Political risk and international investment law.
TABLE OF CONTENTS INTRODUCTION I. THE CONCEPT OF POLITICAL RISK A. A General Definition of Risk B. Defining and Assessing Political Risk (or Not) C. Political Risk in Theory and Practice Today D. Off-the-Shelf Indicators of Political Risk E. Social Scientific Measures of Political Risk Based on Democracy and Veto Points II. INTERNATIONAL LAW AS A POTENTIAL MITIGATOR OF POLITICAL RISK A. An Introduction to BITs B. Empirical Studies of BITs and Foreign Investment Flows C. BITs and Off-the-Shelf Indicators of Political Risk D. The Imperfect Coverage of BITs E. The High Costs and Uncertainty of BIT Litigation F. Imagining a World Without BITs CONCLUDING THOUGHTS
This Article explores some of the conceptual and operational difficulties related to measuring political risk and to international law's arguably rather limited promise to provide meaningful protections against that risk. Part I explores the concept of political risk, providing a working definition of "risk" and examining previous attempts to develop workable definitions of "political risk," which theorists have yet to successfully define. It examines evidence suggesting that, definitional issues aside, businesses only imperfectly attempt to assess political risk and discusses why simple off-the-shelf indexes of political risk and political science indicators of "democracy" and "veto points" are of limited utility for measuring objective political risk. Part II discusses international law as a potential mitigator of political risk. It introduces bilateral investment treaties (BITs) as a major international law mechanism for reducing political risk, reviewing empirical studies of the correlation between BITs and foreign direct investment (FDI) inflows that have attempted to demonstrate that BITs reduce political risk and examining whether BITs are correlated with a prominent off-the-shelf indicator of political risk. It then discusses certain weaknesses of BITs as mitigators of political risk, including their imperfect coverage, the high cost and uncertainty of BIT litigation, and the availability of alternative risk-reducing strategies that might be more effective.
I. THE CONCEPT OF POLITICAL RISK
A. A General Definition of Risk
The standard definition of "risk" is relatively intuitive and easy to grasp. Risk is typically defined as the probability that an event will happen, where the event will have adverse consequences (costs) for the relevant party. (1) Risks may be thought of as greater when the product of probability and costs is higher and as lower when the product is lower. (2) The law review literature on risk sometimes emphasizes a conceptual distinction between risk and uncertainty, where uncertainty is characterized by a lack of knowledge of the true distribution of probabilities of the adverse event. (3) But that conceptual distinction is not of much obvious utility in the present context, as the managers of international businesses will rarely if ever know the true distribution of probabilities, though they may be able to estimate or guess at those distributions with more or less accuracy or bias. (4) Rather, to the extent that managers attempt to assess risk in a mathematical way, they will almost always be working in an environment characterized by significant uncertainty, though they may not have much appreciation of how uncertain their estimates or risk actually are.
B. Defining and Assessing Political Risk (or Not)
The question of which kinds of adverse events a party is interested in predicting can be approached at the ultra-micro level, by attempting to identify any possible adverse event related to a foreign investment that might impose a cost on a business operation. For example, perhaps the company CEO will slip on a banana peel while touring the company subsidiary in Costa Rica, suffering major incapacitating injuries and setting off an internal power struggle that leads to investor worries about succession and causing a sharp drop in the company's share price. Because such speculative risks are more or less infinite, cataloging them, assigning reasonable probabilities to them, and then adjusting corporate behavior in response is an obviously impossible task. Rather, corporate risk assessors are likely to proceed by placing various kinds of risk into conceptual categories, sorting (and discarding) risks based perhaps in part on their perceived importance (e.g., only trying to catalog risks that have a high probability of occurring or that would have major consequences if they were to occur) or on whether they are thought to share similar causes, to have similar adverse consequences, or to be amenable to similar compensating or mitigating behaviors. (5)
The notion of "political risk" is probably one of the most important standard categories of international business-related risk. While the history of thinking explicitly about political risk dates back to at least the 1950s, the 1970s were arguably the heyday of conceptual explorations, with a great many articles appearing in what was then a relatively new field of academic inquiry: "international business studies." (6) In those articles, authors argued that political risk was an important risk for foreign direct investors, especially in the developing world, and offered their own definitions of the concept and, sometimes, their own systems for evaluating it. (7) Probably not surprisingly, political risk emerged as a hot topic in an era characterized by spectacular expropriations of foreign-owned assets. (8)
Perhaps the high point of conceptual explorations of political risk is Stephen Kobrin's influential 1979 article "Political Risk: A Review and Reconsideration." (9) Kobrin reviewed the numerous existing studies, finding that authors adopt widely varying and often conceptually problematic definitions of political risk (or even fail to define it at all). (10) One common approach he identified was to equate political risk with "governmental interference with business operations," a definition, he points out, that has important normative implications about the proper role of government in managing and regulating business affairs. (11) Another approach was to equate political risk with particular events deemed to be "political," such as coups, civilian unrest, or specific acts like expropriation or the imposition of currency controls, that were presumed likely to negatively impact foreign direct investors. (12) In either case, Kobrin viewed political risk as conceptually and operationally problematic. (13) He even suggested doing away with the term completely, though he also offered his own preferred definition, which focuses on whether events "motivated by or [that] have as their objective the maintenance or modification of power or authority relationships at the governmental level.... will reduce returns to the point where the project would be no longer acceptable on the basis of ex ante criteria." (14)
A number of interesting implications arise from Kobrin's article. First, political risk theorists do not agree on what the concept actually entails; moreover, their definitions are often relatively simplistic or even somewhat incoherent. Second, devising a concept that is of obvious operational utility is difficult, even considering Kobrin's preferred definition. In part this is because assigning causality to events based on their motivation (as political) is difficult, especially as the role of government has expanded into all comers of the modern economy (a point that Kobrin recognizes as well). (15) But it is also because existing data, as well as the inherently multi-causal complexities of modern society, will often impede the calculation with any real accuracy of a probability that political event x will happen and if it happens that it will impose costy on a certain project with probability z. Perhaps reflecting these difficulties, Kobrin finds that companies do not really assess political risk with much regularity or sophistication. (16)
C. Political Risk in Theory and Practice Today
Very little progress has been made in the intervening years in either the conceptualization of political risk or the use of political risk assessment by businesses. (17) A recent paper by Export Development Canada (EDC), a government provider of political risk insurance, provides a nice example of both trends. (18)
EDC traces the intellectual history of political risk, identifying four "non-contiguous schools" of thought that, taken together, form a "disconnected" and "fragmented" discipline, (19) so much so that one of these schools has "largely abandoned" its "quest for general theory and systemic explanations." (20) And while recent work in international relations on political instability has embraced general theory and systemic explanations, the focus of that work is largely on international security-type concerns about conflict management and not on political risk as specifically experienced or viewed by business. (21) The result is that "there is neither a generally accepted definition of political risk nor any widely accepted methodologies to evaluate it." (22)
What about the frequency of political risk assessment? EDC confirms Kobrin's earlier finding that "many businesspeople and their financial supporters tend to undertake political risk assessment (PRA) in a superficial, haphazard and subjective way." (23) It cites a study finding that 84% of recently surveyed companies "do not formally integrate social and political risks into their investment decisions," a situation that is "not new." (24) Perhaps the lack of systematic political risk assessment is not surprising. The theoretical literature on decision-making suggests that rational risk assessment of "low-probability/high-consequence events"--which undoubtedly include adverse political risk events like expropriation--is particularly "troublesome" (25) because of such things as the "ambiguities associated with the chances of an event and/or its consequences," the disjunctive between expert and non-expert perceptions of risk, and the individuals (who, of course, make up the corporation) who "do not have clear, stable preferences with respect to... risk." (26)
EDC itself serves as an important exception to the tendency of corporations not to conduct rigorous political risk assessments, though it is also an exception that nonetheless helps explain why the tendency exists. EDC provides insurance for foreign investment projects against loss resulting from expropriation, from political violence, and from foreign exchange or currency restrictions--three classic categories of political risk (though categories that are not obviously closely related conceptually). (27) By EDC's account, it has spent years developing a sophisticated and complex evaluation scheme for each kind of risk at the country and project level. (28) But despite that major effort, EDC sees major problems with its creation, most notably that its model is impossible to validate empirically due to the small number of adverse events that it has encountered (roughly 100 projects funded, with "fewer than five" claims). (29) The small-sample problem is endemic of risk analysis more generally, (30) and it means that EDC is unable to calculate whether its model is "good" in the statistical sense of separating projects that should be funded from those that should not. There is no reason to think that the empirical situation is different for other national providers of political risk insurance, for private providers, or for other kinds of businesses considering investing abroad.
The EDC article illustrates another problem endemic to political risk assessment: the fact that the domestic and international political environments are constantly changing in complex ways, such that a model that accurately predicts political risk in one era may be quite inadequate at predicting political risk in the current era. The article discusses the example of the emergence of a global environmental movement, which has increased pressure on previously reliable national governments to abandon support for projects that the movement deems (genuinely or for strategic reasons) environmentally problematic. (31) EDC's model was simply not constructed to take into account the growing influence of the global environmental movement on national policies. (32)
D. Off-the-Shelf Indicators of Political Risk
It seems unlikely that many businesses thinking of investing abroad would first invest in a system of political risk analysis as sophisticated as EDC's. (33) Yet it seems that many businesspeople do assert that political risk is a major concern. (34) Assuming that they in fact think that it is important and that they act on that perception at least to some extent, how might they go about assessing it on the cheap? (35)
There are a variety of single-number ratings of various kinds of political risk to which businesspeople might turn to get a sense of whether a particular investment opportunity is politically risky. For-profit companies like the PRS Group or BER1 market political risk analysis services, prominent components of which are indexes purporting to measure on an ordinal scale whether a country is risky along a particular dimension. For example, the PRS Group's International Country Risk Guide (ICRG) calculates numerical scores for such things as "government stability," "socioeconomic conditions," "investment profile," and "religious tensions" and weights and combines the scores into a single 100-point "political risk rating." (36) The rating can be used to determine, for instance, that the United States is marginally less politically risky (as ICRG defines things) than the UK and is significantly less politically risky than the Democratic Republic of Congo. The ICRG's quantitative evaluations are performed in-house by staff and are based on political information and financial and economic data, though the ICRG gives no real sense of what data its staff uses or how it transforms that data into index ratings. (37) BERI's approach is similar. It too creates an index of political risk, but it relies on panels of outside experts ("senior bank, company, and government officials"). (38) As additional value added, both companies provide qualitative country guides, and BERI's even state whether the "business environment" in each country "merits investment, contracts for medium and long-term relationships, transaction-by-transaction trade, or no business relations." (39)
But while ratings like those provided by the PRS Group and BERI may provide a comforting basis for corporate decision-making, such ratings suffer from a number of important limitations. Most obviously and also perhaps most importantly, they provide country-level ratings, not estimates of firm- or project-level risk. This distinction is critical, as there is good reason to believe that actual political risk will vary, perhaps tremendously, depending on firm- and project-level characteristics. Take a simple example: a proposed investment in a new nuclear power generation facility in the United States. The United States is a highly-rated (low-risk) country, but the risk of a permit being denied for such a project is probably quite high. To say that the United States is characterized generally by a low level of political risk reveals very little about whether investing in the U.S. nuclear energy sector is politically risky.
Additionally, conceptions of political risk often reflect, at least implicitly, the theory of the "obsolescing bargain" between states and investors, in which state incentives to expropriate or otherwise interfere with investments tend to increase over the life of the investment. (40) But that theory actually suggests that the risk of obsolescing bargains depends on the degree to which the investment's assets are fixed. The implication is that certain sectors (those characterized by a high degree of asset specificity, such as, arguably, hard-rock mining) may be more susceptible to the obsolescing bargain than sectors, such as light manufacturing, in which the real assets (such as access to international distribution networks or to high-level technical knowledge) are easy to either move out of the host country or deny to the state expropriator's new management.
Off-the-shelf indicators of political risk also often do not measure risk as typically defined. Recall the general definition: a probabilistic chance of the occurrence of an event having adverse effects of a particular magnitude on the relevant party. (41) The PRS Group's ICRG index of political risk measures such things as a country's "socioeconomic conditions," including the unemployment rate, consumer confidence, and poverty, which are combined into a 12-point scale that is further combined with other variables into the ultimate (100)-point scale. (42) While socioeconomic factors may indeed be relevant to international businesspeople and may impact the profitability of the foreign operation, the unemployment rate (to take one of the three) is simply not a risk. It is a socioeconomic fact (subject, of course, to measurement error) that may be relevant to a true risk assessment. By itself or combined in a 12-point scale with other socioeconomic statistics, it is hardly an estimate of a probability of some adverse event. Rather, the 12-point scale represents an unformulated theory about the potential impact of socioeconomic conditions on the probabilities of business-relevant adverse events. Socioeconomic factors may in some sense be a proxy for an actual measurement of risk, but they are, by themselves, not risk. To be useful as risk indicators they must still be transformed into actual probabilities of adverse events of particular magnitudes by the end user of the data. This is no small task, to be sure. The relevance of, for instance, unemployment to the risk of an uncompensated expropriation in a certain sector of the economy is a question on which there is, almost certainly, no relevant data.
In risk forecaster Credit Risk International's (CRI) system, there are four indexes of various kinds of business-related risk, each based on a large number of criteria and sub-criteria (100 in all). (43) Its "business environment" index evaluates three major criteria ("management of the economy," "foreign investment climate," and "working conditions"), each of which is measured by an evaluation of numerous sub-criteria. (44) In the case of "foreign investment climate," CRI includes such variables as the "cost of energy," the "price level of local real estate," and "legal restrictions to capital flows" (among seven other variables, all with arbitrary weights). (45) Again, while such variables may be relevant to a theory of political risk (or perhaps serve as reasonable proxies or predictors of risk, at least when included in an empirically verified model), they are not measurements of risk itself. Indeed, many of CRI's variables are statistical facts--the price level of local real estate is something eminently knowable, as is whether a country limits capital flows--and thus they represent the certainty, not the probability, of some state of the world. Furthermore, the state of the world represented by these variables is never linked theoretically or empirically to a business-relevant adverse event. For example, if the probability of expropriation is the risk about which a party is worried, do high real estate prices really mean that a foreign business is more likely to be expropriated?
Off-the-shelf quantitative indicators of political risk thus leave a huge amount of work for the end user to complete in order to conduct an actual risk assessment. (46) Quantitative indicators of risk must be turned into risk measures by linking the indicator to the probability of some particular adverse event by way of (hopefully empirically validated) causal theory. The end user must specify the theory, keeping in mind the specific event being predicted, and use the theory to assign probabilities both to the occurrence of the event and to different magnitudes of adverse impact that may result from the event. While a businessperson could obviously use an off-the-shelf indicator for something without doing this extra work, that use would not be a political risk assessment in any meaningful sense.
As seen above, even quite sophisticated political risk actors like EDC admit that their political risk models are resistant to statistical validation. (47) EDC emphasizes the small-sample problem: it is interested in measuring the probability of a particular adverse event, a claim against an insurance policy, but such claims are relatively rare. (48) Other indicators of political risk, like the PRS Group's, also suffer from validation problems for two reasons. First, the specific adverse events that the indicator is supposed to predict are typically unclear. Is PRS Group's indicator successful if it reliably predicts instances of expropriation? Of expropriation and nationalization? Of the imposition of exchange controls? Of a host state government's failure to provide the investor with "full protection and security" against insurgent attacks? Of its failure to treat the investor "fairly and equitably"? Without a clear definition of political risk (or of the specific adverse events that parties are trying to predict), designing a validation test for a political risk indicator is impossible simply because the dependent variable is unclear.
Even where parties can agree on the adverse event that is to be predicted, in most cases they will find it difficult to identify incidences of those events. Take the simplest event that can be reasonably characterized as representing political risk, the probability of classic expropriation: the seizure of an investor's property by the host state without adequate compensation. While incidences of expropriation are in theory relatively visible (and thus measurable), expropriation has become much less common in recent years than it was in the 1970s. (49) The small sample problem is debilitating. If there are few or no incidences of expropriation, validating a model of the determinants of expropriation (or whether, for example, the PRS Group's political risk index is a good predictor of expropriation) is impossible.
Of course, other relevant adverse events may be more common than expropriation, but they may also be much more difficult to observe. For instance, some scholars suggest that even though most states today have no interest in engaging in classic expropriation, they nonetheless still engage in, and perhaps are increasingly engaged in, "creeping expropriation" (or "regulatory expropriation"), in which the tools of government are used to nibble away at the value of a foreign investment over time, accomplishing over the long run an effective taking of the investor's property. (50) The line between legitimate government regulation and impermissible "creeping expropriation" is blurry, making it difficult to build a database of incidences of "creeping expropriation." In fact, building such a database may be impossible, as indicated by the fact that, to the author's knowledge, such a database has never been built--surprisingly so, given suggestions in the academic literature that "creeping expropriation" is a major source of modern political risk. (51) And if measuring incidences of creeping expropriation is impossible, then so is validating models that are supposed to predict it.
E. Social Scientific Measures of Political Risk Based on Democracy and Veto Points
Parties can attempt to validate measures of risk indirectly by examining, for example, whether a particular indicator is associated not with the adverse event itself but with investment decisions. This is the strategy pursued recently by various social scientists interested in exploring how political institutions might impact political risk. Political scientist Nathan Jensen, for example, argues that democratic institutions make countries less risky, where risk is defined as the probability of changes in government policy. (52) He argues that democracies, compared to autocracies, have more "veto" points, or opportunities for actors to block policy changes that they do not support and that investors presumably value the policy stability that results. (53) Jensen tests his theory by examining whether democracies are more successful than autocracies at attracting foreign investment. (54) He finds that they are, suggesting that democracies are indeed less risky. (55)
Business professor Witold Henisz offers his own version of this thesis, though rather than focus on off-the-shelf indicators of democracy, as does Jensen, Henisz constructs a bespoke database of veto points for a large number of country-years. (56) While Henisz does not examine whether his veto points measure is correlated with FDI flows, he does report correlations with economic growth (57) and foreign investor mode of entry. (58) Those results suggest that his measure of political risk may reflect foreign investor considerations of whether or how to invest in a particular host state.
While such studies are of significant intellectual interest, they suffer from certain limitations as validations of theories of political risk. In part this is because they fail to measure what might be called "objective" risk. By focusing on investor behavior (whether investors are more likely to invest), the studies can at best shed light on whether investors subjectively view countries as politically risky or not. Imagine, for example, that investors actually do accord democracy a prominent place in their investment decision-making process. And imagine that they do so because they view democracy as a good indicator of political risk. An empirical study on whether investors are more likely to invest in democracies will, unsurprisingly, discover a correlation, but that correlation suggests only that investors think democracy is a good indicator of political risk, not that it actually is a good indicator. (59) Given convincing evidence that companies perform political risk assessments haphazardly and sloppily, there would not seem to be much good reason to think that a company's own use of a particular indicator (e.g., democracy) is reliable evidence that the indicator is actually of much use in objectively predicting adverse events.
Whether businesses in fact rely on such indicators naively (e.g. simply as rough measures of "good" and "bad" places to invest) or whether they rely on them as indicators of political risk, understanding the concept with a more sophisticated degree of theoretical coherence and specificity, is also unclear. Neither Jensen nor Henisz provide really convincing evidence that they use them at all. The problem seems particularly relevant to Henisz's indicator. Decent measures of democracy have been available since the 1980s, making it at least possible that companies had access to the data and could have used it as a political risk indicator if they wished. In contrast, Henisz's veto points measure did not exist until Henisz invented and publicized it in the late 1990s and early 2000s. Given the measure's computational complexity, it is very unlikely that companies previously compiled their own, similar measure of veto points for the purposes of political risk assessment. In other words, Henisz's measure of veto points is a reliable indicator of how businesses conceive of and assess political risk only if businesses actually use a veto points measure (or if the veto points measure is somehow correlated with those conceptions and assessments). But because such a measure did not exist until Henisz's work, it seems highly unlikely that businesses actually made investment decisions on the basis of the concept prior to Henisz's development of it. To the extent that Henisz is successful at marketing his indicator as useful for predicting political risk, businesses may in fact start to use it as such. But their use of it does not demonstrate that the indicator is in fact useful for predicting political risk. Validating Henisz's measure poses the same difficulties faced by EDC and other developers of political risk indicators, as discussed above.
Finally, the veto points notion potentially suffers from a conceptual bias similar to the ideological bias against government intervention in the market that Kobrin identified many years ago. In particular, the veto points notion seems to presume that changes in policy are necessarily adverse events and thus that businesses strongly desire policy stability. (60) But policy changes, like revolutions, can herald profits as well as losses. The policy environment may, in its current state, be quite undesirable from a corporate perspective: perhaps tax rates are high, or perhaps an economic sector is closed to foreign participation. Change in undesirable policy toward lower tax rates or allowing foreign participation will be viewed by the foreign investor as a good thing. Veto points decrease the probability of "bad" policy changes but also of "good" ones too.
II. INTERNATIONAL LAW AS A POTENTIAL MITIGATOR OP POLITICAL RISK
A. An Introduction to BITs
Discussions about the benefits of the rule of law often suggest that it helps to promote predictability by limiting the ability of government actors to act arbitrarily. (61) In other words, law can reduce risk by promoting certainty about the future state of the world through the constraint of governmental discretion. In a political system subjected to the rule of law, government actors will predictably act as the law requires them to act and not as they unpredictably may wish to act.
Prom that theoretical starting point, it is easy to imagine that law, or even international law (despite its well-known limitations, particularly the lack of an effective enforcement mechanism), might be useful for reducing the political risks that beset the foreign direct investor. International law's traditional weakness has been somewhat overcome in the world of international investment law through the device of bilateral investment treaties.
BITs provide investors with a package of substantive rights, such as the rights to national and most-favored-nation treatment, to be treated "fairly and equitably," to freely transfer assets and proceeds out of the host state, and to be free from inadequately compensated expropriation. (62) These substantive rights are coupled with the procedural right of guaranteed investor access to international arbitration in cases in which the investor believes that the host state has violated the terms of the treaty. (63) The history of BITs has been told in a number of other articles. (64) Here, it suffices to say that BITs have proven to be enormously popular: thousands have been signed in the past (30) or so years. (65) They have also proven somewhat controversial, as investors are increasingly using them in creative ways to sue governments for alleged misdeeds that, in an era before BITs, would have been unactionable at law, domestic or international. (66)
The popularity of BITs may be premised on their ability (assumed or actual) to protect investors against political risk, however defined. For example, less-developed countries (LDCs) may be motivated to sign BITs because they make investors more likely to invest by ensuring that host state "promises" to investors will be kept. (67) The problem of host state promise-keeping can be viewed as one of political risk.
B. Empirical Studies of BITs and Foreign Investment Flows
The basic idea that BITs may promote foreign investment by using law to reduce political risk has motivated a number of empirical studies of whether BITs are actually effective at doing so. These studies typically use econometric techniques to identify statistically significant correlations between the number of BITs that a host state has signed and the amount of FDI that the host state receives. (68) If more BITs are associated with more FD1, it can be inferred that investors view the treaties as being effective at reducing investment-inhibiting political risk.
In fact, the empirical results are inconclusive. Some empirical studies find that BITs are associated with large increases in FDI, suggesting that the treaties are very effective reducers of risk. (69) Other empirical studies find no effect or only modest effects. (70) There is not a firm empirical basis for inferring that BITs are viewed by investors as meaningfully reducing political risk or that they actually do reduce political risk.
C. BITs and Off-the-Shelf Indicators of Political Risk
Of course, designing a convincing empirical test of the thesis that BITs reduce political risk is quite difficult. Even if such a test were to show a consistent positive correlation between BITs and FD1 flows, this would allow only an inference that BITs actually reduce risk. To show that BITs reduce political risk in an objective sense, defining and operationalizing political risk as a dependent variable would be necessary. But as suggested above, that is challenging.
One might analyze whether one of the off-the-shelf indicators of political risk, like PRTs ICRG index, correlates well with BITs. Figure 1 does just that. On the horizontal axis are plotted the number of BITs that a large number of developing countries had in force with the major capital exporting states (United States, UK, France, Germany, and so on) in the year 2000. On the vertical axis are plotted each country's 2000 ICRG score for "investment profile," the subcategory of risk that, in the ICRG scheme, arguably most closely approximates political risk as opposed to broader conceptions of "country risk." For example, "investment profile" includes a sub-component measuring the risk of expropriation. (71) For the ICRG indicator, a higher score means "less risk," a lower score means "more risk." (72) Only a year of data is presented to simplify the presentation. Figure 1 suggests a weak relationship between ICRG "investment profile" and the number of BITs in force. While the trend line slopes upwards and to the left (indicating less risk as states have more BITs in force), the slope is weak, and the various observations are scattered in a wide band around the trend line. BITs do not appear to be a good predictor of political risk, at least as measured by the ICRG.
But as suggested above, it is not clear that the ICRG risk indexes actually measure risk in a conceptually sound way, let alone an empirically validated way. It is also not clear that the risks that the ICRG measures are necessarily actionable under BITs. Nor is it clear that the best way to aggregate BITs as political risk-reducers is by simply adding up the treaties. Perhaps the actual correlation between BITs and political risk depends on treaty-specific details, on country-specific variables, or on the time period of analysis. A simple scatterplot cannot test these more complicated conjectures. At the same time, however, Figure 1 does suggest that the ICRG risk-rating experts do not obviously give BITs particularly great weight in making their risk determinations. (73)
D. The Imperfect Coverage of BITs
As suggested in the previous section, there is a possibility that the guarantees contained in international investment law only loosely or incompletely address the risks that investors or theorists are likely to lump under the concept of political risk. (74) For example, it seems reasonable to assume that investors concerned about political risk would in part be concerned about the risk that their investments would suffer physical harm in the event of a civil war or insurgency in the host state. Off-the-shelf indicators of political risk will often include some estimate of the probability of political violence, and political risk insurers typically offer insurance policies against damage caused by civil war or other forms of political violence. (75) International investment law does in fact also protect against this kind of political risk but only imperfectly. Most BITs include a clause guaranteeing the investor "full protection and security." (76) The exact meaning of that guarantee is obscure and contested, but the guarantee seems intended, in part, to provide some sort of promise by the host state to physically protect the foreign investment (e.g., through the provision of some minimum standard of police or other protective services). (77) The standard does not seem intended to create strict liability for the host state and may only create a duty for the host state to act with "due diligence" to protect the investor's property. (78) The guarantee is thus not without any useful content (depending on what "due diligence" itself means), but it seems significantly less protective than insurance, which will more closely reflect a strict liability standard under which the policy is triggered regardless of whether the host state acted with diligence.
E. The High Costs and Uncertainty of BIT Litigation
Moreover, the aggrieved investor seeking to enforce a right to due diligence in the provision of full protection and security (or any other right guaranteed under a BIT or customary international law) may have to spend considerable time, effort, and money to convince a tribunal to declare the guarantee violated (79) and further time and effort enforcing the resulting judgment. The high costs of accessing the machinery of international justice may limit international investment law's utility as a political risk-reducing device for claimants without deep pockets. (80) And given that the proper meaning and application of many BIT provisions remain unclear, (81) the success of a legal challenge will itself be uncertain. As a recent study shows, investors win only about half of investor-state arbitrations and rarely win anything close to their claimed damages. (82)
The cost of accessing international investment law and the uncertain meaning of international investment law guarantees mean that international investment law may not meaningfully reduce modern forms of political risk, such as the risk of "regulatory expropriation." Governments routinely intervene in national and international markets, not necessarily as buyers and sellers (though they do that too) but as regulators or even creators of those markets, (83) necessarily suggesting that government action or inaction can have important effects on the profitability of those markets or of individual investments. If the risk of adverse regulatory change is arguably the quintessential modern "political risk," international investment law offers relatively meager protections against it. While BITs routinely offer clear guarantees against inadequately compensated classic expropriation, the extension of expropriation principles to the case of regulatory expropriation is contested, (84) and leaders in BIT practice, such as the United States, seem to be working to restrict the probability that expropriation clauses in BITs will be available to challenge sovereign regulatory decisions that are not discriminatory against foreign investors. (85)
BITs may also contain important "escape clause" provisions that provide host states with excuses and defenses or that exempt certain kinds of investment disputes from mandatory arbitration. (86) For example, the U.S. model BIT contains a number of exceptions designed to enhance the host state's policy space. Article 18 provides a self-judging "essential security" exception that allows the host state to apply otherwise treaty-inconsistent measures "that it considers necessary for the fulfillment of its obligations with respect to the maintenance or restoration of international peace and security, or the protection of its own essential security interests." (87) The self-judging nature of the provision means that the host state's invocation and application of an essential security exception will be difficult or perhaps impossible for an investor to challenge in arbitration. Article 20 provides an exception for prudential measures designed to ensure the "integrity and stability of the financial system." (88) Crucially, the investor's right to challenge state decisions taken under this exception is subject to numerous important limitations in the article's text. (89) Moreover, the U.S. model limits the ability of investors to challenge "taxation measures" as treaty-inconsistent. (90) The net effect of such carve-outs is to reduce the scope of international investment law's applicability to the investor-state relationship, thus reducing its ability to mitigate political risk.
F. Imagining a World Without BITs
When judging the potential effectiveness of international law to reduce political risk, having a sense of baseline expectations is important. When considering BITs as international law, the obvious baseline is a world without BITs. Is that hypothetical world one in which political risk is rampant? In fact, it is probably not. States would have reputational incentives to keep foreign investors happy (at least as long as states desire additional investment in the future). And even where reputational concerns were inadequate (perhaps where ideas about the value of FDI were in flux), investors would have alternative ways to protect themselves against political risk. Some of those alternatives would entail international investment law in a "lite" form. For example, investment contracts with the host state might provide investors with sufficient legal security through arbitration and international law choice-of-law clauses. (91) Or investors might be able to organize their affairs to minimize host state incentives to interfere--by, for example, structuring the investment to maintain dependence on the parent corporation for intellectual property, technology, and management skills. (92)
It is commonplace and even facile to state that low political risk makes investment opportunities attractive. In that view, a principal task for infrastructure investors is to assess host state levels of such risk, just as a task for host states is to convince investors that the level of political risk is in fact low. But as this Article has demonstrated, the concept of political risk is fraught. Analysts do not agree on how it should be defined, operationalized, or measured, and off-the-shelf indicators of it tend to consist of grab bags of variables that are only loosely related either to each other or to a specified theory of political risk. Moreover, existing models of political risk, even the most sophisticated, seem to defy empirical validation, and in any case, many, perhaps most, companies do not make serious attempts to assess political risk prior to investing. Political risk does not seem to be a terribly useful way of thinking about what makes an infrastructure investment attractive or not.
Likewise, this Article has demonstrated that while it is easy to assume that international investment law, as embodied in BITs, is valuable for reducing political risk, in fact BITs probably play, or probably should play, only a limited role in reducing investor perceptions of it. BITs cover only a portion of the adverse events that investors may worry about, accessing BITs is costly and time consuming, and the outcome of BIT arbitrations is uncertain, especially given the vague and ambiguous text of most investment treaties. Moreover, BITs increasingly contain escape clauses, like self-judging national security exceptions, that weaken legal protections that are already somewhat nebulous. Unsurprisingly, some research suggests that investors do not pay much attention to BITs when deciding whether to invest.
To end on a somewhat more constructive note, what does the analysis above suggest that infrastructure investors can or should do when thinking about whether to invest in a particular project (or when thinking about how to think about whether to invest in a particular project)? The investor should appreciate the conceptual and empirical difficulties involved in political risk assessment prior to investing large sums of money to set up a sophisticated internal political-risk-assessment operation. Assuming that investors should assess political risk is easy, but doing so in a theoretically and empirically sound way is a huge task that should not be entered into lightly. Investors should also be aware that political risk assessment "on the cheap" (for example, by using off-the-shelf indicators) is not necessarily any better (or any worse) than political risk assessment done expensively. Assessment on the cheap can provide the decision-maker with a heuristic useful for forcing decisions, but such assessments are not really risk assessments in any sort of rigorous way.
As to investment law, infrastructure investors should be aware of BITs as potentially relevant to their risk assessments, but they should also be suspicious of claims that BITs are highly effective at reducing risk. For sure, BITs provide infrastructure investors with some legal protections, and in some cases infrastructure investors have used the treaties to sue host states for international law violations. (93) On the other hand, in many cases, an infrastructure investor will be better off negotiating with the host state a high-quality investment (concession) contract that provides the primary legal protection against adverse host state action (or inaction). (94)
(1.) See Ephraim Clark & Bernard Marois, Managing Risk in International Business: Techniques and Applications 34 (1996) (providing the typical definition of risk).
(2.) Stanley Kaplan & B. John Garrick, On the Quantitative Definition of Risk, 1 Risk Analysis 11, 13 (1981). Kaplan and Garrick note that this operationalization equates (misleadingly in their view) low-probability/high-cost events with high-probability/low-cost events. Id.
(3.) See Richard A. Posner, Behavioral Finance Before Kahneman, 44 Loy. U. Chi. L.J. 1341, 1345-46 (2013) (discussing the distinction between risk and uncertainty and citing the classical sources developing the distinction).
(4.) See Alain Chevalier & Georges Hirsch, The Assessment of the Political Risk in the Investment Decision, 32 J. Operational Res. Soc'y 599, 606-07 (1981) (arguing that the distinction between uncertainty and risk is often irrelevant to political risk assessments, which are by necessity often based on subjective impressions and on data of "questionable" reliability and accuracy).
(5.) See, e.g., Yadong Luo, Political Risk and Country Risk in International Business: Concepts and Measures, in The Oxford Handbook of International Business 740 (Alan M. Rugman ed., 2d ed. 2009) (breaking business-relevant risk into various categories, including political risk, and subcategories, such as "ownership, operational, and transfer risk").
(6.) See generally Stephen J. Kobrin, Political Risk: A Review and Reconsideration, 10 J. Int'l Bus. Stud. 67 (1979) (providing citations to numerous political risk articles from the 1960s and 1970s).
(7.) Id. (reviewing various definitions and systems for defining and identifying political risk).
(8.) See, e.g., Theodore H. Moran, Multinational Corporations and the Politics of Dependence: Copper in Chile (1974) (analyzing Chile's 1971 expropriation of foreign copper interests); Robin C. A. White, Expropriation of the Libyan Oil Concessions: Two Conflicting International Arbitrations, 30 Int'l & Comp. L.Q. 1 (1981).
(9.) See Kobrin, supra note 6.
(10.) Id. at 67-69.
(11.) Id. at 73.
(12.) Id. at 67.
(13.) Id. at 68.
(14.) Id. at 77.
(15.) Id. at 69.
(16.) Id. at 68 ("[M]ost managers' understanding of the concept of political risk, their assessment and evaluation of politics, and the manner in which they integrate political information into decision making are all rather general, subjective, and superficial."); see also id. at 74 ("[W]ith very few exceptions,... surveys report the absence of any formal or even rigorous and systematic assessment of political environments and their potential impact upon the firm.").
(17.) Witold J. Henisz & Bennet A. Zelner, Legitimacy, Interest Group Pressures, and Change in Emergent Institutions: The Case of Foreign Investors and Host Country Governments, 30 Acad. Mgmt. Rev. 361, 361 (2005) (observing that "the literature has not yet met the challenge posed by Kobrin [in 1979] to identify 'which events matter' and how 'environmental processes affect investor perceptions'" of political risk).
(18.) See Derek Baas, Approaches and Challenges to Political Risk Assessment: The View from Export Development Canada, 12 Risk Mgmt. 135(2010).
(19.) Id at 136-37.
(20.) Id at 138.
(22.) Id. at 137; see also Clark & Marois, supra note 1, at 53-54 ("[Political risk has not received a clear-cut definition.... In fact ... there is no general agreement on exactly what political risk assessment is supposed to measure.").
(23.) Baas, supra note 18, at 136.
(25.) Howard Kunreuther, A Conceptual Framework for Managing Low-Probability Events, in Social Theories of Risk 301, 308-09 (Sheldon Krimsky & Dominic Golding eds., 1992).
(27.) Baas, supra note 18, at 139.
(28.) Id. at 141 (noting that its efforts at political risk assessment began in the early 1980s and continued until the 2000s).
(29.) Id. at 158.
(30.) See Harry Otway, Public Wisdom, Expert Fallibility: Toward a Contextual Theory of Risk, in Social Theories of Risk, supra note 25, at 215, 221 (noting that the "only risks that can be 'measured' are the relatively uninteresting ones for which there are statistical or epidemiological data," whereas most risks of interest address "'trans-scientific' problems, those for which the methods of science can, in theory, provide answers but which, in practice, because of sample-size constraints, cannot be answered").
(31.) Baas, supra note 18, at 158.
(33.) See Chevalier & Hirsch, supra note 4, at 607 ("[A]ny sophisticated methodology [for assessing political risk] is irrelevant [to actual decisions to invest abroad] as it would be inconsistent with the subjective character and the limited information which is typical of such decisions."). Chevalier and Hirsch nonetheless propose in broad form such a methodology, though they warn that "the information available to the analyst [to implement the methodology] is limited." Id. at 609.
(34.) Kobrin, supra note 6, at 74.
(35.) While my discussion below focuses on off-the-shelf quantitative indicators of political risk, businesses can certainly attempt to implement their own, more complex systems of political risk analysis, which may incorporate such indicators. For a review of the main approaches to political risk assessment that notes significant flaws in each approach, see Clark & Marois, supra note 1, at 69-100.
(36.) PRS Grp., International Country Risk Guide Methodology (2012), available at http://www.prsgroup.com/wp-content/uploads/2012/ll/icrgmethodology.pdf.
(38.) BRS, Bus. Env't Risk Intelligence, http://www.beri.com/Publications/BRS.aspx (last visited Feb. 21, 2014).
(39.) PRS Grp., supra note 36; BRS, supra note 38.
(40.) See, e.g., Baas, supra note 18, at 147 (noting that EDC says that its assessment methodology reflects the obsolescing bargain notion); Clark & Marois, supra note 1, at 85 (discussing an element of obsolescing bargain theory, emphasizing in particular the relative bargaining power of the host state and investor). In addition to bargaining power, Clark and Marois also focus on "the strategic dimension of the investment" and "the foreign firm's relation with the host country." Id. at 85. Those variables reemphasize the project- and firm-specific nature of political risk.
(41.) See discussion supra Part I B
(42.) PRS Grp., supra note 36, at 3.
(43.) Clark & Marois, supra note 1, at 70.
(46.) See John Hood & M. Shahid Nawaz, Political Risk Exposure and Management in Multinational Companies: Is There a Role for the Corporate Risk Manager?, 6 RISK Mgmt. 7, 10 (2004) ("[F]or many MNCs possession of the information on the existence of political risk [as measured by readily available indicators] will not solve the key problems for them, ie how do they measure its likely impact on them, and subsequently manage this [impact]? ... What much of the literature [on political risk] remains silent on ... is how MNCs operationalise this information into their risk management strategy.").
(47.) See supra notes 29-30 and accompanying text.
(48.) See supra notes 29-30 and accompanying text.
(49.) See Michael S. Minor, The Demise of Expropriation as an Instrument of LDC Policy, 19801992, 25 J. Int'l Bus. Stud. 177, 180 (1994) (listing annual acts of expropriation from 1860-1992). On the other hand, a number of recent expropriations in Latin America hint at a reversal of this trend. See Yessika Monagas, U.S. Property in Jeopardy: Latin American Expropriations of U.S. Corporations ' Property Abroad, 34 Hous. J. Int'l L. 455, 456-68 (2012).
(50.) Andrew T. Guzman, Why LDCs Sign Treaties That Hurt Them: Explaining the Popularity of Bilateral Investment Treaties, 38 Va. J. Int'l L. 639, 664 (1998) ("Because the costs of outright expropriation are likely to be high, the more moderate course of extracting value from the firm without forcing divestment... may be attractive. This can be done in a wide variety of ways, including changing the tax rate, restricting the repatriation of profits, imposing new labor or local content requirements, and so on. This approach, which is sometimes called 'creeping expropriation,' allows the country to take advantage of the existing management and their skills, thus avoiding the major costs of an outright expropriation, while still extracting value from the enterprise. Creeping expropriation may also be preferred because it is less likely to provoke significant sanctions by the home country of the investor. After all, the firm's assets have not been seized and it is often difficult to identify where the right of a government to set policy crosses over into unreasonable conduct." (footnote omitted)).
(51.) See, e.g., Shain Corey, Note, But Is It Just? The Inability for Current Adjudicatory Standards to Provide "Just Compensation" for Creeping Expropriations, 81 Fordham L. Rev. 973, 976 (2012) ("[T]here has recently been a large increase in the number of indirect expropriations, particularly creeping expropriations."). Corey does not cite any empirical evidence that this is indeed the case, however.
(52.) Nathan M. Jensen, Democratic Governance and Multinational Corporations: Political Regimes and Inflows of Foreign Direct Investment, 57 Int'l Org. 587, 592 (2003).
(53.) Id. at 594.
(54.) Id. at 596-97.
(55.) Id. at 601 ("The empirical results... provide solid evidence of the positive effect of democracy on FDI inflows.").
(56.) Witold J. Henisz, The Institutional Environment for Economic Growth, 12 Econ. & Pol. 1 411 (2000) [hereinafter The Institutional Environment for Economic Growth]; Witold J. Henisz, The Institutional Environment for Multinational Investment, 16 J.L. Econ. & Org. 334, 346-47 (2000) [hereinafter The Institutional Environment for Multinational Investment],
(57.) The Institutional Environment for Economic Growth, supra note 56, at 22-23.
(58.) The Institutional Environment for Multinational Investment, supra note 56, at 350-60.
(59.) Alternatively, a correlation between democracy and foreign investment decisions could indicate that democracy is a proxy for some variable or indicator other than democracy that investors actually do assess and use as a basis for decision-making.
(60.) See, e.g., Kirt C. Butler & Domingo Castelo Joaquin, A Note on Political Risk and the Required Return on Foreign Direct Investment, 29 J. Int'l Bus. Stud. 599, 600 (1998) ("Political risk is often defined as the risk of adverse consequences arising from political events.... However, a definition of political risk that focuses only on negative outcomes is inadequate in identifying the impact of political risk on the firm's cost of capital. For this purpose, we need a definition of political risk that captures both positive and negative consequences of political events."); Hood & Nawaz, supra note 46, at 9 ("A further complication in the definition of political risk is that most approaches are predicated on the notion that such risk is invariably negative. Such approaches fail to recognise that political developments can have a positive effect on MNCs").
(61.) See, e.g., Benjamin K. Guthrie, Beyond Investment Protection: An Examination of the Potential Influence of Investment Treaties on Domestic Rule of Law, 45 N.Y.U. J. Int'l L. & Pol. 1151, 1159-60 (2013).
(62.) See Jason Webb Yackee, Conceptual Difficulties in the Empirical Study of Bilateral Investment Treaties, 33 Brook. J. Int'l L. 405, 416-17 (2008) (providing an overview of BITs).
(64.) For good historical, doctrinal, and theoretical overviews of BITs, see generally United Nations Conference on Trade and Development [UNCTAD], Bilateral Investment Treaties: 1959-1999,
UNCTAD Doc. UNCTAD/ITE/IIA/2 (2000), available at http://unctad.org/en/docs/poiteiiad2.en.pdf; Kenneth J. Vandevelde, Bilateral Investment Treaties: History, Policy, and Interpretation (2010); Kenneth J. Vandevelde, A Brief History of International Investment Agreements, 12 U.C. DAVIS J. Int'l L. & Pol'y 157 (2005).
(65.) Anna T. Katselas, Do Investment Treaties Prescribe a Deferential Standard of Review?, 34 Mich. J. Int'l L. 87, 106 (2012) (noting the existence of "nearly three thousand BITs").
(66.) Stephan W. Schill, Enhancing International Investment Law's Legitimacy: Conceptual and Methodological Foundations of a New Public Law Approach, 52 Va. J. Int'l L. 57, 63-67 (2011) (describing why BITs have proven controversial).
(67.) Guzman, supra note 50, at 688.
(68.) See, e.g., Jason Webb Yackee, Do Bilateral Investment Treaties Promote Foreign Direct Investment? Some Hints from Alternative Evidence, 51 Va. J. Int'l L. 397 (2010).
(69.) See, e.g., Eric Neuntayer & Laura Spess, Do Bilateral Investment Treaties Increase Foreign Direct Investment to Developing Countries?, 33 World Dev. 1567,1582 (2005).
(70.) See, e.g., Jeswald W. Salacuse & Nicholas P. Sullivan, Do BITs Really Work?: An Evaluation of Bilateral Investment Treaties and Their Grand Bargain, 46 Harv. Int'l L.J. 67, 105 (2005); Jason Webb Yackee, Bilateral Investment Treaties, Credible Commitment, and the Rule of (International)
Law: Do BITs Promote Foreign Direct Investment?, 42 Law & Soc'y Rev. 805, 827-28 (2008). There are a number of methodological challenges that afflict empirical studies of the BIT-FDI relationship. A full detailing is beyond the scope of this Article, but at a minimum, it should be appreciated that FDI data quality is notoriously poor, especially across time and across developing countries; that FDI data does not allow sector-specific tests of BIT effectiveness; and that it does not allow much examination of whether BITs increase bilateral (as opposed to aggregate) FDI flows. On some of the weaknesses of FDI data, see Jimmy J. Zhan, FDI Statistics: A Critical Review and Policy Implications (2006), available at http://www.waipa.org/pdf/SurveyResults/Problems_with_FDI_statistics.pdf.
(71.) PRS Grp., supra note 36, at 3-4.
(73.) Showing that BITs are correlated with off-the-shelf indicators of political risk does not show that BITs reduce political risk in an objective sense. Analyses of the kind illustrated in Figure 1 can only shed light on whether the formulators of subjective estimates of risk use the presence or absence of BITs when making their evaluations. In other words, analyses along the lines of Figure 1 have the capacity to reveal what the ICRG experts think about BITs and risk, not whether BITs actually reduce risk. Whether the experts' thoughts are likely to be objectively right depends on how much faith one has in their expertise.
(74.) This is especially a problem where the definition of political risk is quite broad. See, e.g., Butler & Joaquin, supra note 60, at 599 (defining political risk as "the risk that a sovereign host government will unexpectedly change the 'rules of the game' under which businesses operate"). International investment law offers protections against changes in the "rules of the game" only in certain circumstances, however: for example, where the change violates a contractual provision or where it suffers from some other vice, such as impermissible discrimination.
(75.) Leon E. Trakman, The ICSID Under Siege, 45 Cornell Int'l L.J. 603, 655 (2011).
(76.) Stephen E. Blythe, The Advantages of Investor-State Arbitration as a Dispute Resolution Mechanism in Bilateral Investment Treaties, 47 Int'l Law. 273, 275 (2013).
(77.) George K. Foster, Recovering "Protection and Security": The Treaty Standard's Obscure Origins, Forgotten Meaning, and Key Current Significance, 45 Vand. J. Transnat'l L. 1095, 1097-98 (2012).
(78.) Mat 1109-10.
(79.) David Gaukrodger & Kathryn Gordon, Investor-State Dispute Settlement: A Scoping Paper for the Investment Policy Community 19 (Org. for Econ. Co-operation & Dev. Inv. Dev., OECD Working Paper on International Investment No. 2012/03, 2012), available at www.oecd.org/daf/ investment/workingpapers ("[L]egal and arbitration costs for the parties in recent ISDS cases have averaged over USD 8 million with costs exceeding USD 30 million in some cases. In the recent Abaclat decision (which addresses jurisdiction but not the merits), the tribunal noted that the claimants had spent some USD 27 million on their case to date, and that Argentina had spent about USD 12 million." (footnote omitted)).
(80.) Id. at 17 ("Small investors can be [BIT arbitration] claimants.... Far from supporting the view that investment arbitration is not an option for smaller investors, the survey shows that 22% of the claimants in both ICSID and UNCITRAL cases are either individuals or very small corporations with limited foreign operations..."). These results suggest that the high costs of investment treaty arbitration may not pose serious "access to justice" issues. On the other hand, Gaukrodger & Gordon do not examine how successful these claims by small investors actually were. Given the presumably meager legal budgets of such entities, these BIT lawsuits may be quixotic in character.
(81.) Katselas, supra note 65, at 109 ("Given the relative and recent prominence of investment treaty arbitration, the absence of binding precedent, and the fact that tribunals interpret various treaties rather than one statute, it is not surprising that the meanings of the treaty standards are not yet clear....").
(82.) See Susan D. Franck, Empirically Evaluating Claims About Investment Treaty Arbitration, 86 N.C. L. Rev. 1, 49-50 (2007).
(83.) The government's pervasive regulatory role in modern society is captured in Rubin's notion of the "administrative state." See Edward L. Rubin, Beyond Camelot: Rethinking Politics and Law for the Modern State (2005).
(84.) For a review of the debate, see generally Rahim Moloo & Justin Jacinto, Environmental and Health Regulation: Assessing Liability Under Investment Treaties, 29 Berkeley J. Int'l L. 1 (2011).
(85.) See, e.g., U.S. Dep't of State et al., 2012 U.S. Model Bilateral Investment Treaty annex B, at 41, available at http://www.state.gov/documents/organization/188371.pdf (clarifying that expropriation as defined in the treaty is "intended to reflect customary international law" and that nondiscriminatory regulations will only in undefined "rare circumstances" amount to an expropriation); see also Joel C. Beauvais, Note, Regulatory Expropriations Under NAFTA: Emerging Principles and Lingering Doubts, 10 N.Y.U. Envtl. L.J. 245, 248 (2002) (reviewing regulatory takings cases and finding that NAFTA tribunals have approached the doctrine "relatively conservatively, placing significant limitations on the scope of government regulation subject to it"); Moloo & Jacinto, supra note 84, at 24 (describing the standard in practice for finding a regulatory expropriation).
(86.) For an overview of these so-called "non-precluded measures" provisions, see generally William W. Burke-White & Andreas von Staden, Investment Protection in Extraordinary Times: The Interpretation and Application of Non-precluded Measures Provisions in Bilateral Investment Treaties, 48 Va. J. Int'l L. 307 (2008).
(87.) U.S. Dep't of State et al., supra note 85, art. 18.
(88.) Id. art. 20.
(90.) Mart. 21.
(91.) Jason Webb Yackee, Do We Really Need BITs? Toward a Return to Contract in International Investment Law, 3 Asian J. WTO & Int'l Health L. & Pol'y 121, 128-36 (2008) [hereinafter Do We Really Need BITs?]; Jason Webb Yackee, Pacta Sunt Servanda and State Promises to Foreign Investors Before Bilateral Investment Treaties: Myth & Reality, 32 Fordham Int'l L.J. 1550, 1589 (2009).
(92.) Paul E. Comeaux & N. Stephan Kinsella, Protecting Foreign Investment Under International Law: Legal aspects of political Risk 130 (1997) (discussing other risk-mitigation strategies, such as arranging project financing through a host state or international organization, organizing the project so that the investor holds the host state "hostage," or obtaining political risk insurance).
(93.) For example, the International Centre for the Settlement of Investment Disputes' (ICSID) latest case statistics show that eighteen percent of all cases registered under the ICSID Convention or the ICSID Additional Facility Rules involve the electric power and water sectors. ICSID CASELOAD--STAT. (Int'l Ctr. for Settlement of Inv. Disputes, Washington, D.C.), no. 2013-2, 2013, at 12, available at https://icsid.worldbank.org/ICSID/FrontServlet?requestType=ICSIDDocRH&actionVal=Show Document&CaseLoadStatistics=True&language=English42.
(94.) Comeaux & Kinsella, supra note 92, at 133-49 (providing advice on the drafting of investment contracts, which they believe potentially reduce political risk through various mechanisms); see also Do We Really Need BITs?, supra note 91, at 121 (arguing for an international investment law system that relies heavily on contract).
(95.) PRS Grp., International Country Risk Guide Researchers Dataset (2013) (on file with author); IIA Databases, UNITED Nations Conf. on Trade & Dev., http://unctad.org/en/pages/ DIAE/International%20Investment%20Agreements%20(IIA)/IIA-Tools.aspx (last visited June 22, 2014).
Jason Webb Yackee, JD, Duke University School of Law, 2004; PhD, Political Science, University of North Carolina-Chapel Hill, 2007; Associate Professor, University of Wisconsin Law School. I may be contacted at email@example.com.
|Printer friendly Cite/link Email Feedback|
|Author:||Yackee, Jason Webb|
|Publication:||Duke Journal of Comparative & International Law|
|Date:||Mar 22, 2014|
|Previous Article:||Regulating sovereign wealth funds through contract.|
|Next Article:||Comparative criminal justice: an institutional approach.|