (dummy variable = 0), meaning that the results of our analysis measure the common law system against each of the other systems. Thus, a negative coefficient for one of the other legal systems indicates that the common law system is better at attracting FDI. As seen in Model 1, using FEVD, and Model 2, employing OLS with 5-year average, common law systems attract higher levels of FDI relative to the other types of legal systems, in each case at p < .001. Using Model 1 as the standard, we calculate the substantive importance of our results and find that common law systems attract 47% more FDI than civil law systems, 77% more than Islamic, and 101% more than mixed systems. To put these percentages in perspective, if Argentina had been common law in 2006 rather than civil, this would lead to an increase in FDI inflows of $2.54 billion. These results lend strong support to Hypothesis 1 that common law systems attract more FDI than the other types of systems.

View this table:
1

Legal Systems and FDI (Developing Countries 1970–2007)

DV: FDI/GDPt+1 Model 1 FE with Vector DecompositionModel 2 OLS with 5-Year AverageModel 3 Interactions with Distance Years
Civil−0.012 ** (0.006)−0.009 ** (0.003)−0.018 * (0.009)
Islamic−0.016 ** (0.008)−0.022 *** (0.005)−0.012 (0.012)
Mixed−0.018 ** (0.008)−0.017 *** (0.005)−0.032 ** (0.013)
FDI/GDP0.286 *** (0.020)0.273 *** (0.020)
Economic development0.003 (0.005)−0.0002 (0.001)−0.0002 (0.006)
Economic growth0.104 *** (0.018)0.161 *** (0.037)0.099 *** (0.018)
Trade0.010 (0.007)0.031 *** (0.003)0.011 (0.007)
Financial openness0.003 * (0.001)0.002 * (0.001)0.002 (0.001)
Democracy0.003 (0.004)−0.001 (0.004)0.0004 (0.004)
Executive ideology0.002 (0.004)0.005 (0.003)0.003 (0.004)
Political Stability0.0001 (0.0001)0.00001 (0.00007)0.00003 (0.0001)
Resource curse−0.002 (0.004)0.002 (0.004)−0.002 (0.004)
BITs−0.0005 (0.0007)0.0006 (0.001)−0.0004 (0.0007)
Conflict−0.003 (0.003)−0.000004 (0.004)−0.005 (0.004)
Post-Cold War0.009 *** (0.003)0.012 *** (0.003)−0.005 (0.004)
Distance years−0.001 *** (0.0005)
Civil * Distance Years0.0005 (0.0005)
Islamic * Distance Years−0.00007 (0.0005)
Mixed * Distance Years0.001 * (0.0006)
N 2,7505572,750
R 2 .34.30.35
  • (Notes. p values: ***p < .01, **p < .05, *p < .10, two-tailed.)

1

Legal Systems and FDI (Developing Countries 1970–2007)

As mentioned earlier, arguments advanced in the scholarly literature suggest that the various types of legal systems in the world may have become more similar in recent years, especially as between common law and civil law systems. If so, we might expect to find a diminishment over time of the advantage that countries with common law systems have over countries with other types of systems. Similarly, the literature indicates a trend toward nonjudicial arbitration in resolving disputes, and this also might lessen the common law advantage. We account for the effects of these supposed trends in Model 3, using FEVD as our standard, by including the Distance Years variable and interaction terms. The results in Model 3 suggest that the common law advantage has not declined in any real sense over the period covered in our study. First, we expect a negative coefficient if the common law advantage is declining, but only the comparison of common law with Islamic has a negative coefficient. Second, none of the results are statistically significant, so we cannot reject the null hypothesis of no change over time. We graphically present the trend-line results for each type of system in Figure 1(a–c). This allows us to visually appreciate what is going on and also to discover any year-to-year marginal changes that reach statistical significance not captured in the overall model. Following right-to-left on the graph, we see in Figure 1a that, if anything, common law is gaining an advantage over civil law. This is consistent with the positive coefficient reported in Model 3. To discover year-to-year statistical significance, we look to see whether the upper confidence interval line is below the zero line on the y-axis. At about the twenty-first year out, common law increases its advantage over civil law to a degree that is statistically significant and continues on this path through all remaining years. In Figure 1c, we see a similar trend for common law and mixed systems starting in the seventeenth year. For common law and Islamic systems, as presented in Figure 1b, we see that the confidence interval line remains above the zero line, indicating no statistically significant change in the common law advantage over the entire time period of coverage.

1

Marginal effects of legal systems on FDI over distance years

Although not the focus of our paper, but consistent with the prevailing literature, we find in Models 1 and 2 that Economic Growth, Financial Openness, and Post-Cold War are each associated with increased FDI.

Hypothesis 2 postulates that common law countries have greater adherence to the rule of law than do countries with other legal systems. The results reported in Table 2 generally confirm the hypothesis. In Table 2, we find, as expected, that Civil, Islamic, and Mixed are each negatively correlated with Rule of Law, reaching statistical significance at p < .001 for Civil, p < .05 for Islamic, and p < .10 for Mixed. In terms of substantive importance, common law countries garner rule of law scores that are 121% higher than civil law countries, 147% higher than Islamic countries, and 145% higher than mixed countries.

View this table:
2

Legal Systems and Rule of Law (Developing Countries 1996–2007)

DV: Rule of LawModel 1 OLS with 5-Year Average
Civil−0.421 *** (0.083)
Islamic−0.228 ** (0.115)
Mixed−0.235 * (0.132)
Economic development0.395 *** (0.026)
Economic growth−1.729 ** (0.734)
Trade0.097 (0.070)
Democracy0.309 *** (0.077)
Executive ideology0.058 (0.076)
BITs0.016 (0.021)
Conflict−0.215 ** (0.094)
Constant−2.989 *** (0.186)
N 250
R 2 .70
  • (Notes. p values: ***p < .01, ** p < .05, *p < .10, two-tailed.)

2

Legal Systems and Rule of Law (Developing Countries 1996–2007)

Hypothesis 3 predicts that countries that adhere to the rule of law will attract more foreign investor interest. Table 3 provides support for this hypothesis, showing positive and statistically significant results at p < .05 in the DKSEs model, results that are robust with the p < .10 significance levels obtained in both the Newey–West and fixed-effects models. When we calculate scores for substantive importance (using DKSEs values), we see that a one standard deviation increase from the mean of Rule of Law leads to a 59% increase in FDI inflows. Applying this to Argentina in 2006, this amounts to an increase of $3.19 billion.

View this table:
3

Rule of Law and FDI (Developing Countries 1996–2007)

DV: FDI/GDPt+1 Model 1 DKSEsModel 2 Newey–WestModel 3 OLS Fixed Effects
Rule of law0.015 ** (0.007)0.015 * (0.009)0.015 * (0.008)
FDI/GDP0.078 (0.146)0.078 (0.157)0.078 (0.188)
Economic development−0.040 (0.029)−0.040 (0.039)−0.040 (0.039)
Economic growth0.075 (0.083)0.075 (0.081)0.075 (0.082)
Trade0.060 *** (0.022)0.060 ** (0.024)0.060 *** (0.021)
Financial openness0.004 (0.003)0.004 (0.003)0.004 (0.004)
Executive ideology0.002 (0.002)0.002 (0.004)0.002 (0.005)
Political stability0.0008 *** (0.0003)0.0008 *** (0.0003)0.0008 ** (0.0004)
Resource abundance−0.082 (0.179)0.051 * (0.028)0.051 * (0.031)
BITs−0.001 ** (0.0005)−0.001 (0.0008)−0.001 (0.0008)
Conflict0.003 (0.008)0.003 (0.009)0.003 (0.007)
Constant0.493 *** (0.165)0.187 (0.218)0.187 (0.228)
N 1,2621,2621,262
R 2 .47.47
  • (Notes:DKSEs, Driscoll–Kraay standard errors; FDI, foreign direct investment.

  • The results control for country-fixed effects.

  • p values: ***p < .01, **p < .05, *p < .10, two-tailed.)

3

Rule of Law and FDI (Developing Countries 1996–2007)

Discussion and Conclusion

The results reported in our empirical models lend support to our three hypotheses. First, as predicted by Hypothesis 1, we find that developing countries with common law systems attract significantly higher FDI inflows than countries operating under civil law, Islamic, or mixed legal systems. Second, and consistent with Hypothesis 2, we find that common law systems promote much higher levels of adherence to the rule of law than do other types of systems. These results suggest that common law systems are better able to enforce private property rights and contractual commitments, thereby reducing investor risk. Third, we find that countries that adhere to the rule of law attract higher FDI, a result that not only confirms Hypothesis 3, but also supports earlier political economy studies that link rule of law to capital investment (Jensen 2003, 2006; Li and Resnick 2003; Biglaiser and Staats 2010; Staats and Biglaiser 2012).

Countries that respect and adhere to the rule of law provide greater assurance to investors that they will receive fair and effective resolution of disputes not only with the government, but also with private parties. Common law systems have the advantage here. While common law courts tend to have the inclination and independence to render fair and practical results in litigation, the other systems have judges who are imbedded in the state apparatus and required to decide cases based on uncertain legislative mandates, which is true of civil law judges, or are constrained by vertical bonds of authority, weak separation of powers, and limited regard for individual property rights, in the case of Islamic law judges. In such circumstances, civil and Islamic law legal systems place at greater risk the assets and contracts of foreign investors.

Referring to Figure 1, it appears that from about 1987 or 1988, common law systems have edged higher in their advantage over civil law and mixed legal systems, and the trajectory may have started earlier, although we lack findings of statistical significance to make such a statement. The trend as to the advantage over civil law countries is at odds with the suggestion by some legal scholars that common law and civil law systems are converging toward similarity. It may well be that the aspects of convergence are not those that are particularly important for advancing adherence to the rule of law and therefore are not of especial importance to investors. Also, it is likely that most of the convergence is occurring in developed countries rather than the developing countries covered in this paper. Another possibility is that these trend lines are capturing the diminishment of adherence to the rule of law in recent years in a number of Latin American countries, all of which have civil law systems. The mean score on Rule of Law for 17 Latin American countries declined from −0.324 in 1998 to −0.507 in 2007. Two countries stand out, Argentina, which declined from 0.077 to −0.520, and Venezuela, which went from −0.711 to −1.504.

The findings hold important implications. Building on previous work, 17 we present details on how specific legal institutional arrangements can affect FDI. Our work also provides a platform for future research on the role that legal systems play in attracting different types of foreign capital. Finally, our project shows that common law systems have an advantage over the other types of systems in attracting FDI because they support strong and independent judicial systems and adherence to the rule of law. The results suggest that countries should work on improving their adherence to the rule of law and supporting independent judicial systems if for no other reason than to influence aggregate investment decisions of foreign investors.

Footnotes

  • 1 We exclude OECD countries because most FDI inflows into developed countries are motivated by horizontal integration goals while FDI inflows into developing countries normally have vertical interests in mind (Blonigen and Wang 2004).

  • 2 Judges on civil law high courts have considerable prestige, but “it is the kind of respect earned and received by persons in high places elsewhere in the civil service” (Merryman and Pérez-Perdomo 2007:35; see also Mahoney 2001:507).

  • 3 Although we have not discussed mixed legal systems directly, we would expect that countries operating under common law systems to attract more FDI because of the limitations presented under the other legal systems that in part make up mixed systems.

  • 4 The study by Zywicki and Peter Stringham (2010) references and discusses a similar line of argument advanced by a writer in a student comment appearing in the Yale Law Journal.

  • 5 We use this measure rather than FDI in dollar amounts because it standardizes each country's varying economic size (Choi and Samy 2008:90). While we realize that Li (2009b) takes issue with using FDI as a percentage of GDP, we believe this is appropriate because of the importance of market size in determining the amount of investment resources that a country is capable of absorbing and still maintain profitability to investors. As Jensen et al. (2012:6) state: “Foreign investment may take place to supply large local markets, particularly when host countries impose high tariffs that create barriers to market access.” We also note that the great majority of published research both before and after Li (2009b) use FDI/GDP as the FDI metric (for recent works, see Staats and Biglaiser 2012; Jensen et al. 2012).

  • 6 It is plausible that there are underlying political–cultural factors that affect the type of legal system institutionalized in a country and perhaps even FDI inflows. However, we are attempting to ascertain the effects of legal systems on FDI, not the origins of legal systems, which political–cultural factors would most impact. Investigating the political–cultural influences is beyond the scope of this work but worthy of future inquiry.

  • 7 We also considered using GDP as a measure for market size but our dependent variable has GDP in the denominator. The main results remain the same with or without GDP in the models.

  • 8 On the growing importance of BITs, see Büthe and Milner (2008) and Haftel (2010).

  • 9 Another BIT measure is the total number of states that a host country has signed BITs with in a given year. However, using the total number of states is potentially problematic as it would likely overrepresent countries that have signed multiple agreements with smaller countries relative to countries that may have a few agreements but with larger investing countries. Larger investing countries are also prone to have much greater negotiating leverage in bilateral treaties relative to smaller countries, which is an important reason for why the largest global investor by far in the world, the US, has pushed so hard for bilateral agreements. We thus use whether the country has signed an investment treaty with the US as a dummy variable for measuring BITs.

  • 10 We did not use the full range of the Polity IV measure because doing so would amount to measuring the quality of democracy present in a country rather than regime type (democratic or authoritarian). Our rule of law measure captures the quality of democracy and it would be inappropriate to include another measure that overlaps with this. However, we tested our models using the full range of Polity IV and the results obtained were substantially unchanged from what is reported here. Similarly, we do not include Democracy when testing hypothesis 3, because our democracy measure includes many of the same factors that go into defining whether a country adheres to the rule of law and thus democracy should not be included as a control variable (Ray 2003). However, for robustness purposes, we also compared our Polity democracy dummy variable with the democracy measure developed by Cheibub, Gandhi and Raymond Vreeland (2009) and the correlation is .82.

  • 11 Whatever convergence occurring so far, if any, would seem to be insubstantial. As Mattei and Pes (2008, 273) state: “This general observation, certainly pointing to some ‘convergence,’ needs to be understood in context. There is still no comparison between the importance that cases enjoy in the common law as opposed to the civil law where their importance, while increasing is still generally limited. Moreover, the role of judges in the common law tradition is still more important sociologically than that of their civilian colleagues.” Moreover, the fact that much of the common law has been codified does not mean that common law legal systems have become the same or similar to civil law systems. “Common law codes are still exceptions in a panorama of the sources of law in which the grand scheme of legal interpretation is still contained in the common law tradition. One could observe, as to the common law, that should all the statues be abolished overnight, there still would be a viable legal system since all the general doctrines that can be used for principled application of the law to the multitude of facts of life are all judge-made and contained in case law” (Mattei and Pes 2008:273). For the same effect, see Hahto (1968:375–376).

  • 12 We considered but did not use the PRS ICRG Law and Order index because it conflates “law” with maintenance of “order” by whatever means, leading to the strange anomaly that North Korea receives high scores and that Cuba in some recent years is assigned higher law and order scores than Costa Rica and Uruguay, both of which have among the best judicial systems in Latin America (see Staats, Bowler and Hiskey 2005).

  • 13 Because data are available every other year until 2002, we interpolate 1997, 1999, and 2001.

  • 14 Inverse χ2, Inverse normal, Inverse logit t, and Modified inverse χ2: p < .0000, respectively.

  • 15 As is well known, it is not possible to include time-invariant variables (type of legal systems in our case) in a model that is fit by unit-specific intercept (fixed effects). Among the efforts to surmount this difficulty, Plümper and Troeger (2007) proposed fixed effects vector decomposition (FEVD), which is based on three-stage procedure of decomposition. Plümper and Troeger (2007, 125) state that “in the first step, the procedure estimates the unit FE by running a FE estimate of the baseline model. In the second step, the procedure splits the unit effects into an explained and an unexplained part by regressing the unit effects on the time-invariant and/or rarely changing explanatory variables of the original model. Finally, the third stage performs a pooled-OLS estimation of the baseline model by including all explanatory time-variant variables, the time-invariant variables, the rarely changing variables, and the unexplained part of the FE vector.” See Plümper and Troeger (2007:127–129) for more details on this procedure.

  • 16 Our pooled data are calculated on averages for 1970–1974, 1975–1979, 1980–1984, 1985–1989, 1990–1994, 1995–1999, and 2000–2004.

  • 17 See Jensen (2003, 2006), Li and Resnick (2003), Li (2006), and Staats and Biglaiser (2012).

References

View Abstract