\4/PS2372 POLICY RESEARCH WORKING PAPER 2372 Financial Openness, What explains the spread of both democracy and financial Deniocracy, and openness at this time in Redistributive Policy history, given the constraining impact of financial market integration Mansoor Dailami on national policy autonomy? International policy coordination is part of the answer, but not all. Also important is the presence of cost-effective redistributive schemes that provide insurance against the risk of financial instability. The World Bank World Bank Institute Governance, Regulation, and Finance June 2000 POLICY RESEARCH WORKING PAPER 2372 Summary findings The debate about the relationship between democratic develops an analytical framework for a welfare cost- forms of government and the free movement of capital benefit analysis of financial openness to international across borders dates to the 18th century. It has regained capital flows. prominence as capital on a massive scale has become The main welfare benefits of financial openness derive increasingly mobile and as free economies experience from greater economic efficiency and increased continuous pressure from rapidly changing technology, opportunities for risk diversification. The welfare costs market integration, changing consumer preferences, and relate to the cost of insurance used as a mechanism for intensified competition. coping with the risks of financial volatility. These These changes imply greater uncertainty about citizens' insurance costs are the economic losses associated with future income positions, which could prompt them to redistribution, including moral hazard, rent-seeking, and seek insurance through the marketplace or through rent-avoidance. constitutionally arranged income redistribution. A cross-sectional analysis of a large sample of As more countries move toward democracy, the developed and developing countries shows the positive availability of such insurance mechanisms to citizens is correlation between democracy (as defined by political key if political pressure for capital controls is to be and civil liberty) and financial openness. averted and if public support for an open, liberal More rigorous econometric investigation using logit international financial order is to be maintained. analysis and controlling for level of income also shows Dailami briefly reviews how today's international that redistributive social policies are key in determining financial system evolved from one of mostly closed the likelihood that countries can successfully combine an capital accounts immediately after World War II to openness to international capital mobility with today's enormous, largely free-flowing market. democratic forms of government. Drawing on insights from the literature on public choice and constitutional political economy, Dailami This paper - a product of Governance, Regulation, and Finance, World Bank Institute-is part of a broader research effort on "The Quality of Growth." Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact William Nedrow, room J3.283, telephone 202-473-1585, fax 202-334-8350, email address wnedrow@worldbank.org. Policy Research Working Papers are also posted on the Web at www.worldbank.org/research/ workingpapers. The author may be contacted at mdailami@worldbank.org. June 2000. (31 pages) The Policy Researcb Working Paper Series dbsseminates the findcRgs o work in progress to encourage the excange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully p61ished. The papers carry the names of the autbors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the autbors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center FINANCIAL OPENNESS, DEMOCRACY, AND REDISTRIBUTIVE POLICY MANSOOR DAILAMI LEAD ECONOMIST THE WORLD BANK INSTITUTE Mdailai@worldbank.org I would like to thank Dani Kaufinann and Gil Mehrez for their comments and suggestions, and Ilya Lipkovich and Maarten Troost respectively for expert research and editorial assistance. Financial Openness, Democracy, and Redistributive Policy' I. Introduction The observation that all advanced democratic countries are also open to free capital mobility across their national borders invites renewed debate on the link between democracy and financial openness. This debate has a long and distinguished history that goes back to the 18t century, but it has gained prominence recently as capital has become increasingly mobile at a massive scale, and as free economies have come under continuous pressure from rapidly changing technology, market integration, changing consumer preferences, and intensified competition. Reflecting the influence of advances in communication and information technologies, financial innovations, and deliberate government policies to abolish barriers and controls on capital mobility, international financial markets have grown significantly in recent years. The volume in international lending in new medium and long-term bonds and bank loans reached $1.2 trillion in 1997, up from $0.5 in 1988 (BIS 1998). International financial transactions now dwarf world trade at more than five times the value of world GDP. The average daily turnover in foreign exchange markets reached $1.6 trillion in 1995 (up from $0.2 trillion in 1986), compared with the $6.7 trillion a year in trade in goods and services.2 While free capital mobility has its antecedents in 17th and 18t century Europe, the worldwide triumphant spread of democracy as a desirable system of governance is a fairly recent phenomenon bolstered by several developments, including the rise of global civil society and the information age, and the collapse of state socialism. This expansion of democracy, coupled with the corresponding increase in political and civil liberty, means that for the first time in human history, electoral democracy is the world's predominant form of governnent, which represents an historic victory over alternative forms of government (Held, 1995). A recent survey conducted by the Freedom House found that 88 of the world's 191 countries (46 percent) - the largest number ever recorded - were rated as free, meaning that "they, maintain a high degree of political and economic freedom and respect for civil liberties."3 What explains the spread of both democracy and financial openness at this juncture of history, given the constraining impact of financial market integration on national policy autonomy? To be sure, both the goals of a democratic polity and free capital mobility command I This paper was largely inspired by the author's earlier contributions to The Qudity of Grwth, a research project led by Vinod Thomas of the World Bank Institute. In this work, contributors seek to propagate a broader conception of economic growth, embracing environmnental sustainability, governance, financial stabiliy, and human capital as the fundamental ingredients of quality growth and poverty alleviation. In this paper, we examine in greater detail the link between financial openness and democracy as two important aspects of the evolving global govemance structure. 2 The figure reflects the average of world imports and exports of goods and services in 1988 (WDI 2000). 3 Democratically elected govemrnments govem a majority of the world's population. As noted above, 2.354 billion people (40 percent of the world's population) now live in free societies, 1.57 billion (26.5 percent) live in countries that are partially free and 1.984 billion (33.5 percent) live in non-free countries. (Karatnycky, 1999). 1 a high degree of respect today arnong international relations scholars and policy-makers. These goals are commonly viewed as an important ingredient of "liberal democracy." Yet, the practical importance of the liberal component of liberal democracy, when viewed against the backdrop of post-war history, has not been in accord with the tenets of the classical liberal order. Rather, it is more akin to the "embedded liberalism" of the Bretton Woods era, which was composed of a liberal international order in trade and investment with Keynesian-welfare economics on the domestic side. Greatly influenced by the trials of the Great Depression and the Second World War, the architects of the Bretton Woods system favored the use of capital controls by national governments as a tool for preserving national macro-economic policy control, and as a means for defending stable exchange rates and the liberal international trade order. The commitment to protective economic security, which was embedded in this model and provided through a mixture of aggregate demand management, redistributive taxation, and regulation, served the cause of democracy well in combating the threats from fascism and communism. At the same time, multilateralism in trade and investment on the external side contributed to the domestic growth and prosperity upon which policies of income redistribution could be built and maintained. The collapse of the Bretton Woods system of fixed exchange rates in the early 1970s, the rise in oil prices, chronic inflation and slumping global economic conditions, led to a fundamental reappraisal of this policy mix as well as the whole edifice of the political economy of the welfare state. Not least, the redistributive logic of the welfare state came under attack from the forces of globalization of finance, bringing home the realization that with capital mobility and flexible exchange rates, a national government's autonomy in undertaking counter-cyclical macro policy is limited, as France experienced under the socialist govermnent in the early 1980s (McCarthy, 1990). The resurgence of neo-liberalism in the 1980s and 90s, with its emphasis on free capital mobility and a more orthodox fiscal and monetary policy, and the literature related to public choice and constitutional political economy raised serious questions about the cost-benefit calculus of consent for redistribution in a democratic society, as well as the economic costs and the propensity for rent seeking behavior. The institutional solution that is now emerging contains two clearly identifiable trends: first a thrust toward the expansion of the marketplace to areas that were previously under the control of the state, such as the provision of infrastructure services, as well as critical social services such as health and education. Secondly, there is a growing sense and recognition that questions of justice and poverty alleviation are of universal concern and central to the debate on the evolution of the world economy and the sustainability of the global democratic order. The first trend is grounded in the economic logic of efficiency and the better distribution of risk that the private provision of social services seeks to deliver.4 The latter derives its legitimacy from the international redistributive justice doctrine, and a related concern for the social dimensions of the globalization process.5 6 Thus, the centrality of poverty reduction in the contemporary 4 See Dailami and Klein (1998) 5 See Bohman (1999), Beitz (1999a, 1999b), Thomas (2000), Sen (1999), Kapstein (1999) and CEPR (1998) 6 At present, there are approximately 3 billion people living under $2 a day, widt about 1.3 billion living on less than $1 a day. This point is well-explored in 7h Q 4alty of CGmnv, edited by Vinod Thomas. 2 development debate has defined a new mandate for international financial organizations, as articulated most forcefully by James Wolfensohn: "At the center of the issue is poverty; at the center is equity, because whatever be the governments and whatever be the framework, if you have a preponderance of poor people, if you have inequity, you are going to have not only social injustice by definition, but you will have instability. This is something which is increasingly and appropriately recognized and not just as an issue of moral and social conscience in terms of dealing with the question of poverty. It is an issue of good politics. It is an issue of stability and peace."7 With these trends in perspective, how attainable are the twin goals of democracy and financial openness for developing countries? For these countries, democratic aspirations and ideals have been on the rise in recent years, and yet their integration into global international capital markets has been difficult, fraught with risk, and associated with cycles of severe and costly currency, financial and sovereign debt crises, as demonstrated by recent financial crises.8 Experience over the past two decades has been telling: the loss in aggregate domestic output in 1997-99 in affected East Asian countries, measured by the deviation from trends, is estimated to be $500 billion (in 1996 prices and exchange rates), nearly 1.3 times those countries' external debt in 1996 (Dailami, 2000). Latin America lost a decade of economic growth following the debt crisis of the early 1980s, and the international financial community has extended a large sum of financial assistance through multilateral and bilateral rescue loans to crisis-affected countries in the 1 990s. The high social and economic costs associated with financial instability are unacceptable, and provide a strong case for financial innovation and devising better approaches to avoid financial crises in the future and to reduce their severity when they occur. Strengthening domestic regulation and supervision of banks and other intermediaries, rebuilding the information infrastructure of financial markets, including accounting norms, and improving corporate governance constitute the necessary first steps, as has been widely reported in both academic and policy circles. But these will not be sufficient unless they are reinforced by actions to maintain public support for open capital markets. Securing public support for financial openness in democratic countries will require the availability of mechanisms through which insurance is provided to citizens, either through the marketplace or through redistributive policy in the form of public expenditures on education, health, and transfer payments. As with other public goods, public support could suffer from the under-investment problem, meaning that its adequate provision would require addressing the associated agency, moral hazard, and incentive issues. 7 "Development Choices in a Changing World,' Speech made to the American Philosophical Sodety, Philadelphia, Nov. 11, 1999 8 In 1997-1998, a combination of tight liquidity in international capital markets along with austere domestic macroeconomic policy responses led to deep economic and financial crises, which were unprecedented in several important respects, including the extent of the initial depreciation of local currencies, the plunges in asset values on local equity and bond markets, the severe financial distress in finance and industry, and the contraction of economic activity. 3 This article proceeds in three sections. The following section proceeds with a brief narrative of how today's international financial system evolved from one of mostly closed capital movements immediately following the Second World War to the enormous, largely free-flowing market that it is today. The paper will draw on insights from the literature on public choice and constitutional political economy to argue how the logic of economic exit and political voice affects countries' policy choices, both domestically and internationally. Two factors figure prominently: international policy coordination on macroeconomic and regulatory policies; and redistributive policy (health, education, and transfer payments). In this context, the third section begins with a quantitative measurement of financial openness to cross-border capital flows and democracy, and proceeds with some cross-country econometric investigation on the link between financial openness and democracy. The econometric results, which are based on simple correlation and more rigorous logit analysis, provide strength to the argument that redistributive social policies are key in determining the likelihood of countries successfully combining an openness to international capital mobility and democratic forms of government. The fourth section concludes the paper. II. Historical Antecedents Embedded Liberalism and The Bretton Woods Era The relationship between financial openness and democracy appears at first sight to be primarily a function of the level of income, and more precisely, per capita income: rich countries are, with few exceptions, democratic in government, either presidential or parliamentary in type. Rich countries are also open to international capital movement, as they have a high degree of financial sector development, currency convertibility on capital accounts, and enjoy macroeconomic stability, domestic rule of law, and stable institutions that guarantee civil and political liberty. At a deeper level of analysis, the link between democracy and financial openness proves to be complex and subject to historical tradition. For the classical liberal economist, the rise of movable capital in the 1 7th and 1 8h centuries in Europe was seen in a favorable light. As recounted by Hirschman, "...the fact that, with the bill of exchange, a large portion of wealth had become mobile and elusive and was capable of both hiding and expatriation is celebrated as a restraint on the grand coups d 'autorite of the prince and as a positive contribution to good government...."9 This classical liberal ideal, however, broke down in the 1930s with the Great Depression and the Second World War. The post-World War II institutional reconstruction, which created the Bretton Woods system of monetary and exchange rate arrangements, was grounded in the compromise of 9 The fears and hopes aroused by the rise of movable capital in the seventeenth and eighteenth centuries offer many interesting parallels with similarly contradictory perceptions caused quite recently by the rise of the multinational corporation. One could substitute the term "globalization of finance" for "multinational corporation" and make this observation even more relevant to today's world. 4 "embedded liberalism".'I This compromise involved an openness to international arrangements focusing on trade, exchange rates, and investment, combined with national autonomy and discretion in macroeconomic policy and capital flow controls (provided that such controls were not intended to restrict trade). With the Bretton Woods Agreement, capital controls became an accepted norm of the prevailing international monetary system. Indeed, not even the IMF was granted jurisdiction over capital movements. Reflecting the understanding of the time, John Maynard Keynes expressed the issue succinctly in his oft-quoted speech to Parliament, stating that: "Not merely as a feature of the transition, but as a permanent arrangement, the plan accords to every member government the explicit right to control all capital movements. What used to be heresy is now orthodox... It follows that our right to control the domestic capital market is secured on firmer foundations than ever before, and is formally accepted as a part of agreed international agreements." The relative closure of national economies - with a few exceptions - to the free flow of capital in this era afforded governments the scope for deploying the instruments of fiscal and monetary policy, including progressive taxation and public expenditures, in pursuit of national objectives such as full employment and social equity, without fear of the exit of capital. The analytical underpinning of Bretton Woods was the classical open economy models of Mundell and Fleming, according to which countries can attain only two of the following three conditions: capital mobility, fixed exchange rates, and monetary policy autonomy. At the same time, fixed exchange rates facilitated the process of trade liberalization in OECD countries, which was critical for the expansion of world trade and output. According to John Ruggie, the essence of this liberal order - distinct from the classical liberalism of the gold standard - was multilateralism in trade and related exchange arrangements on the external side, combined with unilateral state interventionism in pursuit of legitimate national goals (full employment, national security, and social stability) on the domestic side. This policy mix, reflecting the preeminence of the "embedded liberal" framework of the time, accepted the use of capital controls in Western Europe and Japan in the early post-war era, with the US taking an accommodating and even sympathetic stance. Vital to the "compromise of embedded liberalism", as emphasized by Sally (1998), is an international focus on creating a network of inter-governmental institutions that promote international cooperation and stabilize social contracts, and a domestic focus on cushioning and spreading the costs of adjusting to the measures of international liberalism. By combining fixed exchange rates with capital controls on the external side, and Keynesian welfare state macroeconomics on the domestic side, the Bretton Woods system yielded a stable basis for economic growth and trade expansion. As history attests, this policy mix proved highly successful in at least four key areas: (i) reducing the barriers to trade, which led to a rapid expansion in world trade; (ii) facilitating the move toward the free flow of capital and full capital account convertibility in the major industrial countries; (iii) accommodating the external financing needs of developing countries in the context of their growing economies and increasing integration into the world economy; and (iv) consolidating the move toward I0Coined by John Ruggie (1983), this term connotes a commitment to a liberal order different from both the economic nationalism of the 1930s and the liberalism of the gold standard. For further elaboration, see G. Garrett (1998). R. Sally (1998) also referred to embedded liberalism as "mixed system thinking, or "Smith abroad and Keynes at home". 5 democratic governance and political liberty, initially in Western Europe and as of late, worldwide. The collapse of the Bretton Woods system, the subsequent floating of exchange rates, the rise in oil prices, chronic inflation, and slumping global economic conditions led to intensified currency and interest rate risks in global financial markets during the 1970s and 1980s. These risks instigated responses that were principally "market solution"'" oriented, exemplified by the drive toward the international diversification of capital, and the impressive expansion of derivitive markets (i.e. interest and currency forwards, options and swaps). These steps occurred in tandem with an important shift in the direction of macroeconomic policy away from its traditional focus on full employment and toward price stability. The success of these actions has been considerable on both fronts - derivative markets today provide a broad a range of hedging instruments for managing currency and interest rate risks in major currencies. The total market value of currency and interest rate futures, options, and swaps traded both on exchanges and over the counter was estimated at the end of 1997 to have reached over US$40 trillion, which is considerably larger than the market capitalization value of world stock markets. On the macroeconomic front, industrial countries as well as many developing countries experienced considerable success in attaining stability, with reductions in fiscal deficits and the lowering of inflation and interest rates. Indeed, cross-country empirical research shows that volatility in the main macroeconomic variables, i.e. growth, export and inflation rates (measured by their standard deviation in a sample of 90 developing countries), in the 1990s declined by more than 60 percent as compared with the 1980s. Today, the implementation of measures ensuring financial openness among OECD countries is nearly complete. Progress towards liberalization of capital controls accelerated, particularly in the 1980s, as members' liberalization obligations under The Code of Liberalization of Capital Movements were broadened to include virtually all capital movements including short-term transactions by enterprises and individuals.'2 Thus, the U.K. abolished all exchange controls and achieved capital account convertibility in 1979. Japan completed this in 1980, while the timeline for the rest of the OECD stretched until 1992, when the last group- comprising Ireland, Greece, Portugal, and Spain-completed the abolition of their capital controls. By the early 1990s, the capital accounts of OECD countries were open to a wide range of cross-border financial transactions including capital market securities, money market operations, forward operations, swaps, and other derivatives. This process of liberalization coupled with the internationalization of financial markets means that today in OECD countries borrowers can raise financing in their desired currency at competitive terns, and investors have the opportunity to achieve their desired degree of portfolio diversification in terms of currencies, maturities and risk profile. Regarding emerging market economies, the overall trends have also been towards the reform of local financial markets and the liberalization of cross-border capital movements, but "See Dailami, Mansoor, 'Managing Risks of Capital Mobility," PoicyResmot WokgPaper 2199, World Bank Institute, 1999. 12 See OECD (1990) 6 the progress, the pace, and the scale of liberalization measures has not been even. Domestic reforms in the developing world that have contributed to financial globalization include the privatization of public enterprises, macroeconomic stabilization, and the relaxation of barriers to cross-border trade in financial instruments for both sovereign and private entities, which in turn improved country creditworthiness and expanded investment opportunities. The underlying liberalization trends have been most clear with regard to the rapid increase in the number of countries that have assumed IMF Article VIII, thereby declaring their currencies convertible on current accounts, which often precedes capital account convertibility. In 1970 only 34 countries, or 30 percent of the IMF membership, had declared their currency convertible on current account transactions. By 1997, 143 countries had done so (see Figure 1). In the 1990s alone, 38 countries, including India, Russia, Turkey, Israel, Greece, and the Philippines assumed IMF Article VIII (a complete list of countries having assumed Article VIII, with date of assumption, is given in the Annex). With regard to the liberalization of capital controls in emerging market economies, two sets of indicators are of interest: First, actual flows of capital have witnessed a significant expansion in the 1 990s, with sharp drops in 1997 and 1998, and recovering once again in 1999. Second, the deliberate policies of national governments in the 1990s clearly reflect a considerable degree of easing of exchange restrictions, controls, and barriers to the entry of foreign financial players engaging in conmmercial banking, securities, asset management, and other financial services. More countries open their cvrrent account Figure 1: IMF member countries with convertible currencies on current accounts Countries Assuming IMF Article VilI 200 180 z 160 E140 00120 '6 100m 20 4060S il 20 1970 1975 1980 1985 1990 1997 * Article Vill members * All Member Countries 7 The Neo-Liberal Perspective The ascendancy of a neo-liberal order in the 1980s and 90s, consisting of free capital mobility, orthodox fiscal and monetary policy, and market-oriented economic management, has brought into sharp focus the tension between democratic concerns for protective economic security and the globalization of finance. With the free mobility of capital, monetary policy is ineffective under a regime of fixed exchange rates, and fiscal policy is ineffective under a flexible exchange rate regime. Thus, under either exchange rate regime, a national government's autonomy in choosing and managing macroeconomic policy is restricted. Yet, concern regarding the effect of free capital mobility on democratic order transcends the ambit of macroeconomic management and relates more fundamentally to a broader range of issues, referred to in the literature as the "democratic deficit;" that is, how can democratically elected national governments provide the necessary compensatory mechanisms that could underwrite social stability and democracy within a globalized financial environment. The question has defined the pivotal point around which the debate on the resurgence of neo-liberalism has centered. Within a strict interpretation of contemporary neo-liberalism, priority is assigned to the maximum possible freedom for goods, capital, and investment to move across national borders. This freedom is regarded as the source of economic prosperity, aggregate welfare gains, and wealth and income creation. Thus, from this vantage point, the redistribution logic of the welfare state, which was so critical for the consolidation of democracy in the post-war era, contradicts with the neo-liberal conception of a minimalist state within the current nation-state international relations paradigm. In other words: "The urge for free markets and small governments has created asymmetries in the relationship between the global economy and the national state that have undermined the post-Second War embedded liberal compromise.""(Devetak and Higgot, 1999). The economic benefits of financial openness are both well-articulated and well-known. Thus, it is generally agreed that open capital accounts bring many economic benefits to both individual countries and the global economy as a whole. Major benefits for developing countries include access to a broader menu of investment sources, options and instruments, enhanced efficiency of domestic financial institutions, and the disciplinary impact of capital markets in conducting domestic macroeconomic policy. Additionally, by easing financing constraints, the greater availability of international finance can extend the'time period over which countries can implement needed adjustments.13 For creditor countries, open capital markets can mean broader investment and risk diversification opportunities, particularly as their aging populations and growing pension funds seek higher and safer returns on their investments. From the viewpoint of the global economy, open capital accounts support the multilateral trading system, expanding the opportunities for portfolio diversification and the efficient allocation of global savings and investment (see Fischer, 1998). 13 Markets will be wiling to provide this leeway, however, only if they perceive that countries are truly undertaking adjustments that fundamentally address existing and prospective imbalances. Otherwise, markets will eventually exert their own discipline, in such a way that the time period for adjustment may be brutally shortened (see Dailami and Haque, 1998). 8 This account of the benefits of financial openness also needs to be extended to include a libertarian view emphasizing the freedom of choice and individual liberty that financial openness offers to citizens seeking to invest income and wealth and borrow internationally. This suggests that there is also an important property rights issue involved. As Richard Cooper states, "individuals should be free to dispose of their income and wealth as they see fit, provided their doing so does not harrn others."'' This view has gained prominence in the recent dialogue and narrative of international financial relations. The importance of this liberal paradigm in building a new international financial architecture has also been emphasized by policymakers. According to Lawrence Summers, U.S. Secretary of the Treasury: "We should all be able to agree on the danger of .. . denying a country's own citizens the capacity to convert their own currency and invest abroad. Such measures represent substantial intrusions on freedom." 15 This view assigns particular importance to the goals of openness to international capital movements and democratic governance. These goals command respect today among scholars and policymakers concerned with international finance and the underlying empirical evidence is compelling. The spread of democracy and civil liberty means that for the first time in human.history, electoral democracy is the world's predominant form of government. Furthermore, cross-sectional evidence reveals a robust association between financial openness and political and civil liberty. It can thus be argued that countries more open to international capital flows are also more open to offering political rights and civil liberties to their citizens. There is a concern, however, that the globalization of finance may tend to undermine democratic forms of governance. This conflict between democracy and financial openness is seen to arise from the constraints that globalization in general and the globalization of finance in particular impose on the ability of national governments to deploy redistributive taxation, regulation, and risk-sharing. The consolidation of the move towards democratic governance, in turn, means increased demand by citizens for a political voice, for national economic security, and for social insurance against heightened exposure to international financial volatility. For policymakers in democratic societies, the globalization of finance has brought to the fore the classic challenge - in Hirschman's terminology - of how to balance the threat of exit of capital with the political demands for voice, and the increased political incentives for further goverrnment intervention in cushioning market dislocation. This challenge is greatest for many developing countries with nascent democratic institutions. As recent events in Russia, Brazil and East Asia have demonstrated, the integration of financial markets has increased the potential threat of capital flight and economic insecurity, bolstering the argument that the risks of closer financial integration might be too high. Such risks include a vulnerability to sudden reversals of capital flows, dramatic jumps in the cost of capital, and loss of national policy autonomy. Shifts in investors' sentiment and beliefs, as reflected in a sharp turnaround in capital flows and/or a spike in emerging market economies' borrowing costs, can be caused by coordination failures on the part of creditors. This coordination problem could happen because of incomplete information between creditors that could render their decision to run or flee from a particular country dependent on the behavior of others. This dependence can generate a run, analogous to a 14 Cooper, p. 12 (1998) '5 Speech delivered October 22, 1998, Washington, D.C., Cato Institute's 16th Annual Monetary Conference. 9 "bank run" in domestic settings, adding a non-cooperation premium on top of other country risk premia. 16 Thus, it is increasingly being argued that under some circumstances, i.e., weakness in local financial markets, euphoria and panic behavior of foreign investors, and structural balance of payments problems, there exists a case for deploying capital controls, particularly on short- term flows, to reduce volatility. In this respect, Chile's capital controls experience has attracted considerable interest, partly because of its market-based nature, transparency, and the fact that it is easier to phase out restrictions based on taxation than those on quantitative controls.'7 Indeed, the social costs associated with the recent crises in emerging market economies have been substantial. According to a recent study by Stiglitz and Bhattacharya (1999), unemployment doubled in Thailand and tripled in Korea over the course of a year during the recent crises, while standards of living declined 14% and 22%, respectively; Indonesia also experienced a 25% decrease in standard of living. Not quantifiable are further costs such as lost schooling, malnutrition among some, and political strife.18 In the face of these apparent constraints, then, what accounts for the fact that virtually all modern advanced democracies are today also open to international capital movements and indeed, by the early 1 990s had all achieved capital account convertibility on their currencies?19 It will be argued below that there are two dimensions to this explanation: on the domestic side, redistribution; and on the international side, policy and regulatory coordination. In other words, since global financial risks have both macroeconomic and distributional consequences, the appropriate policy responses would entail both macroeconomic and social policy responses, i.e. redistributive measures such as transfer payments and higher public expenditures on areas such as health and education. At the same time, since financial liberalization implies global welfare gains, there exists a rationale for international financial assistance, or in other words, insurance at the global level. Redistribution As Income Insurance The most relevant aspect of redistribution that is germane to the debate on the link between democracy and the globalization of finance is its recent articulation within the scope of constitutional political economy {Kliemt (1993); Wessels (1993); and Buchanan, Tullock (1 962)}. In contrast to much of traditional Pareto welfare economics, which focuses on aggregate welfare and thus has little to contribute to the theory of redistribution, the literature on constitutional political economy explicitly incorporates a consideration of institutions and the public choice process though which income redistribution can be implemented in a democratic society. Thus, the question of why citizens in a democratic society consent to an involuntary redistribution scheme has generated two sets of justifications: First, is the idea that redistribution 16 See Haldane (1999) for further elaboration of this point. 17 For further analysis, see Sebastian Edwards (1998), Gai Flo, Real Ex&,W Rates, and Capita Con&l Sone Latin Ametrca Expeins. Working Paper from National Bureau of Economic Research, Inc. and Leonardo Hemandez and Klaus Schmidt-Hebbel (1999), Capitl Cls in Gile Efftize? Eff:ici? Enduable? Paper Presented at the World Bank conference on Capital Flows, Financial Crisis and Policies, April 15-16, 1999. 18 For further discussion of the poverty implications of financial instabiltiy see Dailami (2000). 19 See Goodman and Pauly (1993) 10 can be seen as income insurance; that is, so far as citizens are risk averse, they may be willing to give up part of their income today to obtain protection against the risk of becoming poor in the future (Olson, 1987; Wagner, 1986; and Wessels, 1993). This approach draws on the theory of insurance and risk aversion to propel the use of income insurance as a basis for redistribution. Secondly, income redistribution policies are deemed necessary to reduce market inequality and thus promote social stability and cohesion. The motivation for redistribution as income insurance - distinct from altruism and other poverty reduction related motives20 - is induced by volatility and insecurity in underlying economic conditions and when citizens are risk-averse.21 In a world with a high degree of capital mobility across national borders, a dilemma facing open democratic societies is how to balance the threat of exit of capital, made more credible by the opening of capital markets, with the political demands for voice, and increased political incentives for government intervention in cushioning market dislocation. The basic argument can be elaborated more clearly in Hirschman's terminology of exit and voice. Financial market integration has increased the exit potential of capital, as investors' ability to shift their financial resources to other countries has increased by virtue of the openness of capital markets. The threat of exit of capital, in this sense, refers not only to investors' decisions regarding existing foreign assets that can be liquidated or unwinded, but also to domestic liquid assets, such as money, and other liquid assets that can be converted and transferred.22 Financial market integration has also heightened the sense of economic insecurity and risk among a broader section of society. Investors dissatisfied with the host countries' policies or prevailing investment climate find it easier to shift their financial resources to other countries and regions, with a subsequent disproportionate distribution of costs borne by less-mobile factors of production - i.e. labor, agriculture. The political dimension, consequently, becomes important. The counterbalance to the threat of exit of capital is the political voice of citizens, demanding protection against external risks through redistribution, social safety net programs, and other insurance-like measures. In the absence of a market for risks, rational citizens will structure non- market institutions to reduce the welfare losses incurred from volatility in economic conditions. Thus, in this interpretation voice belongs to the political sphere, and how it is exercised is a function of the underlying political institutions and, in particular, the degree of political and civil liberty. We argue that the higher the degree of democracy, the greater the need to balance the threat of capital flight, which is more likely with the opening of capital markets, with political demands, which include the need political incentives for increasing government intervention in cushioning market dislocation. Indeed, critical for easing the tension between politics and financial openness in OECD countries, has been the role of their redistributive policies in 20 See Wessels (1993) for a justification of this type of redistributive scheme, within the realm of methodological individualism. 21 The idea of distribution as insurance has, of course, a long tradition in welfare economnics going back to Lemer (1994); Hasaryi (1953),and Rawls (1971). More recently this issue has been analyzed from the perspective of constitutional political economy, see Mueller (1997); Wessels (1993). 22 The exercise of exit of capital can be seen as a function of the degree of the 'liquidity" of the underlying assets - the more liquid the assets, the less transaction costs involved - and the degree of financial openness of the countty. For fixed assets, such as investments in plants and equipment with high sunk-in costs, the transaction cost for liquidating an investment is much higher. 11 mitigating and redistributing risk, through massive transfer payments and other insurance-type government involvement. In recent years (1991-97), government expenditures in high income OECD countries on health, education, social security, and welfare have averaged about 25 percent of GDP, with smaller open European countries such as Norway, Denmark, and Sweden, spending as much as 30% of their GDP. 3 Arguments about the link between redistribution and financial openness can also be anchored in the median voter model with risk aversion (Bishop, Formby, and Smith, 1991). The median voter model has been used the literature on the politics of income redistribution as a framework for analyzing political choice. Thus, in its simplest form, the median voter model predicts that in a democratic system, the median voter will use the ballot box to facilitate self- interested redistribution. Citizens face a trade-off between the costs and benefits of income redistribution. The main benefits consist of a reduction in the variance of future levels of income; and the costs are the deadweight costs associated with government intervention, as well as the fiscal costs associated with financing such redistributive schemes. Any involuntary redistribution - i.e. not motivated by altruism - is bound to entail deadweight economic costs. One type of costs is the cost related to the government bureaucracy needed to administer transfers - the enforcement costs - or as Benson (1999) categorized them, as rent seeking and rent avoidance. There is also an important dynamic cost - there is less incentive to invest when property rights are not secure and a possibility exists for involuntary income transfers. In addition, redistribution policy is often financed through discretionary taxation, which is associated with fiscal and macroeconomic costs that need to be taken into account. These various costs add up to the claim that the marginal welfare costs of redistribution can be characterized by an upward sloping curve, as shown in figure 2. At the same time, building upon the idea of redistribution as insurance against the risk of financial market volatility, redistribution facilitates financial openness with attendant economic efficiency gains and higher economic growth and welfare, which accompany the process of financial market integration. Thus, there exists a corresponding marginal benefit for redistribution that can be characterized by a downward sloping curve, as depicted in Figure 2. The point of intersection of these two marginal cost and benefit curves determine the level of redistribution that a country could optimally decide upon. 23 Focusing on globalization through trade, Rodrik (1997) also emphasizes the relationship between redistribution and openness. 12 Figure 2: Welfare Cost-Benefit Analysis of Redistribution as Income-insurance marginal X benefit arginal O \ / /marglcost o~~~~ I less more Redistribution T1he Role of International Policy Coordination International policy coordination in macroeconomic policy and in financial regulation and supervision merits special attention. As generally recognized, macroeconomic policy coordination among major industrial countries has been instrumental in reducing payment imbalances, in stabilizing expectations for currency and interest rate movements, and in lessening the volatility of capital flows across their borders. 24 In the same vein, the coordination of international banking regulation in industrial countries has been significant, as exemplified by the Basel Capital Accord of 1992 and the subsequent Core Principles for Effective Banking Supervision. With the memory of the 1980s debt crisis and its prolonged resolution still alive, the international policy and regulatory responses to the 1997-99 crises were prompt. These included an easing of monetary policy in the major industrial countries; extension of large stand- by and direct multilateral and bilateral rescue loans; development of internationally agreed codes and standards of good practice, transparency and disclosure; and establishment of a plethora of high-level committees with the aim of strengthening the safety and soundness of banks and other financial firms involved in international capital flows. The international financial assistance committed from August 1997 to December 1998 to Thailand, Korea, Indonesia, Russia, and Brazil amounted to $190 billion-1.4 times their foreign exchange reserves and 30 percent of the reserves of all developing countries at the end of 1997.25 The Miyazawa Plan, unveiled by the Government of Japan in October 1997, committed US$ 30 billion in yen-denominated assistance 24 See Webb (1995) and Bryant and Hodgkinson (1989) for a discussion of international polcy coordination in rnacroeconornics, and Kapstein (1989), for interational coordination of banking regulation. 25 See Dai[arni (1998) and Eidhengreen (1999) for firther elaboration. 1 3 loans to Asian countries affected by the crisis. Furthermore, the G7 finance ministers and central bank governors endorsed in February 1999 the creation of a Financial Stability Forum to bring together the monetary authorities from the G7, principal regulatory agencies, and the multilaterals, and to serve as a focal point for assessing vulnerabilities affecting the global financial system, and identifying and overseeing the required actions. III. Cross-Country Empirical Analysis Our empirical investigation of the link between democracy and financial openness, as elaborated above, calls for reliable data on democracy, financial openness, and measures of redistribution akin to income-insurance. Such data is not readily available and needs to be compiled from various sources, including constructing indices for countries' openness to international capital flows and democracy. Furthermore, in focusing on the role of redistribution as income insurance, empirical measures of such concepts will not be easy, as data on public expenditures is not typically classified by their risk-reducing characteristics, but by sectoral coverage.26 In an effort to provide a proxy that would directly and effectively incorporate the risk-reducing aspects of public expenditures, we interpret redistribution as defined by the sum of government expenditures on health, education and transfer payments, as a percentage of GDP and averaged over the period from 1991-1997. A. Financial Openness to International Capital Movements Evidence regarding the degree or level of openness of emerging market economies to cross- border capital flows is scanty and fragmented. Informational and methodological problems have hindered the proper development of quantitative measures of the degree of financial openness. In contrast to measures of trade openness (or protection), for which a fair degree of consensus exists on both methodology and systematic data availability, research on measuring the degree of countries' openness to international finance remains in its infancy and requires attention. Most studies in the past have relied on measures of incidence of capital controls, i.e., whether a particular transaction is subject to restrictions or not, rather than the degrees of intensity of such restrictions and controls.27 In practice, controls can take a variety of forms ranging from direct quantitative limitations on certain transactions or associated transfers, to indirect measures - such as withholding taxes or reserves on external assets/liabilities, which is intended to influence the economic incentives of engaging in certain transactions. Such controls could apply to the transfer of funds associated with financial transactions, or the underlying transactions themselves. There exists, of course, no single measure of a country's level of openness. Researchers have relied on a variety of proxies and indicators to assess the degree of openness to capital flows. There are, however, important benchmarks, such as a country's assumption of the IMF's Article VIII, and/or currency convertibility on capital accounts. A viable measure of a country's level of financial openness to international capital markets needs to incorporate, at the least, the 26 For furiher discussion on this issue, see Devarajan and Hammer (1998). 27 See, for instance, Razin and Rose (1994), Alesina, GriDli and MIesi-Ferretti (1994). 14 distinction between the severity of controls and the different types of transactions contributing to capital flows. Table 1 provides preliminary information on such a measure for a sample of 96 countries in 1997. This measure referred to as the Financial Openness Index is constructed using disaggregated measures of capital controls based on the classification and information contained in the IMF Annual Report on Exchange Arrangements and Exchange Restrictions (AEAER), and draws on the coding methodology developed by Quinn and Toyoda (1997). The measure is a composite index of our coding of rules, regulations, and administrative procedures affecting capital flows (both inflows and outflows) for a total of 27 individual transactions in the current and capital accounts of the balance of payments for each country in the sample. Thus out of 96 countries in the sample, it is interesting to note that 46 can be classified, as of 1997, as open and 10 as semi-open, with both of these categories including emerging market economies in Latin America and Eastern Europe.The Financial Openness Index has a correlation of 0.51 with the length of time that a country has accepted the conditions of IMF's Article VIII.28 28Correlation is calculated between the Financial Openness Index and the logarithm of the number of years since a country has assumed RMF's Article VIII. 15 Table 1: Financial Openness Index: selected developed and developing countries, 1997 Open* Largely Open* Partially Closed* Largely Closed* Argentina 1.78 Croatia 1.54 Bahamas, The 1.36 Bangladesh 1.21 Australia 1.77 Ecuador 1.54 Belize 1.44 Barbados 1.28 Austria 1.92 Honduras 1.56 Benin 1.48 Bhutan 1.19 Bahrain 1.73 Israel 1.59 Botswana 1.48 Brazil 1.19 Belgium 1.88 Mongolia 1.56 Bulgaria 1.46 Ethiopia 1.12 Bolivia 1.79 Philippines 1.59 Burkina Faso 1.49 India 1.20 Canada 1.92 Poland 1.54 Burundi 1.39 Malawi 1.26 Denmark 1.92 Slovak Republic 1,58 Cameroon 1.41 Malaysia 1.34 Egypt, Arab Rep. 1.81 Slovenia 1,50 Cape Verde 1.39 Morocco 1.27 El Salvador 1.91 Turkey 1.52 Chile 1.43 Pakistan 1.31 Estonia 1.88 China 1.37 Syrian Arab Rep 1.20 Finland 1.83 Colombia 1.38 France 1.73 Congo, Dem. Rep. 1.42 Germany 1.84 Costa Rica 1.48 Greece 1.91 Czech Republic 1.48 Guatemala 1.73 Dominican Rep. 1.49 Guyana 1.72 Ghana 1.43 Iceland 1.74 Hungary 1.49 Ireland 1.93 Indonesia 1.46 Italy 1.84 Korea, Rep. 1.42 Jamaica 1.76 Lesotho 1.41 Japan 1.73 Mali 1.49 Kuwait 1.77 Malta 1.40 Latvia 1.88 Moldova 1.46 Lithuania 1.85 Mozambique 1.41 Luxembourg 1.93 Namibia 1.33 Mauritius 1.82 Papua New Guinea 1.36 Mexico 1.69 Romania 1.48 Netherlands 1.87 Russian Federation 1.43 New Zealand 1.90 South Africa 1.44 Nicaragua 1.82 Sri Lanka 1.43 Norway 1.83 Thailand 1.46 Panama 1.90 Tunisia 1.39 Paraguay 1.81 Ukraine 1.36 Peru 1.90 Portugal 1.84 Singapore 1.78 Spain 1.82 Sweden 1.86 Switzerland 1.88 16 Trinidad and 1.67 Tobago United Kingdom 1.86 United States 1.85 Uruguay 1.77 Venezuela 1.84 Zambia 1.79 (*) Open: None or minimal regulation for outward and inward transactions together with a generally nondiscriminatory environment. Largely Open: Some regulations are exercised on outward and inward transactions with the need of documentary support but without the need of governmental approval. Partially Closed: Regulation and govemmental approval is required for outward and inward transactions and it is usually granted. Largely Closed: Substantial restrictions and governmental approval is required and seldom granted for outward and inward transactions. B. Democracy Measure Democracy is, of course, a multi-dimensional and poly-significant concept, and difficult to measure empirically in a uniform way. Given the thrust of our analysis, however, on the link between democracy and financial openness, it is possible to rely on a comparative notion of democracy. This is based on the degree to which different countries meet the most commonly accepted criteria of democratic order - i.e. the majority principle, universal suffrage, civil rights and political liberty.29 Thus, our measure of democracy follows the recent literature exploring the role of democracy on economic growth (Helliwell, 1992), income levels (Londregan and Poole, 1996), and wages (Rodrick, 1999). This measure defines democracy as a composite index and draws on the Freedom House measures of political and civil liberty; that is: Democracy _ 14 - civil rights - political rights Democracy - ~12 This index will be defined from 0 to 1, with 0 indicating low democracy and 1 indicating high. Political and civil liberty indices are from the Comparative Survey of Freedom that Freedom House has provided on an annual basis since 1973. The Survey rates each country on a seven-point scale for both political rights and civil liberties (1 representing the most free and 7 the least free). The Survey assesses a country's freedom by examining its record in these two areas: A country grants its citizens political rights when it permits them to form political parties that represent a significant range of voter choice and whose leaders can openly compete for and be elected to positions of power in government. A country upholds its citizens' civil liberties when 29 See Archibugil (1998) for a discussion of various premises for democracy. 17 it respects and protects their religious, ethnic, economic, linguistic, and other rights, including gender and family rights, personal freedoms, and freedoms of the press, belief, and association. The Survey divides the world into three broad categories: "Free" (countries whose ratings average 1-3); "Partly Free" (countries whose ratings average 3-5.5); and "Not Free" (countries whose ratings average 5.5-7). {See Karatnycky, Adrian (1999)} Figure 3 Country Classification: Political Rights and Financial Openness Financial Openness Low High Low ~ 23 9 Low (8.86) (9.09) Democracy High 32 37 High (18.24) (25.59) Drawing on the above measures of democracy and financial openness, Figure 3 provides a 2 by 2 classification of sample countries along the democracy and financial openness axis. The figure shows the number of countries in each category, as well as the mean value for the countries in each quadrant of average government expenditures on health, education and transfer payments as a percentage of GDP for the period 1991-1997 (in parenthesis). A high level democracy means democracy index > 0.6. A "High" level of financial openness is defined by a score of >1.6 in the Financial Openness Index.30 From the matrix, it could be argued that the difference in means between the two groups (High-Democracy & High Financial Openness group versus Low Democracy &-Low Financial Openness) was highly significant at 0.1% level. Table 2 reports the statistical summary of the key variables classified by countries' degree of openness. Of interest here is the wide difference in countries' degree of redistribution as measured by the ratio of government social expenditure to GDP. Thus, countries classified as financially open spend on the average 22.3% of their GDP on social expenditures as compared to 6.7% among financially closed countries. Furthermore, the view that financial openness, democracy, and government social spending go hand in hand is confirmed by the econometric results shown in Figure 3. The cross-country data, with sample size ranging from 70 to 140, shows statistically significant results for all three relationships based on a z-test, with the highest correlation between political liberty and government social expenditures. 30 An.alternate measure of openness of the economic nature is the Fraser Intitute's Economic Freedom of the World rankings. These rankings are strongly correlated with the Financial Openness Index (0.70) and show a correlation of 0.49 with the Democracy Index Thethe 95% confidence interval forthe difference inmeans is 11.15 21.74 18 Table 2: Taxonomy of Financial Openness* Country Grouping by Financial Openness (Open) (Largely Open) (Largely Closed) (Closed) 1 Democracy index' 0.81 0.71 0.63 0.48 2 Civil liberties2 2.28 3.30 3.38 4.55 3 GDP per capita3 13,147 3,051 2,317 1,557 4 Social expenditure (%GDP)4 22.3 23.5 12.5 6.7 5 Total government exp/GDP (%)5 26.0 19.9 23.4 27.7 6 General government consumption 16.1 17.9 15.5 14.7 (% of GDP)6 Number of countries 46 10 34 11 * The table displays the group averages computed for the countries where the data were available. Definition of variables: (1) Democracy index ranges from 0 (lowest) to 1 (highest) and is computed on the bases of pdkl ria?s and ci%d Iirties indices (2) Ciad likrties, see footnote 13. (3) GDP/cap, Gross Domestic Product per capita, average of 1990-97. (4) Social e t7r9 includes the sum of health, education, and social security and welfare; average 1991-97. (5) Total g xwnD.mt expeditkw, average of central government and budgetary accounts plus state or provindal govermment, 1990-97. (5) Geea gow eit 7 xnwnpti^ includes all current expenditures for purchases of goods and services by all levels of government, excluding most government enterprises. It also includes capital expenditure on national defense and security, 1990-197. Figure 4: Relationships among Democracy, Capital mobility, and Government social expenditures Democracy Financial openness r- 0.45 =0.47 7r 0.53 r= 0.70 r= 0.32 Gov't social r= 0.58 Transfers expenditures (% GDP) (% GDP) 19 Note: The cross-country data, with sample size' randing from 70 to 140, shows statistically significant results at 1 % (except correlation between transfers and finnacial openness, which is siginificant at 5%) for all relationships based on a z-test, with the highest correlation between democracy and government social expenditures. C. The Methodology In examining the significance of redistribution on the degree of financial openness and democracy in countries, we deploy the logit model (Amemiya, 1981) to estimate the likelihood that a given country would fall in each of the four classifications defined in Figure 3. For expositional measures we focus first on the determinants of the high-democracy high-financial openness classification by using the binominal logit model. In this model, the dependent variable is defined by a dichotomous random variable y; which takes the value of 1 if country i belongs to the high democracy - high financial openness category, and 0 if it does not. This is given by yi =p, +e, where p is the probability that a given country belongs in the high democracy - high financial openness category, and specified asp = F(ax); where x is a vector of independent variables; and a is the corresponding vector of coefficients; and F(ax) is the cumulative distribution function; and es is an error term assumed to follow the Bernoulli distribution. Expressing the probability for the i-th observation via the logit model, we will obtain: exp(a'x,) ' +exp(a'x) Subsequently, the logit transformation would yield: log Pi = a'xi '-pi This model was estimated by the maximum likelihood method, using cross-country data for a sample of 67 countries for which consistent data on all explanatory variables was available. The estimated results, based on a (binomial) logit model3' of country classifications along the two axes of democracy and financial openness, are reported in Table 3. The results indicate that 31 We obtained the maximizing likelihood using Neuton-Rapson algorithm We used the STATA lo* procedure after 5 iterations. 20 both per capita income and the ratio of social expenditures to GDP have a statistically significant impact in explaining the likelihood that a country falls into the high-high category.32 The model also performs well in predicting the percentage of countries that are correctly classified as belonging to the high-democracy high-financial openness group, i.e. out of 27 countries in the High-High group, 19 were correctly predicted to be in that group (based on the threshold probability of 0.5) , thus producing 70.37 percent correct classification rate. Table 3 Estimation Result of the Binomial Logit Model: the likelihood of countries belonging to the high democracy level andfinancial openness category Independent variable Coefficient Z statistic Marginal effecta Constant -11.234** -4.085 -2.022 Log of social expenditure as % of GDP 1.534* 2.496 0.276 Log of GDP per capita 0.795* 2.496 0.143 Actual number of countries in the target 28 group Predicted number of countries in the target 20 groupb Actual number of countries in other groups 39 Predicted number of countries in other 32 groups Log likelihood -27.744 Note: The dependent variable is coded 1 if the country falls into the financial openness - high democracy, and 0 otherwise. significant at 1% level for a two-tailed z-test significant at 5% level for a two-tailed z-test a marginal effect is marginal change in the probability resulting from an infinitesimal change in the explanatory variable b target group means countries with high level of political rights and high financial openness It is important to note that our estimated results are based on the logit model and the 2x2 country classification along the democracy and financial openness indexes. As with such categorical dependent variable models, the results may be sensitive to the particular standard of classification adopted. 32 Using these estimates, we can also estimate the probabilities of each country belonging to the H:&Hig group, as opposed to all other groups. Given thus probabdity, one can classify a country to the group that has the predicted probability greater or equal than 0.5. This classification can be compared with the actual grouping of the data, and the percentages of correcty predicted countries can be computed for both target and non-target groups. 21 We also used the multi-nominal logit model to estimate the likelihood that a given country would fit into each of the 2x2 classifications (Figure 3). This is an extension of the binominal logistic model, and can be specified as: Pr(y~~ =1~x,)=p~'~ - exp(z (j) 1 + exp(z (2) ) + exp(z(3) ) + exp(zj4f) ) zjJ =aJx;, j =2,3,4 i=l..n where p,j) = the probability that the country i falls in the j h classification; y's are the set of binary response variables, y> =1 if the country i belongs to the group j and yii yy =0, otherwise xi is the vector of explanatory variables for the ith country; and n is total number of countries Note that one of the k groups (here it is the first) has to be fixed as a reference group (or base category). This means that the coefficients of the corresponding equation must be all set to zero, so that we actually estimate the set of k-I equations, which in the case or regular logit model with 2 groups reduces to a single equation with coefficients {ao, a,, a2, ...am}. It can be easily shown then, that exp(zjJ)) expresses the probability that the country i belongs to a group (j) relative to the probability that it belongs to the reference category (here the reference category is the Low-Low group), namely exp(zfj)) = Pr(y, = j) / Pr(yj = 1) .33 The maximum likelihood associated with the above multinominal logit model is given as: InL = (yi ua(J)x; -In es ,under the constraint that a') x = 0 i=l jsl j=l We estimate the equation with the explanatory variables defined as: (i) per capita GDP, and (ii) a log of the rate of social expenditures to GDP, using the same cross-country data as 33 APr1) is the marginal change in the probability Pr(y -1) -= p (j) resulting from an infinitesimal change in the I.th explanatoiy variable. APr(l, j) = a(j) exp(zjj) j = 2,3,4 (I + exp(zi2() + exp(z(3')) + exp(zi(4 ))2 The values of z's are computed at the mean values of the explanatory variables. 22 before. The results are also satisfactory with regard to the low democracy - low financial openness group, but not conclusive with regard to the other two intermediate classifications. Table 4: Multinomial Logit Results Financial Openness (liigh), Financial Openness (Low), Financial Openness (High), Democracy Democracv hig Democracy (Lowl Coefficie Marginal z-statistic Coefficients Marginal z-statistic Coefficients Marginal z-statistic nts Effect Effect Effect Log of Socila 2.348** 0.148 2.664 1.739* 0.169 2.195 -0.523 -0.011 -0.538 Expenditure per GDP Log of 1.2091' 0.076 2.725 0.318 0.031 0.826 1.085* 0.022 2.241 GDP/capita constant -15.209' -0.960 -4.186 -6.430* -0.627 -2.406 -7.860* -0.158 -2.358 significant at 5% level for a two-tailed z test significant at 1% level for a two-tailed z test The reference category is the Financil Opevmss(Low)-denooaa(Loow) group Table 5: Prediction Performance of the Multinomial Logit Model Predicted dependent Variable Total Percent Correct Actual dependent variable 0 1 2 3 0 Financial Openness (Low), 13 2 0 2 17 0.765 Democracy (Low) 1 Financial Openness (Low), 5 3 1 6 15 0.200 Democracy (High) 2 Financial Openness (High), 4 0 1 2 8 0.143 Democracy (Low) 3 Financial Openness (High), 1 2 0 25 28 0.893 Democracy (High) II Total 23 7 2 35 67 0.627 It can be seen that the multinomial logit provided a satisfactory prediction of the two extreme groups (lowlow and bhhg) while underpredicting the other two groups. 23 IV. Conclusion The relationship between financial openness and democracy appears at first sight to be primarily a function of the level of income. Rich countries are open to international capital movement, as they have a high degree of financial sector development, currency convertibility on capital accounts, and enjoy macroeconomic stability, domestic rule of law, and stable institutions that guarantee civil and political liberty. At a deeper level of analysis, however, the link between democracy and financial openness proves to be more complex. This paper discusses the factors linking financial openness and democracy both through a narrative of historical experience as well as cross-country empirical investigation. The narrative section focuses on international policy coordination, redistributive policies, and the embedded liberalism of the Bretton Woods era, which contributed to the survival of democracy in the face of the challenges posed by fascism and commnunism. The Bretton Woods consensus on postwar institutional reconstruction combined multilateralism in trade and related exchange rate arrangements on the external side, with interventionism in pursuit of legitimate national goals (full employment, national security, and social stability) on the domestic side. As history attests, this policy mix proved highly successful. The underlying conditions, trends, and forces at play in global financial markets have changed considerably since 1944. Financial markets have grown significantly in the size, complexity, and range of the services and products they offer. Private capital now dominates development finance, surpassing official flows by 5 to 1 in recent years. And the spread of democracy and civil liberty means that for the first time in human history, electoral democracy is the world's predominant form of government. The consolidation of the move towards democratic governance, in turn, means increased demand by citizens for political voice, for national economic security, and for social insurance against heightened exposure to international financial volatility. Then, if, as in OECD countries, the potential gains from openness to cross- border capital flows are to be realized in developing countries, there will need to be an understanding and appreciation of the interrelationships between domestic politics and a country's openness to international finance. On the international front, since capital account liberalization brings welfare gains globally, there is a justification for international action - policy coordination, prudent financial regulation and supervision, and lender-of-last-resort activity to provide liquidity and emergency financial assistance. For developing countries undergoing democratization in a world where the Bretton Woods system of fixed exchange rates has given way to the challenges posed by an evolving global economy subject to the tenets of neo-liberalism, their integration into global international capital markets has proved to be a process fraught with risk and painful cycles of currency, financial, and sovereign debt crisis. Yet, instead of rolling back liberalization in the face of uncertainties, historic experience points to the benefits of taking complementary action. This paper has argued for the importance of two sets of actions: (i) greater coordination of macroeconomic and regulatory policies at the international level; and (ii) cost-effective redistributive schemes as insurance against the risks of financial instability at the domestic level. The move toward democracy worldwide should facilitate greater policy coordination, as democratic systems foster a greater degree of regime stability, policy credibility, and security of 24 property rights than authoritarian ones. At the same time, the consolidation of the move towards democratic governance, in turn, means increased demand by citizens for political voice, for national economic security, and for social insurance against heightened exposure to international financial volatility. Among OECD countries, capital mobility as a policy objective gained currency and support only after significant trade liberalization had taken place, and only in the context of democratic countries in which the state's ability to respond to citizen's demands for national economic security had been established. This system must be recreated on a global level, and the experience of OECD countries must be taken into account when considering the status of developing countries moving toward democratization and full capital account convertibility in today's international financial environment. This paper presented information suggesting that there is a positive and statistically significant correlation between democracy, open capital flows and redistributive domestic social policies. This correlation, notwithstanding some data limitation, is significant in unveiling the positive forces at work in generating the convergent trends towards greater democracy and financial openness, as well as the critical role of redistribution-confirmed also by the more rigorous multinomial logit model. Further research clearly needs to move from the analysis of correlations to reflect on the causalities involved in the nexus of market openness, democracy, and redistributive policies. The thrust of the analytical framework sketched in this paper is that countries choose their degree of openness to international finance in order to maximize social welfare while safeguarding against the risks of capital flow volatility. Democratic governance plays an important role in framing the specifics of the social welfare function in a manner that is consistent with citizens' desire for redistribution and their degree of risk aversion.34 34This analytical framework is described more fully in an upconiing paper. 25 ANNEX Date that selected IMF member countries have assumed Article VIII El Salvador 11/06/46 Indonesia 05/07/88 Mexico 11/12/46 Portugal 09/12/88 Panama 11/26/46 Republic of Korea 11/01/88 United States 12/10/46 Turkey 03/22/90 Honduras 07/01/50 Thailand 05/04/90 Canada 03/25/52 Switzerland 05/29/92 Dominican Republic 08/01/53 Greece 07/22/92 Tunisia 01/06/93 Belgium 02/15/61 Morocco 01/21/93 France 02/15/61 Israel 09/21/93 Germany 02/15/61 Mauritius 09/29/93 Ireland 02/15/61 Barbados 11/03/93 Luxembourg 02/15/61 Trinidad and Tobago 12/13/93 Netherlands 02/15/61 Ghana 02/02/94 Sweden 02/15/61 Sri Lanka 03/15/94 Italy 02/15/61 Bangladesh 04/11/94 United Kingdom 02/15/61 Lithuania 05/03/94 Austria 08/01/62 Latvia 06/10/94 Jamaica 02/22/63 Pakistan 07/01/94 Kuwait 04/05/63 Estonia 08/15/94 Japan 04/01/64 India 08/20/94 Nicaragua 07/30/64 Paraguay 08/23/94 Costa Rica 02/01/65 Malta 11/30/94 Australia 07/01/65 Croatia 05/29/95 Guyana 12/27/66 Poland 06/01/95 Denmark 05/01/67 Moldova 06/30/95 Norway 05/11/67 Slovenia 09/01/95 Bolivia 06/05167 Philippines 09/08/95 Argentina 05/14/68 Czech Republic 10/01/95 Singapore 11/09/68 Slovak Republic 10/01/95 Malaysia 11/11/68 Botswana 11/17/95 Malawi 12/07/95 Ecuador 08/31/70 Hungary 01/01/96 Bahrain 03/20/73 Mongolia 02/01/96 South Africa 09/15/73 Benin 06/01/96 The Bahamas 12/05/73 Burkina Faso 06/01/96 Papua New Guinea 12/04/75 Cameroon 06/01/96 Venezuela 07/01/76 Mali 06/01/96 Chile 07/27/77 Russian Federation 06/01/96 Namibia 09/20/96 Uruguay 05/02/80 Romania 03/25/98 New Zealand 08/05/82 FYR Macedonia 06/19/98 Belize 06/14/83 Bulgaria 09/24/00 Iceland 09/19/83 Rwanda 12/10/98 Spain 07/15/86 Source: Exchange Arrangements and Exchange Restrictions, Annual Report 1999 26 REFERENCES Alesina, Alberto, Vittorino Grilli, and Gian Maria Milesi-Ferretti, "The Political Economy of Capital Controls, " in Capital Mobility: The Impact on Consumption, Investment, and Growth, ed. Leonardo Leiderman and Assaf Razin, (Cambridge: Cambridge University Press, 1994), pp. 289-328. Amemiya, Takeshi, Qualitative Response Models: A Survey Beitz, Charles R., "International Liberalism and Distributive Justice: A Survey of Recent Thought," in World Politics, Vol. 51, No.2, January 1999, pp. 269-296. Beitz, Charles R., "Social and Cosmopolitan Liberalism," in International Affairs, Vol. 75, No.3, July 1999, pp. 515-530. Bishop, John A., John P. Formby, and W. James Smith, "Incomplete Information, Income Redistribution and Risk Averse Median Voter Behavior,' in Public Choice, No. 68, 1991, pp.41- 55. Bohman, James, "International Regimes and Democratic Governance: Political Equality and Influence in Institutions," in International Affairs, Vol. 75, No.3, July 1999, pp. 499-514. Brunetti, Aymo and Beatrice Weder, "Political Credibility and Economic Growth in Less Developed Countries," in Constitutional Political Economy, Vol. 5, No. 1, 1994, pp. 23-44. Bryant, Ralph C. and Edith Hodgkinson, "Problems of International Cooperation," in Can Nations Agree?: Issues in International Economic Cooperation, (Washington, D.C.: Brookings Institution, 1989), pp. 1-11. Calleo, David P., "Since 1961: American Power in a New World Economy," in Economics and World Power: An Assessment of American Diplomacy Since 1789, ed. William H. Becker and Samuel F. Wells, Jr., (New York: Columbia University Press, 1984). Calvo, Guillermo A. and Enrique G. Mendoza, "Petty Crime and Cruel Punishment: Lessons from the Mexican Debacle," in American Economic Review, Papers and Proceedings, Vol. 86, No. 2, Fall 1998, pp. 1-3, 6. Cooper, Richard, "Should Capital-Account Convertibility be a World Objective?, " in Essays in International Finance Series, 207, (Princeton: International Finance Section, Department of Economics, May 1998), pp. 11-19. Dailami, Mansoor, "Euphoria and Panic: Developing Countries' Relationship to Private Finance," in EDI Forum, (Washington, D.C.: Economic Development Institute of the World Bank), Vol. 3, No. 2, Fall 1998, pp. 1-3, 6. 27 Dailami, Mansoor, "Managing Risks of Capital Mobility," Policy Research Working Paper 2199, World Bank, October 1999. Dailami, Mansoor and Michael Klein, "Government Support to Private Infrastructure Projects in Emerging Markets," in Dealing with Public Risk in Private Infrastructure, ed. Timothy Irwin, Michael Klein, Guillermo Perry, Mateen Thobani, World Bank, 1998. Dailami, Mansoor, and Nadeem ul Haque, "What Macroeconomic Policies Are 'Sound?'," World Bank and IMF Policy Research Working Paper 1995, World Bank, October 1998). Devarajan, Shantanyanan, and Jeffrey S. Hammer, "Risk Reduction and Public Spending," Policy Research Working Paper 1869, (Washington, D.C.: World Bank, Public Economics, Development Research Group, January 1998). Eichengreen, Barry, Globalizing Capital: A History of the International Monetary System, (Princeton, Princeton University Press, 1996). Fischer, Stanley, "Capital Account Liberalization and the Role of the IMF," in Essays in International Finance Series 207, (Princeton: International Finance Section, Department of Economics, May 1998), pp. 1-10. Fraser Institute, Economic Freedom of the World 2000, (Vancouver, The Fraser Institute, 2000), available at http://www.fraserinstitute.ca Garrett, Geoffrey, "Global Markets and National Politics: Collision Course or Virtuous Circle?," in International Organization, Vol. 52, No. 4, Autumn 1998, pp. 787-824. Gold, Joseph, International Capital Movements Under the Law of the International Monetary Fund, (Washington, D.C.: IMF, 1977). Goodman, John B. and Louis W. Pauly, "The Obsolescence of Capital Controls?," in World Politics, Vol. 45, October 1993, pp. 50-82. Haggard, Stephan and Sylvia Maxfield, "The Political Economy of Financial Internationalization in the Developing World," in Internationalization and Domestic Politics, ed. Robert 0. Keohane and Helen V. Milner, (Cambridge: Carnbridge University Press, 1996), pp. 209-239. Haldane, A., "Private Sector Involvement in Financial Crisis," Financial Stability Review (November 1999), Bank of England. Held, David, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance, (Stanford, CA, Stanford University Press, 1995). Helleiner, Eric, States and the Reemergence of Global Finance: From Bretton Woods to the 1990s, (Ithaca and London: Cornell University Press, 1994). 28 Helliwell, John F., "Empirical Linkages Between Democracy and Economic Growth," in NBER Working Paper Series, No.4066, May 1992. Hirschman, Albert O., Essays in Trespassing: Economics to Politics and Beyond, (Cambridge: Cambridge University Press, 1981). , Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations and States. Harvard University Press. Kapstein, Ethan B., "Resolving the Regulator's Dilemma: International Coordination of Banking Regulations," in International Organization, Vol. 43, No. 2, Spring 1989, pp.323-347. Kapstein, Ethan B., "Distributive Justice as an International Public Good: A Historical Perspective," in Global Public Goods: International Cooperation in the 21s" Century, ed. Inge Kaul, Isabelle Grunberg, Marc A. Stem, (New York and Oxford, Oxford University Press, 1999), pp.88-115. Karatnycky, Adrian, "The 1998 Freedom House Survey: The Decline of Illiberal Democracy," in Journal of Democracy, Vol. 10, No. 1, January 1999, pp. 112-125. Leidermann, Leonardo and Assaf Razin ed., Capital Mobility: The Impact on Consumption, Investment and Growth, (Cambridge: Cambridge University Press, 1994). Londregan, John and Keith T. Poole, "Does High Income Promote Democracy?" in World Politics, Vol. 49, No. 1, October 1996, pp. 1-30. Mathieson, Donald J. and Liliana Rojas-Suarez, "Capital Controls and Capital Account Liberalisation in Industrial Countries," in Capital Mobility: The Impact on Consumption, Investment and Growth, ed. Leonardo Leidermann and Assaf Razin, (Cambridge: Cambridge University Press, 1994), pp. 329-350. Mueller, Dennis C., "Constitutional Constraints on Governments in a Global Economy," in Constitutional Political Economy, 1998: 9, pp. 171-186. Olson, M, "Why Some Welfare-State Redistribution is a Great Idea" in Democracy and Public Choice, ed. C.K. Rowley, (Oxford and New York, 1987). Pauly, Louis W., Who Elected the Bankers?, Ithaca and London: Cornell University Press, 1997). Pasour, E.C., "Redistribution and Constitutional Political Economy," in Constitutional Political Economy, Vol. 5, No. 1, Winter 1994, pp. 81-98. Przeworski, Adam and Fernando Limongi, "Political Regimes and Economic Growth," in Journal of Economic Perspectives, Vol. 7, No. 3, pp. 51-69. 29 ,"Capital Mobility, State Autonomy, and Political Legitimacy," in Journal of International Affairs, Vol. 48, No. 2, Winter 1995, pp. 369-3 88. Quinn, Dennis, "The Correlates of Change in International Financial Regulation," in American Political Science Review, Vol. 91, No.3, September 1997, pp. 531-551. Radelet, Steven and Jeffrey Sachs, "The East Asian Financial Crisis: Diagnosis, Remedies, Prospects," in Brookings Papers on Economic Activity, No. 1, ed. William C. Brainard and George L. Perry, (Washington, D.C.: Brookings Institution, 1998), pp. 1-90. Rodrik, Dani, Has Globalization Gone Too Far?, Institute for International Economics. Rodrik, Dani, "Democracies Pay Higher Wages," in The Quarterly Journal of Economics, Vol. CXIV, Issue 3, August 1999. Ruggie, John G., "International Regimes, Transactions, and Change: Embedded Liberalism in the Postwar Economic Order," in International Regimes, ed. Stephen D. Krasner, (Ithaca: Cornell University Press, 1983), pp. 195-231. Sachs, Jeffrey, Aaron Tomell and Andres Velasco, "Financial Crises in Emerging Markets: The Lessons from 1995," in Discussion Paper Number 1759, (Cambridge, Massachusetts: Harvard Institute of Economic Research, Harvard University, 1996). Sally, Razeen, "Classical Liberalism and International Economic Order: An Advance Sketch," in Constitutional Political Economy, Vol. 9, No. 1, March 1998, pp. 19-44. Sen, Amatya, "Global Justice: Beyond International Equity," in Global Public Goods. International Cooperation in the 21' Century, ed. Inge Kaul, Isabelle Grunberg, Marc A. Stem, (New York and Oxford, Oxford University Press, 1999), pp. 116-125. Thomas, Vinod, ed. The Quality of Growth, World Bank (forthcoming). Verdier, D., "Domestic Responses to Capital Market Internationalization Under the Gold Standard, 1870-1914," in International Organization, Vol. 52, No. 1, Winter 1998. Wagner, R.E., "Wealth Transfers in a Rent Seeking Polity," in Cato Journal 6, 1986, pp. 155- 171. Webb, Michael C., The Political Economy of Policy Coordination: International Adjustment Since 1945, (Ithaca and London: Cornell University Press, 1994). Wessels, J. H., "Redistribution from a Constitutional Perspective," Constitutional Political Economy, Vol. 4, No. 3, pp. 425-448. 30 Williamson, John and Molly Mahar, "A Survey of Financial Liberalisation," in Essays in International Financial Series 211, (Princeton: International Finance Section, Department of Economics, Novernber 1998). World Bank, Global Development Finance, (Washington, D.C.: The World Bank, 1998). 31 Policy Research Working Paper Series Contact Title Author Date for paper WPS2352 How Stronger Patent Protection in Carsten Fink May 2000 L.Tabada India Might Affect the Behavior 36896 of Transnational Pharmaceutical Industries WPS2353 The Sao Mateus-Jabaquara Jorge Rebelo May 2000 S. Van Veldhuizen Trolleybusway Concession in Brazil Pedro Machado 38722 WPS2354 When the Bureaucrats Move out of Yi Chen May 2000 M. Yafi Business: A Cost-Benefit Ishac Diwan 34649 Assessment of Labor Retrenchment in China WPS2355 Greed and Grievance in Civil War Paul Collier May 2000 A. Kitson-Walters Anke Hoeffler 33712 WPS2356 Bureaucratic Delegation and Political Philip Keefer June 2000 P. Sintim-Aboagye Institutions: When Are Independent David Stasavage 37644 Central Banks Irrelevant? WPS2357 Evaluating Carbon Offsets from Kenneth M. Chomitz June 2000 J. Ancrum Forestry and Energy Projects: 33512 How Do They Compare? WPS2358 Why Infrastructure Financing Daniela Klingebiel June 2000 M. Salehi Facilities Often Fall Short of Their Jeff Ruster 37157 Objectives WPS2359 Reducing Carbon Dioxide Emissions Will Martin June 2000 L. Tabada Through Joint Implementation of 36896 Projects WPS2360 Corruption and the Composition of Beata K. Smarzynska June 2000 H. Sladovich Foreign Direct Investment: Shang-Jin Wel 37698 Firm-Level Evidence WPS2361 The Welfare Effects of Private Sector George Clarke June 2000 H. Sladovich Participation in Guinea's Urban Water Claude Menard 37698 Supply Ana Maria Zuluaga WPS2362 A Transitory Regime: Water Supply Claude Menard June 2000 H. Sladovich In Conakry, Guinea George Clarke 37698 WPS2363 Would Collective Action Clauses Barry Eichengreen June 2000 S. Kpundeh Raise Borrowing Costs? An Update Ashoka Mody 39591 and Additional Results WPS2364 Perverse Effects of a Ratings- Patrick Honohan June 2000 A. Yaptenco Related Capital Adequacy System 38526 Policy Research Working Paper Series Contact Title Author Date for paper WPS2365 Leading Indicator Project: Lithuania Stephen S. Everhart June 2000 M. Geller Robert Duval-Hernandez 85155 WPS2366 Fiscal Constraints, Collection Costs, Keiko Kubota June 2000 L. Tabada and Trade Policies 36896 WPS2367 Gender, Poverty, and Nonfarm Constance Newman June 2000 M. Clarke Employment in Ghana and Uganda Sudharshan Canagarajan 31752 WPS2368 Seeds of Corruption: Do Market Harry G. Broadman June 2000 S. Craig Institutions Matter? Francesca Racanatini 33160 WPS2369 How the Proposed Basel Guidelines Giovanni Ferri June 2000 E. Mekhova On Rating-Agency Assessments Li-Gang Liu 85984 Would Affect Developing Countries Giovanni Majnoni WPS2370 A New Model for Market-Based Marcelo Giugale June 2000 M. Geller Regulation of Subnational Borrowing: Adam Korobow 85155 The Mexican Approach Steven Webb WPS2371 Shock Persistence and the Choice Marcelo Giugale June 2000 M. Geller of Foreign Exchange Regime: An Adam Korobow 85155 Empirical Note from Mexico