Saylor.org's Comparative Politics/The Rise of the Democratic State and the “Third Wave”

NOTE: The three waves of democracy include 1) the early 19th century when white men were given the vote in the United States, 2) after WWII, and 3) since 1974, and includes Latin America, the Asian Pacific, and eastern Europe.

=The Economics of Young Democracies:Policies and Performance= By Nathan Converse and Ethan B. Kapstein, Center for Global Development Working Paper, March 2006

Introduction
The second half of the twentieth century may eventually become known as the “age of democracy.” Since the “third wave” of democratization began in 1974, nearly 100 states have adopted democratic forms of government, including, of course, most of the former Soviet bloc nations (Huntington 1991). America policy-makers have expressed the hope that this democratic wave will extend even further, to the Middle East and onward, perhaps reaching China’s shores in the not-too-distant future (“Rice Plans to Encourage Democratization,” 11 October 2005, www.state.gov).

Why do policy-makers care about advancing the cause of democracy? Beyond its contributions to civil liberties and international peace, democratic governance is increasingly seen as being necessary to the achievement of sustained economic development and a better standard of living (Bardhan 1999; Diamond 2005; Halperin, Siegle, and Weinstein 2005). According to Larry Diamond, “a high quality democracy…will be used to generate public goods that stimulate investment and commerce and raise the quality of life” (Diamond 2005). And in a recently published study, Halperin, Siegle and Weinstein claim that, “democracies generally experience more rapid and consistent improvements in the well being of their populations than do autocracies” (Halperin, Siegle and Weinstein 2005, 65).

On the basis of such claims, these and other scholars have advanced the policy recommendation that democratization must play a central role, and perhaps the central role—in both domestic policies and foreign assistance programs that seek to promote sustained economic growth among other objectives. In fact, it appears as if policy-makers have heeded that advice, with Secretary of State Condoleezza Rice repeatedly emphasizing the need to promote democracy and only mentioning poverty relief in passing in her speech announcing the creation of a new Director of Foreign Assistance position in her Department (“Remarks At the U.S. Agency for International Development On Foreign Assistance,” 19 January 2006, www.state.gov). Likewise, the new foreign aid program of the United States, the Millennium Challenge Account (MCA), provides funds only to those countries that “rule justly,” and it is plausible that the World Bank under the leadership of Paul Wolfowitz will also move in that direction, tying development assistance to democracy-building more directly. Understanding therelationship between democracy and development therefore goes well beyond an “academic exercise.” Despite the west’s political commitment to support democratization, the durability of the new democratic age remains an open question. By some accounts, at least half of the world’s youngest democracies—often referred to in the academic literature as being “unconsolidated” or “fragile”—are still struggling to develop their political institutions, and several have reverted back to authoritarian rule. Among the countries in the early stages of democratic institution building are states vital to U.S. national security interests, including Afghanistan and Iraq. Ensuring that the developing world’s democracies survive and prosper is therefore a crucial policy challenge. The ability of fledgling democracies to maintain popular support depends, at least to some extent, on the ability of their governments to deliver economic policies that meet with widespread approval. But what sorts of economic policies are these, and are they necessarily the same as the policies required for tackling difficult issues of economic stabilization and reform? Conversely, what sorts of economic policies are most likely to spark a backlash against young and fragile democratic regimes? Do the leaders of young democracies face trade-offs as they ponder their electoral and economic strategies?

These are among the questions we explore in this paper, which aims to provide an overview of the monograph we are currently writing on the economics of young democracies. As already noted, we believe these issues assume significant policy relevance in light of recent U.S. (among other bilateral and multilateral) foreign aid initiatives explicitly aimed at promoting democratic consolidation. Evidence that economic performance significantly affects the probability of democratic survival in a given country renders the question of how well the new democracies manage their economies all the more pressing from a policy-making standpoint. Adam Przeworski and colleagues, for example, have found the probability that a young democracy will be overthrown grows ten-fold when the economy contracts two years in a row (Przeworski, et al., 2000:109). The idea that poor economic performance poses a threat to democracy has been supported by a great deal of quantitative work. In an oft-cited study, Rueschemyer, Stephens, and Stephens (1992) demonstrate that economic factors played a decisive role in nearly all the democratic reversals in Latin America between 1920 and 1980. Evidence from opinion surveys also shows that public support for democracy has frayed when economic conditions deteriorate (Haggard and Kaufman 1995).

This paper examines much of what we know about the theory and empirics of economic policy and performance in the consolidation and reversal of newly democratic states. The paper is in four sections. First, by way of background, we examine the hypothesized relationships between democratic politics and economic policy-making. Second, we analyze the findings of several important empirical studies with respect to the economic performance of young democracies around the world. In the third section, we provide some descriptive statistics which indicate how young democracies have fared and are faring in practice, making use of the new dataset that we have compiled (which, among other things, is more up-to-date than the others cited herein). We are particularly interested in comparing the economic performance of those young democracies that successfully consolidate as opposed to those that ultimately fail. Do the data reveal any distinctive economic patterns with respect to democratic consolidation and reversal? We will show that they do. In particular, we find that deteriorating or stagnant economic performance constitutes a red flag or warning signal that the country is at risk of democratic reversal. Moreover, we observe considerable variation in economic performance, suggesting that the design of political institutions in new democracies may have a significant influence on the probability of their survival. The fourth section provides our conclusions and policy recommendations, while a discussion of our dataset and some summary tables are found in the appendices.

I. The Political Economy of Democracy
In recent years a growing number of scholars and policy-makers have expressed optimism that democracy and development are mutually reinforcing. In the spring of 2005, for instance, Larry Diamond told an audience in Ghana, “Africa cannot develop without democracy” and asserted that the literature points “clearly” to “a causal effect of democracy on economic growth….” As a consequence, he urged his listeners to shun any thoughts of authoritarian solutions to their economic problems. He suggested that the East Asian miracle, for example, “took place in a historic and regional context that is unlikely to be repeated,” and that it therefore failed to provide a relevant developmental model for contemporary political leaders (Diamond 2005). What are the channels through which democracy (our definition and list of country cases are found in the appendices) influences economic development? The literature posits several channels or pathways by which politics influences economics that we explore in this section:

First, democracy may influence economic policy through the channels provided by the electoral process;

Second, that influence may be felt through the channels provided by political parties that serve to aggregate voter interests;

Third, economic policy may be shaped through the channels provided by the institutional structure—e.g. presidential vs. parliamentary—that a particular democracy adopts;

Fourth, the age of a democratic regime may influence economic performance, since democratic institutions may take time to develop and mature.

Overall, these inter-locking attributes of a democracy—its electoral system (who gets to vote and which offices are contested), its political parties (how many, which interests are aggregated, and how well they are organized), its institutions (what type), and its age (how long these have been in place)—define its “constitutional political economy.” Indeed, we should emphasize that while many cross-country studies of democracy and economic performance define democracies at the most aggregate or abstract level, the variance among democracies along these four attributes may be of great significance for economic outcomes. In short, the type of democracy may matter no less than democracy itself in terms of delivering sustainedeconomic growth.

Voters, Elections and Economic Policy
Competitive elections play a key role in all mature democracies, and central to democratic electoral theory is the presence of a median voter, or “swing voter” whose support is required if a politician is to be elected to office (Downs 1957). By definition, the median voter, who occupies the fifth and sixth income deciles, owns less capital than the average for the nation as a whole, or to put this in the starkest terms we might say that the median voter is poorer than the average voter. This fact is consequential for economic policy and performance, for it suggests that the median voter, whose political support is decisive in contested elections, will demand income redistribution from the rich from her representatives as the price of her support.

Indeed, the demand for income redistribution is at the heart of contemporary political-economy accounts of political regime change, specifically the shift from authoritarian to democratic regimes (Acemoglu and Robinson 2006). In Acemoglu and Robinson (2006),democracy is modeled as a game-theoretic solution to interactions between rich and poor, in that it provides credible, limited redistribution to the poor through the franchise while preserving the property rights of the rich. One empirical problem with this argument, however, is that we do not observe such redistribution in practice, particularly in the world’s young democracies.

Returning to the world of theory, scholars have argued that since the median voter has a powerful incentive to demand redistribution, those with wealth in a democracy will have less incentive to save and to invest, and they may even decide to take their capital to friendlier locales, leaving the country poorer. One frequently posited outcome of this interaction between rich and poor is that democracies inevitably emphasize short-run consumption at the expense of long-run investment, with the result that a country’s growth potential goes unfulfilled (Huntington 1968; Rao 1984). In short, the economic policies of democracies are redistributive toward the median voter and growth is negatively affected as a consequence. One might expect these pressures to be particularly acute in developing countries where the gap between the median voter and those in the upper income deciles in likely to be enormous, and where the median voter may even be lacking in some of the basic necessities of life. This is a claim made by Haggard and Kaufman (1995) for example, and we will return to it below in our discussion of age of democracy.

The electoral influence on economic policy does not simply lead to income redistribution toward the median voter, but it has widespread macroeconomic effects as well. In the process of gaining the support of voters, politicians may be tempted loosen monetary policy and avoid fiscal discipline in order to provide infrastructure and jobs. From this perspective, a balanced budget is seen as a collective good, and no politician is motivated to contribute to it. Instead, she wants other politicians to practice fiscal restraint while she provides her own constituents with both “pork” and low taxes. Since every politician thinks in the same way, the result in a democracy is high spending and low taxation, and for this reason James Buchanan famously argued that democracies will tend toward the creation of budget deficits. Democratic governments inevitably become bloated, and it follows that “a democratic society…will tend to resort to an excessive use of deficit finance…” (Buchanan and Wagner 1978).

The general tendency of democracies to adopt lax fiscal and monetary policies has been more precisely associated with the electoral process through the so-called “political business cycle” (PBC). Economists have long observed that democratically elected politicians have an incentive to stimulate the economy artificially by loosening monetary policy and abandoning fiscal rectitude during the run-up to elections, and this behavior has been formally modeled by opportunistic PBC theories (as opposed to partisan PBC theory, discussed in the next section)Early theories of the opportunistic political business cycle focused on the opportunities arising from an exploitable Phillips curve and a myopic electorate (Alesina et al. 1997: 15), suggesting that competitive elections will generate volatility in employment, inflation, and growth rates. Later work (for example, Rogoff 1990) made the questionable assumption of a myopic electorate unnecessary by positing asymmetric information regarding the incumbent’s competence. These rational opportunistic models note that while both competent and incompetent incumbents can stimulate the economy by stepping-up government spending and expanding the money supply, only the more capable policy-maker can bring about additional economic improvements using superior managerial skills. Thus, the PBC allows competent incumbents to signal their abilities to voters. While the long-run effects of political business cycles on economic performance are unclear, they presumably undermine the level of investment and hence the growth rate by introducing greater volatility into the economy. It worth noting that PBC theories implicitly assume competitive elections, in which a degree of uncertainty exists regarding the outcome—a leader who will certainly be reelected has no reason to signal his managerial competence by stimulating the economy (Block 2002: 6). By implication, a shift from a less democratic government to a more democratic government with greater contestation of public offices implies that political business cycles may be magnified in newly democratizing countries.

Academic models that emphasize the influence of the median voter on economic policy have won a prized spot in the political economy pantheon. But empirical support for the median voter hypothesis is mixed at best, and it has been queried from a theoretical perspective as well, for various reasons. For one thing, since the interests of voters tend to be multi-peaked rather than single-peaked, their ability to hold politicians accountable on particular issues may be doubted. But even in the presence of single-peaked voting, voters may be unable to induce good behavior from their politicians, particularly if the official believes she is likely to lose the next election in any case, say due to exogenous shocks (e.g. an increase in oil prices) that undermine economic growth.

More prominently, Olson famously questioned the ability of individuals to advance their material interests effectively, given the free-rider temptation in the face of collective goods provision (Olson 1965, 1982). For example, while the median voter might wish for income redistribution, she might not be bothered to go out and vote, assuming that others will perform that task and produce the desired outcome. Since individuals face severe organizational impediments due to their preference for free-riding, their political impact might not be very great after all as compared to that of organized interests. Olson suggested instead that small groupswith a shared purpose would be most successful in advancing their political and economic agenda to the extent they had a strong material interest in a particular policy outcome. And among these small, politically active interest groups, private sector business lobbies have won particular attention from scholars.

The relationship between electorally motivated politicians and rent-seeking private sector lobbyists has been summarized in a simple equation by Grossman and Helpman (2002). They argue that:

V = C + aW

Where:

V = votes,

C = campaign contributions, and

a = a parameter term that suggests the trade-off between C and aggregate welfare (W).

Thus, politicians need to win votes but their campaigns are paid for by lobbyists who make contributions in return for desired policies (e.g. tariffs, contracts, or subsidies). By kowtowing too much to the lobbyists, however, the politicians generate welfare-reducing policies that cost them the votes of everyday citizens. They must therefore strike an appropriate balance between lobbyists and voters; still, the private sector interest will be both prominent in policy-making and generally welfare-reducing from an economic standpoint (Grossman and Helpman emphasize the formation of trade policy in their own work).

Olson asserted that these lobbies can play an extremely negative role in the economic performance of nations; indeed, he assigned them a central role in the economic decline of democratic states (1982). For Olson, it was organized interest groups that prevented politicians from implementing welfare-enhancing reforms. Interestingly, whereas Olson tended to view interest group power as being weakest in young democracies, Hellman has demonstrated how organized groups effectively captured the reform process in post-Communist Russia, blocking those reforms that would have led to more openness and competition (Hellman 1998). This observation leads to a more general, empirical set of questions regarding how well organized special interest groups actually are in the world’s young democracies, how much influence they have over policy-making, and what economic policies they seek. This is an area in which much work remains to be done.

In contrast to Olson, some theorists have tried to demonstrate that democracy ameliorates a society’s distributive conflicts and as a result this type of political system is efficiency-enhancing. Rodrik (1999; 2000), for example, elaborates a model in which two interest groups with divergent policy preferences repeatedly interact to formulate policy. If the groups face uncertainty about which will have a superior bargaining position in the future, they can reach an equilibrium in which they compromise, since the costs of making concessions is offset by the future benefits of concessions by one’s opponent when that group is in power. This outcome depends not only on repeated interactions, but also on uncertainty about the groups' future ability to impose their will on their opponent. By introducing such uncertainty through competitive elections, democracy promotes compromise, reducing policy volatility and facilitating better economic performance.

Competitive elections are also expected to be efficiency-enhancing for other reasons. For example, popular discourse, as well as economic theory, has generally viewed the public sector as less efficient than the private due to a dearth of competition in the former. However, democracy has the potential to introduce competition into the public sector, albeit indirectly. Lake and Baum (2001) lay out a theoretical model in which the government is a monopoly provider of public services, and thus restricts supply to drive up price in the form of rents extracted by government officials. They view democracy as a way of rendering the market for public services "contestable," with candidates for political office being potential entrants threatening to undercut the monopoly provider. In short, by introducing competitive pressures into the public sector, democracy forces the state to supply more and better public services.1

To summarize, one of the central characteristics of democracy is its reliance on competitive elections for choosing political leaders who are accountable to voters. That process has distinctive implications for economic policy, at least some of which have been held by scholars to be quite negative (Bardhan and Yang 2004). In the world’s young democracies, these negative attributes might even be accentuated as political leaders seek to shore up support and build credibility with voters (Keefer 2005). Yet it is not just through the median voter that the economic policies of democracies are shaped; these policies may also be a function of the way in which political parties and institutions aggregate contending interests. We turn to these perspectives in the following sections.

Political Parties and Economic Policy
There is a longstanding tradition in the political economy literature linking the political party in power with distinctive approaches to economic policy; this has been called the “partisanship school” of political economy (for reviews, see Drazen 2001 and Rueda n.d.). Briefly, the partisanship school holds that left-wing or social democratic governments will protect the interests of labor and the “working class” while right-wing or conservative governments will protect the interests of those who have accumulated capital. The distinctive economic policy approaches are allegedly made manifest in macroeconomic performance, with social democracies having higher levels of taxation, government spending, and inflation than conservative governments. A related literature modeling partisan political business cycles views the different preferences of left- and right-wing parties for unemployment and inflation as producing macroeconomic fluctuations (Alesina et al. 1997: 45).

As Rueda notes (n.d.), critical to such approaches is the assumption that political parties have “core constituencies” whose interests are paramount in setting the platform and in policy-making should the party form a government. A party’s main challenge, therefore, is to build a “winning coalition” around this core constituency. On the other hand, if the party follows solely the preferences of its core it may well lose election after election (that, of course, could be a perverse outcome of internal party politics). While the winning strategy of coalition-building may temper the more extreme strands of a given economic strategy the overall trajectory remains clear to voters, at least in contrast to that of the rival party. Thus, parties provide voters with clear choices. By “unpacking” political parties, this analysis of electoral politics allows for centrifugal forces within parties, which counteract the drive to appease the median voter, and endogenizes parties’ actual platforms.

It should further be emphasized that the partisanship approach is firmly materialistic, in that it associates core constituencies with economic classes. So far as we are aware, the partisanship school has not applied its model of party behavior to other constituencies, such as religious or ethnic groups. But such non-class-based allegiances are a distinctive feature of many new democracies, and these divisions may become the basis for competitive politics within the democratic setting. In Iraq, for example, political parties seem to be based largely on ethnic groupings, though of course within each ethnic faction one finds further political divisions that are also organized into parties. We note that the effects of these party groupings on economic policy and performance have been little studied; Haggard and Kaufman (1995), for example, rely on traditional, class-based partisanship models in their analysis of the political economy of democratic transitions. This is an area that will certainly attract more research from socialscientists in the years ahead and may lead to a very different view of the interests that political parties represent in domestic politics, and the policies they pursue once in government.

Still other theorists have emphasized that the ideological orientation of the political party in power matters less than the level of institutionalization that characterizes the parties that compete for office. In their study of the political economy of economic reform in 13 countries, Williamson and Haggard (1994) reject the idea that right-wing governments are better suited or even more likely to implement market-oriented reforms. By contrast, they find that the governments are more likely to carry out economic reforms when the party in power enjoys a broad support base. Similarly, Haggard and Kaufman stress the crucial importance of institutions that are capable of channeling conflicts over distribution and policy for the formulation of high-quality economic policy, and conclude that broad-based, cohesive political parties (or stable coalitions of parties) are particularly suited to perform this function. In contrast, they find that fragmented or polarized party systems generally fail to implement welfare-enhancing economic policies (Haggard and Kaufman 1995). We note that developing countries, including young democracies, vary enormously in terms of the degree of institutionalization of political parties and the relationship between party organizations and those who govern. One might say, in short, that what matters from an economic policy perspective is the capacity of parties to link politicians and large groups of voters in a way that enhances policy credibility.2 And that observation, in turn, leads directly to our discussion of democratic institutions and economic policy.

Democratic Institutions and Economic Policy
Underlying electoral processes and party structures are the basic democratic institutions established by a nation’s constitution. In discussing the implications of institutional structure for the formation of economic policies, analysts have emphasized the ways in which institutions can check the power of the state to prevent abuses, as well as examining how institutional structure influences policy outcomes. A separate literature has suggested that more abstract democratic institutions—personal freedom, an unfettered press, etc.—generate a dynamism conducive to efficiency and economic development.

Numerous theorists have focused on the way in which democracy’s institutional features positively influence economic performance by putting in place checks and balances. One check results from the fact that incumbents (or their parties) can be held accountable by the public in the next election. Whereas under autocracy, leaders may arbitrarily expropriate property for their own benefit, or government officials may demand bribes without fear of punishment, democracy generates incentives to refrain from such behavior, leading to stronger property rights, greater economic efficiency and less uncertainty (North 1990). Rodrik (1999) builds a model that produces similar results.

Moreover, the separation of powers present in most democracies introduces a further source of checks and balances. Whereas voters can only hold politicians accountable ex post,3 legislatures may block executive excesses ex ante. Particularly relevant to a discussion of economic policy is the existence of unelected institutions designed to buffer certain sets of policies from immediate domestic political pressures. Independent central banks, whose directors are often appointed for relatively long terms of office, insulate macroeconomic policy from potentially destabilizing popular demands. Supreme courts, whose members may serve for life, serve—in theory—as a further bulwark against abuses by legislatures and executives.

Analysis of institutions and their impact on economic policy have extensively explored the impact on policy of differences in the relationship between the fundamental bodies of executive and legislative authority. In particular, parliamentary democracies may be contrasted with presidential systems, in which the chief executive is elected by direct voting and given a more or less independent (from the legislature) power base and set of responsibilities. These differences could have important economic effects. Persson and Tabellini (2002) have pushed this analysis the farthest, developing a number of testable propositions that compare and contrast parliamentary and presidential regimes, viz.:

(1) “Parliamentary regimes will have larger governments than presidential regimes.”

(2) “Presidential regimes should be associated with less rent extraction and lower taxation than parliamentary regimes.”

(3) “Presidential regimes…should also be associated with more targeted programs at the expense of broad spending programs” (Persson and Tabellini 2002, 30).

The rationale for this distinction between presidential and parliamentary regimes in terms of economic policy and performance in part stems from the different checks and balances they introduce. Persson and Tabellini argue that political systems with “a larger concentration of power in parliamentary regimes makes it easier for politicians to collude with each other at the voters’ expense…” (Persson and Tabellini 2001, 4). However, others argue that parliamentary regimes are best suited to check the abusive power of the president, especially in developing countries where legislatures are often weak. As a consequence, developing countries have been urged to avoid strong, presidential models in favor of parliamentary democracy. Parliamentary systems with coalition governments have been held to be the most growth-friendly, as the various factions keep check on each other’s extreme spending tendencies; they therefore “internalize” the benefits of the separation of powers model. Overall, however, it is not clear which regime type—presidential or parliamentary—generates higher levels of long-run growth

Beyond electoral processes and formal institutional structures, other theorists point to the role of more diffuse democratic institutions in promoting market efficiency by inducing the free flow of capital, labor, or information. Early versions of this idea can be found in the work of Marx (Giddens 1971: 51), though it is most often associated with Hayek, Friedman (and Drucker in his 1939 End of Economic Man) and the Austrian School more generally. More recently and formally, North (1990) has hypothesized that markets function most efficiently in democratic societies due to the personal freedom they allow. Similarly, Sen (1994) highlights the way in which the free transmission of information in democracies facilitates efficient outcomes. This line of analysis, then, takes us beyond the formal constitutional political economy of democratic polities and reminds us of the importance of the many attributes associated with liberal democracy in particular, including civil liberties, a free press, and so forth, all of which may be in short supply in the young democracies of the developing world.

Work showing a relationship between democracy and higher productivity provides empirical support to this hypothesis. Przeworski el al. (2000) find that while dictatorships have higher rates of labor force growth, democracies exhibit more rapid total factor productivity growth. Evidence of an association between democracy and higher productivity can also be found in Minier (1998), Krieckhaus (2005), and Rivera-Batiz (2002). That democracies use existing factor endowments more efficiently would appear to support the idea that some of the more diffuse institutions of democracy serve to bolster economic performance.

Democracies Young and Old: Does Age Matter?
Much of the literature relating democratic polities to economic policies and outcomes was written in the context of the advanced industrial states. But young democracies in the developing world initial conditions and political circumstances that could render these models less relevant. Young democracies will by definition have less well-developed democratic structures: electoral processes, institutional mandates, and the composition of political parties may all be fluid and heavily contested in the period following democratization. Furthermore, young democracies may lack the traditions of a free press that informs voters with respect to the performance of their politicians, and the educational system may be weak, with entry—especially to higher education—restricted to the elite. Theoretical analysis as well as previous empirical work suggests that while democratization may provide some immediate benefits, some of its salutary impact on economic performance will be mitigated early on, as governments in such countries may have a particularly hard time formulating and executing consistent policies that provide the cornerstones for sustained growth.

Analysis of economic management in young democracies has focused to a large extent on the effects of the uncertainty present in such circumstances. Haggard and Kaufman stress that “uncertainties with respect to the stability of the new democratic order. . . affect the time horizons of both private and public actors,” leading politicians to pursue “policies aimed at managing opposition in the short run while discounting the gains from less popular reform measures” out of concern that political opponents will resort to non-democratic means (Haggard and Kaufman 1995, 152). Keefer (2005) highlights another source of uncertainty: the fact that politicians and political parties in new democracies lack track records, and thus have difficulty making credible campaign promises. This “credibility of commitment” problem fosters clientelism, as incumbents seek the support of any existing patron-client networks and target spending to benefit small groups, with which credibility is more easily built.

Other theorists have noted the time needed to build democratic institutions beneficial to growth, and suggested that young democracies will consequently reap fewer economic rewards. As Papaioannou and Siourounis point out, the adherents of the Austrian school mentioned in the previous section cautioned, “the benefits of democracy will show themselves only in the long run, while its more immediate achievements may well be inferior.” (Hayek quoted in Papaioannou and Siourounis 2004, 4). In Rodrik’s (1999; 2000) model of social cooperation discussed above, democracy facilitates policy stability only after repeated interactions by groups with competing agendas. This perspective implies an initial absence of norms of compromise, which may amplify the challenges arising because “new democratic governments faceexceptionally strong distributive pressures, both from groups re-entering the political arena after long periods of repression and from established interests demanding reassurance” (Haggard and Kaufman 1995, 152). We would therefore expect policy-making in new democracies to be characterized by acrimonious debate and unwillingness to compromise, resulting in policy volatility and impaired economic performance.

Surprisingly, relatively few studies have investigated the relationship between the age of democratic governments and economic performance. This lacuna stems partly from the preference among analysts for pooling data and thus looking at cross-country differences and at the relationship between long periods of democracy and at averages of economic indicators over time (Papaioannou and Siourounis 2004). Consequently, this gap is being filled as time series analysis becomes widespread. Although Przeworski et al. (2000) have a well-known finding that the older a democracy, the less likely it is to be reversed, he and his colleagues find that the age of democracy is not significantly related to economic performance.

Two more recent studies have found that young democracies in fact perform as well or better than more mature regimes. Controlling for other determinants of economic performance, Gasiorowski finds that average growth and inflation in young democracies do not differ significantly from “mature” democracies, no matter what threshold is used to separate these two groups (2000).4 Rodrik and Wacziarg (2004), on the other hand, find that a dummy variable for the first five years following democratic transition is significantly associated with higher economic growth and with lower variance in growth rates.5 Still another study suggests that states may experience slower growth in the first two years following democratization before seeing acceleration in growth (Papaioannou and Siourounis 2004). The fact that the latter analysis looks only at countries that permanently democratized, however, limits its policy relevance, since many of the greatest economic challenges occur in countries that eventually fell back into dictatorship.

Turning to other indicators of economic performance, Keefer finds that longer periods of uninterrupted democracy are associated with lower corruption, public investment, and public employment, as well as with higher provision of public goods ranging from the rule of law tosecondary education (2005).6 By contrast, Lake and Baum, running time series regressions, find that an increase in a given country's level of democracy results in a statistically significant and rapid jump in public service provision, measured by educational enrollment, life expectancy, and a host of other indicators (2001). The different conclusions reached by these different studies demonstrate that how one asks questions about the economic performance of new democracies, as well as the methodology used to answer that question, affects the results obtained.

Summary
As a political system, democracy has a number of distinctive attributes that have been posited by scholars to have important economic effects. By way of summary, we highlight the following hypotheses, some of which point us in contradictory directions:

(1) Owing to the influence of the median voter, democracies are more likely to redistribute income than non-democracies. This, in turn, reduces the growth potential of these polities.

(2) Political competition in democracies drives politicians to reduce taxes and increase expenditures; democracies thus tend toward budget deficits.

(3) Elections generate incentives for politically motivated manipulation of economic policy, resulting in a political business cycle.

(4) Owing to the inability of individual voters to organize effectively, public policy in democracies will be shaped powerfully by rent-seeking and welfare-reducing special interest groups.

(5) Democracies will exhibit greater policy consistency, since repeated negotiations over policy facilitate compromise.

(6) By introducing a modicum of competitive pressure into the monopolistic public sector, democracy will result in the provision of more and better public goods.

(7) In order to maintain the support of their core constituencies, political parties will adopt distinctive economic policies both in their platforms and when they achieve power. Left of center parties will favor the interests of labor; right of center parties those of capital.

(8) The presence of broad-based, cohesive political parties is conducive to high-quality economic management, since these are better able to mediate distributional conflicts.

(9) The separation of powers associated with democracy mitigates the potential for governmental abuse, since the different branches of government can check one another.

(10) The economic policies and performance of democracies depends upon whether they have adopted presidential or parliamentary systems of governance. Parliamentary systems will tend to support larger governments than presidential systems, but may reduce the ability of executives to accumulate and abuse power.

(11) The younger the democracy, the more unstable its economic policies and performance.

In the following section, we explore these hypotheses through the lens of the particular public policies that democracies actually adopt.

II. Economic Policy in Young Democracies
While democracy may have a direct effect on GDP growth, most studies of the overall democracy and growth relationship envision the former as having various positive or negative impacts on the latter through a variety of different channels, including those highlighted in the previous section: electoral politics, interest groups, institutions, and the age of the regime. This list, however, is far from exhaustive. Scholars have examined numerous other channels through which democracy affects growth (for in-depth discussions of this approach, see Tavares and Wacziarg 2001; Krieckhaus 2005; and Bourgignon and Verdier 2000). The advantages of an analytical approach that focuses on the precise pathways by which democracy influences economic policy and performance are twofold.

First, examining specific channels through which democracy impacts growth avoids some of the methodological pitfalls associated with simply plugging a variable measuring democracy into a standard growth regression. As Baum and Lake note, most studies adopting that approach do put forward some of the theoretical arguments regarding the democracy-growth relationship, but fail to account for the implied multi-colinearity in their econometric work (2003, 335). They cite as an example the argument that democracy encourages growth by rendering property rights more secure. This implies that measures of democracy will be correlated with investment—presumably another variable in the growth regression—since investors face less uncertainty about their returns.

Investigating the relationship between democracy and specific policies, which in turn impact growth, helps to avoid this misspecification and thus may offer clearer indications of a causal relationship between democracy and economic growth. Second, this methodology provides more concrete guidance to policy-makers, both in developing countries and in industrialized countries seeking to promote development through foreign aid and other policies. Consequently, in this section we examine the specific economic policies that the leaders of young democracies seem to pursue through the lens provided by recent empirical research. We find that this approach indeed produces more clear-cut findings.

Fiscal Policy
As already noted in the first section of this paper, it has long been held that democratically elected politicians will seek to minimize taxation while maximizing government spending, and that as a result democracies will tend toward budget deficits. But empirical research suggests that the relationship between democracy and fiscal policy is complex, with some studies finding that democratic governments do not differ in their fiscal policies from non-democratic regimes, while others find that democracies, particularly younger democracies, do tend to have looser fiscal policies.

In a pioneering study that compares and contrasts the economic policies of democracies and non-democracies, Mulligan, Gil, and Sala-i-Martin (2004) find no significant relationship between countries' level of democracy and their average education spending, social security spending, and corporate tax rates, although they do find that more democratic states tend to have somewhat lower average government revenue than less democratic ones.7 Similarly, Satyanath and Subramanian (2004), find no robust relationship between levels of democracy and government consumption as a percent of GDP.8

However, in their broader study of the channels by which democracy affects economic growth, Tavares and Wacziarg (2001) find that the level of democracy is associated with higher government spending9. Similarly, Block (2002) finds that countries with competitive elections tend to exhibit higher current government expenditure than in countries without competitive elections. Further complicating the picture, Keefer (2005) finds a significant negative relationship between the number of years of uninterrupted democracy in a country and the average public sector wage bill as well as public investment (both as a percentage of GDP) over the period.

In fact, these results may not necessarily conflict with one another, due to the differences in the analytical techniques used and in the variables examined. The results of Tavares and Wacziarg (2001) and Keefer (2005) appear most directly relevant to the study of new democracies, the former because their time series regression takes account of within-country variation, and the latter because Keefer employs the age of democracy as an independent variable in his cross country regression. Thus, it would appear that young democracies may exhibit initial increases in public spending possibly followed by a longer-term decline.

Monetary Policy
Given its importance to economic well-being at both the macro- and micro-levels, few topics have attracted more research in political economy than the making of monetary policy (Drazen 2001). Monetary policy presents numerous puzzles for researchers, including the failure of some states to control inflation or to regulate the banking system and its money creation function.

Further, researchers have been attracted to normative questions of institutional design: how does one enhance the credibility of commitment to monetary stabilization in those cases where policy-makers have a poor track record? In seeking to answer both these positive andnormative issues, political economists have explored a wide-range of variables that could influence the making of monetary policy, including the presence or absence of political business cycles and of democratic institutions that constrain (or not) financial profligacy.

In their broad analysis of the relationship between macroeconomic stability and democracy, for example, Satyanath and Subramanian (2004) find that the average level of democracy in a country has a significant and substantial negative association with average money supply growth over the period 1960 to 1990. Following the line of analysis put forward by Rodrik (2000), they attribute this relationship to the superior ability of democratic institutions to mediate distributive conflicts; indeed, more generally, democracy may be conceptualized as a technology for the peaceful resolution of distributive conflicts, which contributes to growth in myriad ways.

More narrowly, in studying monetary policy in Africa, Guillaume and Stasavage (2005) similarly find that democratic institutions enhance the quality of monetary policy-making. They note that monetary policy credibility depends crucially not simply on central bank independence, but on the extent of checks and balances in place to guarantee Central Bank independence. Formal structures such as constitutionally established vetoes can help to put such checks in place, but so can more informal institutions, such as multi-party democracy and the existence of coalition governments. Thus, while they do not test the relationship between better democracy and monetary policy econometrically, the find that higher levels of democracy are one factor facilitating the development of a credible monetary policy.

Although these studies do not directly address the conduct of monetary policy in new democracies, some suggestive implications can be drawn. Both, for example, stress the importance of effective institutions; in the case of Satyanath and Subramanian (2004), these institutions mediate distributive conflicts, while in the case of Guillaume and Stasavage (2005) they check political actors seeking to manipulate policy. Consequently, while democracy may facilitate better monetary policy over the medium and long term, new democracies with nascent institutions could be expected to experience difficulties controlling monetary management. We explore this question in greater detail below as we explore other economic policies that democratic leaders might seek to manipulate for political reasons.

Exchange Rate Policy
The choice of exchange rates has become one of the most difficult issues of economic policy facing governments in the post-Bretton Woods era. Faced with the “unholy Trinity” of trade-offs among competing economic objectives, nations must seemingly adopt the least of the evils and hope for the best. The near-continuous cycle of currency crises and exchange rate collapses since the 1970s suggests that the world’s economic managers are far from achieving a stable equilibrium.

Why do states find it so difficult to commit to credible exchange rate policies? Willett (2004) suggests that any analysis of that question must focus on politics. In particular, he asserts that owing to conflicting political pressures—say between interest groups and the median voter—“governments will not always choose policy regimes that are ideal from the standpoint of generating economic efficiency and avoiding crisis…” Alesina and Wagner (2004) find empirical support for this proposition in their study of the relationship between broad measures of institutional quality and exchange rate policy. Specifically, they find that countries with poor institutions tend to attempt to peg their exchange rates, but then fail to maintain the peg. Nonetheless, more fine-grained analysis may help illuminate how political parties and institutions aggregate competing interests within the cockpit of electoral politics, and the exchange rate policies that are generated as a result. We are unaware of any empirical work that takes the sort of theoretical sketch provided by Willett and others and examines it on either a case study or cross-country basis.

Instead, most research compares and contrasts “democracies” and “non-democracies” at the most abstract and aggregate level. Satyanath and Subramanian (2004), for example, use annual average changes in the nominal parallel market exchange rate as a broad measure of macroeconomic stability, noting that this variable reflects not only incorporates domestic factors such as inflation, but also external factors, such as problems with the balance of payments and debt. They find that the level of democracy has a significant and robust negative association with exchange rate volatility, with much larger coefficients than other variables tested, including income inequality and openness to trade. While Satyanath and Subramanian (2004) do not examine exchange rate policy specifically, Block et al. (2002) do attempt to determine whether political considerations impinge on exchange rate management. They find evidence that devaluations are postponed in pre-election periods, presumably to avoid provoking public dissatisfaction by increasing the price of imports.

Tavares and Wacziarg (2001), in contrast, observe no significant relationship between the level of democracy in a country and the black market premium on the exchange rate,10 which they employ as a proxy variable to measure distortion economic policies more broadly. Consequently, while exchange rate volatility may be lower in democracies, it is not because their exchange rate policies are more or less distorted. Clearly, more nuanced research is needed to advance our understanding of the political economy of exchange rate management.

Public Goods Provision
If voters have any direct influence in democracy, one might expect it to be felt in the area of public goods provision. Normally, voters want the state to provide education, health care, a clean environment, infrastructure, and an ever-expanding list of goods and services. In a major survey linking democracy and public goods, Lake and Baum (2001) find that public goods provision is not only greater in democracies but is generally qualitatively superior to that in less democratic countries as well. In cross sectional regressions, they find that higher levels of democracy are associated with better education, as measured by a host of indicators including literacy, primary school student-teacher ratios, and the level of enrollment at all grades, and with better health, as measured by life expectancy, mortality, vaccinations, and population per physician, as well as access to healthcare and clean water.11 Lake and Baum (2001) also run time series regressions and find that an increase in a given country's level of democracy results in a statistically significant and rapid jump in public service provision.

These results are also supported by descriptive statistics included in Papaioannou and Siourounis (2004), indicating that democratization yields rapid increases in life expectancy and schooling. Likewise, Tavares and Wacziarg (2001) find that higher levels of democracy are associated with higher average years of secondary schooling. Keefer (2005) also finds that longer periods of uninterrupted democracy are characterized by higher average secondary school enrollment, further bolstering the claim that democracy is associated with better public service provision.

That a positive relationship exists between democracy and the provision of public goods, specifically education and healthcare, appears straightforward. Both cross country and timeseries analyses indicate that greater democracy is associated with more and better educational and health outcomes. The evidence also suggests that even in new democracies, performance in this area tends to improve fairly rapidly. The posited theoretical reason for this apparently strong relationship is the demand of the median voter for such public goods, which of course are paid for by the polity at large. This could also be consistent with a broader political strategy of consolidating young democracies through the delivery of basic services, as well as the Aristotelian idea that since democracies rely on a cadre of educated citizens, they should be expected to invest in human capital. We note that while Persson and Tabellini posit that parliamentary regimes in particular should be relatively more effective at delivering public goods, we are unaware of any explicit comparisons of performance across different institutional structures in the context of the newly democratic states.

Trade Policy
Few public policies are more contested in democracies than foreign trade, and probably none have been subject to more research by political economists. With respect to developing countries, one of the major puzzles that has been examined in the literature is the shift to more liberal trade policies during the 1980s and 1990s, which seems surprising in view of the much-touted abilities of protectionist interest groups to block policy changes. In their analysis of this puzzle, Milner and Kubota (2005) find that higher levels of democracy are associated with more liberal trade regimes, particularly in the newer, developing-world democracies.12 Their results are robust to the inclusion of variables measuring other factors frequently cited as causing trade liberalization, such as economic crisis and external pressure. The conclusion they draw follows Mayer (1984) and Rodrik (1998) in suggesting the importance of the median voter. Since the median voter in the developing world profits from intensive use of her labor, she may have a strong interest in increasing labor-intensive exports and capital-intensive imports. While this analysis is plausible, it must be emphasized that these authors do not empirically investigate whether or not the median voter does in fact benefit from trade liberalization. Indeed, evidence on this point is mixed. While Bourguignon and Morrison (1990) find that openness is associated with a reduction and inequality, Kapstein and Milanovic (2003) have shown that the medianvoter in the poorest countries does not benefit economically from globalization, a finding that poses a puzzle for this line of analysis. Papaioannou and Siourounis include in their examination of the democracy-growth relationship descriptive statistics showing that "democratizations are followed by a substantial increase in international trade . . . " (2004: 10), indicating that the policy changes encountered by Milner and Kubota (2005) produce real increases in trade as a percentage of GDP. This finding is confirmed by Tavares and Wacziarg, who show that a greater degree of democracy is significantly associated with a greater trade-to-GDP ratio. Other research encountering a positive relationship between democracy and trade liberalization includes Dethier et al.’s (1999) study of economic reform in post-Communist countries. In a pooled OLS regression they find that countries' average scores on the Freedom House index are significantly associated with an index of liberalization that includes trade liberalization, domestic price liberalization, and privatization.

The robust finding that more democratic governments tend to have more liberal trade policies suggests that young democracies are likely to liberalize trade with a resulting increase in their trade-to-GDP ratios. However, the reasons for this shift in trade policy and performance have yet to be conclusively determined.

Privatization
Privatization of state-owned enterprises has also proved controversial, which is not surprising given that it generally means worker displacement and rising unemployment, as new owners of formerly state-owned enterprises seek to make organizational changes that will enhance efficiency. Does this mean that young democracies, which could be especially motivated to avoid increasing unemployment, are likely to put privatization decisions on hold? Surprisingly, that does not seem to be the case.

In their study of 76 developing countries during the period 1987 to 1994, Biglaiser and Danis (2002) find that the presence of democracy has a substantial and significant positive association with the extent of privatization.13 Indeed, simple averages for each year in theirsample show that democracies privatized more than non-democracies in every year. This relationship holds even when controlling for numerous other factors generally cited as partial determinants of privatization, such as the level of development, the fiscal deficit, and the current account position. They explain their findings by noting that while proposals for privatization in democracies may be met with highly visible protests on the part of such affected groups as workers, within authoritarian regimes it is the executives at state-owned enterprises who will effectively exert pressure from behind closed doors to block privatization. The association between democracy and privatization is supported by the above-mentioned study of post-Communist countries by Dethier et al. (1999), which employs as a dependent variable an index that includes as one of its components privatization of state-owned enterprises.

Further, as Kapstein and Milanovic (2003) have shown, in many new democracies privatization has been associated with the rise of social spending targeted at displaced workers. Rather than assisting the poor or those who are most vulnerable, political actors have sought to win support for privatization and globalization by providing side-payments in the form of unemployment insurance and pension reform, thus permitting them to make the transition to the new, private sector-oriented economies in which their skills may be in high demand. Thus, privatization and other economically disruptive policies may succeed particularly in those democratic settings where they are associated with more generous safety net provision.

Institutional Quality
Among their many attributes, democracies are often portrayed as less corrupt and administratively more efficient than non-democracies, and older democracies are posited to be better managed than younger democratic states. Keefer (2005) tests the relationship between the age of democracy and variables he sees as symptomatic of clientelistic behavior, and finds that longer periods of democracy are associated with lower corruption, higher bureaucratic quality, and the rule of law (as well as with higher secondary school enrollment and lower public sector spending, as mentioned above). Numerous other studies support the hypothesis that democracy is associated with lower corruption (Treisman 2000)14 and higher institutional quality (Rivera-Batiz 2002).15 Similarly, Rigobon and Rodrik (2004) find that democracy and the rule of law are mutually reinforcing.16

While much of this literature finds that democracy improves institutional quality only after an extended period of time (Keefer 2005, Treisman 2000), other work has found its effect to be more immediate. In a study of economic performance in post-Communist countries, Beck and Laeven (2005) find that the level of democracy in 1992 is significantly and positively associated with the level institutional development in 1996.17 The theoretical analyses discussed in the previous section, which see democracy as checking abusive government and introducing competitive pressures into the public sector, provide some indication of the causal mechanisms behind these findings. However, further exploration into how and when democratic government produces better institutions, if that is indeed the case, is need.

The Political Business Cycle
As discussed in Section I, political business cycle (PBC) theory makes the general claim that electoral schedules influence a country’s economic policies. While the theory of political business cycles suggests that politicians will attempt to manipulate growth to further their chances of reelection, in practice their ability to do this may be severely limited, particularly in developing countries. While increasing the rate of GDP growth with sufficient precision to affect the outcome of an election is a tall order for officials in even the most developed economies, the ability of governments in developing countries to do this is extremely limited due to relatively low levels of monetization, the small size of government expenditure to GDP, and greater vulnerability to external shocks (Schuknecht 1996; Brender and Drazen 2004). It is thus unsurprising that no studies have found that PBCs in developing countries cause growth to change in the periods before or after elections. For example, Schuknecht (1996) finds no association between election periods and changes in output growth.18 Similarly, in a study ofAfrican economies, Block el al. (2002) mention in passing that they find no association between multi-party or founding elections and growth.

Other evidence suggests that this failure to impact aggregate economic performance is not for lack of trying. Block (2002) finds that current expenditure picks up in election years, but only in countries holding competitive elections.19 Consequently, a shift from dictatorship to democracy implies the emergence of a political budget cycles.20 Whereas Block (2002) compares democratic and non-democratic regimes, Block et al. (2003) contrast new democracies with more established ones, and find that political budget cycles—manifested through increased public spending, net claims on government, and seignorage—are more pronounced under newer democratic governments.21 They find that these impacts are amplified when the election in question is a founding election, defined as the first competitive election after a period of non-competitive or no elections. Moreover, founding elections are associated with acceleration of the growth of the money supply (specifically M1). They outline a number of reasons why PBCs might be more pronounced in new democracies. Not only do incumbents who have not been competitively elected likely have more discretion to manipulate the economy, but such leaders have an incentive to fend off any challengers early in order to minimize future competition (Block et al. 2002: 9,12).

The association between new democracies and political business cycles is confirmed by Brender and Drazen (2004), who find that political budget cycles are present only in new democracies. In a sample of 68 democracies from 1960 to 2001, they find that if the sample is narrowed to exclude the first four elections following democratization, all traces of a political budget cycle disappear. By contrast, election year dummies are significantly associated with higher deficits and greater expenditure during the first four elections following democratization.

Including separate dummy variables for the first, second, third, and fourth election following democratization shows that the significance of the election-year declines as the democracy ages.

These studies of the political business cycle suggest that new democracies will undergo an increase in the variability of rates of inflation and money supply growth and in the level and composition of government spending. In addition to the disruptions that these variations cause, they will also likely retard growth in the years after democratization. Ramey and Ramey (1995: 1148) note that volatility in monetary growth and fiscal policy— precisely the variables affected by the political business cycle—significantly lowers growth. Exploring this relationship in detail, Aizenman and Marion (1999) find that volatility in these variables is associated with lower private investment, which presumably hampers growth.

As noted in Section I above, the political business cycle may theoretically enhance economic efficiency, since it allows the electorate to select the most competent candidate, although this result depends on whether or not one assumes that politicians know whether or not they are competent (Persson and Tabellini 2000).22 To date, no one has found evidence of this theoretically possible “efficient PBC.” Consequently, the presence of political business cycles in new democracies, with the concomitantly higher volatility of monetary and fiscal variables and the resulting slower growth, suggests that new democracies may face particular challenges in the formulation of economic policy.

Summary
Just as democracies have distinctive political attributes, they also seem to generate some distinctive public policies. Further, the age of democracy seems to matter to economic policy-making, with older democracies performing better than younger democracies across a range of measures. To highlight some of the major findings of the empirical literature, we note that:

(1) Democracies, including young democracies, appear to provide more and better public services to their citizens.

(2) Democracies, including young democracies, are more open to trade than non-democracies, perhaps due to the median voter’s influence.

(3) Democracies, including young democracies, are more likely to privatize their state-owned enterprises than non-democracies.

(4) Democracies have better institutional quality than non-democracies, but older democracies fare better along this dimension than do young democracies.

(5) Politicians in young democracies appear to make extensive use of the political business cycle, introducing macroeconomic volatility into the growth equation.

In the following section, we examine the extent to which these findings are supported by our new data set, which is more up-to-date and includes more cases than most data sets used in the studies cited herein.

III. Economic Performance of Young Democracies
In this section we provide an overview of the economic performance of young democracies since 1960. Overall, since 1960 we count 114 total episodes of democratization, 44 of which were reversed at some point; among those that were reversed, several then later underwent second and even third democratization episodes, as indicated by Table 1 below (see Appendix 1 for precise definitions and methodology). Overall, nearly 38 percent of the democratizations that occurred between 1960 and 2003 eventually underwent reversals, ceasing to be democracies for some period of time.23 Cases in which countries underwent democratization for the first time in the post-1960 period had nearly the same rate of reversal (38.6 percent), as those undergoing democratization for the second time (41.7 percent). In contrast, only one of the six cases in which a country democratized for a third time was reversed, while in the single case in which a country underwent a fourth democratization, it was not reversed.

The question that we ask in this section is whether differences can be discerned among those that successfully democratize and those that undergo total reversals. We also wish to see if any regional differences can be found among the world’s young democracies in terms of their economic performance. Overall, our objective is to see whether the descriptive statistics yield any patterns that could be amenable to more precise testing. To this end, we examine the average economic performance of countries in each of the ten years following democratization. As we will see, we believe that some intriguing patterns do indeed emerge. We leave aside issues of specific causality in this working paper, with the intention of addressing this question in our monograph.

Before describing our findings, a brief discussion of our methodology is warranted. Our unit of analysis is young democracies, and as mentioned above, some countries have undergone more than one democratization in the period under study. We have counted each episode of democracy as a separate case. The analysis in this section compares the average performance of cases in which democratization did not undergo a reversal to that of those in which it was ultimately reversed. As is evident from a look at the underlying data included in Appendix 2, the size of the group of new democracies that eventually undergo reversals falls over time, as each year a few countries cease to be democracies. Because we are interested in how new democratic governments manage the economy, it would not make sense to include years in which a country’s government was not democratic.24

The example of Argentina provides an illustration of this methodology. Argentina democratized twice during the period: once in 1973 and again in 1983. These two episodes count in our dataset as two separate young democracies. In the former case, democracy was overthrown in 1976. As a consequence, Argentina from 1973 to 1975 is included in the subgroup of young democracies that are ultimately reversed, but drops out after the second year (i.e. 1973 is year zero, the year of democratization; 1974 is year one; 1975 is year two). The episode of democracy in Argentina that began in 1983 is included in the subgroup of young democracies that were not reversed (with 1983 counting as year zero of democracy, 1984 as year one, and so on).

Growth
Contrasting the economic performance of successful new democracies with that of democratizations that were ultimately reversed reveals that growth in the former trends upwards, while the latter shows no clear trend, as Figure 1 demonstrates. Because the initial difference in the average growth rates is statistically significant as are the differences in later years,25 it would appear that growth in unreversed democracies is on average lower in the years immediately following democratization, but overtakes that of reversed democracies around the seventh year after democratization. This finding is particularly interesting in light of those theories discussed in Section I suggesting that democracy takes time to yield economic benefits (Haggard and Kaufman 1995; Keefer 2005, Rodrik 1999, 2000). Young democracies that ultimately undergo total reversals may cut corners early on, avoiding some of the challenges of democratization but forgoing its later benefits.

Figure 2 suggests that rates of investment are also significantly higher in unreversed new democracies than in countries where democracy is overthrown, although in both cases it trends upward. Moreover, the variance of investment falls steadily in unreversed new democracies, while exhibiting no clear trend in countries where democratization is ultimately reversed. The literature focusing on the demands for redistribution unleashed by democracy (Acemoglu and Robinson 2006) casts this pattern in an interesting light. Lower investment may signal misgivings on the part of those holding capital regarding the desirability of democracy, doubts that later translate into support for a return to authoritarianism. In short, after an initial honeymoon period, it appears that investors in some democracies quickly become aware of the fragility or perhaps undesirability of their political system and investment levels stagnate.

Inflation
Analysis of inflation in young democracies is complicated by a handful of cases that experienced periods of hyperinflation, with the annual change in consumer prices reaching into the thousands of percent. In the interest of uncovering general trends, we here exclude rates of inflation more than five standard deviations above the average for each subgroup for each year. To cite the most extreme example, Peru in 1990 (ten years after democratization) recorded inflation of nearly 7500 percent. As this is more than eight hundred standard deviations above the average for the “reversals” subgroup ten years after democratization, we exclude this value.26 The trends discussed here are robust to changes in the rule used to exclude extreme values.

Applying this methodology, consumer price inflation in unreversed new democracies is initially high, but trends downward, while inflation in new democracies undergoing reversals shows no clear trend, fluctuating around 15 percent throughout the period. Average rates of inflation converge around the seventh year after democratization, as Figure 3 illustrates. As would be expected, data on the average rate of money supply growth and the average change in the official exchange rate show very similar trends.

These trends are broadly in line with the findings of Block et al. (2003) that democratization is associated with acceleration in the growth of the money supply, but that this slows again as democracy ages. This pattern also lends also credence to the idea, discussed in Section I, that thebenefits of democracy take time to materialize, either because of the uncertainty present under a new political regime or because institution-building takes time. Where these difficulties are confronted early on, economic performance may be initially inferior but later improve, while in cases where they are suppressed, the seeds for later democratic reversal may be planted.

Government Spending
Government spending as measured by public consumption as a percentage of GDP is initially about three percentage points higher in democracies that do not undergo total reversals than in those that do, a statistically significant difference. However, while spending trends downward in the former, it rises in the latter, and the two converge roughly nine years after democratization, as can be seen in Figure 4. That no political budget cycle is evident in this data is not surprising, since election schedules differ between the countries in the sample.

These trends could be interpreted to mean that while the state in successful new democracies initially responds to demands for resources from the newly empowered population with relative largesse, it later grows more capable of managing such demands in a fiscally responsible manner. By contrast, governments in countries ultimately undergoing another reversal apparently become less able to assuage such demands, and become gradually more profligate.

Openness
Figure 5 suggests that successful new democracies do on average exhibit greater openness to trade than new democracies that are later reversed, as measured by the ratio of imports plus exports over GDP, and once again, the difference is statistically significant, with the null hypothesis of equal average openness rejected in all 11 years. Both groups show an upward trend in openness over the ten years following democratization. This finding is in line with other work indicating that democracy facilitates trade liberalization as well as economic reform more broadly (for example, Milner and Kubota 2005).

Ethnic Divisions
Numerous studies have found that a higher degree of ethnic fractionalization is associated with slower economic growth, higher corruption, and low institutional quality (Easterly, Ritzen, and Woolcock 2005; Alesina et al. 2002), suggesting that countries with an above-average degree of ethnic division may face particular difficulties when undergoing democratization. Our data appear to support this proposition. Table 2 suggests that while new democracies with lower levels of ethnic fractionalization on average have a significantly higher rate of economic growth during the first five years prior to democratization, the average growth rates of the two groups converge in the second five-year period after democratization. This would appear to support theories that view democracy's ability to facilitate conflict resolution as salutary for growth. Interestingly, whereas growth rises with democratization in countries with below-average ethnic fractionalization, it falls with democratization in the more ethnically diverse countries, perhaps due to the disruption caused by the airing of ethnic conflicts suppressed during periods of authoritarian rule.

Similarly, young democracies with a lower than average degree of ethnic fractionalization have average rates of investment that are 1-2 percent of GDP higher than those with relatively greater fractionalization. Again, there is some trend towards convergence toward the end of the first ten years following democratization, although the differences between the average rates of investment are statistically significant at the 1 percent level in every year. Analysis of inflation data also lends support to the following hypothesis: that governments of states with higher than average levels of ethnic fractionalization have difficulty managing the economy in the years immediately following democratization, but to some extent master these difficulties over time.

Thus, our data suggest that the level of ethnic fractionalization plays a role in determining how well young democracies are able to deal with the challenges of political development and economic management, with states characterized by a high degree of ethnic division having more difficulty managing the economy in the initial period following the transition from authoritarianism.

Regional Variations
Changes in key economic indicators in new democracies follow quite different trajectories in different regions, a finding in line with those of Krieckhaus's (2005) contention that democracy's economic impact varies by region. While we will not pursue here the reasons for this variation in detail—that topic will be fully developed in our monograph on the economics of young democracies—we might simply assert that we do not believe that these differences can be ascribed solely to such “initial conditions” as geography. As Steven Fish has written, “polities undergoing regime change do not necessarily suffer under a tyranny of initial conditions. They make their own fates” (Fish 2001, 82). What this suggests is that democratic institutions and economic policy decisions also matter greatly, and that these are not merely a function of a region’s prior history. They are also a reflection of the ideas that leaders hold, the institutions—e.g. independent central banks—that they do or do not establish, the support they receive from foreign donors, and the possibility of anchoring their economies in a larger entity like the European Union. These hypotheses will be pursued in the monograph version of our work. For now, we will simply summarize the variation that we observe.

In terms of growth, the performance of Eastern European countries differs significantly from that of other new democracies, with output contracting in the first three years after democratization and growth resuming four years after democratization, as is apparent from Figure 6. This sharply different growth trajectory, presumably caused by the specific circumstances of the transition from communism, demonstrates the difficulty in drawing overall conclusions about the economic performance of young democracies and their survival or failure.

Interestingly, average growth for new democracies in Sub-Saharan Africa generally matches and even outperforms that of the other countries in the sample, as Figure 7 illustrates. This result is likely a product of the strong performance of newly independent democracies in the region during the 1960s, a period of stronger growth worldwide. Indeed, if cases classified as new democracies at independence are excluded from the African group, average growth falls by a percentage point or more.

Trends in inflation exhibit similar regional divergences.27 While new democracies in Latin America are plagued by high and volatile inflation, Eastern European nations faced high and variable inflation immediately after the transition to democracy, but this has since moderated. By contrast, average inflation in Africa rarely exceeds 10 percent, as Figure 9 demonstrates. It is worth noting that Africa’s relatively low average inflation was not solely a result of 14 countries’ membership in the CFA Franc zone. While new democracies in this monetary union do exhibit low inflation—under 10 percent on average, consumer price inflation in Sub-Saharan African countries outside of the CFA Franc Zone has averaged under 20 percent, well below the averages seen in Latin America and Eastern European countries in the early years of democracy.

Once again, it would appear that the high inflation in new Eastern European democracies was a consequence of the transition from communism, which they were able to overcome during the first five years following democratization. The Latin American countries' inability to reign in inflation on a consistent basis could be taken as evidence that the region's young democracies have failed to mediate between groups demanding resources, instead resorting to expansionary policies that have periodically produced hyperinflation.

Regional differences are also apparent in public spending, with the four regions with substantial numbers of new democracies having fairly different levels, as is apparent from Figure 9. Few clear trends are apparent, although the persistent regional differences suggest one explanation for the lack of clear conclusions in the literature on public policy and democracy, since many such studies do not include regional dummies (for example, Rivera-Batiz 2002).

Summary
This analysis bears out the implications drawn from broader literature on growth and democracy, in particular that new democracies that ultimately undergo reversals under-perform relative to their successful counterparts. Although growth is initially lower in the sustained democratizers, it continually rises over time. By contrast, countries in which democracy is ultimately reversed show a trend of stagnant or falling growth. Similarly, new democracies that do not undergo reversals have higher inflation in the early years, but improve over time. Investment is also lower in democracies that experience reversals, as is trade openness. Government spending also trends upwards in countries where democratization is reversed and trends downward in successful democratizers. Overall, it would appear that establishing economic credibility early in the democratic mandate is a difficult task, but once such credibililty is gained, growth accelerates and inflation stabilizes. Where credibility is absent, these variables show no such positive trend. This suggests that discovering the deeper sources of economic policy credibility in young democracies should be a major target of research.

This analysis also demonstrates the need for detailed regional case studies. Growth and inflation evolve in markedly different ways in new democracies in different regions, while regional differences in public spending greatly overshadow changes in spending during the ten years following democratization. Understanding the reasons behind this regional variation will be one of our major tasks in the monograph that we are also writing.

IV. Conclusions and Policy Recommendations
There are few greater policy challenges today than ensuring the success of the world’s many unconsolidated democracies. As we have seen, the economic performance of new democracies—and hence the chances of regime survival—has varied along a number of critical variables, including their ability to maintain investment levels, to control the money supply and budget deficits, and to open their economies to foreign trade. We have also seen that regional differences are significant, with Latin American democracies in particular experiencing the most volatility in terms of economic performance. Pursuing the sources of these differences is clearly a crucial task for research.

From a policy perspective, our work makes several points that we believe are relevant to public officials in both developing countries and the industrial world:

First, the evidence suggests that leaders in young democracies often face trade-offs between their electoral and economic reform objectives. The tendency of leaders to exploit the political business cycle in the run-up to elections indicates that in some settings political competition may induce behavior that is inconsistent with the building of strong economic foundations for sustained growth.

Second, the capacity of leaders in new democracies to pilot their economies varies a great deal across countries. This means that “one size fits all” policy solutions are unlikely to be appropriate where democratic consolidation is a fundamental medium-term objective, and that particular country circumstances must be taken into account as the two tracks of economic reform and democracy-building are pursued.

Third, foreign aid donors must recognize the richness of the democratic experience around the world and the various forms that it takes. While some of the academic literature suggests, for example, that parliamentary regimes may be more growth-enhancing than presidential systems, that work hardly provides a conclusive basis for preferring one type of democracy over the other. Still, the potential costs and benefits of different electoral and institutional systems are worth full consideration as new democracies shape their constitutional political economy.

Fourth, our analysis of the economic performance of new democracies strongly suggests that deteriorating or stagnant economic performance constitutes a red flag or warning signal that the country is at risk of political reversal. Poor economic performance—low growth, high inflation, etc.—during the first five years following a democratization is not necessarily cause for alarm. However, if that performance does not pick up or deteriorates, democracy is at greater risk of being overthrown. Although the direction of causality remains to be settled—does failure to improve economic performance discredit the regime, or is such performance a symptom of weak institutions?—new democracies plagued by negative growth and high inflation, and which fail to undertake such necessary reforms as trade liberalization and privatization, merit the attention of foreign aid donors who seek to promote democratic consolidation.

TABLES AND APPENDICES:

http://www.cgdev.org/files/6583_file_WP_85.pdf

NOTES

1 This argument, it will be observed, rings of the “fiscal federalism” literature which argues that federalist systems introduce competition into the public sector by allowing economic agents to “vote with their feet,” forcing local governors to restrain their personal rent-seeking and provide the public services that people actually want. However, questions may be raised about the economic benefits of political decentralization in those developing countries where government institutions are too weak to provide many of the public goods that citizens demand.

2 We thank Phil Keefer for emphasizing this point.

3 Notwithstanding recent instances of electorates unwilling to wait until the next election to unseat incumbents, as in Ecuador, Bolivia, Argentina, Thailand, and the Philippines.

4 Gasiorowski (2000) employs a dataset of his own construction, which classifies 97 countries as democratic, semi-democratic, or authoritarian in each year from independence through 1992.

5 Rodrik and Wacziarg (2004) analyze data from 154 countries. They look specifically at the average rate of GDP growth in the five years following democratization.

6 Rather than treating each country in each year as an observation, Keefer treats each episode of uninterrupted democracy as an observation and averages the values of the dependent variables over the episode. He uses data from 1975 to 2000. He classifies as democracies regimes with a top score on the Legislative and Executive Indices of Electoral Competition from the World Bank’s Database on Political Institutions. Indicators of corruption, bureaucratic quality, and the rule of law are from Political Risk Service’s International Country Risk Guide.

7 Mulligan, Gil, and Sala-i-Martin (2004) average the country's Polity IV score over the 30-year period to measure democracy.

8 Satyanath and Subramanian (2004) employ the average of the Polity IV variable measuring the level of constraints on the executive to measure the level of democracy in a country during the period in question.

9 Note that they use Freedom House ratings.

10 The premium is defined as the difference between the black market rate and the official rate, divided by the official rate.

11 Baum and Lake (2001) use the overall Polity III score to measure the level of democracy.

12 Milner and Kubota (2005) measure democracy using the Polity IV score, and run time series cross-sectional regressions to test its relationship with openness, as measured by both the unweighted average tariff rate and the Sachs-Warner dummy variable. They analyze a sample of 100 countries over the period 1970 to 1999.

13 Biglaiser and Danis (2002) use the Polity IV score to construct a dummy variable that classifies countries as either democratic or non-democratic. They measure the level of privatization in a country with a variable consisting of thevalue of privatizations in a given year divided by the state-owned enterprises' share of GDP in the previous year multiplied by the current year’s GDP.

14 Treisman (2000) uses Transparency International’s perceived corruption index to measure corruption. He obtains like results employing the Freedom House rating and Przeworski et al.’s classification.

15 Rivera-Batiz (2002) studies this relationship using a sample of 115 countries and employing pooled OLS regression, averaging the variables' values over the period 1950 to 1990. He measures democracy with the Freedom House index, and institutional quality using the index constructed by Hall and Jones (1999), which combines the Political Risk Service institutional quality index with the Sachs-Warner dichotomous measure of openness.

16 Rigobon and Rodrik (2004) define democracy in terms of the Polity IV score.

17 Beck and Laeven (2005) use the Polity IV score for constraints on the executive in 1992 as their measure of democracy. They assess institutional quality using an index that includes corruption, the rule of law, regulatory quality, and government effectiveness, among other things

18 Schuknecht (1996) examines a sample of 35 developing countries between 1970 and 1992.

19 Block examines a sample of 69 developing countries. As independent variables, he selects indices of electoral competitiveness and democracy (both taken from the World Bank’s Database of Political Institutions) as well as a dummy variable for election years. On the assumption that leaders facing reelection will boost current expenditure, which tends to more immediately benefit voters, Block employs current expenditure as a share of total central government expenditure as the dependent variable.

20 Interestingly, whereas Keefer (2005) finds that the proportion of public spending devoted to investment will fall as democracies mature, Block (2002) finds evidence that electoral contestation is associated with a fall in public investment as a share of GDP. These results are not necessarily contradictory, since investment as a share of the government budget may rise while falling as a percentage of GDP. Nonetheless, their logic conflicts on a theoretical level: Keefer views public investment as a prime avenue for clientelistic channeling of jobs and money to supporters, whereas Block sees current expenditure as the easiest way to direct public resources to supporters.

21 Block et al. (2002) use a sample of 44 Sub-Saharan African countries for the period 1980 to 1995 and employ as independent variables an election-year dummy and a dummy for competitive elections.

22 In this model, if politicians think they are competent, then they will attempt to ramp up spending or stimulate the economy. By contrast, if a politician knows he is incompetent the costs of macroeconomic manipulation to him personally will outweigh the gains from trying to trick voters into believing he is competent.

23 Some of the new democracies that were not totally reversed did undergo minor democratic reversals. While we do not discuss such cases separately here, we will do so in our monograph. For a more detailed discussion of this issue, see Appendix 1.

24 This may at first glance generate concerns about attrition bias; however, this would only undermine our findings if the democracies undergoing reversals during the period in question exhibited better economic performance than those countries that remained democracies for the first ten years after democratization. That various studies (most notably Przeworski et al. 2000) have found that the countries with low or negative growth are more likely to undergo reversals suggests that our findings are not driven by attrition bias. Indeed, if only countries in which democracy endured for at least 10 years, but was eventually reversed, are included in the reversals group, the general relationships described continue to hold.

25 The null hypothesis that the average growth rates for the two groups are equal can be rejected at the 5 percent level in 6 of the 11 years. References to statistical significance in this section refer to t-tests of the difference between means of independent samples.

26 The instances of hyperinflation excluded are: Argentina (1984-5,88-91), Bolivia (1984-5), Brazil (1988–94), Nicaragua (1990-1), Peru (1988–90), Armenia (1994), Belarus (1993-4), Bulgaria (1997), Lithuania (1993), Romania (1993, 1997), and Ukraine (1993-5).

27 Whereas we excluded cases of hyperinflation from the discussion of inflation above, in this subsection, we have included these episodes. The fact hyperinflation occurred in two particular regions suggests it is an important part of the regional story, and should not be left out.

REFERENCES

Acemoglu, Daron and James Robinson. 2006. Economic Origins of Democracy and Dictatorship (New York: Cambridge University Press).

Aizenman, Joshua and Marion, Nancy. 1999. Volatility and Investment: Interpreting Evidence from Developing Countries. Economica 66: 157-79.

Alesina, Alberto., Devleeschauwer, Arnaud, Easterly, William, Kurlat, Sergio and Wacziarg, Romain. 2002. Journal of Economic Growth 8: 155-194.

Alesina, Alberto and Rodrik, Dani. 1994. Distributive Politics and Economic Growth Quarterly Journal of Economics 109 (2):: 465-90.

Alesina, Alberto, Roubini, Nouriel with Cohen, Gerald. 1997. Political Cycles and the Macroeconomy. Cambridge, MA: MIT Press.

Bardhan, Pranab and Tsung-Tao Yang. 2004. Political Competition in Economic Perspective. BREAD Working Paper no. 78.

Baum, Matthew. and Lake, David 2003. The Political Economy of Growth: Democracy and Human Capital. American Journal of Political Science 47(2): 333-347.

Beck, Thorsten. and Laeven, Luc. 2005. Institution Building and Growth in Transition Economies. World Bank Policy Research Working Paper 3657.

Biglaiser, Glenn and Danis, Michelle 2002. Privatization and Democracy: The Effects of Regime Type in the Developing World. Comparative Political Studies 35(1): 83-102.

Block, Steven A. 2002. Elections, Electoral Competitiveness, and Political Budget Cycles in Developing Countries. Harvard University Center for International Development Working Paper No. 78.

Block, Steven A., K. Ferree, and S. Singh. 2003. Multiparty Competition, Founding Elections, and Political Business Cycles in Africa. Journal of African Economies 12(3).

Borner, Silvio, Aymo Brunetti, and Beatrice Weder. 1995. Political Credibility and Economic Development. London: Macmillan.

Bourguignon, Francois. and Morrison, Christian. 1990. Income Distribution, Development and Foreign Trade. European Economic Review 34: 1113-1132.

Bourguignon, Francois and Verdier, Thierry. 2000. Oligarchy, Democracy, Inequality and Growth. Journal of Development Economics 62: 285-313.

Brender, Adi and Drazen, Allan. 2004. Political Budget Cycles in New Versus Established Economies. NBER Working Paper 10539.

Bresser Pereira Luis Carlos, J.M. Maravall, and A. Przeworski, Eds. 1993. Economic Reforms in New Democracies: A Social Democratic Approach. Cambridge: Cambridge University Press.

Buchanan, James and Richard Wagner. 1978. Democracy and Keynesian Constitutions: Political Biases and Economic Consequences. London: Institute for Economic Affairs.

Dethier, Jean-Jacques. Ghanem, Hafez, and Zoli, Edda. 1999. Does Democracy Facilitate the Economic Transition? World Bank Policy Research Working Paper 2194.

Diamond, Larry. 2005. Democracy, Development and Good Governance: The Inseparable Links. Annual Democracy and Governance Lecture, Center for Democratic Development, Accra, Ghana.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harpers.

Drazen, Allan. 2001. The Political Economy of Macroeconomics. Princeton: Princeton University Press.

Easterly, William and Levine, Ross. 1997. “Africa's Growth Tragedy: Policies and Ethnic Divisions.” Quarterly Journal of Economics 111(4): 1203-1250.

Easterly, William, Ritzen, Jozef and Woolcock, Michael. 2005 Social Cohesion, Institutions and Growth. New York University Development Research Institute working paper 17.

Fernandez, Raquel and Rodrik, Dani. 1991. Resistance to Reform. American Economic Review 81: 1146-55.

Fish, M. Steven. The Dynamics of Democratic Erosion. In Richard D. Anderson, Jr., et al. Postcommunism and the Theory of Democracy. Princeton: Princeton University Press, 2001.

Gasiorowski, Mark J. 2000. “Democracy and Macroeconomic Performance in Underdeveloped Countries: An Empirical Analysis.” Comparative Political Studies 33(3): 319-349.

Giddens, Anthony. 1971. Capitalism and Modern Social Theory: An Analysis of the writings of Marx, Durkheim, and Max Weber. Cambridge: Cambridge University Press.

Grossman, Gene and Elhanan Helpman. 2002. Interest Groups and Trade Policy. Princeton: Princeton University Press.

Guillaume, Dominique and Stasavage, David. 1999. Making and Breaking Monetary Policy Rules: The Experience of African Countries. Centre for the Study of African Economies Working Paper 99-2 Oxford University.

Haggard, Stephan and Robert R. Kaufman. 1992. Economic Adjustment and the Prospects for Democracy. The Politics of Economic Adjustment: International Constraints, Distributive Conflicts, and the State. Princeton: Princeton University Press.

Haggard, Stephan and Robert Kaufman. 1995. The Political Economy of Democratic Transitions. Princeton: Princeton University Press.

Halperin, Morton; Joseph Siegle, Michael Weinstein. 2005. The Democracy Advantage (New York: Routledge).

Hellman, Joel. 1998. Winner Takes All: The Politics of Partial Reform in Postcommunist Transition. World Politics 50: 203-234.

Huntington, Samuel P. 1968. Political Order in Changing Societies. New Haven: Yale University Press.

Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth Century (Norman: University of Oklahoma Press).

Kapstein, Ethan B. and Branko Milanovic. 2003. Income and Influence: Social Policy in Emerging Market Economies. Kalamazoo, Mi.: Upjohn Institute.

Keefer, Philip. 2005. Democratization and clientelism: Why are Young Democracies Badly Governed? World Bank Policy Research Working Paper 3594.

Keefer, Philip. 2005. Beyond Elections: Politics, Development, and the Poor Performance of Poor Democracies. Unpublished manuscript.

Krieckhaus, Jonathan. 2005. Democracy and Growth: How Regional Context Influences Regime Effects. British Journal of Political Science 35.

Lake, David and Baum, Matthew. 2001. The Invisible Hand of Democracy: Political Control and the Provision of Public Services. Comparative Political Studies 34(6): 587-621.

Marshall, Monty G. and Jaggers, Keith. 2002. Polity IV Project: Political Regime Characteristics and Transactions, 1800-2002. University of Maryland.

Mayer, Wolfgang. 1984. Endogenous Tariff Formation. American Economic Review 74: 970-985.

Milanovic, Branco. Forthcoming. Relationship between Income and Democracy Re-examined, 1820-2000: A Non-Paramentric Approach.

Milner, Helen. and Kubota, Keiko. 2005. Why the Move to Free Trade? Democracy and Trade Policy in the Developing Countries. International Organizations 59: 157-193.

Minier, Jenny 1998. Democracy and Growth: Alternative Approaches. Journal of Economic Growth 3:241-66.

Mulligan, Casey, Gil, R., and Sala-i-Martin, X. 2004. Do Democracies Have Different Public Policies than Non-democracies? Columbia University Department of Economics Discussion Paper No.: 0304-14.

Munck, Geraldo L. and Verkuilen, Jay. 2002. Conceptualizing and Measuring Democracy: Evaluating Alternative Indices. Comparative Political Studies 35(1): 5-34.

North, Douglas. 1990. Institutions, Institutional Change, and Economic Performance. New York: Cambridge University Press.

Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, Ma.: Harvard University Press.

Olson, Mancur. 1982. The Rise and Decline of Nations. New Haven, Ct.: Yale University Press.

Papaioannou, Elias and Gregorios Siourounis. 2004. Democratization and Growth. London Business School Working Paper Series.

Persson, Torsten and Guido Tabellini. 2002. The Economic Effects of Constitutions. Unpublished manuscript.

Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub and Fernando Limongi. 1996. What Makes Democracy Endure? Journal of Democracy 7(1): 39.

Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub and Fernando Limongi.. 2000. Democracy and Development: Political Institutions and Well-being in the World, 1950–1990. (New York: Cambridge University Press).

Przeworski, Adam. and Limongi, Fernando. 1993. Political Regimes and Economic Growth. Journal of Economic Perspectives 7: 51-69.

Ramey, G. and Ramey V.A. 1995. Cross-country Evidence on the Link Between Volatility and Growth. American Economic Review 85: 1138-51.

Rao, Vaman. 1984. Democracy and Economic Development. Studies in Comparative International Development 19(4): 67-81.

Rigobon, Roberto and Dani Rodrik. 2004 Rule of Law, Democracy, Openness, and Income: Estimating the Interrelationships. NBER Working Paper 10750.

Rivera-Batiz, Francisco L. 2002. Democracy, Governance, and Economic Growth: Theory and Evidence. Review of Development Economics 6(2): 225-47.

Rodrik, Dani. 1998. The Rush to Free Trade in the Developing World. In eds. Sturznegger, F. and Tommasi, M. The Political Economy of Economic Reform. Cambridge, Ma. MIT Press.

Rodrik, Dani. 1999. Where did all the Growth Go? External Shocks, Social Conflict, and Growth Collapses. Journal of Economic Growth.

Rodrik, Dani. 2000. Participatory Politics, Social Cooperation, and Economic Stability. American Economic Review 90(2): 140-144.

Rodrik, Dani and Romain Wacziarg. 2004. Do Democratic Transitions Produce Bad Economic Outcomes? http://ksghome.harvard.edu/~drodrik/democracy.pdf.

Rogoff, Kenneth., 1990. Equilibrium Political Budget Cycles. American Economic Review (1990) 80:21-36.

Rudea, David. n.d. Political Parties and Economic Policy in Industrialized Democracies: An Insider-Outsider Partisanship Model. Unpublished Manuscript, Cornell University.

Rueschemeyer, Dietrich. Stephens, Evelyne. and Stephens, John. 1992. Development and Democracy. Chicago: University of Chicago Press.

Satyanath, Shanker. and Subramanian, Arvind. 2004. What Determines Long-Run Macroeconomic stability? Democratic Institutions. Unpublished manuscript.

Schuknecht, Ludger, 1996. Political business cycles and fiscal policies in developing countries. Kyklos 49: 155-170.

Sen, Amartya. 1994. Poverty and Famines. Oxford: Oxford University Press.

Tavares, José and Romain Wacziarg. 2001. How Democracy Affects Growth. European Economic Review 45(8): 1341-1379.

Treisman, Daniel (2000). “The Causes of Corruption: A Cross-National Study.” Journal of Public Economics 76(3): 225-56.

Willett, Thomas D. 2004. The Political Economy of Exchange Rate Regimes and Currency Crises. Background Paper for the Claremon Conference on the Political Economy of Exchange Rates.

Williamson, John. and Haggard, Stephan. 1994. The Political Conditions of Economic Reform. The Political economy of policy reform. Ed. Williamson, J. Washington: Insititute for International Economics.