KEY CONCEPTS Published Barbara Adam, Time Alan Aldridge, Consumption Colin Barnes and Geof Mercer, Disability Mildred Blaxter, Health Steve Bruce, Fundamentalism Anthony Elliott, Concepts of the Self Steve Fenton, Ethnicity Michael Freeman, Human Rights Anne Mette Kjasr, Governance Michael Saward, Democracy John Scott, Power Anthony D. Smith, Nationalism 7 Governance and the World Bank The World Bank ignited a large debate on governance when it identified bad governance as a cause of economic crisis in many Third World countries. It called for greater transparency, efficiency and accountability in the countries to which it gave loans. However, scholars started applying these criteria to the World Bank itself: they asked whether the Bank as an international organization was accountable or followed transparent procedures. Thus, the issue of the governance of international organizations was added to the debate. Finally, the World Bank is also increasingly part of global governance processes as it takes part in redistribution from rich to poor and as it analyses the consequences of rich countries' trade policies to the poor countries. Three ways of how governance and the World Bank interrelate can thus be identified: 'good governance' as policy condition; governance of the World Bank as organization; and the World Bank as part of global governance. The World Bank programmes are an illustration of the difficulty of transferring particular models of governance to other countries. The World Bank as an organization illustrates how increased globalization and fragmentation has rendered accountability structures more complex. The World Bank is thus a case of the relevance of governance in contemporary global politics. Governance and the World Bank 173 'Good governance' as policy condition The World Bank contributed significantly to the emanation of the concept of governance in the 1990s. It introduced the concept of governance in a 1989 report pointing to corruption, bad policies and nepotism as severely hampering development. With the report, the World Bank sparked a debate in the international development community on what good governance was about, on how to promote good governance, and on whether exporting the Western systems of accountability and democracy is feasible or indeed realistic. The World Bank does not operate with one single definition of governance. Rather, the usage varies according to the particular report at hand. For example, in the 1989 report, governance is defined as 'the manner in which power is exercised in the management of a country's economic and social resources for development' (World Bank, 1989: 60). In a later report, 'governance is the institutional capability of public organizations to provide the public and other goods demanded by a country's citizens or their representatives in an effective, transparent, impartial, and accountable manner, subject to resource constraints' (World Bank, 2000a: 48). In spite of the varying definitions, the core features of the Bank's operational notion of governance refer to reducing corruption and strengthening rule-bound behaviour.1 This understanding entails four components of good governance that the World Bank has adopted as policy guidelines in recipient countries. Public sector management entails civil service reform and privatization initiatives. Legal framework for development is about making and enforcing rules that can make a market work, such as private property rights. Accountability aims at strengthening institutions to hold the government accountable, as for example an ombudsman, the Auditor General or parliamentary public accounts committees. Finally, transparency and information are keywords for programmes that support a free media or help the government publicize statistics, such as publishing the public budget annually. The World Bank's support for good governance was new in the sense that it was much more political than had so far 174 Governance and the World Bank been seen. Good governance programmes touched upon sensitive issues of distribution of power and resources, and some of them raised protests locally. For example, privatization programmes mean that the state sells off many of its enterprises and therefore loses an important avenue for gaining support through appointments to positions in these organizations. Additionally, civil service reform often involves laying off thousands of public employees. It also means removing essential functions to semi-autonomous bodies, which reduces the degree of control that the central government can exercise. Thus resistance to governance programmes is likely to emerge among groups who lose from the reforms, such as officials who were employed in state-owned enterprises or government staff in fear of losing their jobs (World Bank, 1997). The record of these programmes was mixed. It became clear that supporting institutional reforms was much more difficult than supporting specific projects, such as building a road or a hospital (Israel, 1987). The rise of the governance agenda in the development community thus raised key questions of the dynamics at stake in processes of nation- and state-building. It was argued that the 'external factor' should be taken seriously in the study of domestic politics (Doornbos, 1995). Changes in regime structures, for example, did not take place in complete insulation from the outside world. They could be triggered by the demands from international financial institutions for economic austerity, leading to social protests that could then cause the government to introduce reforms (Sorensen, 1993a). Or the failure to comply with aid conditions could lead to a cut in aid, and hence undermine the patronage that had so far bought critical support for the regime (Grindle, 1996). It is widely agreed that drastic cuts in aid at the end of the cold war caused some regimes, for example the Somalian and the Zairean regimes, to collapse (Ciapham, 1996). World Bank-supported programmes may also play a part in domestic politics in more subtle ways, such as strengthening some parts of the government (the ministries of finance) while weakening others (the ministries of planning). The adoption of governance programmes, then, promoted a more explicit recognition within academia of the role of external actors in processes of state-building. Governance and the World Bank 175 In the early 2000s, the governance concept seems to have gradually lost importance in the World Bank. It has proved extremely difficult to monitor recipient country compliance with all the specific conditions attached to good governance programmes. A World Bank report in 1998 assessed the effectiveness of aid in a large number of countries, and its main conclusion was that aid worked best in countries with a good policy environment (World Bank, 1998b, van der Hoeven, 2001}. So, from demanding good governance ex ante, there has been a movement towards ex post selectivity: good governance is now considered more a criterion against which to decide who qualifies for assistance (Doornbos, 2001). While good governance has not been abandoned by the World Bank, its usage has thus changed from being something to promote in developing countries to being a requirement before recipient countries are allowed loans in the first place. In sum, the practice of international institutions such as the World Bank triggered a debate in comparative politics about how external forces influence domestic politics, and whether it was at all possible to transfer models of 'good governance' built on Western ideas to non-Western settings (see chapters 5 and 6). One factor is the World Bank's emphasis on good governance in creditor countries. Another is whether the Bank itself, as an organization, lives up to norms of good governance, such as accountability and transparent procedures. Whereas the World Bank's governance programmes have been subject of debate in comparative politics and in theories about the role of the state in development, this latter question of accountability relates more to the debate in international relations regarding ways in which international organizations can be held accountable. Governance of the World Bank Is the World Bank itself subject to the norms of accountability, responsiveness and transparency that it requires from recipient governments? That question is not easily answered. The World Bank is a multi-purpose organization, and 176 Governance and the World Bank therefore, it is difficult to identify exactly to whom the Bank should be accountable. The difficulty posed by this question justifies an outline of the Bank's history, its purposes, and the interests represented by the World Bank. Such a discussion shows that the World Bank must be accountable to its major shareholders but that it also has a global constituency consisting of, among others, an increasingly assertive group of international non-governmental organizations. This latter group will be discussed on pages 183-6 in the third section on global governance. The International Bank for Development and Reconstruction (IBRD) was established in the late 1940s in the aftermath of the Second World War, which had left large parts of Europe and Japan in ruins. The Bank's purpose was to provide capital for development and reconstruction in capital-poor areas. By the beginning of the 1960s, Europe's reconstruction had proved a success, and, in principle, it was no longer a capital-poor area. In the meantime, however, more and more colonies in Africa and South Asia achieved independence from their colonial powers, and it was natural that the Bank started focusing on the emerging world. It typically financed large infrastructural projects, such as power plants or road networks. Often, these projects would not have been undertaken for private capital, since private investors would consider them too risky. In 1960, the International Development Agency (IDA) was established in order to give credit on very favourable terms, because it became difficult to find projects with returns big enough to satisfy IBRD conditions. IDA loans thus have a large grant element and they go to the very poorest countries. The World Bank is thus first and foremost a bank. The core institution, the IBRD, borrows on private capital markets and lends at low rates to needing members. It can lend at quite favourable terms because its member states provide guarantees for the loans it takes. The Bank, like any bank, must be accountable towards its shareholders, which are its members. The United States is the largest shareholder, with 16.62 per cent of the shares. The next largest holder is Germany, with less than 5 per cent of the shares. The United States therefore has a lot of influence on Bank decisions, and has a veto power on constitutional issues. The US executive Governance and the World Bank 177 The World Bank between the US and the world American Interests: •Opening markets • Removing staie regulation • Liberalization of capital markets $ The World Bank p "Global" Interests: •Careful protection of some industries •Some state regulation •Some regulation of financial and capital markets Capital injections Legitimacy Figure 7.1 The World Bank between the US and the world director is known to be extremely influential to the extent that any policy that displeases him or her will not be implemented {Woods, 2000, Wade, 2001). The United States also effectively chooses the World Bank's president, and it makes the single biggest contribution to IDA, which is funded by the surplus from IBRD and member-state contributions. So, like any bank, the IBRD must be accountable to its major shareholder, and IDA towards its major source of income. But, on the other hand, the World Bank derives much legitimacy from being a multi-lateral institution acting in an impartial manner. This legitimacy is crucial in order for the Bank to enjoy wide acceptance and cooperation. Figure 7.1 illustrates how accountabilities are split between US and 'global' interests - US in order to get capital injections, and 'global' in order to gain legitimacy. Obviously, there is no such thing as 'global' interests, but the term is used here to illustrate that a number of states in the global community have had different experiences of economic development from those of the United States. The Scandinavian countries, for example, have long argued that the state could 178 Governance and the World Bank play a -role in development and provide a safety-net for the poor in order for economic growth to become more balanced. Likewise, Japan and the East Asian tigers boosted industrial development by acting as investors when private agents found the risks too high. Also, they funded crucial research and development in new industries, and protected these new industries by imposing tariffs until they were considered viable enough to compete on world market terms (Evans, 1996; Weiss, 1998). These countries have had positive experiences with state regulation and are not, like the US and Britain, strong advocates of deregulation.2 The term 'global' therefore reflects a more positive view of state intervention. Sometimes the global interests have influenced Bank processes more than others: Japan pushed for the publication of the 1993 report The East Asian Miracle, which acknowledged the important role of the state in the East Asian economic boom, but, in general, free-market ideology was never abandoned (Wade, 1996). The World Bank and its presidency, then, have to maintain a difficult balancing act and, in this balancing act, accountabilities must be turned in more than one direction: to its major shareholder, the USA, and to the global community of states, and even, as we will see on pages 183-6, to an emerging global civil society. This difficult situation is reflected in many aspects of the Bank's work. Three examples are worth mentioning. The first illustration of how the Bank must please contradicting interests concerns its struggles to define the governance concept in an impartial manner. According to the first World Bank report (1992) entirely devoted to governance, the concept has three important aspects: 1 The form of political regime. 2 The process by which authority is exercised in the management of a country's economic and social resources for development. 3 The capacity of government to design, formulate, and implement policies and discharge functions. The report explains that it confines itself to the latter two, because its mandate forbids it to get involved in the domes- Governance and the World Bank 179 tic politics of creditor countries.3 The global community of states is not eager to let the Bank interfere with countries' internal affairs. The Bank has, however, been under pressure from the United States and other Western donors to broaden its agenda to include matters of human rights and democracy. Its recent adoption of human rights as an important part of development should be seen in this light (Hyden and Court, 2001; Kendal, 1995; World Bank, 1998a). When understanding governance as the 'exercise of power for development', as the Bank does, it is very difficult not to touch upon sensitive political issues. Indeed, it is often argued that politics is about power, about 'the capacity of social agents, agencies and institutions to maintain or transform their environment, social or physical' (Held, 1987: 275). As noted, World Bank governance programmes often affect domestic agents, and their relative capacities to transform their environment. In other words, the Bank does affect domestic political matters, whether it desires to or not. It does, however, continue to insist upon a neutral definition of governance, not involving aspects of the political regime, and, therefore, the Bank's work on governance reflects the tension between American and 'global' interests. A second illustration of how interests collide in the Bank and render accountability more complex is the publication Attacking Poverty (World Bank, 2000a). The background to the publication was that in 1998 the Bank's chief economist, Joseph Stíglitz, publicly criticized the way the International Monetary Fund had tackled, and in some respects possibly triggered, the East Asian financial crisis in 1997. Pressure from the International Monetary Fund (IMF) had forced the East Asian countries to liberalize their financial and capital markets. The crisis emerged when investors pulled out of an unsustainable real estate boom, partly caused by IMF pressure to remove restrictions on bank lending on real estate (Stiglitz, 2001: 178). The IMF's response was to impose tight fiscal policies, but, since these countries already had budgetary surpluses starving the economy of much-needed investment in infrastructure and education, the result was to further deepen rather than end the crisis (Wade, 2001}.4 Since the American Treasury had a strong interest in opening financial and capital markets, Stiglitz's critique of the 180 Governance and the World Bank IMF was not well received. The US deputy secretary in the Treasury therefore asked the World Bank's president, James Wolfensohn, to persuade Stiglitz to stop his criticism. Wolfensohn reluctantly did so, and Joseph Stiglitz chose to resign. In the meantime, the publication on poverty was under way. The person in charge of the report-team was Ravi Kanpur, Stiglitz's appointee, who had been chief economist for the Bank's Africa section. At the time of Stiglitz's resignation, the draft poverty report was circulated for comments. According to Robert Wade (2001), the draft was in many ways anathema to Treasury thinking: It had a controversial section on empowerment of poor people, requiring the provision of social safety-nets, and therefore entailed a larger role of the state in social development. This section drew on completely new material that had been researched by the Bank in relation to its 'Voices of the Poor' project, in which tens of thousands of people were interviewed about their own life chances. In the draft report, providing safety-nets was therefore prioritized above free-market reforms. Also, there was a section on how to democratize and regulate capital and financial markets. The US Treasury asked for major revisions of the report in a more neo-liberal direction, revisions that Kanpur felt it would be irresponsible to undertake, so he too resigned. The final report represented a compromise, with a larger neo-liberal element than the draft. It contained a new chapter on growth and poverty, and the chapter on markets and the poor no longer emphasized pre-establishment of safety-nets. The section on empowerment was moved to the latter half of the report, while the need for controlling capital and financial flows was substantially watered down. Wade's conclusion is that the US has too much control over the Bank. Of course, that is a normative view. But the case illustrates the delicate balancing act the Bank has to carry out between American and 'global' interests. The third and final illustration to be mentioned here of how the Bank strikes a balance between a neo-liberal (American) approach and a 'global' view more benevolent to state direction is contained in a speech of 6 March 2002 by the president, James Wolfensohn, in which he talks about partnerships in development (table 7.1). Wolfensohn is Governance and the World Bank 181 known to be a strong president, who has been able to accommodate a lot of the criticism of the Bank and who has strengthened Bank autonomy vis-ä-vis the US Treasury. His speeches, however, reflect both views. In neo-liberalism, the goal is to maximize individual and social well-being through market-led economic growth. Therefore, the main actors are individual entrepreneurs and private firms. Economic relations are seen as a positive-sum game in which cooperation is to the benefit of all (Jackson and Sorensen, 1999). State regulation or intervention in the economy is seen as a potential threat to growth because it distorts prices and incentives. It should therefore be minimized whenever possible. A 'global' view, as used here, reflects a more positive view on the role of the state; although basically agreeing with the dominance of the market, it acknowledges that some protection of infant industries may be in order. It also sees economic relations as potentially conflictual and zero-sum, because some individuals or groups may get richer at the expense of others. The main actors, therefore, are not only individuals and private firms but also classes and states. The quotes listed in table 7.1 represent both views. In the left-hand column, quotes reflecting the benefit of the market are listed, while the right-hand column quotes reflect the view that capitalism can create inequality and that redistribution on a global scale is necessary. The duality that the speech expresses is a cause of wide-ranging criticisms of the Bank. The left has criticized the Bank for being an imperialist institution with a bad record on poverty reduction and social issues. The right has argued that the Bank should be shut down because it is overly interventionist, destroys free market operation and is overly political (Gilbert and Vines, 2000), Some (such as Wade) argue that the World Bank is much too dominated by the United States' neo-liberal views, while others, such as Ngaire Woods, acknowledge a large US influence but also emphasize the varying ability of different World Bank presidents to assert their autonomy (Woods, 2000). It seems that, even if the Bank has become more nuanced and is not merely a neo-liberal institution, there is a difference between rhetoric and practice. At the practical level, many recommendations and programmes reflect a neo-liberal bend. Table 7.1 Neo-liberal and 'global' views in a Wolfensohn speech Quotes reflecting neo-liberal views Quotes reflecting 'global' views All peoples are partners in making the world a safe place There is no wall. We are linked by trade, investment, finance, by travel and communications Over the past 40 years, life expectancy at birth in developing countries has increased by 20 years - about as much as was achieved in all of human history prior to the middle of the twentieth century [Good governance policies] have generated growth led by the private sector... By building a more favourable environment for productivity and development, they are creating jobs, encouraging growth in domestic savings and investment, while also spurring increases in foreign direct investment flows [Western leaders] must move forward on the issue of trade openness, recognizing that without market access poor countries cannot fulfil their potential no matter how well their policies Rich nations must also take action to cut agricultural subsidies We have learned that we must focus on the conditions for investment and entrepreneurship, particularly for smaller enterprises and farms. Poverty can lead to exclusion, anger, even conflict Belief in that wall has for too long allowed us to > view as normal a world where less than 20 per cent of the population ... dominates the world's wealth ... and takes 80 per cent of its dollar income Belief in that wall has allowed us for too long to view the violence, disenfranchisement, and inequality in the world as the problem of poor, weak countries and not our own Too much inequity ..., too much exclusion, too many wars ... and now AIDS threatening to reverse many of the gains made over the last 40 years We know that there are conditions which foster successful development: Education and health programs to build the human capacity of the country. . . We have learned that debt-reduction for the most highly indebted poor countries is necessary Rich countries must recognize that even with action on trade ... there is still a fundamental need to boost resources for developing countries.. .to roughly 0.5 per cent of GNP ... But that is not enough for pro-poor growth: we must also promote investment in people, empowering them to make their own choices Source: 'A Partnership for Development and Peace', speech delivered at the Woodrow Wilson International Center by James D. Wolfensohn, president of the World Bank Group, Washington DC, 6 March 2002; available at: www.worldbank.org/News/Speeches *0 O « O 'i o w CR O Cn EU 2 3 ni n> ErS. 3 ja n p O" 3. 3u n 3," » O O c1 n ^3 s c er H 3i o.^ tr* ^ j=r 3 2 o s 8 Q 8 § g rS' £u O 3 "> O £u O co B CTQ e- 3 rí i «^ HQ 3 O 3 Suť w 3 n É2. E < s í? n> o •-s O •-* 0 3 0 CTQ Oj n 9B. ^ C 3 cT£ O K £ rr a £L » ^o £L tu ^ o. ff . O D- N 3 S. 3 .*0 J=- EU ju r, ST 3 •-"■>-' m" m 2 na 3 3~ o £U ja ^ ^ íl eu "a <-i Q. « g CTQ g qq_ 3 61 ft t» 3 ct ° tr ctil ju f? n> n. O O) -^ í^» írS ÉTs en r-f ■ ' t/> ,___, 3- n> íu o SJ i-s g- i:l. § ET-g n co JJ 3- C <-+ S íl tu o rro* 3 Q- EU ž 2*- S3 í=r C. " £u N m o H 2 fu 3 Cu cr 3 n> f O EU "5 » EJ. J11 co D- KJ 3 EU ►1 n> ju 3 S >0 " Ei. £ _ P 3 ^ N EU O c*5 3 *- cr a " o O O EU n> O i-s tro g 3 13 I... r-t- ■—■ i-í r-ľ H* o 9_ Eu fjq O 5-1 n O tu i/) O n O .____ EU £1) 60 •" EU 3-2 í? o B- ■- c K a o 3 o> 3" O.^ lí rl P §-S" ss ^•sS113 P? Q S^S o CT ÍL 3 §•1 £u 3 1—'OQ 3 £U EU ta 1-1 rt b: * n rš g » 3 H -^ S .w i« P n' CD 0) 7T OJ Q. O O" 0>_ n <" O n fl>" ro 3 m »-i CfQ B" OQ qq_ o" cr £L o w fii Ef. 3 O O. 3 n a> vi O O CfQ Oq o o" < erí £u 3 EU N (-j -^ 3 R n 3 Oi CTQ r-t .-i :CfQ O 3-3 eu n> ju ~" 3 E-o' 5^ 3 M- '---' tr >7 SÍ ^ O ja o O n> fu í? ^ 3 O 3 ^ OQ S O (T) 3 co r& 3 1-1 C/) ►1 3 l~^1 3 0Q Éo O 1— o~ Eu "" K-- 55 EU ö 2-3 Eu O- n O g ►*~ ju ^ EU ""^ 3 O' D. 3 3-EU £u O a. . to w EU t-. 3 H^ tu 3 ?r O 3 r-t 3~ rt) 3 Si-3 K-* l-r1 rt 3 !-t 0:3 n" fí ^ £ i< 3 ^ T3 ?T td n EU P 3 7r 3 f!) O O f Wřf 3 ^ P ^ 2 r p 3-2 S ?r fu . 3 Eu cr« EX .T^ 00 > Q. 3-o' 3- 1-i 3- r-f O > a. ^ 9- a o S" s EU £g v> EU 3 rt P. ^3 w 3 D- r-t 3 EI. 2 o S 3 o . n> •O n> en rt 3 O O 3 1-1 crq m >-t ŕ» 3 „ Q 3 3^ fu C ■-<• 3 W £u N. P On řr 5*3 n 3 ,__1 w £~ 3 cr rt O a 3 í" ^ »-n cr cr fu 3 5" a. rc CTQ 3 É n 3 t-1 O Eu 3 CO D- 3 O 3 Ľ • 3 p-}Crq 3- 3- rt> VO O ' CfQ id n> cr o p- tu 3' crq . o fu i-i 0- o Eu 3 - S3 O 3 n n> Qj Q. Ol 00 ÜJ 184 Governance and the World Bank testers come mainly from rich countries. Some of them represent Northern non-governmental organizations. And some of these NGOs collaborate with sister organizations in the South, who may represent smaller grassroots movements of people directly affected by World Bank-supported projects. Local-global civil-society advocacy networks have pressured for the World Bank to apply good governance norms to its own organization in order to become more transparent and accountable in relation to, for example, human rights and environmental issues. A central actor in the 'anti-globalization' movement is the 'People's Global Action' group, who criticizes the World Bank for, among other things, maintaining its privatization policies and the demand for expenditure cuts in Third World countries. An Eastern European-based international NGO, 'Bankwatch', monitors the operations of the World Bank and IMF to prevent environmentally and socially harmful impacts of development finance.5 The emergence of this kind of global civil society could not be ignored by the World Bank, which has not been entirely unresponsive to the demands for change. Pressure from civil-society groups has contributed to institutional measures to increase the Bank's accountability towards groups affected by its programmes. For example, in environmental policies, during the 1980s there was a growing awareness of rainforest destruction and indigenous peoples' rights. NGOs began highlighting cases of World Bank-sponsored projects, in which rainforests had been destroyed or indigenous peoples had been adversely affected (Fox and Brown, 1998). These campaigns, together with internal critique in the Bank, resulted in the creation of an environmental department as part of a reorganization in 1987. A controversial case in which indigenous people were affected was Nepal's Arun III Hydroelectric Dam. There had been much protest against the project because it involved the resettlement of tens of thousands of people. The new president of the Bank, James Wolfensohn, decided to cancel Bank support of the project, thus responding to NGO demands, overruling his own senior management in this regard, and shocking both critics and supporters (Fox and Brown, 1998: 3). The Arun controversy indirectly triggered the establishment of a new accountability mechanism: the Bank's Inspec- Governance and the World Bank 185 tion Panel in 1994. Through the Inspection Panel, citizens of developing countries can directly communicate grievances regarding the environmental and social costs of World Bank projects. The Panel, which is independent and does not employ World Bank staff, then investigates whether the Bank has lived up to its own environmental and social standards in a specific project. It is widely agreed that the Inspection Panel was created in response to sustained advocacy campaigns by coalitions of non-governmental organizations in the North, with grassroots in the South (Fox, 2000). In the US, NGOs collaborated with members of Congress to pressure for reforms in the Bank. Since Congress can control funding of the IDA, it was a powerful channel of influence. Jonathan Fox and L. David Brown quote a Bank official: I was in charge of trying to ... help ... to raise money for IDA and you all know it's gotten very difficult to do so. One thing I learned very quickly is that we need the support of NGOs in the North in order to do that. It also became very clear very quickly that the NGOs in the North are very closely related to their work and experience with the NGOs in the South. And so it became very quickly clear that we had to build better bridges to the NGOs in the South.6 The establishment of the Inspection Panel is one indicator that the World Bank has become more responsive towards its surroundings. An indicator of increased transparency is the publication of a huge number of documents on the Internet. Bank procedures and Bank collected data, as well as country reports, are publicly accessible. In addition, the World Bank tries to respond to NGO criticism by setting up space for debate on the Internet. For example, in connection with the 2002 Oslo conference on development economics, the Bank responds on its website to claims by NGOs (represented by the Oslo 2002 movement).7 The Poverty Reduction Strategy Papers that the Bank now requires recipient governments to implement, in collaboration with stakeholders and civil-society groups, may be seen as another example of how the Bank has focused on openness and responsiveness. In sum, the World Bank has become more accountable to a global constituency represented by international NGOs - a constituency that the World Bank considers to have a legiti- 186 Governance and the World Bank mate interest in its policies. However, the question is whether the Bank has become more accountable in practice or merely in policy. Jonathan Fox has analysed the cases taken up by the Inspection Panel since it was established, and he points out that member governments often block further investigation. In the case of Brazil's Itaparica Dam, for example, the Brazilian government promised irrigation to displaced farmers in order to get the Bank's board to reject the claim. In doing so, it avoided further and more serious investigation of how people were adversely affected by the project. Fox (2000) thus argues that the extent of real change as a result of the establishment of the panel is quite small. Conclusion The World Bank and governance interrelate in three ways. The first has been a matter of debate in comparative politics because it is about the conditions the Bank attaches to loans to poor countries, and therefore affects the discussion regarding the best model of economic governance and how Bank conditions affect state and politics in general. In the 1990s, the Bank supported good governance programmes, but since the turn of the twenty-first century it has become more selective, tending to prefer to give loans to countries that already have good governance. The second way in which governance and the Bank interrelate has more to do with the debate in international relations regarding how to hold large international organizations to account. In this respect, the World Bank is accountable towards its largest shareholder, the United States. But it must also be accountable towards the rest of its member states in order to be accepted as a legitimate institution. Because of globalization, global civil society pressures international organizations such as the World Bank to apply good governance principles to their own organization. The third way the Bank and governance interrelate is that it is a part of global governance, and, therefore, the Bank increasingly becomes accountable to a global public in addition to being accountable towards its members and shareholders. The World Bank, Governance and the World Bank 187 particularly under the Wolfensohn presidency, has taken some steps to improve accountability and transparency. The extent to which these efforts will change Bank practice remains to be seen. But since the United States is the largest shareholder, future changes are not likely to be entirely anathema to the interests of the American government. Governance of the World Bank must still strike a balance between diverging constituencies. The World Bank is thus a case of complex accountability structures and blurred boundaries between domestic and international in a more globalized world. Conclusion The army of governance theorists is so disparate that one is led to think that the word, governance, itself is like a label placed on a whole batch of bottles which are then distributed among diverse producers each of whom fills them with the drink of his choice. The consumer has to look carefully. Baudin, 1942A-51 The statement above was written about the concept of corporatism, but the very same could be said about governance. As the introduction to this book makes clear, governance is used in many different contexts and has many different definitions. For example, in public administration and public policy, one usage of governance relates it to network steering while a broader perception of the concept refers to the management of all kinds of rules and practices affecting policy-making, be they of a hierarchic, market- or network-dominated character. And in European studies, governance is used both as referring to multi-level policy-making and as referring to something broader, as more than government but not necessarily of a multi-level type. However, it is possible to discern a core of governance which is common to the different usages. This core has to do with the conception of governance as referring to something Conclusion 189 more than government. Governance processes include state as well as non-state actors who are bound together in a plurality of networks. Governance theories share a broad institutional background, and they are all reactions to perceived inadequacies of existing approaches within their sub-fields. Governance, therefore, is not just old wine in new bottles, but denotes a new approach to the study of politics. By summarizing the preceding chapters about governance in public administration and public policy, in international relations and in comparative politics, this concluding section outlines the common features of governance as used in the three sub-fields of political science. It then discusses the consequences of insights from governance theory for the boundaries of the sub-fields and explores the different usages of governance between the sub-fields. It returns to the issues of democracy and accountability, ending with a discussion of common problems that occur with a governance approach. Summarizing the chapters In the introductory chapter, governance was defined in terms of rules, where rules include norms and formal and informal codes of behaviour. Governance refers to the setting, application, and enforcement of rules. In governance theory, concern with both the input-side (democratic procedures) and the output-side (efficient and effective institutions) can be discerned, although the latter has clearly been debated most. There is also a common concern with rules of the game, but the focus varies between setting, application or enforcement. In chapter 2, the debate on governance in public administration and public policy was sketched. Governance was introduced as a consequence of changes in the public sector during the 1980s, which had been characterized by a wave of new public management reforms. In the aftermath of the reforms, an increasing number of policy networks have emerged. These often prove to be efficient deliverers of service, but they may also lead to fragmentation, and they certainly provide a challenge to overall coordination. CD > "P fÖ "*] J- (C m a .y £ !tľí 0 "5 u a 0 > "■P m —* b m l/l a v e jp 0 "5 u a E a. &_ S > 3 O LU D) ~ň c .2 "P c V) C .2 L. ".M (U (D ■M C 0) i_ c o ■*-> nj > b. U +j .S2 "Ô "E a I o ■a ľa 3 .y Q. JQ ■u 3 C a. (D -J Q. C 3 a -M O 3 Q. "D C S +-> a -M =3 O 51 fO O e oj O >. u ■M c 3 cu a +-■ 'u 3 it O LU C a» 'u LU "D C ■—' >, .i c? .y £ ífc cu LU T3 ľ) (D C *-> g -P "= 2 g ■t; °- E ■K í .ť c Q. fO íľ _ O fö Š I .tť t; ■M 3 l/l s— c: +ť o -s r=. -ti ro _ .r: íj O QJ > i/> ■— >> .2 -a -M C +-1 m 3 OJ ■h .y ■p > Q. 'u 3 O it LU c 1! — 1/1 >. u tc E |P CTI 3 U í« CU O p q! _J LU a. m (U 3 ■o 2; ■&1 £ E. 3 i/T OJ OJ — O tu > _0J -a .y '•p II 3 2 £ *£ >■ ^1 in 1/1 C 03 C 1— o O) 13 ta w o .2 „- +j +■» .ií m 5 « C C O "TS QJ § S o i; c 01 S.™ .5-š I« a u o u Conclusion 191 Government is only one of many actors in the delivery of services and, as a consequence, it needs to strengthen its coordinating role. Scholars thus argue that governments must accept they can no longer steer directly, but must learn how to manage networks in an indirect way to enable an efficient service delivery, for example by including important groups in negotiations about policy decisions. There are two major conceptions of governance in public administration and public policy. One is narrow, relating governance to the management of networks, the other broad, referring to the process whereby rules of public policymaking and implementation are set, applied and enforced. These rules (or institutions) can stress a hierarchic, network-oriented or market-oriented organization. While the primary concern in the literature has been with service delivery, it remains an open question whether networks play a large role in all policy sectors. In other words, the role of the state in governance varies according to sector. The focus in governance theory has been on the efficiency of public policy-making and implementation. Legitimacy is thus mainly seen to derive from the output-side: from effective performance. However, some scholars have additionally raised the issue of input-oriented legitimacy: in a situation where much policy-making and implementation takes place within relatively closed networks, the question of how democratic accountability is ensured becomes urgent. Legitimacy and accountability may be increased by including all stakeholders in the policy network; however, the need to make sure that policies are in the interest of all in an aggregated sense, rather than dominated by particularistic interests, calls for the establishment of procedural rules, such as deliberative forums. In international relations, the concern with governance emerged as a result of evidence that complex interdependence had accelerated. Globalization, with intensified global interactions, implies that there is an identified need for regulation at the global level. New international institutions for global problem-solving have emerged, while existing global institutions have taken on new tasks. These changes have led to other actors than the state playing an important role on the global scene. For example, global transnational networks 192 Conclusion have emerged around issues such as the environment, human rights o*r landmines. The concern with governance by IR scholars has been concentrated mainly among liberalise and solidarists within the English School. Realists and pluralists do not consider the construction of institutions for global governance to be feasible. A common definition of global governance is provided by James Rosenau, referring to systems of rule at all levels of human activity - from the family to the international organization - in which the pursuit of goals through the exercise of control has transnational repercussions. Chapter 3 defines global governance along the same lines, as referring to the process whereby rules of global public policy-making are set, applied and enforced. An important question for IR scholars is the effectiveness of international institutions and whether nation-states can be bound at all by international rules. Realists argue that rules can only bind states as long as the rules are in accordance with the interests of that state. Pluralists argue that the rules of sovereignty and non-intervention indeed serve to maintain order most of the time. Liberalists (such as Keohane) often argue that more international rules should be established in order to provide efficient global policy-making. Strong liberalists (such as Rosenau) go further by arguing that the main question is not only whether states will be bound by rules, but is also about recognizing the myriad of actors already engaged in affecting norms and rules in global politics. IR scholars should therefore abandon the presumption that states should always and necessarily be the basic analytical unit. The concern in IR theory has mainly been on the output-side and with the need to strengthen the effectiveness of global institutions. However, there is also an increasing concern with the democratic accountability of such institutions, and thus about the input-side as well: there is a lively debate whether it is possible to establish procedures on the global level to ensure more democracy, or whether it is more realistic to concentrate on existing institutions and, if so, how their accountability could be increased. Studying the European Community was for a long period mainly a concern of international relations scholars, Conclusion 193 as described in chapter 4. The governance approach has the virtue of moving European Studies beyond the debate about integration as a supra-national or intergovernmental affair. Instead, the governance approach takes our attention to what happens, not only at the intergovernmental conferences where treaties are signed, but in the European polity. Rather than asking the IR question about the extent of integration (how far has it proceeded?), the governance approach identifies the actors involved in the EU policy process and analyses the impact on national policy-making (what are its consequences?). Some EU scholars identify multi-level governance (MLG) with a specific institutional set~up, involving networks of actors at three levels: the regional, the state and the EU level. Others have a broader conception of governance as the setting, application and enforcement of rules for European policy-making, be they networks, markets or hierarchies. The specific institutional set-up varies according to policy sector: Cohesion policy, for example, is characterized by networks of supra-national and national actors, as well as sub-national actors, whereas in regulatory policies, both hierarchic and network steering are commonplace. Studies of governance in the EU have mainly been concerned with the efficiency of European policy-making, for example, the methods of securing compliance by member states. The EU's legitimacy is thus seen to derive mainly from the output-side. But the democratic deficit of the EU has also been raised by governance scholars. They provide the insights that, precisely because many actors and levels are involved in European processes of governance, efforts to increase democratic legitimacy have to be directed at multiple arenas. Chapters 5 and 6 outlined governance in relation to two prominent debates in comparative politics: one about the role of the state in development, the other about democratization. The first is perhaps most appropriately termed comparative political economy, and overlaps slightly with some of the issues touched upon in chapter 2, because economic policy obviously is also of interest to scholars in comparative public policy. Theories about the state and economic development discuss the optimal model for development. The classical 194 Conclusion liberal model emphasized market-led development; the Key-nesian model a more regulated market-capitalism; and the neo-liberal model has argued for deregulation and a dismantling of the state. The dichotomy between market and state in these models tends to ignore other ways in which the economy can be governed. Self-governing networks are important in this context, because they span across the public-private divide. Governance refers to how institutions for economic policy-making and implementation are set up, and what their consequences are. Governance does not refer to one particular model of development, but many scholars used insights from studying developmental states in East Asia to criticize the particular governance model applied by Western donors. The main concern of students of governance and development has not been democratic-input procedures. It has been implicitly assumed that institutions generate legitimacy through effective outputs. To the extent that democracy has been debated, it has been part of an argument that democracy is not a necessary prerequisite for the developmental state, as the East Asian experience demonstrates. However, democratic procedures may be an intrinsic part of the developmental states elsewhere, which is illustrated by the examples of Botswana and Mauritius. Democratization theories were initially dominated by the modernization approach that searched for the structural prerequisites of democracy. Transition theories, on the contrary, studied actors in transition processes, and assumed they would carry through a transition process that eventually led to democracy. Governance treads a middle ground and acknowledges institutional constraints on individual behaviour, yet, at the same time, when institutions are in flux, individuals and groups have a potentially large influence on defining new rules. Governance is about the setting, application and enforcement of regime rules, but the focus is on the setting of rules, precisely because a transition implies a move away from one institutional set-up towards another. Governance in democratization theory is thus, in a sense, meta-policy-making: it refers to the setting of rules that guide rule-making. Contrary to transition theory, governance does not assume that transitions will result in democracy; rather, democracy'may be one outcome of governance processes. Conclusion 195 Governance is less loaded than democracy and is therefore capable of covering broader processes of change. In all, governance in comparative politics involves the study of different institutional models and processes that are relevant to economic development and regime change. The difference between the two is on the type of rules in focus: theories about the state in development focus on setting an institutional framework for economic policy-making, while democratization theories focus on setting an institutional framework for the political regime - a constitution. Both study the conditions of governance and debate how some economies and societies are less governable than others. Because governability varies, one unique model of governance cannot be applied universally, and this is true for political as well as economic constitutions. States take up the challenges posed by globalization in different ways. Weaker states tend to depend heavily on the international financial institutions and therefore carry out neo-liberal programmes that may not be suitable to their economic and political situation. International trade rules do not allow weak states to follow the same rather protectionist strategies that the developed countries followed before them. Stronger states, such as the developmental states in East Asia, tend to be more autonomous towards external actors and have experimented with their own, more state-directed, models. Thus, although globalization may involve a fragmentation or dislocation of power away from the nation-state, state institutions remain important filters for global processes, resulting in different outcomes for different national settings. The two traditions in comparative politics have remained distinct. While there is a literature on how democratization affects economic reform, there is still a need for research about how processes of regime governance and economic governance interact. Finally, chapter 7 demonstrated how studying an international organization such as the World Bank can involve drawing upon more than one way to use the concept of governance; in this case governance in relation to development models, but also as a framework for analysing regime transitions. Also, governance as a global process is relevant, because the Bank is an important part of global decisionmaking structures regarding economic redistribution, envi- 196 Conclusion ronmental issues and, also, the promotion of 'good governance'. 'Finally, the World Bank is an institution that illustrates the complex accountability structures that characterize global politics today. Disciplinary boundaries Governance theories have developed out of different theoretical debates, and they remain quite insulated. With a few exceptions (especially between Rosenau and Rhodes), there are not many cross-references between the governance literature in different sub-fields. Yet the very focus of governance theory, i.e. to investigate the political implications of - and responses to - social and economic change, indicates that sharp boundaries between political science sub-fields cannot be upheld. Borders between nation-states were never watertight. Indeed, a long-standing tradition in political sociology, represented by authors such as Anthony Giddens and Michael Mann, maintains that the nation-state was never entirely a domestic creation. The spread of the modern nation-state across the globe is seen by these authors as a defining feature of globalization. Globalization entailed the consolidation of, rather than the erosion of, states. With accelerated globalization since the 1980s caused by, among other things, the deregulation of capital flows and a new electronic communication infrastructure, social relations extend even more across borders. The only possible exceptions to the global reach of markets are some parts of Africa and South Asia. If we recognize that loci of authority extend beyond the nation-state, the study of politics has to let go of artificial boundaries between comparative politics and international relations, or between comparative political economy and international political economy. In this book, these boundaries are upheld to maintain structure and clarity, but another important reason is that the debates have largely remained with the distinct sub-fields. If the purpose, as in this volume, is to relate governance to existing debates, the distinction between sub-fields is therefore maintained. Reading through the chap- Condusion 197 ters should make clear that there is a great need for sub-disciplinary interchange (although there are some exceptions regarding this distinction between sub-fields in recent publications, for example, Pierre, 2000). Since there are many common concerns in governance theory, there are also common problems, an issue we shall turn to now. More cross-disciplinary collaboration among governance theorists could mean that such problems would be more adequately addressed. Key discussions and common problems in governance theory In conclusion, we shall turn to the key features currently occupying all governance theorists regardless of sub-field. These are networks, reciprocity, accountability and democracy. They are discussed from different angles and with varying attention, but they remain vital concerns. This final section also addresses some problems in governance theory, mainly concerning the lack of attention to issues such as power, interests and conflict. Networks and reciprocity To a certain extent, all governance scholars study relations of reciprocity, whether inside networks or across networks. To illustrate, governance in relation to democratization theory deals with the fundamental issue of generating legitimacy for the public realm by establishing democratic procedures. These procedures involve reciprocal action by state as well as societal actors. Theories about economic governance stress the reciprocal interaction between state and economic agents in networks that increase the efficiency and implementation of economic policy. The theory about European governance stresses the involvement in policy-making of networks of national representatives, commission officials and non-state actors. Governance in public administration and public policy developed out of the debate on policy networks, in which horizontal reciprocity among the members of a 198 Conclusion network is seen as essential. Governance in international relations is also occupied with how to govern international networks. Many IR scholars believe that reciprocity and trust, spanning across nations, is crucial for the establishment of international institutions of governance. There is a difference with regard to scale, of course. Whereas units of the networks may be individuals in local governance, they are most often organizations, or individuals representing organizations, at the national level. At the global level, the units are states and NGOs. The larger the network and the wider it expands, the more likelihood there is of difficulty in locating the core of authority. In the case of a policy process in a medium-sized town, networks may render the identification of authority more difficult, but, in a network comprising many international NGOs and other actors, it can be almost impossible to identify who is responsible for a certain action. The difference of scale thus gives rise to different consequences for accountability and democracy, as we shall see on the following pages. There is also a difference with regard to whether networks are viewed in positive or negative terms. In public policy, for example, networks are usually viewed in a positive manner, because they increase policy-making efficiency. They can generate information and ease implementation immensely. However, some also highlight the negative implications: networks can function as barriers that impede effective implementation. In theories of development and democratization, networks are seen to have more ambiguous consequences. On the one hand, horizontal networks help to generate trust and reciprocity across regions and ethnic groups. On the other, vertical networks may reinforce cleavages and inter-group conflict. In international relations, the democratic potential of networks as pressure groups on governments is recognized, but their unaccountability is viewed as a serious problem. Finally, although negative as well as positive implications are discussed, there is still very little emphasis on conflict within or among networks. Self-governing networks are based on reciprocity and consensus. Yet there has been very little discussion (with a few important exceptions) of a failure to reach consensus. What happens when conflicting interests within the network cannot be reconciled? In much governance Conclusion 199 theory, relations seem to run smoothly and there are few frictions. However, the essence of politics concerns the determination and allocation of values, and decision-making cannot escape conflicting interests with regard to how values are allocated. The democratic state (as pointed out by Paul Hirst (2000)) was designed to contain conflict; networks were not. It seems, then, that states will still have an important role as providers of the stability which is a pre-condition for effective decision-making in the first place. Accountability The common concern with networks additionally implies a preoccupation with accountability. Governance has been identified both narrowly with network management and more broadly as referring to the management of rules. In both senses, there is recognition that policy-making is no longer strictly confined to the nation-state. The search for accountability characterizes most sub-fields, although in varying ways. In public administration and policy, the main focus of the debate on the question of accountability has been not so much with regard to democratic control, but rather as it pertains to bureaucratic or hierarchic control. The expansion of self-organizing, self-regulating networks in service delivery implies a loss of direct control. If public care for the elderly, for example, is contracted out to a private organization that again contracts out specific service functions to yet another set of private contractors, the accountability chain may break. The control over the use of money may be weakened, because the degree of separation between the government and the services it funds become larger. The preoccupation with accountability, in other words, has related more to a concern with how to control the use of resources in order to perform efficiently rather than democratically. This is true for international relations, too, in the sense that the rise of international regimes, NGOs and other multinational actors means a fragmentation of policy-making authority. States no longer have (if they ever had) full control over regulation and decision-making. Above the state there is no authoritative body, nor likelihood of one, which can ensure that the mul- 200 Conclusion tiple actors in global politics can be responsible for their actions. In addition, the case of the World Bank clearly illustrates how accountability has become more complex and multi-directional. The main debates have concerned ways to ensure efficiency and transparency, and thereby ameliorate corruption problems. When discussing the role of the state in economic development, the emphasis on performance accountability can be discerned as well. The developmental state was accountable in many ways, despite the fact that it was not a democracy in the liberal sense. Financial and outcome responsibility was high, and the public authorities were often part of networks to promote technological innovation. Hence, networks were accountable in the sense that they were responsible for producing results. When discussing governance and regime transitions, accountability becomes central not only in terms of performance but also in terms of fairness and equality. Network management is a strategy to build trust and reciprocity, i.e. social capital, because inclusion will promote interaction across regional or ethnically based groups. Accountability is a crucial part of such a strategy because, without it, social capital may be destroyed. In governance theory in comparative politics, it is therefore assumed that accountability increases with increased reciprocity between state and societal actors. This will help build democracy and, essentially, it will strengthen the state's ability to formulate and implement policies that promote economic and social development. Although international relations and public administration scholars see the rise of networks as enhancing flexibility and as potentially enhancing efficiency, they also see them as threatening accountability because they are out of reach of the state. This is not always the case in theories of governance in comparative politics, where the state may derive its strength from formal or informal policy networks. In economic governance, the state may gain from inducing networks of economic agents to self-governance. The questions of state control and state strength are not looked upon in the same light within the different sub-fields and a further exploration of these issues is arguably needed. Conclusion 201 In sum, the type of accountability most in focus in governance theory has been bureaucratic, or performance-related accountability, with the exception of the literature on democratization. In all sub-fields, however, there is an increasing concern with democratic accountability, which has important, though different, implications at different levels of governance. Democratic governance at different levels At the local level, the rise of community networks and voluntary organizations in service delivery has had positive as well as negative implications. On the positive side, there is more flexibility, and service deliverers have been observed to be more responsive towards clients' needs. In addition, participation has broadened and individual members of networks have been empowered. On the negative side, there is more fragmentation, less control, and the risked exclusion. Remedies to these problems have been put forward, one suggestion being that local councillors could be more involved in the networks. However, the role of local councillors should also be to ensure the aggregate interest of the whole community, so there is a danger that local councillors might identify excessively with the interest of one particular network. Hence, a procedural solution could be the establishment of deliberative forums in which decisions are discussed with a representative group of citizens. At the national level, it has been argued that the emergence of networks, with the increased complexity and globalization of economy and society, narrows both the reach and the degree of control left in the nation-state. Although we have argued that abandoning the state as an important locus of authority and power is premature, it can be maintained that the conditions under which democracy is exercised have changed. When the state administration loses its reach, national parliaments can no longer hold it to account. However, even in such a situation, national parliaments may still play a role as democratic inputs to supra-national organs, as has been debated in the case of the EU. The implications 202 Conclusion of globalization and network society for democratic governance thus need to be examined further. Some democratization theorists have not yet taken globalization and the challenges it poses for democracy into account, but others, such as Andrew McGrew, David Held (cosmopolitan democracy) and Paul Hirst (associational democracy), argue for the necessity of experimenting with democratic models more fitting to the complex and fluid power structures that characterize modern politics. Although different, both models support democratization at many levels and they argue for the importance of democratizing self-governing organizations at the lowest levels. Functions that can be carried out by such organizations should be decentralized, although national parliaments would still be important as watchdogs and forums, in which decisions regarding the allocation of public funding could be taken, On the regional level, the European Union is an example of a supra-national organization that in many ways resembles national political systems, and governance theory studies it as such. Multi-level and transgovernmental policy networks increase efficiency in many areas, but they also render accountability mechanisms more complicated. Negotiations take place in these networks which are beyond the control of national parliaments. The European Parliament could play a role, but its influence remains limited. In some sense, then, policy-making in the EU is beyond the control of European citizens, and it is possible to talk about a democratic deficit in the EU. The insights from governance theory are that the multiple loci of policy-making and implementation imply that democratization cannot be introduced at one level only; it involves greater openness and transparency at multiple levels. At the global level, the implications of globalization, networks and complexity for democratic accountability are more difficult to discuss. At the local level, the democratic problem posed by networks seems easier to solve than the democratic challenge provided by networks and regimes at the international level. There is no global government and it is naive to assume such government could be designed in the near future. The best we can do is to improve the accountability of existing institutions, and a number of actors are making efforts in that respect. As chapter 7 on the World Conclusion 203 Bank made clear, international NGOs have started to demand more accountability from international organizations. In addition, some boycotts of big multinationals have actually led to changes in the way companies act. McDonald's or Nike are among examples of large corporations that have been targeted by such actions. Such cases make James Rosenau (2000: 195) argue that 'nascent forms of democratic governance can be discerned in the labyrinth of globalized space'. However, a global democracy based on the same principles as national democracy is not feasible, for the simple reason that power is diffused and exercised in a myriad of loci. In addition, making international institutions more accountable to NGOs will not suffice because, typically, the NGOs themselves are not accountable to any well-defined constituency. Therefore, further exploration of existing or possible accountability mechanisms at the global level is needed. The insights from governance theory thus have different implications for democratic governance at the various levels. At the local, national, and possibly even at the regional level, measures to strengthen accountability are more realistic than at the global level, where there is no recognized elected forum for decision-making. Continued importance of state and hierarchy? The emphasis on networks in governance theory has a flip side. It tends to ignore the continued importance of hierarchy. And it tends to ignore the interplay between hierarchy and networks. As outlined in chapter 2, network governance may take place in the shadow of hierarchy. Networks may have representatives from different state-organizations. If interactions within the network become more frequent than interaction with the mother-organization, hierarchical accountability may suffer. This is what often occurs in weak states, where public officials are generally more loyal to their own ethnic group than to their employer. Further consideration of the interaction between hierarchies and networks is clearly needed, together with an examination of the conditions under which hierarchical solutions would be preferable to networks. Our discussion suggested that when benefactors 204 Conclusion from a, policy are highly concentrated, interests may be skewed in one direction, with the result that a network could be dominated by these interests. Hierarchy would be needed to ensure that the outcome was in the interest of the majority. In public administration and public policy, and to some extent in the governance literature in IR, there is a tendency to argue that governance now takes place without government. However, as discussed, the state arguably remains important, both in a number of policy sectors and on the international scene. States are key filters through which global processes are moulded. As channels of democratic input to supra-national organizations, states remain sources of legitimacy in modern governance.2 States are better placed to contain conflict than networks. And, although states may lose sovereignty to supra-national organizations, they also gain strength from them - for example, EU member states sometimes attribute blame to the EU when implementing unpopular policies. In comparative politics, the concern with the state is still prominent. This is probably because the attention is often on weak states in the developing world that are in the midst of state-building. Governance therefore implies the setting of meta-rules, i.e. rules about establishing a political order. In such settings, it is not the hollowing out of the state that is relevant, but rather how to strengthen a state that is under pressure from both the global economy and Western donors, as well as from societies that are very difficult to govern. In all, governance does not take place without government, and governance theory should leave the role of the state open to empirical investigation rather than simply assume that the role is declining. Under-emphasis on power Finally, the attention on networks in governance theory tends to ignore a fundamental aspect of power: networks typically have limited membership. Although they span across the state-society divide, they include some actors and exclude others. For example, weak states are typically char- Condusion 205 acterized by closed ethnic networks, in which case governance is concerned with setting rules that avoid conflict between them. Power often has a winner-takes-all character, because the ethnic group in power may favour its own members at the cost of other ethnic groups. This problem of network management is not as serious in stronger states, but it clearly illustrates the danger of letting particularistic interests dominate policy-making. The focus on state-society synergy and on a win-win situation tends to underemphasize the cases in which the state is overly repressive or in which society needs to assert its autonomy against the state. Focusing on win-win situations is important, but there is a tendency to ignore who loses and how losers could be compensated. In international relations, too, the realist insight that state power and national interests may dominate institutions for global problem-solving provides a sobering angle on global governance enthusiasts. It could also be appropriate to further develop themes of neoMarxist arguments regarding corporate power. Hence, practically all usages of governance lack a discussion of power and interests. By analysing cases of state-society synergy and win-win situations, governance theory offers a new angle on the rules of policy-making, whether on a sub-national, national, transnational or supra-national scale. However, this should not allow us to ignore the fact that some actors have more power and may therefore dominate the allocation of values. Despite these problems, governance is a concept that is here to stay in political science. Its reference to a process involving more than government indicates far more than a reworking of old ideas. It concerns the political implications of social change, applying a fresh focus on political institutions in a changing world.