Political risk for multinational companies:

Empirical evidence from a new dataset

by
Jo Jakobsen

Jo Jakobsen, PhD Associate Professor Department of Sociology and Political Science Norwegian University of Science and Technology NO-7491 Trondheim Telephone: +47 735 51141 Fax: +47 735 91564 [email protected]

Abstract There is much to suggest that political risk for foreign direct investors has become an increasingly complex phenomenon. To vindicate this claim, though, we need systematic empirical evidence. This paper offers just that. It seeks to increase our knowledge of political risk by means of presenting a new, explorative dataset. The data, which are based on prominent news reports, contain information on 332 cases of realized political risk in developing economies, 19982005. The paper first develops a theoretical model of the causal relationships between sources and effects. Therafter, the dataset is carefully examined and descriptive statistics are presented. The 332 reported events of political risk show such a variety, with respect to effects, sources, actors, and industries, that a multidimensional framework appears highly warranted. In particular, it seems that the host government nowadays has at its disposal a wide range of fairly subtle means by which it can secure for itself or the nation a bigger slice of the FDI revenue pie.

Introduction
Political risk for multinational companies is a salient phenomenon. Consider, for example, the recent nationalizations and forced contract renegotiations faced by foreign investors in Bolivia and Venezuela; or the kidnappings that are routinely suffered by foreign employees in places like Nigeria, Yemen, Colombia, and the Gulf of Aden; or the lackluster judicial protection that multinational companies (MNCs) enjoy in many parts of the world. Consider also events in the realm of global politics, like international terrorism, or the sanctions against resilient Iran that have recently hurt a large number of energy multinationals. So the risks to international business are clearly still there, as this paper will argue. But the paper also argues that the contents of the phenomenon of political risk have changed from the first half of the 1970s (the natural point of reference for a political risk analyst). Whereas foreign investors used to worry about expropriations and nationalizations first and foremost, their worries have increased over the years in the sense that political risk has evolved to become much a more complex characteristic of the (arguably much more complex) international business environment. The events causing harm to investors the political risk effects vary greatly, as do the actors responsible for realizing the risk, and as do the root sources of the risks. Political risk, I hold, is a multidimensional phenomenon. That is the argument and point of departure of the present paper. Yet establishing some reasonably firm testimony to the effect that socio-political risk is multifaceted surely requires us to go beyond the mere presentation of scattered anecdotal examples. Systematic empirical evidence is key. The paper offers just that. It seeks to increase our knowledge of political risk by means of presenting a new, explorative dataset. The data, which are based on prominent news reports, contain information on 332 cases of realized political risk in developing countries, from 1998 to 2005. In addition to exhibiting information about the risk
2

effects, I also coded the probable source or sources of each event, along with actors and affected companies or industries. After the next main sections presentation of the theoretical model, this dataset is carefully examined (as is the methodology employed), and descriptive statistics are presented. The results of the empirical analysis generally meet the expectations put forth. The 332 reported events of political risk show such a variety, with respect to effects, sources, actors, and industries, that a multidimensional framework seems highly warranted. Regarding effects or actual outcomes, for example, some 48 percent of the incidents are associated with government intervention; 39 percent relate to war and social unrest; while 13 percent of risks were realized courtesy of actions taken by either activists (like NGOs) or MNC partners or competitors. Within each these main categories, moreover, several different forms of intervention are recorded. In particular, it seems that the host government nowadays has at its disposal a wide range of fairly subtle means by which it can secure for itself or the nation a bigger slice of the FDI revenue pie. Today, creeping expropriation is surely more widespread than the outright version of government takeovers of foreign-owned businesses. The sources of these events are also varied: socio-political instability and grievances, political institutions, and the preferences and attitudes of key host-country players all seem to be prominent causes of risk. What is more, these sources often combine to produce a specific outcome, lending further credence to the view that a broad-based framework is appropriate.

Political risk a causal framework

Political risk a definition and an example Political risk can be defined in a concise way as the impact of politics on markets (Bremmer, 2005, p. 52), or as political events and processes that can negatively affect doing
3

business (Alon & Herbert, 2009, p. 127). Alternatively, we can employ a less succinct definition, but one that more or less encompasses a similar wide range of events, as others do (for example, Jakobsen, 2010) and as this paper does: here, political risk is conceived as those events, actions, processes, or characteristics of a socio-political nature that have the potential to directly or indirectly significantly and negatively affect the goals of foreign direct investors. Applying such a broad definition ensures that the concept of political risk embraces a multitude of incidents. Thus, cases of political risk will differ markedly with respect to the actors involved, the particularities of each harmful event, the root sources of investor trouble, and the firms or industries affected. Take, for example, the notorious case of Bolivia. On May 1st 2006, Bolivias left-wing president, Evo Morales, kick-started a process of nationalization of foreign-owned petroleum firms that has continued to this day.1 This came about after the country, three years earlier, had been rocked by a series of violent riots and demonstrations by ordinary, poor citizens and radical leftist groups alike. The unrest forced the resignation of the nations pro-American president, Gonzalo Snchez de Lozada. The purported bone of contention was a planned project under which the landlocked countrys abundant natural-gas reserves were to be exported to the United States via Chile (Bolivias arch enemy) by multinational companies.2 This project was soon put on hold, though, courtesy of the sociopolitical turmoil.3 Moreover, in a July 2004 referendum on the gas issue, the people of Bolivia

Now its the peoples gas, The Economist, 4th May 2006. Highly flammable, The Economist, 11th September 2003; On the brink, The Economist, 16th October 2003;

An angry people bring down their president, The Economist, 20th October 2003.
3

Local dissent hampers export plans, Lloyds List, 24th October 2005.

voted for a 32 percentage point increase in royalties for foreign companies.4 Since that vote, petroleum firms have seen their tax bills increase and ownership shares diminish even further.5 Similar troubles have later been bestowed upon mining and power firms.6 In Bolivia, therefore, a series of discrete yet interconnected events proved troublesome for the international companies. The political risk effects or outcomes that is, the acts of intervention that directly harmed the foreign companies consisted of tax and royalty increases, forced contract revisions, the sale or transfer of ownership, and project delays. The actors responsible for realizing the risk also varied. It was ultimately the politicians that is, the president and the Congress who ordered the tax hikes and the nationalizations. But other socio-political actors were also highly influential in the process, like left-wing radicals and even ordinary citizens. The riots, to be sure, were but one of many root causes or sources of government intervention against the foreign companies operating in the country. Another source was the frequent changes of political leadership, where each change of president led to increased pressure against investors. More abstract forces were also at play, like Bolivians hostility toward Chile. But perhaps the most salient issue at stake was, and indeed still is, the extent to which foreigners should be allowed to control or exploit Bolivias natural resources the nations patrimony. Sentiments of economic nationalism loomed large and surely helped realize the risk for foreign companies. Yet, while economic nationalism, frequent changes in political leadership, and foreign policy grievances can undoubtedly

Counter-reform or muddle along, The Economist, 15th July 2004; Bolivias gas vote fires fierce debate,

BBC News, 17th July 2004; Mesas middle ground, The Economist, 22nd July 2004.
5

Protests fail to stop Bolivia law, BBC News, 18th May 2005; Mocking pilate, The Economist, 19th May

2005.
6

The government renationalises a big tin smelter, The Economist, 19th February 2007; Another Bolivian

nationalisation, The Economist, 6th May 2010.

precipitate, and ultimately cause, harmful intervention against foreign investors as they clearly have done in Bolivia these characteristics and issues do not by themselves create trouble for foreign investors. They can, however, be regarded as typical independent variables in political risk analysis. The dependent variables are the actual interventions by influential host-country actors. For analytical purposes, this distinction between independent and dependent variables or between sources and effects is vital.

The causal chain in political risk analysis Each possible country example of political risk usually exhibits its own unique combination of sources, effects, and actors. In Bolivia, the combination was particularly complex. In other nations, we find different mixes: In Nigeria, kidnappings and acts of sabotage directed against the oil industry by rebel groups or grievance-ridden local communities constitute the main risk; in Iran, it is currently mostly about economic sanctions being implemented by foreign states because of Tehrans purported nuclear ambitions; in Burma, a military dictatorship, the actions of (often foreign-based) activists have caused severe problems for Western multinationals. Apart from emphasizing the distinction between sources of risk and risk effects, the main point I am trying to make is that political or country risk is a multidimensional phenomenon. The events causing harm to investors the risk effects vary greatly, as do the actors responsible for realizing the risk, and as do the root sources of the events.

---------------------------------------- FIGURE 1 IN HERE ---------------------------------------------

Figure 1 illustrates in a simple manner what I mean. To the right in the figure are the political or country risk effects or outcomes. Even if this concept covers a broad range of
6

incidents, all of which directly affect the profits or other goals a company may have, I find it expedient to separate between three broad categories. Each of the categories essentially denotes both the particular types of intervention in question and the main actor(s) at play: (a) government intervention in or regulation of business (including intervention by the parliament, bureaucracy, and judiciary); (b) acts of intervention relating to war, terrorism, and social unrest; and (c) other acts of intervention committed by non-state actors (for example, NGO activism and fraudulent bahavior by domestic businesses).

Political risk effects Government intervention Many scholars seem to acknowledge that policy changes and government intervention constitute the most important class of political risk outcomes (Comeaux & Kinsella, 1997; Graham, 1996; Henisz, 2000; Jensen, 2006; OSullivan, 2005). The most well-known form of government intervention is surely expropriations and nationalizations. The period ranging from the latter half of the 1960s until the mid-1970s the expropriation era is particularly (in)famous (Jodice, 1980; Kobrin, 1984). This period did not last for long, however. As the 1980s went by, expropriations and nationalizations seemed gradually to lose their appeal as a policy tool for LDC governments (Cass, 2007; Minor, 1994). This trend has more or less continued until today. The few recent instances of outright takings of MNC property cue Bolivia and Venezuela should perhaps be regarded as expections to the rule that most contemporary acts of government intervention in the affairs of multinationals are of a more subtle kind than during the heyday of forced takeovers (Bthe & Milner, 2008, pp. 741-744; Oetzel, 2005; Ratner, 2008). Indeed, many governments quite simply aim to secure for themselves or their nation a larger share of the proceeds from the foreign investment once assets are sunk in the host country (Bthe & Milner, 2008, pp. 743-744; Jensen, 2006, p. 47).
7

This can be done in a variety of ways, for example by breaching contracts or demanding that they be renegotiated; by increasing corporate taxes or royalties; by imposing bans or taxes on trade, production, or investment; or, as in Argentina after the 20012002 financial crisis, by implementing rigid price controls on specific industries.

Acts relating to war, terrorism, and social unrest The second category of political risk effects acts of intervention relating to war, terrorism, and social unrest is also a broad category that encompasses a wide variety of incidents. Moreover, data from Overseas Private Insurance Corporation (OPIC), a U.S. political risk insurance entity, reveal that political violence claims have predominated in recent years, accounting for 59 percent of claims paid in the business from 1991 to 2004 (OSullivan, 2005, p. 30). Nowhere is this issue more topical than in the Niger Delta, where recent years have seen hundreds of oil workers being kidnapped by armed rebels and gangs of local youths allegedly angry with the lack of an equitable distribution of the petroleum wealth. Similar problems continue to haunt politically unstable countries like Iraq, Yemen, Niger, Sudan, and Colombia (Guqueta, 2003; Richani, 2005) as well as shipping traffic in the waters outside Somalia (Warren, 2011).

Other acts committed by non-governmental actors The last category of political risk effects a residual class consists of other acts of intervention committed by non-state actors, like activists and domestic businesses. The risk that stems from other firms, usually indigenous ones, can be regarded as political insofar as the firms seek to exploit superior knowledge of and access to political or judicial decisionmaking processes (Khanna, Palepu, & Sinha, 2005). Neither should one underestimate what is often described as a new risk that has particularly come to the fore with the globalization of
8

media and the advent of the Internet. This is threat of activism from non-governmental organizations (NGOs), consumers, shareholders, local communities, workers and labor unions, and the media (Skippari & Pajunen, 2010; Vachani, Doh, & Teegen 2009). Often such actors operate in concert, creating a uniform pressure that eventually leads to serious reputational damage to the company and/or a decision to divest from the host country.

Sources of political risk Yet, information on effects and the preferences and motivations of actors is far from sufficient; one must also acquire an understanding of the root sources of political risk effects and how these sources in the last instance can affect the company. As Figure 1 illustrates, four main categories of root sources of realized political risk can be extracted from the literature: (a) the obsolescing bargain mechanism; (b) socio-political instability and grievances; (c) political institutions; and (d) preferences and attitudes.

The obsolescing bargain mechanism The first key theoretical contribution stems from the literature that deals with changes in the bargaining relationship between the firm and the host country. It was Raymond Vernon (1971) who, some four decades ago, pinpointed the mechanisms that often make it rational for host governments to change terms of contract ex post. The so-called obsolescing bargain model (OBM) states that large foreign direct investment (FDI) deals are prone to later revisions that hurt multinationals. This is due to the hostage effect that ensues once capital is sunk: the firm cannot, before an investment is made, make credible threats that it will withdraw if the government reneges on pre-investment promises given the relative immovability of its assets. Therefore, the relative bargaining power of the firm decreases over time (Moran, 1974; Vernon, 1971).
9

Numerous real-world examples support this notion. The most renowned treatment of the OBM is arguably Morans (1974) study of the nationalization of the copper industry in Chile. Vernons model also offer a valid (part) explanation of many of the forced takovers in the 1960s and 1970s (Jodice, 1980; Kobrin, 1984). Others have used the model to explain interventions in specific sectors of the economy, like manufacturing (Vachani, 1995) and infrastructure (Ramamurti, 2001; Wint, 2005). Recent events in petroleum-rich nations like Venezuela, Bolivia, Ecuador, Russia, and Chad perhaps bolster the claim that, in principle, the mechanism is valid irrespective of time period and host-government ideology (Jakobsen, 2006; Ramamurti, 2001; Wint, 2005).

Socio-political instability and grievances The second main source of political risk socio-political instability and grievances is a fairly broad category encompassing civil war, terrorism, ethnic and religious tensions, projectspecific grievances, international conflict, and interstate warfare. Although firms with investments in nations that are particularly exposed by virtue of their rogueness (Iran is a prominent example) risk being caught in the web of international disputes, intrastate political instability is surely a more common and relevant phenomenon for most MNCs (Harbom & Wallensteen, 2005, p. 627). Most risky are all-out civil wars. These can affect multinationals in both direct and indirect ways. The impact is indirect when a conflict causes a general reduction in economic activity in the host country; the impact is direct when acts of organized violence consciously target economic actors. The latter is often the case in countries where rebel groups financial viability becomes dependent on criminal activities such as looting, kidnapping, and sabotage (with MNCs being forced to pay ransoms or protection money) (Collier & Hoeffler, 2004; Richani, 2005; Ross, 2006). In other cases, the ultimate goal of a rebel group is (or eventually becomes) to gain control over some valuable natural resource
10

that is controlled by foreigners (Regan, 2003; Ross, 2004, 2006; Snyder & Bhavnani, 2005). Such outcomes are explained by the so-named greed theory of collective violance by reference to what are believed to be typical motivations of rebels (Collier & Hoeffler, 2004; de Soysa, 2002; Fearon & Laitin, 2003). According to this school of thought, rebellions commence and last first and foremost because they are beneficial to specific groups in society. To assemble and sustain armed organizations is a costly undertaking that requires proper financing; hence the lure of targeting resourceful multinationals (Le Billon, 2001; Richani, 2005; Ross, 2004) .

Political institutions Thirdly, political institutions perhaps constitute the dimension of sources of risk that the literature overall has attended to most closely in recent years. A widespread argument that draws on the more general veto-player theory (Leeds, 1999; Leeds, Mattes, & Vogel, 2009; Tsebelis, 1995) states that a poorly developed institutional framework reduces the credibility of governments promises to protect private propery rights (Ahlquist, 2006; Fowler, 2006; Henisz, 2000; Jensen, 2006, 2008; MacIntyre, 2001). Such institutional traits as democracy, separation of powers, rule of law, and bureaucratic efficiency reduce transaction costs and, not least, the risk of government intervention. More specifically, one could argue, as many do, that in countries where the political system is based on the principle of checks and balances, the executive power enjoys little scope to undertake arbitrary or unlawful intrusions into the affairs of foreign businesses. A range of recent empirical studies lend support to the vetoplayer explanation, primarily by way of illustrating that international investors, ceteris paribus, prefer to locate in democratic countries with well-functioning power-separation mechanisms (Ahlquist, 2006; Henisz, 2000; Jensen 2006, 2008; MacIntyre, 2001).

11

Attitudes and preferences The fourth main dimension of sources of political risk may be labeled attitudes and preferences. The inclinations and sentiments of prominent host-country actors clearly affect how foreign investors will, or should expect to be treated. Of course, the generel trend in developing nations goes in the direction of economic liberalization and investor friendliness (Kobrin, 2005; Simmons, Dobbin, & Garrett, 2006), though one should not exaggerate the impact this has on the level of political risk. This is so not least because trends, attitudes, and preferences tend to change quicker than political institutions and socio-political instability, as recent developments in a handful of South American countries indeed indicate (Margheritis & Pereira, 2007). To this it can be added that economic nationalism does not seem to vanish with growth and development. One explanation of this can be that the mechanisms that ensure that indigenous firms are usually preferred over foreign ones often have much to do with the narrow and enduring interests of specific rent-seeking groups in society, in particular domestic capital (Crystal, 1998; Gilpin, 1987; Jakobsen & de Soysa, 2006; Jakobsen & Jakobsen, 2011; Milner & Kubota, 2005). To domestic firms, in particularly the laggard ones, the greatest threat should come in the form of efficient multinationals with superior ownership advantages. It may therefore be perfectly rational for domestic capital owners even in rich countries to try and lobby their government for protection; and in so far as they succeed, economic nationalism will ensure (Crystal, 1998; Oetzel, 2005). Figure 1 and the subsequent discussion of sources and effects provides the backbone and point of departure of the coding scheme that was used to guide the process of data collection for this study. The next main section elaborates on this, outlining the coding practice in more detail. The main section thereafter presents the results from an analysis of 332 cases of realized political risk in developing countries, 19982005.
12

Methodology
This section presents the methodology and coding practice used to create the 332-case dataset. The data compilation strategy draws on other studies that use similar designs but whose scope and focus are narrower with respect to risk variables, industries, or countries (Crowley & Loviscek, 2002; Henisz & Zelner, 2005; Jagersma & van Gorp, 2003; Jodice, 1980; Kobrin, 1980, 1984). While casting a broader net than the aforecited studies, the present paper also relies on secondary sources for its data collection. Specifically, two prominent news journals that are truly global in scope BBC News World (online edition)7 and The Economist8(online edition) are systematically scanned with the aim of creating a reasonably representative sample of political risk for multinational companies.9 The resulting dataset yields information on political risk effects, sources of political risk, actors involved in realizing the risk, industrial sector, the companies affected, and the year(s) in question. I only focus on developing economies, as defined by the United Nations Conference on Trade and Development in 2002, roughly the mid-period of the data sample (UNCTAD, 2002).10

Coding guidelines Several decision rules were employed in the coding process. To start with, the basic unit of analysis is the act (of intervention by a socio-political player). That is, when the host

I used BBC Newss search engine at http://www.bbc.co.uk/news/world/#blq-search. I collected the data from the The Economists subscription-based website at http://www.economist.com. The majority of cases included in my dataset essentially stem from reports by the BBC News. Offering a more

comprehensive treatment of many of the events, The Economist still provides much valuable supplementary information.
10

The dataset can be obtained from the author upon request.

13

government, a rebel group, an NGO, or some other relevant actor intervenes and thereby obstructs the goals of one or more multinational companies, this counts as an act and is duly recorded. This follows Kobrins (1980, 1984) expropriation dataset in that each affected firm is not necessarily treated as a separate case, so as to avoid gross over-counting that does little to enhance our understanding of the political risk phenomenon. The news sources are systematically scanned for incidents of downside political risk. If an event is to be included in the dataset, a basic requirement is that it must have been unequivocally reported by the BBC News or The Economist, or both. If deemed necessary, however, once an act of intervention or dispute is recorded, I supplement the information yielded by these two base sources with reports from other news agencies as well. Furthermore, following the expropriation literature I also include cases that are later resolved, provided that they initially seemed to present major problems for the affected firms. The decision by an MNC to withdraw from the host country is also counted, even if the firms profits or employees are not directly hurt by the event in question, so long as the divestment is essentially involuntary and related to socio-political sources or actions. I also include in my data threats to multinationals, provided that they prompt some sort of significant response (notably, a withdrawal) from the investor. A specific incident of political risk is assigned only one code or value. This is not the case, though, when I record sources and actors. A certain political risk effect can surely have more than one source, and several host-country actors can be involved in producing a particular effect. As for the categories, two-stage hierarchies are employed; that is, the coding scheme is divided into main dimensions as well as sub-categories. Loosely following Poynter (1985, p. 32), I specifically consider as a case of realized political risk any event, action, or dispute that exhibits any inherent (socio-)political content (with regard to source, effect, or actor) and ultimately obstructs the goals of a foreign direct
14

investor. Defined in this manner, political risk becomes a broad term encompassing sundry and disparate events. Still, there will always be some borderline cases that are difficult to classify. Briefly stated, cases were excluded from the dataset if one or more of the following applied: only scant information on the particularities of the event is present; the available evidence suggests that the risk or dispute should be deemed commercial or market-related rather than political; the risks are essentially not man-made (see Miller, 1992); the evidence indicates that political motives are absent or negligible; no change in the MNC-host country interface has taken place, even if foreign investors are ultimately affected (the mere existence of investor-unfriendly regulation is therefore by itself insufficient to warrant inclusion in the dataset); the affected firm is not operating across national borders; and multinational companies are only indirectly affected by an event (like the badly hurt airlines or tourism industry after 9/11).

Results and analysis

Political risk by country This section presents descriptive statistics of the data. Table 1 ranks the countries according to the total number of events (only countries with six or more incidents are shown). Of course, most reported cases occurred in countries with a comparatively large stock of FDI and, hence, a bigger pool of potential victims of political risk. Table 1 should therefore be interpreted with caution as it does not weigh the number of incidents in a country by that countrys total stock of FDI.

------------------------------------------- TABLE 1 IN HERE --------------------------------------------

15

Nigeria tops the chart with 35 reported incidents of realized political risk in the period, nearly all of which involved oil firms and were directly associated with war, rebel activity, or social unrest (that is, the category labelled War in the table). Foreign companies in Iraq faced similar problems: a total of 20 war-related incidents are recorded for the period, most of them occurring in 2003 and 2004, the first two years of U.S. occupation (see also Table 2). My sample also suggests that kidnappings and acts of sabotage directed toward foreign investors were relatively common in Colombia (12 reported cases) and Indonesia (10). Concerning the distribution of incidents of government intervention, which my data show to be the most common form of risk effect, China tops the table. Notable is the wide variety of interventions that characterizes this prime FDI destination: in the period under investigation, investors have suffered price controls, caps on profits, corporate tax increases, bans on investment, ex-post ownership restrictions, and contract breaches. These are all events that are relatively low-key and perhaps reflective of a complex, unfamiliar, and highly bureaucratic investment environment in which the Chinese government and the Communist Party are by far the most important stakeholders (Fan, Morck, & Xu, 2009). Likewise, interventions by the host government in the affairs of multinationals have been quite common in Venezuela (12 registered cases), India (11), Indonesia (10), Russia (10), Argentina (9), Iraq (9), Iran (7), Bolivia (6), and South Africa (6). Myanmar (Burma) tops the residual list of risks where non-governmental activists or other businesses are the main culprits. The Nongov category by and large denotes incidents that are directly related neither to the actions of host governments nor to war, terrorism, or social unrest. My sample includes five cases from Myanmar where the investors were either forced to withdraw from the host country following severe pressure from activists (this concerned Premier Oil, British American Tobacco, PricewaterhouseCoopers, and Triumph International) or had to undergo costly litigation processes (Unocal and Total). Oil companies
16

operating in Sudan bowed to similar pressure: in 2003, both Canadas Talisman and Austrias OMV chose to pull out from the war-ridden country after having long been accused of indirectly fueling the civil conflict.11

-------------------------------------------- TABLE 2 IN HERE -------------------------------------------

The sample distribution of political risk effects among LDCs is largely as expected and seemingly conforms to popular notions about the relative riskiness of different FDI destinations. An expected pattern also emerges in Table 2. In the far-right column, the country with the largest number of recorded events in a given year is shown. Reflective of increased guerrilla activity by left-wing groups FARC and ELN, Colombia tops the statistics for 1998. The year after, Nigeria recorded the largest number of incidents, while Indonesia under the unstable post-Suharto transition period had the most cases of realized political risk in 2000 and 2001. In fact, Indonesia suffered four consecutive years of net FDI outflows from 1998 to 2001 (UNCTAD, 2004, p. 370). Argentinas FDI numbers also dwindled considerably in the early years of the new century, following the financial collapse and the many cases of government intervention in the infrastructure sector. Then, in 2005, after Iraq had topped the chart the two previous years, foreign investors in Venezuela really started feeling the effects of the Bolivarian Revolution, as President Hugo Chvez proceeded to increase taxes and slash foreign ownership in a number of industries.

The distribution of political risk effects

Oil giant exits Sudan, BBC News, 31st October 2002; Talisman pulls out of Sudan, BBC News, 10th March 2003; Oil firm flees Sudan, BBC News, 2nd September 2003.
11

17

Figure 2 indicates that government intervention is the most common broad type of effect, accounting for nearly half of all incidents. This result corresponds with what has become an increasingly widespread notion among scholars, namely that policy changes and government interference in MNC operations probably constitute the most important class of political risk outcomes (Comeaux & Kinsella, 1997; Graham, 1996; Henisz, 2000; Jensen, 2006; OSullivan, 2005).

---------------------------------------------- FIGURE 2 IN HERE ----------------------------------------

Government intervention Table 3 presents more detailed information on specific forms of government intervention. The data suggest, first, that host governments and politicians have at their disposal a wide array of policy tools that can be used in order to increase their nations (and sometimes even their own) share of the spoils of FDI. A trend that Kobrin (1984) alleged had become discernible as early as in the mid-1970s, developing countries, confident in their own ability to gain from the bargaining game of FDI, seem to prefer more refined and subtle regulation to clear-cut expropriation. The data actually contain only six cases of outright nationalization, expropriation, or confiscation. And while an additional 12 cases involve the imposition of ownership restrictions and a further 14 events concern foreign companies that were blocked from entering the host country in the first place (blocking of investment) 126 of the cases have little or nothing to do with questions of ownership per se.

-------------------------------------------- TABLE 3 IN HERE -------------------------------------------

18

At the top of the list, with 31 registered cases, towers the sub-category of breach, termination, or suspension of a contract or license. (In addition, five cases involving forced or unwanted contract renegotiation are recorded.) This is perhaps indicative of the ostensible tendency of host governments to intervene through subtler means. It is nonetheless important to keep in mind that, in many cases, breach of contract is not too dissimilar to expropriation given that an expropriation will normally also involve breaking a contract between the host and the foreign investor, while a breach of contract will often ultimately affect ownership. The consequences for the investor can, in any case, be severe. Still, interventions are oftentimes of a non-flagrant nature, instead corresponding more closely to the concept of creeping expropriation. This term denotes risks that gradually yet significantly erode firms property rights, even though the investor may retain nominal ownership (Comeaux & Kinsella, 1997, p. 8). Table 3 contains several examples of just this. Perhaps the most conspicuous cases are those involving oil and gas companies in Venezuela and Bolivia. These firms have suffered several corporate tax and royalty increases over the last few years (in addition, of course, to acts of a more direct expropriatory nature). The subcategory of regulations/taxes/bans on trade, production, investment, sales, or withdrawal of funds is also illustrative in that none of these cases involved firms that were forced to relinquish large chunks of ownership or withdraw from the host nation altogether. Cases of bureaucratic or political delays also conform to common definitions of creeping expropriation, as do arbitrary back-tax claims or disputed tax claims, of which investors in Russia have often been victims. Other indirect expropriatory measures include the implementation of price controls, tariff freezes, or caps on profits; the imposition of a fine; government/public control over operations/board/management; and disputed/arbitrary chanrges, court rulings, or legal process.

19

Acts relating to war, terrorism, and social unrest Table 4 depicts the distribution of 131 acts relating to war, terrorism, and social unrest among six sub-categories. In the vast majority of cases, the main actors through which these risks are realized are rebel groups, terrorist organizations, and local community stakeholders. One thing to note is that around three-quarters of these cases concern either acts of terrorism (including sabotage and armed attacks by rebels) or kidnapping of workers. With regard to the sub-category of sabotage and terrorism/armed attack, a useful distinction can probably be made between: (a) massive bomb attacks (like the November 2002 suicide attack that caused such carnage at the Israeli-owned Paradise Hotel in Mombasa, Kenya12); (b) smaller-scale attacks against physical targets (like the recurrent bombings of foreign-owned pipelines in Colombia and Yemen); and (c) armed assaults on company employees (like the April 2003 killing by Islamic terrorists of a British geologist in Eritrea13).

--------------------------------------------- TABLE 4 IN HERE -----------------------------------------

Most acts of violence against multinational companies are committed by rebel groups or terrorist organizations and take place in countries that already suffer from internal armed conflicts (Lia & Kjk, 2004). Nevertheless, many such events are not related to civil or interstate war. Disaffected local stakeholders sometimes resort to acts of sabotage in an attempt to force multinationals to pay heed to their grievances. Conflicts with local stakeholders can undoubtedly become very expensive. In the first seven months of 2001, for
12

Kenya terror strikes target Israelis, BBC News, 28th November 2002; Israel vows revenge, The Economist,

29th November 2002; Mombasa bomb suspects go to trial, BBC News, 2nd February 2004.
13

UK geologist killed by Eritrea rebels, BBC News, 17th April 2003; British geologist killed by Islamists

with al-Qaida links, The Independent, 19th April 2003.

20

example, U.S. oil producer Caltex suffered estimated losses of $400 million due to social unrest in the northern Indonesian province of Riau.14 The sub-category of severe social unrest contains similar incidents where firms have been forced to suspend operations or withdraw from the host country altogether because of social instability. Likewise, the category of war threat or severe instability and threat to physical safety includes 19 cases of involuntary retreat by multinationals due to looming or actual war; the data contain examples from countries like Afghanistan, Iraq, Sudan, the Solomon Islands, the Democratic Republic of Congo, Pakistan, Angola, Burundi, and Nigeria. The last sub-category contains fewer cases and comprises events related to protests, demonstrations, or blockades against a company. Usually staged by local stakeholders, such incidents entail direct action by the players involved and can cause serious delays to projects.

Other acts committed by non-governmental actors The last main dimension other acts committed by non-governmental actors encompasses incidents that are directly related neither to the actions of host governments nor to war, terrorism, or social unrest. The risks placed under this heading do not result in any physical damage being inflicted on workers or the investment itself, nor do they involve direct action by local stakeholders (like, for instance, blockades). These risks are essentially generated by one of two types of actors: either activists (typically NGOs) or domestic businesses.

-------------------------------------------- TABLE 5 IN HERE ------------------------------------------

14

Caltex faces threat in Indonesia, BBC News, 7th August 2001; Caltex may avoid Indonesia blockade, BBC

News, 8th August 2001.

21

Although stakeholder activism can take on many different forms and involve a variety of actors, in most such cases the risk is realized in full or in part courtesy of the intervention of one or more NGOs. As Table 5 shows, the data contain 25 cases of NGO activism (including involvement by other activists). This sub-category is by necessity fairly broad. Most cases still share one fundamental trait, namely the exertion of strong pressure on multinationals by NGOs (through the conscious use of the media). Facing severe and lasting pressure to divest, the MNC may have little choice but to withdraw from its investment, thereby losing expected future profits (as well as existing assets). Note that I single out for special treatment one particular form of effect, namely potentially detrimental lawsuits related to activism, a sub-category containing ten reported cases. According to my data, what often characterizes such cases is that the suit itself is brought before a court in the United States a country whose laws are often applied extraterritorially irrespective of the country in which the alleged misconduct actually took place (Olsen, 2002; Puckett, 2008; Putnam, 2009). While I certainly believe that vested business interests constitute an important source of political risk effects, my sample suggests that only rarely are indigenous firms directly responsible for realizing risk. The sub-category of dispute with or fraudulent behavior by other company (which largely involves domestic companies exploiting political institutions to their own advantage) only includes five cases. I also chose to specify one additional category comprising three incidents of corruption (where the receiver of the bribe payment was not a host-government official or entity) that could not usefully and logically be placed elsewhere.

Political risk effects, by industry Figure 3 exhibits the distribution of political risk effects among different industries. It is noteworthy that nearly two-thirds of all cases involve firms in either the extractive sector or in
22

infrastructure/utilities. Since these industries do not account for an equally large share of total FDI flows to the developing world, the evidence implicitly indicates that political risk is essentially industry-, firm-, or even project-specific. The risk profile of an oil company, for example, is significantly different from that of a trading firm. Whereas the former deploys large sunk assets that may become a tempting target for host-government appropriation, the latter does not; and while the oil firm is highly visible in the host country and can easily stir nationalist emotions simply by extracting a natural resource, the trading company operates in a non-strategic sector and is therefore less likely to receive negative attention. These notions are supported by the data: only four cases of realized political risk involving wholesale or retail trade are recorded, whereas 113 of the interventions targeted the petroleum industry. The results also conform closely to the obsolescing bargain model, which holds that investment deals involving the deployment of large fixed assets are particuarly vulnerable to later revision demands by the host government, largely due to the so-called hostage effect (Vernon, 1971). Extractive-sector firms, of course, usually conduct sizable and highly visible and profitable investments that quite often tend to spur nationalist feelings among the hostcountry population. Public utilities and infrastructure companies are also susceptible to the obsolescing bargain mechanism (Ramamurti, 2001; Wint, 2005).

------------------------------------ FIGURE 3 IN HERE -------------------------------------------------

Sources of political risk Figure 4 shows the prevalence of the main sources or causes of political risk effects (in addition to the two supplementary residual sources of macroeconomic/financial

23

performance and company performance15). At this point I must stress that results should be interpreted with caution, as coding proved very demanding and also involved a substantial amount of discretion. The basic decision rule was to record causes of risk only if these were suggested by the news reports themselves. In most cases, multiple sources were registered. Note also that the count does not include the obsolescing bargain mechanism. While I do expect this mechanism to be almost ever-present in the 332 cases of realized political risk, it is virtually impossible to infer directly from the data that a shift in bargaining power is the main causal variable at work. The previous sub-section nonetheless indicated the probable pervasiveness of the OBM.

------------------------------------------------- FIGURE X IN HERE -----------------------------------

Socio-political instability and grievances The first thing to note is that the category of socio-political instability and grievances is associated with 220 or roughly two-thirds of the recorded cases of realized political risk. My data especially support the link between armed conflict and natural-resource extraction. Of the 91 cases associated with war or rebel activity, 59 percent involved petroleum or mining firms. This finding might lend some credence to the view that many conflicts are essentially greed-based; armed insurgencies, some claim, often result from economic opportunities or motivations (Collier & Hoeffler, 2004; de Soysa, 2002; Fearon & Laitin, 2003; Ross, 2004). In this respect, natural resource-extracting firms are especially likely targets of rebel attacks given that they invest in assets that can provide ethnic groups with a possible financial basis for independence as well as yield immediate revenue through extortion.

15

The category of company performance includes first and foremost cases where conflict arises after claims that the firm has failed to honor specifics of the investment contract.

24

Others emphasize the grievance explanation, arguing that most civil wars are caused by feelings of injustice among certain segments of society (Cederman & Girardin, 2007; Gurr, 1970; Saxton, 2005). While the objectives of rebels are virtually impossible to identify if we base the analysis solely on the information conveyed by the news reports, the grievances of local community stakeholders in connection with specific projects are more easily detected. I count scores of cases where feelings of injustice and resentment among locals, in some way or another, are reported to be important causes of MNC trouble. The cases range from kidnappings and acts of sabotage in Nigeria (where local communities and rebels alike complain that the oil firms do little to protect the environment or ensure that the petroleum wealth is equitably distributed); to contract cancellations in Bolivia and Indonesia following claims that foreign utilities set prices too high; to the boisterous divestment pressure faced by multinationals operating in the human-rights breaching regimes of Myanmar and Sudan; and to property invasions in Malawi and Argentina by disaffected indigenous groups contending that the foreign-controlled land is theirs by right. In short, the data support the contention that todays investors often run into trouble simply because they are unable to establish or sustain sound relations with local communities (Jenkins, 2004), or because they behave and operate in a fashion that ultimately invites protests, social unrest, acts of sabotage, or even lawsuits (Banfield, Haufler, & Lilly, 2005; Reed, 2002).

Political institutions A total of 142 cases are judged to be associated with political institutions. Furthermore, closer (unreported) scrutiny of the data reveals that the vast preponderance of these cases ultimately resulted in an act of intervention by the government. This is not at all surprising. Recent studies of political risk that highlight the role of the host nations institutions basically argue that a poorly developed institutional framework reduces the credibility of host governments
25

promises to protect investors property rights (Ahlquist, 2006; Fowler, 2006; Henisz, 2000; Jensen, 2006, 2008; MacIntyre, 2001). Such institutional traits as democracy, a separation of powers (that is, the existence of veto players), rule of law, and bureaucratic efficiency work to decrease transaction costs as well as the risk of harmful government intervention, not least because they mitigate the obsolescing bargain mechanism (Jakobsen, 2006). A quite abstract phenomenon, the institutional framework does not easily lend itself to identification. It is particularly difficult to investigate empirically the veto player and democracy hypotheses, considering that the news reports rarely yield any clear-cut proof of a link between these institutional characteristics, on the one hand, and political risk effects on the other. However, the data do suggest that a substantial number of cases could be deemed a result of a lack of rule of law. This includes numerous interventions (by the host government) that are strongly contested by firm, with the latter often referring to the interference as unlawful. The category hence comprises, among other things, causal avenues that end in unilateral contract abrogations, including several that affected infrastructure firms in Argentina, Bolivia, and Iran. Cases of political risk involving the judiciary are also placed under the heading of political institutions, as are incidences related to corruption and cronyism.

Preferences and attitudes 97 of the interventions can, to some extent, be ascribed to the preferences and attitudes of important economic players. Although I did certainly encounter significant problems with respect to the process of coding the cases and delimiting sub-categories (not shown), the data nonetheless reveal some very useful information. First, 43 cases are placed under the broad sub-heading of economic nationalism and anti-foreign capital sentiments. This category signifies that the actors involved, as reflected by the news reports, made some sort of
26

reference to preferring increased domestic control over the project(s) in question (for example, with respect to ownership or revenues). Five of these cases involved multinationals in Russia, a country that has turned increasingly nationalistic on issues relating to its naturalresource industries. Second, I was able to identify ten cases where the key actors cited national security concerns as a major reason for the intervention (such cases were assigned to this category irrespective of whether or not these concerns seemed legitimate). Third, 18 of the cases were related to political ideologies of various kinds, but only half of these incidents including five occurring in Venezuela were decisively connected to socialist or Marxist philosophy. Fourth, I recorded 30 cases where vested business interests helped precipitate the intervention (26 of these cases involved host-country indigenous firms; the four remaining incidents involved other foreign companies). These results are testament to the view that for indigenous firms and industries, the main threat often comes in the form of FDI, and they therefore ought to favor a more restrictive policy vis--vis multinationals (Crystal, 1998). Defensive economic nationalism and rent-seeking behavior by indigenous firms and industries can be a highly attractive way to maintain profitability for domestic businesses that cannot normally survive competition from more efficient foreign-owned firms (Jakobsen & de Soysa, 2006; Gilpin, 1987, p. 33; Tollison, 1982).

Multiple and conjunctural causes The preceding discussion has treated the different sources of political risk on a one-by-one basis. But what should also be considered are all those cases where several distinct causes operate in concert to produce a specific political risk effect. Behind the May 2006 Bolivian oil nationalization and the subsequent tightening of the states grip on foreign investors, for example, lay a volatile, multifaceted mix of popular disaffection, social turmoil, nationalist
27

pressures, and foreign policy grievances. Overall, the data do lend support to the notion that political risk is a multidimensional phenomenon. Table 6 exemplifies.

----------------------------------------- TABLE 6 IN HERE --------------------------------------------

What concern us most here are those instances where a risk effect is judged to have been caused by factors pertaining to different main dimensions. The results for preferences and attitudes (PREFERENCES) are particularly conspicuous in that this dimension is rarely identified as the sole source of MNC trouble; it usually occurs in combination with one or both of the other dimensions. Take, for instance, the sub-category of vested business interests. Domestic firms are quite often able to take advantage of malfunctioning political institutions in the host country, as this helps them outperform foreign investors. Russia and Indonesia are especially infamous in this respect: crony capitalism and unfair competition are features that presumably permeate(d) the business environments in these two countries. It might, thus, be fair to conclude that the field of political risk analysis is no exception to the rule that most socio-political outcomes are engendered by multiple and/or conjunctural causes (Ragin, 1987). The phenomenon of political risk is not only multidimensional in the sense that many different single-source causal avenues exist (multiple causation); it is also quite often the case that sources coincide to produce a specific outcome (conjunctural causation).

Summary and conclusions

By way of presenting and analyzing a new, explorative dataset on political risk effects in developing countries, I have sought to illustrate many of the theoretical and conceptual ideas
28

that were outlined in the theoretical section. Results from the analysis were largely as expected, suggesting that profit-curbing events in the socio-political sphere can take on many forms and involve a wide range of actors. My data also indicate that the sources or causes of these effects are plentiful. A holistic approach to political risk analysis therefore seems warranted. The first general finding is that the profit-curbing events themselves the risk effects or outcomes differ markedly from one another. Some 48 percent of the recorded incidents are associated with government intervention; 39 percent are acts of intervention related to war, terrorism, or social unrest; while 13 percent of risks were realized courtesy of actions taken by activists or MNC partners or competitors. But even within each of these broad categories, the actual incidents that ultimately cause corporate losses vary greatly. For example, the data suggest that host governments and politicians have at their disposal a wide array of policy tools that can be used in order to increase their nations share of the proceeds from FDI. Acts of forced divestments of ownership, for their part, are relatively rare occurrences nowadays, as others have pointed out (Cass, 2007; Minor, 1994). It seems that contemporary governments prefer to intervene in subtler and less extreme ways. This is possibly because the costs including reputational losses associated with throwing the foreigners out altogether are greater than the benefits, thereby enouraging smaller-scale regulatory moves instead. Co-existing with expropriation-prone developing nations, the early literature on political risk tended to focus primarily on the host (central) government. Later scholars have also emphasized that the government is the main actor through which risks are realized. While my data clearly lend some support to this view, it is nevertheless true that the majority of incidents recorded herein are associated with interventions by non-state actors. In particular, 39 percent of the interventions are judged to be related to war, terrorism, or social unrest, the
29

second main category of political risk effects. I count 54 cases of sabotage or terrorism; 42 kidnappings of MNC employees; and 30 additional cases where multinationals have been directly affected by war or severe social unrest (for instance, by being forced to divest). In addition, my data encompass 43 cases where non-governmental activists or other businesses are usually the main culprits. But not only do the risk effects (and the actors responsible for realizing the risk) vary greatly; so, too, do the root sources or causes of the profit-curbing incidents. The theoretical section specifically identified four main dimensions of such sources. All these four major types of sources are judged, by the literature at large, to be of importance in that they each account for specific yet often complex causal avenues that ultimately lead to the realization of political risk. And results from the empirical analysis do lend support to this argument. The paper has stressed the importance of separating between sources of political risk and political risk effects. These are all vital and in principle perfectly distinct parts of the political risk landscape. Theoretically, our dependent variable is losses or trouble incurred by foreign direct investors due to the intervention of some socio-political actor(s). But interventions in the affairs of MNCs do not take place without reason. Whether abstract or palpable, one or more basic sources or causes always lie behind a decision to intervene. For the multinational company concerned with political risk, it is paramount that such deeperlying traits and sources also be scrutinized, especially since most FDI is undertaken with a long-term view. What may seem like a rather innocuous issue or attribute today (like Bolivias old foreign-policy grievances) might very well incite the host government, or other actors, to take action against MNCs tomorrow. And contrarily, what many investors seemingly believe is a particularly risky feature like, for instance, the strength of leftist political parties, which does not appear frequently as a source in my data may in reality pose

30

little true risk. Political risk forecasting, thus, obviously requires knowledge about all the elements of the causal sequence (and of the relative importance of these elements).

31

References Ahlquist, J. S. (2006). Economic policy, institutions, and capital flows: Portfolio and direct investment flows in developing countries. International Studies Quarterly, 50(3), 681-704. Alon, I., & Herbert, T. T. (2009). A stranger in a strange land: Micro political risk and the multinational firm. Business Horizons, 52(2), 127-137. Banfield, J., Haufler, V., & Lilly, D. (2005). Transnational corporations in conflict-prone zones: Public policy responses and a framework for action. Oxford Development Studies, 33(1), 133-147. Bremmer, I. (2005). Managing risk in an unstable world. Harvard Business Review, 82(6), 5160. Bthe, T., & Milner, H. V. (2008). The politics of foreign direct investment into developing countries: Increasing FDI through international trade agreements? American Journal of Political Science, 52(4), 741-762. Cass, F. (2007). Attracting FDI to transition countries: The use of incentives and promotion agencies. Transnational Corporations, 16(2), 77-122. Cederman, L. E., & Girardin, L. (2007). Beyond fractionalization: Mapping ethnicity onto nationalist insurgencies. American Political Science Review, 101(1), 173-185. Collier, P., & Hoeffler, A. (2004). Greed and grievance in civil war. Oxford Economic Papers, 56(4), 563-595. Comeaux, P. E., & Kinsella, N. S. (1997). Protecting foreign investment under international law: Legal aspects of political risk. Dobbs Ferry, NY: Oceana Publications. Crowley, F. D., & Loviscek, A. L. (2002). Assessing the impact of political unrest on currency returns: A look at Latin America. Quarterly Review of Economics and Finance, 42(1), 143-153. Crystal, J. (1998). A new kind of competition: How American producers respond to incoming foreign direct investment. International Studies Quarterly, 42(3): 513-543. de Soysa, I. (2002). Paradise is a bazaar? Greed, creed, and governance in civil war, 198999. Journal of Peace Research, 39(4), 395-416. Fan, J. P. H., Morck, R. & Xu, L. C. (2009). Institutions and foreign direct investment: China versus the rest of the world. World Development, 37(4), 852-865. Fearon, J. D., & Laitin, D.D. (2003). Ethnicity, insurgency, and civil war. American Political Science Review, 97(1), 75-90. Fowler, J. H. (2006). Elections and markets: The effect of partisanship, policy risk, and electoral margins on the economy. Journal of Politics, 68(1), 89-103.

32

Gilpin, R. (1987). The political economy of international relations. Princeton, NJ: Princeton University Press. Graham, E. M. (1996). Global corporations and national governments. Washington, DC: Institute for International Economics. Guqueta, A. (2003). The Colombian conflict: Political and economic dimensions. In K. Ballentine & J. Sherman (Eds.), The political economy of armed conflict: Beyond greed & grievance (pp. 73-106). Boulder, CO: Lynne Rienner. Gurr, T. R. (1970) Why men rebel. Princeton, NJ: Princeton University Press. Harbom, L., & Wallensteen, P. (2005). Armed conflict and its international dimensions, 19462004. Journal of Peace Research, 42(5), 623-635. Henisz, W. J. (2000). The institutional environment for multilateral investment. Journal of Law, Economics and Organization, 16(2), 334-364. Henisz, W. J., & Zelner, B. A. (2005). Legitimacy, interest group pressures, and change in emergent institutions: The case of foreign investors and host country governments. Academy of Management Review, 30(2), 361-382. Jagersma, P. K., & van Gorp, D. M. (2003). International divestments an empirical perspective. Business Horizons, 46(6), 61-69. Jakobsen, J. (2006). Does democracy moderate the obsolescing bargain mechanism? an empirical analysis, 19832001. Transnational Corporations, 15(3), 65-104. Jakobsen, J. (2010). Old problems remain, new ones crop up: Political risk in the 21st century. Business Horizons, 53(5), 481-490. Jakobsen, J., & de Soysa, I. (2006). Do foreign investors punish democracy? Theory and empirics, 19842001. Kyklos, 59(3), 383-410. Jakobsen, J., & Jakobsen, T. G. (2011). Economic nationalism and FDI: The impact of public opinion on foreign direct investment in emerging markets, 19902005. Society and Business Review, 6(1), 61-76. Jenkins, H. (2004). Corporate social responsibility and the mining industry. Corporate Social Responsibility and Environmental Management, 11(1), 23-34. Jensen, N. M. (2006). Nation-states and the multinational corporation. Princeton, NJ: Princeton University Press. Jensen, N. M. (2008). Political risk, democratic institutions, and foreign direct investment. Journal of Politics, 70(4), 1040-1052. Jodice, D. A. (1980). Sources of change in Third World regimes for foreign direct Investment, 19681976. International Organization, 34(2), 177-206.

33

Khanna, T., Palepu, K. G., & Sinha, J. (2005). Strategies that fit emerging markets. Harvard Business Review, 82(6), 63-76. Kobrin, S. J. (1980). Foreign enterprise and forced divestment in LDCs. International Organization, 34(1), 65-88. Kobrin, S. J. (1984). Expropriation as an attempt to control foreign firms in LDCs: Trends from 19601979. International Studies Quarterly, 28(3), 329-348. Kobrin, S. J. (2005). The determinants of liberalization of FDI policy in developing countries: A cross-sectional analysis, 19922001. Transnational Corporations, 14(1), 67-104. Le Billon, P. (2001). Angolas political economy of war: The role of oil and diamonds, 1975 2000. African Affairs, 100(398), 55-80. Leeds, B. A. (1999). Domestic political institutions, credible commitments, and international cooperation. American Journal of Political Science, 43(4), 979-1002. Leeds, B. A., Mattes, M., & Vogel, J. S. (2009). Interests, institutions, and the reliability of international commitments. American Journal of Political Science, 53(2), 462-476. Lia, B. & Kjk, . (2004). Energy supply as terrorist targets? Patterns of petroleum terrorism 196899. In D. Heradstveit & H. Hveem (Eds.), Oil in the Gulf: Obstacles to Democracy and Development (pp. 100-124) . Aldershot: Ashgate. MacIntyre, A. (2001). Institutions and investors: The politics of the economic crisis in Southeast Asia. International Organization, 55(1), 81-122. Margheritis, A., & Pereira, A. W. (2007). The neoliberal turn in Latin America: The cycle of ideas and the search for an alternative. Latin American Perspectives, 34(3), 25-48. Miller, K. D. (1992). A framework for integrated risk management in international business. Journal of International Business Studies, 23(2), 311-331. Milner, H., & Kubota, K. (2005). Why the move to free trade? Democracy and trade policy in the developing countries. International Organization, 59(1), 107-143. Minor, M. S. (1994). The demise of expropriation as an instrument of LDC policy, 1980 1992. Journal of International Business Studies, 25(1), 177-188. Moran, T. H. (1974). Multinational corporations and the politics of dependence: Copper in Chile. Princeton, NJ: Princeton University Press. Oetzel, J. M. (2005). Smaller may be beautiful but is it more risky? Assessing and managing political and economic risk in Costa Rica. International Business Review, 14(6), 765-790. Olsen, J. E. (2002). Global ethics and the Alien Tort Claims Act: A summary of three cases within the oil and gas industry. Management Decision, 40(7/8), 720-724.

34

OSullivan, R. C. (2005). Learning from OPICs experience with claims and arbitration. In T. H. Moran & G. T. West (Eds.), International Political Risk Management: Looking to the Future (pp. 30-74). Washington DC: World Bank. Poynter, T. A. (1985). Multinational enterprises & government intervention. New York: St. Martins Press. Puckett, B. (2008). The foreign corrupt practices act, OPIC, and the retreat from transparency. Indiana Journal of Global Legal Studies, 15(1), 149-176. Putnam, T. L. (2009). Courts without borders: Domestic sources of U.S. extraterritoriality in the regulatory sphere. International Organization, 63(3), 459-490. Ragin, C. C. (1987). The comparative method: Moving beyond qualitative and quantititive strategies. Berkeley, CA: University of California Press. Ramamurti, R. (2001). The obsolescing bargaining model? MNC-host developing country relations revisited. Journal of International Business Studies, 32(1), 23-39. Ratner, S. R. (2008). Regulatory takings in institutional context: Beyond the fear of fragmented international law. Journal of International Law, 102(3), 475-528. Reed, D. (2002). Resource extraction industries in developing countries. Journal of Business Ethics, 39(3), 199-226. Regan, A. J. (2003). The Bougainville conflict: Political and economic agendas. In K. Ballentine & J. Sherman (Eds.), The Political Economy of Armed Conflict: Beyond Greed & Grievance (pp. 133-166). Boulder, CO: Lynne Rienner. Richani, N. (2005). Multinational corporations, rentier capitalism, and the war system in Colombia. Latin American Politics and Society, 47(3), 113-144. Ross, M. L. (2004). How do natural resources influence civil war? Evidence from thirteen cases. International Organization, 58(1), 35-67. Ross, M. L. (2006). A closer look at oil, diamonds, and civil war. Annual Review of Political Science, 9(1), 265-300. Saxton, G. D. (2005). Repression, grievances, mobilization, and rebellion: A new test of Gurrs model of ethnopolitical rebellion. International Interactions, 31(1), 87-116. Simmons, B. A., Dobbin, F., & Garrett, G. (2006). The international diffusion of liberalism. International Organization, 60(4), 781-810. Skippari, M., & Pajunen, K. (2010). MNE-NGO-host government relationships in the escalation of an FDI conflict. Business & Society, 49(4), 619-651. Snyder, R., & Bhavnani, R. (2005). Diamonds, blood, and taxes: A revenue-centered framework for explaining political order. Journal of Conflict Resolution, 49(4), 563-597. Tollison, R. D. (1982). Rent seeking: A survey. Kyklos, 35(4), 575-602.
35

Tsebelis, G. (1995). Decision making in political systems: Veto players in presidentialism, parliamentarism, multicameralism and multipartyism. British Journal of Political Science, 25(3), 289-325. United Nations Conference on Trade and Development (UNCTAD) (2002). World investment report 2002: Transnational corporations and export competitiveness. New York & Geneva: United Nations. United Nations Conference on Trade and Development (UNCTAD) (2004). World investment report 2004: The shift towards services. New York & Geneva: United Nations. Vachani, S. (1995). Enhancing the obsolescing bargain theory: A longitudinal study of foreign ownership of U.S. and European multinationals. Journal of International Business Studies, 26(1), 159-180. Vachani, S., Doh, J. P., & Teegen, H. (2009). NGOs influence on MNEs social development strategies in varying institutional contexts: A transaction cost perspective. International Business Review, 18(5), 446-456. Vernon, R. (1971). Sovereignty at bay: The multinational spread of U.S. enterprises. New York: Basic Books. Warren, R. C. (2011). Piracy and shipowners ethical dilemmas. Society and Business Review, 6(1), 49-60. Wint, A. G. (2005). Has the obsolescing bargain obsolesced? Negotiating with foreign investors. In R. Grosse (Ed.), International Business and Government Relations in the 21st Century (pp. 315-338). Cambridge: Cambridge University Press.

36

Table 1. Developing countries ranked by political risk effects, 19982005

Country Nigeria Iraq Indonesia China India Venezuela Colombia Russia Argentina Iran Peru Bolivia Ecuador Myanmar South Africa Pakistan Sudan Turkey Congo, DR Total 35 30 23 15 15 14 13 13 12 9 8 7 7 7 7 6 6 6 6 Gov 3 9 10 13 11 12 0 10 9 7 4 6 1 2 6 2 2 3 3 War 30 20 10 0 3 2 12 0 0 1 2 1 5 0 0 3 2 1 2 Nongov 2 1 3 2 1 0 1 3 3 1 2 0 1 5 1 1 2 2 1

Notes: Total denotes total incidents of political risk effects in the period; Gov denotes incidents of government intervention; War denotes acts relating to war, terrorism, or social unrest; Nongov denotes acts committed by non-governmental actors other than those relating to war, terrorism, or social unrest; only countries with six or more total events are shown.

37

Table 2. Political risk effects, by year

Year Effects total Effects government intervention 16 9 16 22 20 19 30 26 158 Effects war and unrest 15 17 14 15 17 14 30 9 131 Effects other non-gov. actors 2 4 4 7 13 7 6 0 43 Country with largest no. of incidents Colombia Nigeria Indonesia Indonesia Argentina Iraq Iraq Venezuela Nigeria

1998 1999 2000 2001 2002 2003 2004 2005 Total

33 30 34 44 50 40 66 35 332

38

Table 3. Government intervention in business, by incidents

Effect government intervention Breach/termination/suspension of contract or license Blocking of investment (temporary or final) Intervention/sanctions by foreign government Ownership restrictions Regulations/taxes/bans on trade, production, investment, etc. Price controls/tariffs freeze/cap on profits Corporate tax/royalty increases Bureaucratic/political delays Corruption Other intervention, policy change, or dispute Nationalization, expropriation, or confiscation Disputed/arbitrary charges, court ruling, or legal process Back-tax claims or disputed tax claims Forced or unwanted contract renegotiation/revision/review Other general or industry-wide policy changes Government/public control over operations/board/management Fine Problems related to business climate in general Total Count 31 14 14 12 12 9 8 8 8 7 6 6 5 5 5 4 2 2 158 Percent 19.6 8.9 8.9 7.6 7.6 5.7 5.1 5.1 5.1 4.4 3.8 3.8 3.2 3.2 3.2 2.5 1.3 1.3 100.3

Notes: Percentages are in proportion to the total number of cases of government intervention; due to rounding, percent total does not sum exactly to 100.

39

Table 4. Acts relating to war, terrorism or social unrest, by incidents

Effect war and unrest Sabotage and terrorism/armed attack Kidnapping or hostage-taking War threat or severe instability and threat to physical safety Severe social unrest Protests/demonstrations/blockades against company Total Count 54 42 19 11 5 131 Percent 41.2 32.1 14.5 8.4 3.8 100.0

Note: Percentages are in proportion to the total number of cases of acts relating to war, terrorism, or social unrest.

40

Table 5. Other acts committed by non-governmental actors, by incidents

Effect other non-governmental actors NGO activism (including involvement by other activists) Potentially detrimental lawsuit related to activism Dispute with or fraudulent behavior by other company Corruption (including MNC-initiated bribes) Total Count 25 10 5 3 43 Percent 58.1 23.3 11.6 7.0 100.0

Note: Percentages are in proportion to the total number of cases of other acts committed by non-governmental actors.

41

Table 6. Combination of sources (main dimensions) contributing to political risk effects

Sources SOCPOL POLINST PREFERENCES SOCPOL + POLINST SOCPOL + PREFERENCES POLINST + PREFERENCES SOCPOL + POLINST + PREFERENCES Total Count 144 42 15 39 21 45 16 322 Percent of total cases 43.4 12.7 4.5 11.7 6.3 13.6 4.8 97.0

Notes: SOCPOL denotes socio-political instability and grievances; POLINST denotes political institutions; PREFERENCES denotes preferences and attitudes; included in the numbers are also instances where one or more main dimensions appear in combination with macroeconomic/financial performance and/or company performance; the ten cases that do not involve any of the three main dimensions of causes are excluded.

42

Figure 1. The causal framework

(ACTORS)

SOURCES
Obsolescing bargain mechanism Socio-political instability and grievances Political institutions Preferences and attitudes

EFFECTS
Government intervention Acts relating to war/terrorism/unrest Interventions by other non-state actors

43

Figure 2. Distribution of political risk effects, by main dimension (N=332)

Non-governmental actors 13 %

Government intervention 48 %

War and unrest 39 %

44

Figure 3. Distribution of political risk effects, by industry (N=332)

Finance/insurance Trade Construction 3 % 1% 3% Agriculture/forestry 5% > 1 industry 6% Oil and gas 34 %

Services 8%

Mining 9%

Manufacturing 10 %

Utilities/infrastructure 21 %

45

Figure 4. Sources contributing to political risk effects, by main dimension

Source Macroeconomic/financial performance 11 (3.3%)

Company performance

42 (12.7%)

Preferences and attitudes

97 (29.2%)

Political institutions

142 (42.8%)

Socio-political instability and grievances 0 50 100 150 200

220 (66.3%)

250

Political risk effects - number of incidents

Notes: Due to the presence of multiple sources, the total number of sources recorded exceeds the number of cases in the dataset; the numbers in the table reflect the count and percentage of cases where the source in question is regarded as having contributed to the realization of political risk; percentages are in proportion to the total number of cases in the dataset.

46