How Do International Organizations Promote Quality of Government? Contestation, Integration, and the Limits of IO Power

Transcription

1 International Studies Review (2012) 14, How Do International Organizations Promote Quality of Government? Contestation, Integration, and the Limits of IO Power Monika Bauhr G oteborg University AND Naghmeh Nasiritousi Link oping University How do international organizations (IOs) promote quality of government (QoG) and reduce corruption? IOs play a central role in most accounts of power in international relations. However, our understanding of how IOs exercise power seldom moves beyond the traditional material normative dimensions of power. We suggest that an important dimension to understand IO power is the contestation integration dimension, where IOs can exercise power either by integrating countries into networks of cultural exchange or by contesting existing orders. By analyzing multilateral aid data and building on recent advances in our understanding of the effectiveness of IO anti-corruption work, we apply this framework to show how the contestation integration dimension helps us understand the success or failure of anti-corruption strategies. We show that when IOs contest existing orders using governance rankings and aid conditionality, they suffer from ideational shortcomings, including lack of objective data and contested policy advice. In contrast, measures based on integration, such as the membership process of IOs or interaction with IOs, are more likely to suffer from internal procedural shortcomings, such as IOs failing to internalize and mainstream the norms that they seek to promote. Our findings have implications for both understanding conditions that limit the diffusion of the international anti-corruption agenda and advancing our knowledge of IO power and its limits. The detrimental effect of bad government institutions is well established (see, for example, Mauro 1995; Gupta, Davoodi, and Tiongson 2000; OECD 2001; Rothstein and Teorell 2008; Holmberg, Rothstein, and Nasiritousi 2009). Uncorrupted, impartial, and competent government institutions are central to dealing with some of the most important problems of our time, including poverty alleviation, climate change, and mistrust. 1 In the early 1990s, the World Bank placed anti-corruption and good governance on its agenda and soon other international organizations (IOs), such as the IMF and the 1 In 1998, UN Secretary-General Kofi Annan identified good governance as perhaps the single most important factor in eradicating poverty and promoting development (Annan 1998). Bauhr, Monika and Naghmeh Nasiritousi. (2012) How Do International Organizations Promote Quality of Government? Contestation, Integration, and the Limits of IO Power. International Studies Review, doi: /misr Ó 2012 International Studies Association

2 542 Contestation, Integration, and the Limits of IO Power European Union, followed suit. The declaration Partnership for Sustainable Global Growth that was adopted by the IMF in 1996 identifies promoting good governance in all its aspects, including ensuring the rule of law, improving the efficiency and accountability of the public sector, and tackling corruption as an essential element in promoting economic development (IMF 1996). The basic question asked in this study is simple: How do IOs promote quality of government (QoG)? QoG is here used in its broad sense, including corruption, government effectiveness, rule of law, and impartial decision making. In order to answer this question, we propose a comprehensive theoretical framework for the influence of IOs on domestic institutions. We subsequently use this framework to suggest conditions under which IOs effectively socialize states. In particular, our framework suggests that the level of integration involved in the way IOs exercise power, that is, the extent to which IOs integrate countries into networks of cultural exchange as opposed to contesting domestic orders, is key to understanding both the effects of IO involvement and how internal developments within IOs influence these effects. The analysis helps us understand the success or failure of strategies to improve QoG. We suggest that when IOs contest existing orders using governance rankings and aid conditionality, they suffer from ideational shortcomings, including lack of objective data and contested policy advice. In contrast, measures based on integration, such as the membership process of IOs or interaction with IOs, are more likely to suffer from internal procedural shortcomings, such as IOs failing to internalize and mainstream the norms that they seek to promote. The article thereby seeks to make several important contributions. Our framework develops an underexplored dimension of IO power, the contestation integration dimension. Important conceptual and empirical debates about power in international relations are structured around the material normative dimensions of power. IOs can convince member states of the rightfulness and appropriateness of particular international norms, or seek to incentivize states into adopting these norms based on rational calculation of functional benefits. Although the material normative dimension offers essential insights into IO power, it is insufficient for understanding the exercise of IO power as well as conditions under which it has desired effects. More specifically, whereas the material normative dimension can be seen as primarily distinguishing between the type of pressures that IOs put on governments, the contestation integration dimension of power pertains more to the means used to put pressure on governments. IOs can either challenge and contest existing orders in society or integrate governments in international settings. Furthermore, our framework provides much-needed attention to internal weaknesses in the strategies that IOs employ (Alderson 2001; see also Pevehouse 2002). We use it here to explain the dissonance between positive accounts of how IOs can contribute to reduce corruption and promote better government institutions on the one hand and the numerous empirical accounts that fail to find such a positive correlation on the other. While broad empirical studies have confirmed the link between international integration and reduced levels of corruption (Sandholtz and Gray 2003), numerous case studies fail to find a positive link between IO engagement and better government institutions (Kelley 2004; Schimmelfennig 2005). Past studies tend to explain failures in IOs norm diffusion by looking at conditions that are external to the approaches employed, including domestic opposition to government reforms, level of development, or lack of domestic leadership (Dollar and Svensson 2000; World Bank 2000). Such explanations, albeit valid, are insufficient for explaining failures by IOs to promote norms. This is because most countries that are the target of IO interventions for promoting good governance norms are characterized by a low

3 Monika Bauhr and Naghmeh Nasiritousi 543 political will for reforms due to high vested interests, weak institutions through which to implement reforms, and limited resources for carrying out comprehensive reforms (Fjeldstad and Isaksen 2008). In other words, pointing out that IO strategies failed to meet their objectives because of unfavorable domestic conditions leads to a very pessimistic view of the potential transformative power of IOs. In this paper, we show that even if perfect external conditions were to exist (such as no domestic opposition to good governance reform), the strategies themselves have drawbacks that reduce their potential for promoting norms. Thus, our theoretical framework does not only provide a way to understand IO power, it also helps explain why IO involvement may not contribute to positive domestic developments and what type of internal improvement would enhance IO influence. We proceed as follows. In section I, we present our theoretical framework for how IOs influence government institutions. This section discusses the dimensions of IO power, our typology of means used by IOs to influence government institutions, and the relationship between the type of power employed and internal limits to IO power. Section II unpacks one by one the distinct means that IOs use to promote QoG norms in member states and explore their respective internal weaknesses. Next, we provide a short illustration of the use of the framework by considering the European Union s role in promoting anti-corruption in neighboring countries. The final section concludes and examines how this theoretical framework can contribute to important conceptual and empirical debates. Dimensions of IO Power: How International Organizations Promote Quality of Government The long-standing debate about the power of IOs in IR has traditionally been structured along a material normative dimension. Using the terminology of March and Olsen, IOs can influence norms in member states by appealing either to the logic of appropriateness or to the logic of consequentiality (March and Olsen 1989). IOs can convince member states of the rightfulness and appropriateness of particular international norms, or seek to incentivize states into adopting these norms based on rational calculation of functional benefits. While most scholars would contend that there is no such thing as pure material or pure normative power, the relative weight or primacy that is given to material and normative power varies considerably between different approaches in IR. The concept of norms is primarily used in constructivist approaches to international relations, while most neorealist and neoinstitutionalist approaches assign a greater importance to material factors in attempts to understand international relations (Heinmiller 2007:657). While the material normative dimension structures and offers essential insights into important aspects of IO power, it is insufficient for understanding the exercise of IO power or the conditions under which IOs have desired effects. In addition to the material normative dimension, we add here a contestation integration dimension of power. We conceive of contestation as IO power that builds on opportunities to formulate and put forward their preferences to the governments that they try to influence, and that the primary means by which they do so is to challenge and contest existing domestic orders. 2 Integration, on 2 Contestation should be seen as also including IOs efforts to encourage positive elements within the country, which indirectly challenges negative elements (such as corruption among officials, widespread poverty, or abuses of human rights). In other words, contestation does not imply that IOs contest the entire country or even specific actors within it. Rather IOs contest structures and institutions that are perceived to foster these undesirable effects.

4 544 Contestation, Integration, and the Limits of IO Power the other hand, is defined as IO power that builds on assimilating governments in international settings. 3 Although both the material normative and the contestation integration continuum can be seen as dimensions of IO power, the nature of these dimensions varies. Specifically, while the normative material dimension focuses on what type of persuasive powers IOs possess, the contestation integration dimension focuses on the tactical powers that IOs use to influence states. IOs can use material incentives such as threatening to withdraw aid, promising investments or membership, or promoting a transformative, ideological, or moral discourse on the superiority of good government institutions (Moravcsik 2000). Thus, we see IOs as having not only the power to classify, fix meanings, and articulate desirable norms and a material power to signal material incentives and possibly even coercion, as suggested by the material normative distinction. They also have the power to diffuse these norms (cf. Barnett and Finnemore 1999). We suggest that these norms can be diffused either through contestation or through integration. Thus, the distinction between contestation and integration complements the distinction between material and normative power, by facilitating an understanding of the methods used to exercise IO power to promote norms. 4 Although there is substantial disagreement on how norms diffuse, constructivist scholarship has offered a number of important insights into the process by which norms gain ground. These include processes of coercion, persuasion, learning, and mimicry (Haas 1992; Meyer, Boli, Thomas, and Fransisco 1997; Finnemore and Sikkink 1998; Checkel 2001). Socialization theory, for instance, treats IOs as social environments wherein states and state agents may be socialized through participation (Checkel 2005:815). Furthermore, socialization has been defined as a process of inducting actors into the norms and rules of a given community (Checkel 2005:804). Although these processes all relate to our taxonomy of strategies and dimensions of IO power, the focus of our framework is different. It focuses on how IOs exercise power and how they use it to promote policy and institutional change, rather than on the processes of change. Most of the processes of socialization are relevant to all the methods included in our theoretical framework. Thus, both contestation and integration can involve processes of coercion, persuasion, learning, and mimicry. Figure 1 illustrates these two dimensions and places four of the strategies used to influence governments within this framework. The approaches that will be discussed in this paper have been adopted from Alderson s classification of the most important mechanisms for international socialization (Alderson 2001:432). IOs can promote QoG reforms by (i) producing rankings and country evaluations, (ii) placing conditions on economic assistance, (iii) promoting interaction with IOs and between member states, and (iv) through membership processes in connection with the enlargement of IOs. We maintain that the first strategy rankings and country evaluations is based on contestation and is normative in its approach. For example, rankings of government performance, whether based on level of corruption, environmental performance, or economic liberalization, can be placed in this category. Social rankings and hierarchies are used by outside actors to contest and challenge existing orders in that state. Therefore, we see their primary power in the contestation end of the continuum. In a certain sense, the pressure invoked by social rankings can be seen as promoting state 3 Thus, our use of the terms contestation and inclusiveness does not correspond to Dahl s (1971) use of these terms, since it is not used to depict aspects of democracy but aspects or strategies for IO influence (cf. Dahl 1971). 4 Our definition of power is wider than the traditional realist conception of power as the ability of IOs (or states acting through IOs) to use material resources to get states to do what they otherwise would not do. It also includes institutional forces and social relations that impact on actors abilities to determine their own fate (Barnett and Duvall 2005).

5 Monika Bauhr and Naghmeh Nasiritousi 545 FIG 1. Dimensions of IO Power and Means Used to Influence Domestic Institutions efforts to be integrated into communities of proper or civilized states. However, our framework focuses on how IOs use these rankings, and not on the transformative power of rankings in general. Thus, although the domestic pressure invoked by rankings may well be integrative, IOs use rankings primarily to contest existing domestic orders (cf. Towns 2010). Furthermore, social rankings provide standards of behavior for states (Katzenstein 1996). Thus, although changing positions in state hierarchies can lead to material benefits, such as foreign investments and growth, such benefits are often not immediate, and therefore, the power of rankings and country evaluations is primarily normative. The second category conditions on economic assistance is similar to rankings in that it is used to contest existing policies and institutions in states. However, the direct means of doing so is primarily material, most commonly to withdraw aid to misbehaving states. IOs use lending and aid as direct material incentives to government reforms. Thus, while the direct power of rankings is primarily normative (by defining rules for appropriate government behavior), the direct power of conditionality is primarily material (by linking material resources to expected behavior). In other words, while the rankings and country evaluation strategy is based on defining a normative goal that states need to conform to, the conditions on economic assistance strategy (in the form of conditionality for example) are used to provide material incentives for countries to conduct reforms that would presumably lead to that goal. Lending and aid are thus used as tools with a direct approach taken to provide technical advice to bring about reform. In contrast to these two strategies, both interaction with IOs and the enlargement of IOs are based on integrating states into networks of cultural exchange. Interaction with IOs refers to the involvement of government representatives in international settings, which allows IOs to socialize these government officials or bureaucrats into international norms. This could be through workshops, capacity-building programs, or other interactions. The Enlargement of IOs strategy, on the other hand, is based on IOs exercising pressure for domestic change in candidate countries in exchange for the primarily material benefits of membership. While members may gain both material and nonmaterial benefits from membership in IOs, such as the European Union, the main tool for bringing about intended reforms in new member states is through offering substantial

6 546 Contestation, Integration, and the Limits of IO Power and often immediate material benefits. 5 The Enlargement of IOs strategy is therefore a potentially powerful tool, as it allows governments in aspiring member states to credibly commit to governance reforms and weaken domestic opposition with the promise of membership benefits. This is in line with scholarship that highlights the impact of international norms on influencing domestic policy choices by appealing to government officials and interest groups (Cortell and Davis 1996). Thus, the different strategies can be placed alongside two continuums based on their primary type and means of power. It is important to note here that there is often a causal relationship between the different strategies that IOs use to promote QoG. Ranking and evaluations are, for instance, an important input to policies placing conditions on economic assistance, and conditionality can also be used in the membership process of IOs. Thus, while it is important to keep the strategies used by IOs apart for analytical reasons, there are important links between them, and they are often used simultaneously. Internal Limits to IO Power: Understanding Success and Failure in Affecting Change In addition to providing insights into the different ways in which IOs exercise power, the framework can also provide an understanding of the conditions under which IOs influence domestic institutions. 6 While influencing domestic institutions is generally not the main purpose of IOs, several IOs have interpreted their mission as including a mandate for affecting governance reform in member states either as a way to alleviate poverty, improve development outcomes, or increase competitiveness (see, for example, IMF 1997; World Bank 2007). Some IOs have placed significant financial and reputational capital behind the drive to reform governments. An analysis of aid data shows that IOs efforts toward strengthening governance has involved considerable and growing funds in the last decade. In Figure 2, we show the spending of four IOs over time on QoG-related projects, which include, for example, government administration, legal and judicial development, and economic and development policy planning. We see that the World Bank has invested heavily in promoting governance reform in member states. Figure 3 shows a breakdown of the data for the World FIG 2. IO Aid for Strengthening Governance, Source. aiddata.org. Constant (2000) prices, commitments, in USD millions. 5 See, for example, Schimmelfennig (2005). 6 See Table 1.

7 Monika Bauhr and Naghmeh Nasiritousi 547 FIG 3. World Bank Aid for Strengthening Governance, Source. aiddata.org. Constant (2000) prices, commitments, in USD millions. Bank into categories of aid activities for the same time period. The broad categories include, for example, public administration reform and anti-corruption aid. While the sums are significant, the statistics only shows a partial picture of IOs efforts to influence domestic institutions as it only represents one set of tools that IOs can use to affect change, that is, aid. The other means of influence identified above are difficult to quantify but need to be added to the total figure. Nevertheless, the aid data give us an indication of the considerable resources that are committed to governance reform by IOs. By virtue of their financial and political strength, therefore, IOs can be expected to influence domestic institutions. However, an analysis of the World Governance Indicators over time shows a more pessimistic picture of QoG reform around the world. Over the last decade, few countries have achieved a statistically significant improvement in their governance scores (Kaufmann 2010). Looking at one of the most discussed governance problems corruption indicates that overall results in this area have been weak. Figure 4 graphs countries control of corruption scores in 1998 on the horizontal axis and 2009 on the vertical axis. Higher scores indicate a higher level of control of corruption, meaning that those countries above the line have improved their scores while those below the line have weakened their scores in control of corruption over the time period. Overall, few countries have improved their control of corruption scores significantly. Countries like Liberia, Rwanda, Cape Verde, St. Lucia, and Qatar show improvements while Zimbabwe, Eritrea, and Myanmar show deteriorations. If we only look at countries that have received anti-corruption aid, it does not markedly change the results. Figure 5 graphs the control of corruption scores in 1998 on the horizontal axis and 2009 on the vertical axis for countries that have received anti-corruption aid from IOs. Liberia and Rwanda remain as countries showing improvements in control of corruption, which may indicate that IO aid has had an effect in those countries. Zimbabwe however shows deterioration, despite receiving anti-corruption aid. Moreover, countries such as Cape Verde, St. Lucia, and Qatar that have shown improvements according to Figure 4 have done so despite not having received significant anti-corruption aid. On the whole, therefore, the figures indicate that IOs push for improving QoG in member states has had mixed results.

8 548 Contestation, Integration, and the Limits of IO Power By focusing directly on the amount of anti-corruption aid rather than overall aid levels, we see that the relationship between support for anti-corruption reforms in member states is more nuanced than previous studies on IO anti-corruption work indicate. For example, while Charron (2011) finds that multilateral aid has had a significant impact on improving corruption, the figures presented here illustrate that the effect of anti-corruption aid channeled through IOs shows great variation. These results are only indicative, however, as they do not factor in the amount of anti-corruption aid received in relation to the country s GDP or other control variables. Nevertheless, they point to modest and mixed results in terms of QoG reform in the area of control of corruption in member states an area that has been the focus of much IO work in the last decade. Because much of the anti-corruption aid channeled through IOs is conditioned, the focus on this type of aid is illustrative of our contestation dimension. Through this type of aid, IOs contest existing orders. The results indicate that there is variation in the successfulness of this type of intervention by IOs. Similar results are found for our integration dimension. Here, we operationalize integration as number of IO memberships. Figure 6 shows the relationship between IO membership and control of corruption. We see that there is a weak association between these two variables, with more integration only weakly associated with higher control of corruption. 7 The same result is obtained when using other measures of QoG, such as the ICRG s QoG measure (not shown here, contact corresponding author for results). Again, the purpose of these figures is in no way to perform a systematic test of the influence of IOs. Their purpose is merely to illustrate a) that changes in overall corruption, at least as measured by these indices, has for most countries been low over the last 10 years and b) that there is not necessarily a homogeneous FIG 4. Change in corruption levels from 1998 to The figure shows the change in corruption levels using the World Bank Governance Indicators for 1998 and 2009 (Kaufmann et al. 2009). The line is reference line and indicates no change in the level of corruption. R 2 =.85. Data accessed through Teorell, Jan, Marcus Samanni, Sören Holmberg and Bo Rothstein (2011). The Quality of Government (QoG) Dataset, version 6 April 11. University of Gothenburg: The QoG Institute, This association becomes insignificant, however, when controlling for GDP, democratic quality, and level of inequality; results available upon request.

9 Monika Bauhr and Naghmeh Nasiritousi 549 FIG 5. Change in corruption levels from 1998 to 2009 for countries receiving anti-corruption aid. The figure shows the change in corruption levels among countries receiving anti-corruption aid from 2000 to 2010 using the World Bank Governance Indicators for 1998 and 2009 (Kaufmann et al. 2009). The line is reference line and indicates no change in the level of corruption. R 2 =.60. Data on control of corruption accessed through Teorell, Jan, Marcus Samanni, Sören Holmberg and Bo Rothstein (2011). The Quality of Government (QoG) Dataset, version 6 April University of Gothenburg: The QoG Institute, Data on anti-corruption aid accessed through aiddata.org pattern of change in QoG in countries that IOs engage with, either through contestation (measured here as anti-corruption aid) or through integration (measured here as number of memberships in IOs) and c) to stimulate a debate about these patterns and encourage empirical research along these lines, to complement research seeking to understand the aggregate effects of IO involvement. This is important to understand how to improve the effectiveness of IOs in promoting QoG. Thus, while the data presented here may not adequately capture the full extent of QoG improvements, 8 the question nevertheless remains: Why have efforts to spearhead QoG reforms around the world had so little success and what accounts for the mixed results? Several studies seek to understand why IO involvement does not always have the desired effects. Many of these focus on the level of domestic resistance against reforms, including lack of domestic leadership, lack of resonance of international norms with domestic audience, lack of domestic norm entrepreneurs, and weak institutional capacity to implement reforms (Morrissey 1995a,b; Acharya 2004; Parmentier 2006; Radelet 2006; Stevenson 2010). The literature on the influence of IOs on the promotion of QoG is no exception. Several studies point at domestic reasons for the failure of international anti-corruption campaigns and the traditional anti-corruption tools used by IOs (Dollar and Svensson 2000; World Bank 2000; Neuman and Calland 2007; Persson, Rothstein, and Teorell 2010). While an understanding of domestic resistance against government reforms adds important insights to our understanding of why QoG reforms may fail, our focus here is on the internal limits to IO power, that is, 8 See Section II below on the inadequacy of governance scores. The World Governance Indicators are however widely used for research purposes.

10 550 Contestation, Integration, and the Limits of IO Power FIG 6. The relationship between number of IO membership and control of corruption The figure shows the relationship between number of IO memberships and levels of control of corruption in 2009 using the World Bank Governance Indicators (Kaufmann et al. 2009, accessed through Teorell, Jan, Marcus Samanni, Sören Holmberg and Bo Rothstein The Quality of Government (QoG) Dataset, version 6 April University of Gothenburg: The QoG Institute, Data on IO memberships from hampering factors within the strategies that IOs use to influence domestic institutions. In other words, rather than attributing successful or unsuccessful norm diffusion to the level of domestic resistance against reforms, we instead look at the internal qualities and shortcomings of the strategies themselves. The internal limits to IO power can be of at least three main types: motivational, ideational, and procedural. Motivational limits imply a weak prioritization and commitment (for example, in the form of low financial or personnel commitment and motivation) to promoting a cause. Ideational shortcomings in IO strategies are defined as limits in the knowledge and expertise that IOs possess. In other words, ideational shortcomings pertain to the advice that IOs give to governments and the quality of that advice. Procedural shortcomings refer not to what IOs do but how they do it, that is, the procedure by which decisions are made within IOs and in their interaction with governments. A growing body of studies suggests that IO procedures do matter for their effectiveness in promoting norms (Woods 2000; Grigorescu 2002). While motivational-, ideational-, and procedural-based explanations for IO power is a recurrent theme in the norm diffusion literature, these explanations are primarily used to explain the nature of IOs or their role in international policy coordination (Haas 1992; Barnett and Finnemore 1999). We suggest here that they can be used to explain why IO strategies fail or succeed in influencing domestic institutions and norms. We suggest that as strategies move toward the integration end of the contestation integration scale, officials experiences with the internal procedural aspects of IOs play an increasingly important role. Conversely, strategies primarily based on contestation may be more sensitive to the quality of IO advice, including the objectivity and overall quality of the policy advice given. Motivational factors are, on the other hand, independent of the type of strategy used and can therefore affect to the same extent strategies of contestation and integration alike. Our theoretical framework thus develops

11 Monika Bauhr and Naghmeh Nasiritousi 551 previous literature on sources of IO performance 9 by specifying when certain internal shortcomings are likely to matter more. In short, we suggest that while all types and means of IO power suffer more or less equally from a lack of IO motivation to promote QoG, strategies based on contestation are more sensitive to ideational shortcomings while strategies based on integration are more sensitive to procedural shortcomings. This is because strategies involving contestation need to provide a compelling alternative for reform and thus require a sound ideational- and knowledge-based foundation to be effective in promoting reform. For strategies based on integration, on the other hand, the organizational procedural aspects are more likely to influence outcome as the key for integration to work is a sound process of integration. The extent to which IOs embody the norms that they seek to promote, such as transparency and control of corruption, may also be more important for the domestic effects of IO strategies based on integration, since integration involves more interactions with IOs and thereby provides experiences of the merits and drawbacks of these norms and decision-making procedures as applied in international settings. An analysis of internal limits to IO power may thereby offer an explanation for why studies arrive at conflicting conclusions on IOs effectiveness in reducing corruption and promoting better QoG. For example, Sandholtz and Gray s empirical study of 150 countries suggests that being integrated in international networks of communication through IOs reduces the level of corruption (Sandholtz and Gray 2003). They identify the World Bank as the most important IO teacher of norms in the field of good governance (Sandholtz and Gray 2003:769). However, their conclusion does not always find support in other empirical studies on the effect of international integration and government reforms. Research on reform agendas in Eastern Europe, for instance, finds that IO strategies that do not offer both strong political and economic incentives for countries to adopt better government institutions often fail (Kelley 2004). Thus, while IOs have a potentially important role to play in bringing about reforms in countries with weak QoG, their effectiveness depends on the strategies that they use and how they employ them. In the following, we unpack the different strategies that IOs employ to influence domestic institutions and link their effectiveness to motivational, ideational, and procedural limits. Rankings and Country Evaluations Normative Contestation One of the central ordering factors in international society builds on social ranking or hierarchies of states. Rankings and country evaluations both generate and draw upon these social rank or hierarchies, that is, the ordering of states as superior and inferior to one another (Towns 2010). These rankings foster a naming and shaming process whereby states are pressured into improving their governments as a way to better their image in the world and are therefore an example of normative contestation. There are a range of indicators and indices that attempt to measure government quality, of which the best known is the World Bank s Worldwide Governance Indicators. Other partial systems of governance ratings, on which the World Bank s indicators are based, include, for instance, Transparency International s Corruption Perceptions Index and Freedom House s Freedom in the World ratings. Rankings are not exclusively produced to assess the QoG of states, however, with other rankings existing in areas such as environmental performance and level of development See, for example, Gutner and Thompson (2010) and Barnett and Finnemore (2004). 10 See, for example, the Environmental Performance Index and the Human Development Index.

12 552 Contestation, Integration, and the Limits of IO Power The benefits of rankings and country evaluations to encourage and produce government reforms are frequently acclaimed. It has been argued that these tools serve an important role in encouraging government reforms. According to Rotberg, Transparency International s corruption rankings have managed to shame countries and rulers in Africa and Asia to reduce corruption at the national level, leading more nations to seek to be perceived as less corrupt (Rotberg :73). The rationale for rankings is thus that they can draw attention to the state of governments in different countries and provide normative pressure for states to introduce reform. They may also contribute to embolden civil society organizations that work with these issues so that they can be more effective in their efforts (Rotberg :73). Thus, in theory, rankings and evaluations should work positively in encouraging QoG reforms. In practice, however, internal limits in the exercise of IO power could work against this strategy. The literature invokes in particular two types of internal shortcomings of IO effectiveness in enhancing normative pressures through QoG rankings: imprecise data and market pressures. While the first internal shortcoming is ideational and based on the criticism of the objectivity and usefulness of the data and ranks produced, the second is motivational as IOs can be driven by other goals than improving governments when making investments. In particular, the difficulty in producing objective indicators may work against this goal of improving the QoG and reducing corruption. This is because current rating systems are based on perceptions and can therefore appear as being biased (Rotberg :72; de Haan and Everest-Phillips 2007). Other studies have similarly pointed to problems in the construction and use of rankings, such as the lack of transparency in some indicators that can lead to the misuse of indicators when conclusions drawn from these data are translated into policies (Arndt and Oman 2006; Kaufmann and Kraay 2008). 11 Thus, as no rigorous objective criteria to assess QoG exist, countries at the bottom of the rankings may discard them as being political. Low-ranked countries that are strong enough to withstand the normative pressures that this strategy brings into force may choose to largely ignore them. Such an example could be seen in certain Middle Eastern countries, where many rankings are dismissed as a case of Western bias against their governments. This could therefore weaken the positive effect as described in the theory. Delapalme, for example, identifies the lack of reliable data for assessing QoG as a major obstacle for donors to implement evidence-based reform policies: Therein lies the most immediate challenge linked to the strengthening of governance: the ability, for governments, citizens, partners, and media alike, to effectively define measurable goals and then assess results. If governance is crucial for achieving better and quicker development, data are crucial for building and achieving good governance (Delapalme 2011:2). A lack of reliable data thus hinders effective IO policy. For the ranking mechanism to work at its full potential, then, the indicators need to be both accurate and authoritative qualities that are hard to achieve in practice. Rotberg, for example, calls for the setting up of an independent NGO to collect and draw up reliable measures of good governance (Rotberg :84). Although this could potentially solve some of the above-mentioned problems, it is not certain that the normative pressure of government rankings will be as effective as envisaged in theory. The data would, for example, need to address the concern among many countries that QoG reforms often take a long time and that governance rankings often fail to reflect steps taken in a positive direction (Johnston 2001). 11 For an analysis of how government rankings could become a self-fulfilling prophecy, see Warren and Laufer (2009).

13 Monika Bauhr and Naghmeh Nasiritousi 553 Credit-rating systems and other measures that guide investors is another form of ranking that can give rise to interstate competition. These types of country evaluations may lead to higher investments, but are largely normative in that they are not related to a direct material benefit. Studies on foreign direct investment (FDI) argue that government quality is important in attracting investors to a country. The World Bank has produced research that points at investment benefits in countries with high QoG (see, for example, Isham, Kaufmann, and Pritchett 1997:237). Furthermore, Blanton and Blanton (2007) argue that respect for human rights is positively related to increased FDI inflows, and Busse (2004) finds that democratic rights and FDI have a positive correlation. Thus, in theory, if the World Bank predominantly invested in developing countries that have taken steps toward improving their governments and reducing corruption, the rate of return on those investments would be greater. Such a strategy would hence discipline countries into improving their governments so as to attract investment. In practice, however, the Bank has a mandate that precludes it from taking political considerations into account when making investment decisions. This may explain why World Bank investments also go to countries with a very bad government record. As the redefinition of the term corruption from being a political issue to being an economic issue showed, however, the Bank has scope to direct its lending as it sees fit. 12 The actual problem may be that QoG variables play a lesser role in investment decisions than some research has suggested. Bénassy-Quéré et al. have, for example, shown that while government quality matters, investments are predominately directed by economic factors such as market size (Bénassy-Quéré, Coupet, and Mayer 2005). Similarly, an UNCTAD report on what governments can do to attract FDI mentions things like improving infrastructure and skills, but does not propose that countries increase accountability or strengthen anti-corruption work (UNCTAD 2008). Ultimately, investment decisions by IOs rest with member states that may have their own priorities for targeting investments even to countries with low QoG. 13 The most obvious cases when QoG and profitable investments clash are in many resource-rich countries. When natural resources are involved, the positive link between QoG and profitable investments often breaks down. In other words, resource-rich countries may not feel a pressure to improve their country evaluations if they perceive that investments will be forthcoming in spite of low rankings. 14 These problems thus once again highlight that the disciplining effects of rankings and evaluations to promote QoG reforms can be hard to obtain in practice. As will be shown below, these internal contradictions also exist in the three other strategies for promoting international norms. Conditions on Economic Assistance Material Contestation Conditions on economic assistance can be described as methods involving financial assistance with strings attached. These include structural adjustment programs and conditionality schemes and can be described as involving material contestation. Such strategies have frequently been practiced by both states (that 12 This will be discussed further in the next section. An interesting analysis of how the World Bank s anti-corruption work has contributed to a politicization of the Bank is provided by Marquette (2003). 13 For an analysis of the political nature of investment decisions by the IMF, see Thacker (1999). See also Kilby (2006, 2011). 14 The World Bank has made some attempts to reconcile this conflict by agreeing with the resource-rich country that some of the profits will be used to improve government institutions and reduce corruption in the country. An example of this is the Chad Cameroon oil pipeline, where the World Bank employed this innovative approach in an attempt to break the resource curse. The idea was that with World Bank assistance, the oil venture would result in reducing poverty and improving governance in Chad. However, the result of this project has shown that once contracts have been signed, the country has little incentive to comply with this agreement (Pegg 2006).

14 554 Contestation, Integration, and the Limits of IO Power is, the Millennium Challenge Account 15 ) and IOs such as the World Bank and the IMF. Conditionality became controversial as international financial institutions used policy-based lending to make developing countries adjust to the debt crisis of the 1980s (Koeberle, Bedoya, Silarszky, and Verheyen 2005). The objective of conditionality and structural adjustment programs is to enable donors to question aid-recipient countries policy structures and processes and to get them to alter them according to universal criteria and conditions established by the donors (Doornbos 2003:11). Key reform areas include public sector budgets, the legal system, civil service reform, and civil society support. As will be shown below, it is the contested nature of this policy advice that can reduce the effectiveness of this strategy. Thus, similarly to normative contestation of ranking, the internal shortcomings are mainly ideational. When IOs set conditions on economic assistance, the idea is that objective advice is given to governments who wish to receive foreign aid or borrow from international financial institutions and that withdrawing aid or the threat of doing so contributes to this end. These financial carrots consequently hinge on good performance, so as to ensure that the disbursed resources come to best use (IMF 1997:8 9). In practice, however, good performance is hard to achieve. The main problem is similar to the difficulties involved in producing objective indicators for what should count as good QoG. The theory on which this approach builds assumes that objective advice is unproblematic. Thus, an important internal limit to the effectiveness of conditionality as a means to reduce corruption and promote QoG is ideational, that is, the lack of solid knowledge on the effectiveness of the QoG reform measures suggested. Weaknesses in IO advice have, for example, previously been studied in relation to anti-corruption reforms. Most of the traditional measures used to reduce corruption have been the subject of controversies, and there is a lack of agreement on whether traditional anti-corruption advice does in fact reduce corruption (Fjeldstad and Isaksen 2008:13). Measures that IOs propose to contain corruption have been built on the premise that corruption is typically a problem of information asymmetries and rent-seeking. This view has its foundations in principal agent theory. Although this theory comes in a number of forms, the pioneering works of Becker and Stigler and Rose-Ackerman present corruption as a problem whereby the principal (government bodies or the citizens) has to find incentives to ensure that the agent (government officials) can be trusted to act honestly (Becker and Stigler 1974; Rose-Ackerman 1978). 16 Because the principal cannot monitor the agent s work at all times, the agent has the opportunity to engage in rent-seeking. Corruption thus arises when the benefits of the corrupt act outweigh the costs of possible detection and punishment. According to this view, corruption can be understood according to Klitgaard s formula discretion plus monopoly minus accountability equals corruption (Klitgaard 1988). In practice, this advice has resulted in public sector reform, including decentralization and a push for democratization and transparency (World Bank 2000). The basic idea of all these traditional anti-corruption measures has been to increase the chances of corrupt actors being detected and remove power from corrupt institutions. However, each of these different policy advices has been subject to important critique. For instance, fiscal and political decentralization has a compelling logic of 15 The Millennium Challenge Account is a bilateral aid program created by the Bush administration in It works according to the principle that developing countries with good governance are able to make better use of the funds received. Countries that are eligible for development aid through the Millennium Challenge Account are therefore selected according to a number of criteria, including Civil Liberties, Voice and Accountability, and Rule of Law ( 16 Another version of this principal-agent framework considers the public as the principals and the government as the agent.

15 Monika Bauhr and Naghmeh Nasiritousi 555 allocative efficiency, but one that is not always born out when implemented in practice, due to differences in levels of bureaucratic quality and information asymmetries (Tanzi 1996). Similarly, transparency is often believed to offer a solution against corruption and misconduct by officials; however, recent studies suggest that transparency alone does not increase accountability and may in fact increase corruption (Bac 2001; Olken 2004; Lindstedt and Naurin 2010; Bauhr and Nasiritousi 2012). One plausible reason for this is a failure to understand the social dilemma character of the problem of corruption (Persson et al. 2010), making measures based on the principal agent framework largely ineffective. In other words, the policy advice offered by IOs has ideational knowledge-based shortcomings. What is attempted in placing conditions on economic assistance is in fact a form of social engineering that is difficult to achieve in practice. Grindle, for example, argues that the governance advice offered to developing states is often built on a weak theoretical framework that lacks contextual clarity (Grindle 2004). Other critics argue that IOs too often focus on policies that are based in a particular ideology that aims at rightsizing the state, including trade liberalization and privatization, over other institutional factors (Campbell 2001; Khan 2002). Gerring and Thacker (2005) show that some of the neoliberal policies that have been associated with organizations such as the World Bank and the IMF may be linked with lower corruption, but that downsizing the state does not have an effect on corruption. The authors therefore conclude that policymakers should be wary of policy prescriptions that are overaggregated and that do not look at distinctions or make explicit necessary qualifications (Gerring and Thacker 2005:251). The problematic policy advice by IOs is also raised in a review of the World Bank s anti-corruption strategy. According to Fjeldstad and Isaksen, the World Bank s policy advice in this field suffers from two weaknesses: First, the anti-corruption strategies and policies seem to assume that the Bank knows what works or not, which is not the case. Second, it sometimes appears that the Bank uses as an argument (often superficially) that controversial aspects of its general macroeconomic and sector policies also help reducing corruption (Fjeldstad and Isaksen 2008:13). Thus, an extensive body of literature contends that inappropriate policy recommendations have made conditions on economic assistance strategies ineffective, and at worst even counter-productive. Three examples of counter-productive use of economic incentives have been pointed out as being particularly significant. First, structural adjustment programs can undermine democratic accountability as governments have to conform to the demands of the creditors (Whitfield 2005; Ksenia 2008). Second, to the extent that structural adjustment programs require reductions in national budgets, corruption could increase due to lower wages for civil servants (Ksenia 2008). Third, privatization undertaken in areas with weak institutions can increase corruption by increasing the opportunities of rent-seeking (Ksenia 2008). 17 One solution to these problems that is often recited in the literature is to increase country ownership, so as to give greater voice to domestic stakeholders. However, as Kapur and Naim point out, in a country where democratic checks and balances are weak, ownership may be little more than a sign of the stranglehold that vested interests have on national policy (Kapur and Naim 2005:91). In other words, ownership does not guarantee better policies. If governments are corrupt or captured by vested interests, ownership may make policies neither better nor more legitimate. 17 See also Rose-Ackerman (1996): Although privatizing state-owned enterprises reduces opportunities for corruption, the privatization process itself can create corrupt incentives. A firm may pay to be included in the list of qualified bidders or to restrict their number. It may pay to obtain a low assessment of the public property to be leased or sold off, or to be favoured in the selection process. Before reforms to the process in Argentina, for example, privatizations allegedly favoured those with inside information and connections, as have some privatizations in the former Eastern bloc.