Journal of Business Ethics (2009) 88:841849 DOI 10.

1007/s10551-009-0316-5

Springer 2010

Are Corruption Indices a Self-Fullling Prophecy? A Social Labeling Perspective of Corruption

Danielle E. Warren William S. Laufer

ABSTRACT. Rankings of countries by perceived corruption have emerged over the past decade as leading indicators of governance and development. Designed to highlight countries that are known to be corrupt, their objective is to encourage transparency and good governance. High rankings on corruption, it is argued, will serve as a strong incentive for reform. The practice of ranking and labeling countries corrupt, however, may have a perverse effect. Consistent with Social Labeling Theory, we argue that perceptual indices can encourage the loss of needed investment and, thus, contribute to higher rates of corruption within unfavorably ranked countries. In effect, corruption indices may inhibit foreign direct investment, the effect of which is to encourage the status quo in terms of corruption ranking. Using an experimental study design, we test the effects of country corruption rankings on the assessment of country investment desirability and nd ranking exposure causes shifts in country investment desirability for 10 of 12 countries studied. These ndings suggest that corruption rankings, which are based on perceptions of corruption,

may cause country isolation and a reduction in legitimate means of investment. KEY WORDS: corruption, collective action, network governance

The issue of corruption is vitally important. It affects the economic viability of nation states, endangers the physical well-being of untold thousands and unjustly enriches supremely mendacious individuals. (Dunfee and Hess, 2001, p. 489)

Introduction As multinational corporations seek new business opportunities, there is an increasing need for information regarding the extent of perceived corruption, commitment to the rule of law, and maturity of governance standards in prospective countries. Firms need to understand the investment risks associated with the expansion into new consumer markets, adoption of foreign supply chain partners, and decisions about desirable plant relocations. Many scholars argue that high rates of corruption present signicant risks for investors and adversely affect economic outcomes (Dunfee and Hess, 2001; Goldsmith, 1999; Habib and Zurawicki, 2002; Hess and Dunfee, 2000). Thus, valid data on country corruption levels are strongly desired by rms. In recent years, corruption indices, which assess perceptions of businesspeople aggregated at the country level, have increased in popularity, serving as tools for raising awareness about international corruption (Dunfee and Hess, 2001; Goldsmith, 1999; Hess and Dunfee, 2000, Johnston, 2004; Laufer, 2006; PR

Danielle E. Warren is an assistant professor at Rutgers Business School Newark & New Brunswick. Her research focuses on constructive and destructive deviance. More recent work examines the effects of informal and formal sanctions as well as moral emotions on deviant behaviors. She holds a Ph.D. from The Wharton School, University of Pennsylvania. William S. Laufer is Julian Aresty Professor of Legal Studies and Business Ethics, Sociology, and Criminology at the Wharton School of the University of Pennsylvania. He is Director, Carol and Lawrence Zicklin Center of Business Ethics Research, where he works in collaboration with multilateral institutions and non-governmental organizations such as the World Bank, United Nations, and the Center for Political Accountability. He received the B.A. in Social and Behavioral Sciences at The Johns Hopkins University, the J.D. at Northeastern University School of Law, and the Ph.D. at Rutgers University.

842

Danielle E. Warren and William S. Laufer people regardless of the behavior exhibited (Becker, 1963). Becker states that we cannot assume that these people have actually committed a deviant act or broken some rules because the process of labeling may not be infallible; some may be labeled deviant who in fact have not broken a rule (Becker, 1963, p. 9). Thus, those who deviate have not necessarily exhibited deviant behaviors. The notion that deviance is merely a matter of social construction presents considerable concern for those involved in evaluating corruption.

Newswire, 2008; The Economist, 2008; Wilhelm, 2002). Transparency International (TI), a non-governmental organization responsible for a number of international corruption and bribery rankings, constructs the Corruption Perceptions Index (CPI), one of the most commonly cited sources of aggregate, country-level perceptions (Alesina and Weder, 2002; Habib and Zurawicki, 2002; Steensma et al., 2005). TI has also gained popularity in the popular press (PR Newswire, 2008; The Economist, 2008). Just recently, The Dow Jones, a major US media outlet, chose to partner with TI in combating corruption (PR Newswire, 2008). Such a partnership will only expand awareness of the corruption index. We assert that widespread awareness of country corruption is useful to the extent that (1) the indices accurately depict relative differences in country corruption and (2) labeling countries by perceptions of corruption does not cause isolation, diminished incentives to invest, and further corruption. In this article, we examine one of the most popular indices TIs s Corruption Perception Index (CPI) and discuss the potential inaccuracies of perceptual measures of corruption. After considering the shortcomings of the TIs s CPI, we use an experimental study to demonstrate the investment effects of exposure to the CPI ratings. The ndings suggest that ratings of investment desirability are indeed inuenced by corruption index data. Drawing on Beckers Social Labeling Theory, we maintain that indices that publicly label countries as corrupt can cause isolation from further investments and opportunities for development. In effect, the corruption indices serve as self-fullling prophecies that reinforce, rather than reform, developing economies.

Corruption The social labeling tradition typically focuses on individuals and communities and pays particular attention to those who are labeled criminals and stigmatized by such labels (Becker, 1963; Goffman, 1963). In the business literature, Warren (2007) has considered the damaging effects of stigmas suffered by employees when their organizations are involved in scandals. While this tradition does not typically focus on countries, we believe that the labeling phenomena and related theory extend to countries that are labeled corrupt. Similar to labeling theorists concern for the truth behind a label, several business scholars note that common perceptions of corruption may not gauge the same degree or type of practices within countries (Johnston, 2004; Laufer, 2006). The variance in conceptualizations of corruption is best realized after reviewing the academic literature, which varies greatly in denitions of corruption (Habib and Zurawicki, 2002; Husted, 1999; Rodriguez et al., 2005). Habib and Zurawicki (2002) refer to corruption as acts of improbity, which includes behaviors that are illegal as well as improper. Husted (1999) notes the variance in corruption denitions but settles on a narrower one that describes corruption in terms of a specic exchange involving two parties. In contrast, Rodriguez et al. (2005) recognize corruption as involving both specic transactions as well as a general relationship between the public and private sector but the authors ultimately choose an aggregate perspective for purposes of theory building. Little is written about the application of social labeling theory to countries, but recent corruption

Deviance Deviance scholars have long questioned the manner in which society determines who is a deviant. In the business literature, deviance is often characterized as a relative concept, requiring a target for comparison (Warren, 2003). This perspective nds its roots in the sociological literature on social labeling. One school of thought asserts that the term deviance is nothing more than a label that attaches itself to

Are Corruption Indices a Self-Fullling Prophecy? rankings have provided opportunities to apply social labeling theory at the country level. In one of the few articles that considers social labeling in terms of country corruption, Nelken and Levi (1996) analyze the political motivations behind labeling practices as corrupt. More directly, they question whether or not accusations of corruption against certain regimes are not merely political tactics used to rise to power because, many times, the new regimes eventually engage in similar practices. This sentiment is shared by Becker (1963, p. 9) who explains, social groups create deviance by making the rules whose infractions constitute deviance, and by applying those rules to particular people and labeling them as outsiders. For instance, Johnston (2004) notes that certain forms of corruption such as political corruption do not receive the same level of attention as bribery when countries are rated for corruption indices. In short, people vary in their conceptualization of corruption which affects their ranking of countries. This disparity, however, is not addressed by corruption rankings, even those rankings that depend on perceptual measures. Next, we will outline our concerns with corruption rankings that use perceptual measure.

843

(2004, p. 275) notes, The scope, origins, and effects of corruption are often assumed to differ by degree, but not in kind, across a wide range of societies suggesting, in effect, that corruption in Denmark, Korea, and Bangladesh varies only in terms of its extent. This leads to another criticism which relates to the way TI computes its single dimension of corruption. Many country ofcials criticize the lack of transparency in the creation of the TI rankings (Hess and Dunfee, 2000; Maragay, 2006). Other criticisms note that the ranking of countries by corruption levels downplays the variance of conditions within countries (Johnston, 2004, Laufer, 2006). As a means to highlighting differences in corruption levels within a country, Laufer (2006) examined convictions of local, state, and federal ofcials for public corruption in the United States. In addition to displaying the variance in corruption levels within one country, Laufers (2006) analysis sheds light on the substantial levels of political corruption in those countries that promote corruption rankings. Distorted shifts in ranks Rankings depend upon the number of countries included in the index. As the number of countries uctuates, rankings uctuate. Since the better ranked countries also have the highest level of economic development, their rankings are rarely affected by the inclusion of new countries. The poorly ranked countries are most affected by the inclusion of new countries because their corruption levels may appear to have declined even though perceptions of corruption did not shift. For instance, a country that held the last place in the ranking will appear to suffer a rise in corruption if its rank shifts three points even though its change in rank is caused by the addition of three more countries to the index. Despite the shortcomings of perceptual measures of corruption, many defend the rigor of the measure. Wilhelm (2002) empirically compared TIs CPI to two other sources of data that relate to country corruption (Black Market Activity Index and Excess Regulation Index) and found that the CPI moved in conjunction with the other measures. Wilhelm (2002) asserts that these ndings provide evidence of the CPIs validity, but the other indices are also perceptual measures, and so it is difcult to claim that such correlations provide evidence that specic

Corruption rankings Our critique of the corruption rankings fall into two categories: (1) shortcomings of perceptual measures and (2) distortions of rankings due to sample uctuation. Perceptual measures As we already noted, denitions of corruption vary within the business literature. The individuals who comprise the study populations for perceptual corruption indices also possess varied denitions of corruption and the behaviors that constitute corruption. When survey respondents rate countries on corrupt behaviors, the respondents judgments of a countrys level of corruption will depend on their particular denition of corruption. After reviewing an index such as CPI, it is difcult to discern what, if any, behaviors are prevalent in a particular country. A similar criticism suggests indices wrongly misrepresent corruption as a single dimension. Johnston

844

Danielle E. Warren and William S. Laufer Here, the corruption itself is not questioned, but its hindrance to development is accepted. Finally, the author of the CPI makes a strong case that corruption reduces the ratio of investment to GDP (Lambsdorff, 2003). Implicit in this conclusion is the position that those countries ranked high on corruption will make an effort to improve their rating (Lambsdorff, 2003). We take a different approach in considering the effects of the index itself. In the development of hypotheses, we address the rst part of the selffullling prophecy theory, the isolating effects of corruption rankings.

forms of corruption do indeed occur within the ranked countries. In the next section, we consider the effects of labeling a country as corrupt.

Theory Theorists assert that the initial labeling of deviance can be a self-fullling prophecy such that those who are labeled deviant inevitably behave deviant (Becker, 1963). This outcome occurs because the initial label of deviant is thought to cause isolation from groups and society and such isolation reduces access to legitimate sources of resources and afliations. The lack of legitimate resources leads those who were initially only labeled deviant to secure resources from illegitimate sources or maintain afliations with other deviants. Thus, those who may have been deviant in name only start to embark on a pattern of deviance; the label begets a new behavioral pattern. While social labeling theory is typically used to describe members of society and has been used rarely in reference to countries, we believe the theory of self-fullling prophecy offers insight into the effects of corruption rankings. In particular, we assert that the theory provides an understanding of how the processes of labeling countries as corrupt, specically those related to perceptual corruption rankings, affect a countrys future opportunities and actions. The rankings, we argue, create a stigma that discredits the business environment within a particular country (Goffman, 1963). Similar to the spoiled organizational identities in business scandals (Warren, 2007), stigma due to corruption rankings can produce irreparable harm to a countrys opportunities for investment. Not only will the corrupt label reduce legitimate sources of funding, it may attract those who desire more risky ventures (Lambsdorff, 2003). While other scholars have mentioned the downward spiral of corrupt forces but such perspectives typically start with a position that assumes that the initial level of corruption exists and is not mislabeled. For instance, OHiggins (2006) writes about corrupt governments that control valuable extractive resources and divert funds to benet the government which prevents future economic development.

Isolation A substantial body of literature supports the theory that corruption causes isolation in terms of investment. Habib and Zurawicki (2002), in a study of 89 countries, compared 1999 CPI data to the three years of foreign direct investment and found a correlation between the index ratings and investments. Similarly, Goldsmith (1999) found favorable CPI data for emerging economies correlated with improved economic and political conditions. Alesina and Weder (2002) explained that some countries, such as Scandinavian countries, provide less aid to countries that are considered corrupt. Organizations such as the World Bank have openly stated that they will not provide aid to countries deemed corrupt (Alesina and Weder, 2002). These empirical studies suggest an association between country corruption and changes in investment. These studies, however, are not focused on the effect of the corruption index itself. Most studies are constructed in such a way that the CPI is used as a proxy for country corruption instead of an independent predictor of country investments. Here, we empirically examine the specic effects of the corruption index itself as an isolating factor. The premise of this research is that investors are sensitive to the risks suggested by country corruption rankings. Therefore, our rst hypothesis captures our main assertion.
Hypothesis 1:

Exposure to country corruption rankings affects country investment desirability.

Are Corruption Indices a Self-Fullling Prophecy? The initial hypothesis is multidirectional because it gauges the general relationship between the index and investments. The next two hypotheses unpack the relationship between country corruption indices by examining the effects of favorable rankings as well as unfavorable rankings on investment desirability. Past studies using the CPI typically focus on the full index rather than addressing how relationships between rankings and investment may be driven by a subset of the population. Our approach allows for a more ne-grained analysis of ranking effects. The relationship between corruption rankings and investment desirability could be a result of stigmatization of those countries with unfavorable rankings so we chose to separately theorize the effects of favorable rankings on desirability of country investment. Most theorists would not expect favorable corruption rankings to harm a countrys investments. At worst, the rankings would have no effect, but most would assume that the rankings would improve the countrys reputation and attract investments. Here, we adopt the dominant view and predict that favorable rankings will increase investment desirability.
Hypothesis 2:

845

Exposure to favorable corruption rankings relates positively to investment desirability.

the ability to conduct business in an illegitimate manner. The latter assertion is often voiced by those who rank lowest on the corruption indices. More specically, some leaders of lower-ranked countries assert that the better ranked countries essentially export their corruption to the lower ranked countries (Hess and Dunfee, 2000). Support for this claim appears in Transparencys ranking of countries by bribe payers. Hess and Dunfee (2000) note that Singapore was ranked third in the CPI but eleventh in the Bribe Payers Index (BPI), which suggests that Singaporean businesses behave differently at home than when operating overseas. Thus, some evidence suggests that an unfavorable ranking will increase investment desirability. Even if this relationship is found to be true, it would not undermine the application of social labeling theory. A positive relationship between unfavorable ranking and investments may indicate a rise in investments from illegitimate sources. Once the deviant label is assigned, legitimate sources of resources and afliations cease to exist and other illegitimate sources of funds may rise, but we assert that these funds will only bolster corruption within the country and harm economic reform.
Hypothesis 3:

Since the relationship between corruption rankings and investments may be driven by a halo effect for those countries with favorable rankings, we present separate theory on the effects of unfavorable rankings. Two theories explain the link between unfavorable rankings and investments. The dominant perspective suggests that exposure to a countrys unfavorable ranking will reduce investment in that country (Lambsdorff, 2003). Investments in poorly ranked countries may fall due to risks associated with corrupt political regimes that divert funds to personal projects. Firms may want to avoid pressure to conduct business in ways that violate extant country laws such as the Foreign Corrupt Practices Act. Thus, the dominant perspective suggests unfavorable ranking will cause isolation and a decrease in desirability of country investments. A competing, but less popular, perspective suggests that investors may seek business in countries that are unfavorably ranked because the investors desire opportunities to earn above-average returns or

Exposure to unfavorable corruption rankings relates negatively to investment desirability.

Methods We conducted an experimental study to test the effects of TI rankings on perceptions of a countrys desirability in terms of business expansion and willingness to recommend a country for investment. The sample was composed of 127 undergraduate business students at a large northeastern university.

Manipulation Corruption rankings We used TIs s CPI for 20022006 which is a composite index of more than 10 polls and perception surveys from independent rms and institutions (Transparency International, 2006). We split TI rankings into four quadrants for ve years of data

846

Danielle E. Warren and William S. Laufer fall in. The results of the manipulation check indicate that respondents who received the rankings were better able to choose the correct quadrant for each of the 12 countries. The mean number of correct answers was 11.90 for those who received the TI rankings and 7.51 for those who did not receive the TI ranks (p < 0.001).

and chose three countries from each quadrant. We included ve years of rankings to give subjects a sense of progress and/or stability in corruption assessments. Our criteria for choosing the 12 countries involved regional representation as well as placement in the rankings. We ensured that there were at least three countries per quadrant, and that at least one country appeared from a specic region. The quadrants were calculated by dividing the TI ranking in the base year, 2002, by four. For the subsequent rankings, we divided the list by four and noted the quadrant in which our 12 countries fell. Our study included two conditions (No rankings, Rankings). Those in the control condition (No Rankings) were told that countries are ranked based upon corruption, but the study participants were not provided ranking data. Manipulation check Respondents received a map and were asked to place a countrys quadrant of TI ranking next to the countrys name (I, II, III, or IV where I is least corrupt and IV is most corrupt). In order to provide consistency across conditions, we also offered the study respondents in the control condition a map of the world and asked them to note which TI quadrant of corruption they believed each country would

Dependent variable Country desirability Respondents were told they were the Director of Operations of a multinational corporation and given the task of recommending to the Board of Directors where to expand the rms business. Respondents were asked to rate the desirability of expansion in the 12 countries using a 7-point scale where 1 represents undesirable and 7 represents desirable.

Results Our rst hypothesis, that exposure to country corruption indices affects country investment desirability, was supported for 10 of the 12 countries (Table I). Bangladesh (t = 5.07, p < 0.001),

TABLE I Comparison of mean country desirability for investments No TI ranking (s.d.) Favorable ranking Switzerland Finland South Korea Morocco Portugal Slovenia Unfavorable ranking Haiti Colombia Pakistan Bangladesh Albania Senegal TI ranking (s.d.) (1.34) (1.38) (1.24) (1.20) (1.26) (1.57) (1.61) (1.30) (1.25) (1.76) (1.52) (1.29)

5.77 5.41 4.33 3.89 4.69 3.70

(1.52) (1.43) (1.87) (1.35) (1.38) (1.24)

6.41 6.34 5.03 4.40 6.12 4.98 2.02 3.17 2.41 2.17 2.57 3.43

2.64 (1.56) 2.70.(172) 3.27 (1.62) 3.73 (1.65) 3.69 (1.33) 3.33 (1.46)

N (No rankings) = 65 and N (TI rankings) = 59. p<0.01, p<0.05, p<0.001

Are Corruption Indices a Self-Fullling Prophecy? Portugal (t = -5.98, p < 0.001), South Korea (t = -2.43, p < 0.05), Haiti (t = 2.18, p < 0.05), Slovenia (t = -5.01, p < 0.001), Pakistan (t = 3.21, p < 0.01), Switzerland (t = -2.49, p < 0.05), Albania (t = 4.33, p < 0.001), Morocco (t = -2.18, p < 0.05), and Finland (t = -3.68, p < 0.001). We found no statistical support for a ranking exposure effect on investment desirability for Senegal (t = -.41, p = 0.681) or Colombia (t = -1.69, p = 0.094). Hypothesis 2 stated that exposure to favorable rankings would increase investment desirability. For purposes of this analysis, we dened favorable rankings as the top half of the index and unfavorable rankings as the bottom half. All of the six countries (Switzerland, Finland, Morocco, Portugal, South Korea, and Slovenia) in the top half received higher, statistically signicant, ratings of investment desirability when respondents were exposed to the corruption rankings. Hypothesis 3 stated that exposure to unfavorable rankings would decrease investment desirability. Of the six countries that started in the bottom half (Senegal, Pakistan, Bangladesh, Albania, Colombia, and Haiti), four received lower, statistically signicant, ratings of investment desirability when respondents were exposed to the corruption rankings.

847

Discussion Given that corruption indices are growing in popularity within the research community and among practitioners (Alesina and Weder, 2002; Habib and Zurawicki, 2002; Steensma et al., 2005; PR Newswire, 2008; The Economist, 2008), it is important to understand their shortcomings and effects. In this study, we take the rst step in testing the isolating effects of perceptual corruption country rankings. We investigate the degree to which corruption rankings affect an individuals assessment of a countrys investment desirability. Our ndings suggest that individuals investment assessments are indeed affected by exposure to corruption indices for certain countries. To date, most empirical research suggests that countries with higher corruption rates receive less foreign investment (Habib and Zurawicki, 2002; Lambsdorff, 2003). This literature, however, offers no direct link between corruption indices themselves and investment decisions. Here,

we nd this relationship holds true for four of the six countries with unfavorable ratings. Furthermore, by studying the unfavorably ranked countries separate from the favorably ranked countries, we found evidence that countries can experience a boost in investment desirability from a favorable corruption ranking. In the presentation of theory, we mentioned that the main tenets of social labeling theory support a rise in investment from illegitimate sources after a country is labeled deviant. Future studies could capture the rankings effects by investment source through the manipulation of the nature of the investor (e.g., aid provider, private bank). Such a study will provide a more nuanced analysis of corruption ranking effects on isolation from legitimate sources of funding. Another contribution of this article includes connecting the corruption literature to wellregarded theory in the sociological literature. To date, the literature lacks a comprehensive theoretical framework (for an exception, see Rodriguez et al., 2005). Social labeling theory suggests that country corruption rankings dampen future legitimate business opportunities for those labeled as corrupt and that this label, in effect, reinforces corrupt means of business. While we could not address the full theory within a single empirical study, we believe other studies could systematically address different aspects of the theory. As we mentioned earlier, new countries are added to the index every year and differences in these countries funding could be examined alongside similar countries that were already in the index as well as those which were not. Such a study would allow for some causal evidence of ranking effects. Our ndings coupled with future studies on patterns of country corruption may have far reaching implications for country development and reform. Finally, there are very few critics of the CPI and, more generally, the idea of corruption rankings. One notable exception, Galtung (2005), makes research on the possible labeling effects of corruption rankings seem all the more important. Galtung (2005) writes about the artifacts of an annual ranking of country level perceptions. Movement on such rankings, according to this author, is unlikely to result from internal anti-corruption reforms alone. He offers numerous examples, including the fact that

848

Danielle E. Warren and William S. Laufer any subsequent isolation from legitimate investments due to these rankings may be unwarranted. Acknowledgments
We would like to thank Gem Guiang for her excellent research assistance and the Zicklin Center for Business Ethics Center for funding this project. Tom Dunfee affected our development as academics. This article reects Toms inuence on our scholarship in two ways. It advances the discussion of corruption, a topic that was an ongoing concern of Toms. It also reects Toms approach to scholarship to thoughtfully question even those concepts that are widely accepted and respected. Here we question the legitimacy and inuence of popular corruption rankings because of their empirical shortcomings and potential damage to developing economies. We will sorely miss hearing Toms critique of our work and seeing a warm grin spread across his face as we exchange ideas. More than anything, we will miss Toms mentorship, leadership and unwavering integrity.

changes in ranking often require the deterioration of ranking of other countries in proximity. Improvement of countries in a region may also affect rank ordering. By publishing the CPI on an annual basis, Galtung (2005, p. 14) concludes, TI wittingly sustains this negative race to the bottom.The CPI cannot reward reformers because the standardization technique emphasizes rank ordering over internal reforms. It thereby reinforces negative perceptions about certain countries (and the positive images other, already well-to-do countries). Notably, Galtungs (2005) critique of perception rankings extends well beyond artifacts that might reinforce negative perceptions of corrupt countries. His nal criticism centers on the use of corruption rankings, such as CPI, in bilateral aid negotiation with countries. Galtungs (2005) specic concern is with aid conditionality, and he provides evidence that countries are disqualied from receiving aid from, for example, the Millennium Challenge Account (MCA) due to its perceived corruption. The MCA Board is directed to consider the CPI as additional information in determining eligibility for donor aid. That a countrys ranking may be used instrumentally in aid determinations adds additional signicance to our already existing concerns with labeling effects. Additional research should also explore the relationship between aid conditionality and corruption rankings. Shortcomings associated with this study include the absence of actual investment decisions and the use of student respondents. Also, our study did not address the second portion of the self-fullling theory in that we cannot gauge how countries respond to the loss of legitimate means of funding due to receiving the label of corrupt.

References
Alesina, A. and B. Weder: 2002, Do Corrupt Governments Receive Less Foreign Aid?, The American Economic Review 92, 11261137. Becker, H. S.: 1963, Outsiders: Studies in the Sociology of Deviance (The Free Press, New York). Dunfee, T. W. and D. Hess: 2001, Getting from Salbu to the Tipping Point: The Role of Corporate Action Within A Portfolio of Anti-Corruption Strategies, Northwestern Journal of International Law & Business 21(2), 471490. Galtung, F.: 2005, Measuring the Immeasurable: Boundaries and Functions of (Macro) Corruption Indices, in F. Galtung and C. Sampford (eds.), Measuring Corruption (Ashgate, London). Goffman, E.: 1963, Stigma: Notes on the Management of Spoiled Identity (Prentice-Hall, Englewood Cliffs, NJ). Goldsmith, A. A.: 1999, Slapping the Grasping Hand: correlates of Political Corruption in Emerging Markets, American Journal of Economics & Sociology 58, 865883. Habib, M. and L. Zurawicki: 2002, Corruption and Foreign Direct Investment, Journal of International Business Studies 33, 291307. Hess, D. and T. W. Dunfee: 2000, Fighting Corruption: A Principled Approach, Cornell International Law Journal 33(3), 593626.

Conclusion In this article, we took the rst step in empirically testing the effects of corruption rankings on investment desirability using a controlled experiment. The ndings suggest that unfavorable rankings reduce investment desirability for many countries while favorable corruption rankings boost investment desirability. To the extent that the corruption index contains measurement aws, a countrys ranking relative to other countries may be inaccurate, and

Are Corruption Indices a Self-Fullling Prophecy?

Husted, B. W.: 1999, Wealth, Culture, and Corruption, Journal of International Business Studies 30, 339359. Johnston, M.: 2004, Comparing Corruption: Participation, Institutions, and Development, in W. C. Heffernan and J. Kleinig (eds.), Private and Public Corruption (Rowman & Littleeld Publishers, Inc, Lanham, Maryland), pp. 275322. Lambsdorff, J. G.: 2003, How Corruption Affects Persistent Capital Flows, Economics of Governance 4, 229243. Laufer, W. S.: 2006, The Importance of Cynicism and Humility: Anti-Corruption Partnerships with the Private Sector, Development Outreach 8, 1821. Maragay, F. V.: 2006, Corruption Survey Lacks Transparency, Manila Standard, November 10, 2006. Nelken, D. and M. Levi: 1996, The Corruption of Politics and the Politics of Corruption: An Overview, Journal of Law and Society 23, 117. OHiggins, E. R. E.: 2006, Corruption, Underdevelopment, and Extractive Resource Industries: Addressing the Vicious Cycle, Business Ethics Quarterly 16, 235254. PR Newswire: 2008, Dow Jones Announces Partnership with Transparency International, October 23, 2008. 8:00 AM GMT. Rodriguez, P., K. Uhlenbruck and L. Eden: 2005, Government Corruption and the Entry Strategies of Multinationals, Academy of Management Review 30, 383396. Steensma, H. K., L. Tihanyi, M. A. Lyles and C. Dhanaraj: 2005, The Evolving Value of Foreign Partnerships in Transitioning Economies, Academy of Management Journal 48, 213235.

849

The Economist: 2008, Asia: The graft-busters strike against Indonesias anti-corruption drive, September 27, 2008. Transparency International: 2006, Corruption Perceptions Index (Transparency International, Berlin). Warren, D. E.: 2003, Constructive and destructive deviance in organizations, Academy of Management Review 28, 622632. Warren, D. E.: 2007, Corporate Scandals and Spoiled Identities: How Organizations Shift Stigma to Employees, Business Ethics Quarterly 17, 477496. Wilhelm, P. G.: 2002, International Validation of the Corruption Perception Index: Implications for Business Ethics and Entrepreneurship Education, Journal of Business Ethics 35, 177189.

Danielle E. Warren Rutgers Business School, Newark & New Brunswick, 1 Washington Park, Newark, NJ, U.S.A. E-mail: [email protected] William S. Laufer The Wharton School, University of Pennsylvania, 3730 Walnut Street, Room 670 Jon M. Huntsman Hall, Philadelphia, PA 19104-6340, U.S.A. E-mail: [email protected]