Economic Origins of Dictatorship and Democracy
Page 8
How does this affect the relationship between inter-group inequality and transition to democracy? The most important implication is that as inequality increases and democracy becomes more costly for the elites, repression becomes more attractive. Therefore, greater inter-group inequality may also discourage democratization.
Putting these two pieces of the story together, we find that there is a nonmonotonic (i.e., inverted U-shaped) relationship between inter-group inequality and the likelihood of transition to democracy. In the most equal societies, revolution and social unrest are not sufficiently attractive for the citizens; either there are no challenges to nondemocratic systems or any challenges can be met by temporary measures, such as some limited redistribution. In other words, in these fairly equal societies, the citizens are already benefiting from the productive resources of the economy or even perhaps from the growth process, so they do not make further strong demands. This may be the reason why democracy arrived late in a number of equal and rapidly growing economies, such as South Korea and Taiwan, and has yet to fully arrive in Singapore. In stark contrast, in the most unequal societies (e.g., South Africa prior to 1994), the citizens have great reason to be unhappy and often try to rise up against the authority of nondemocracy. Now, however, the elites have a lot to lose from abandoning the system that looks after their own interests and transitioning into one that will place a greater redistributive burden on them. Thus, instead of democracy, a highly unequal society is likely to result in a repressive nondemocracy- or, sometimes when repression is not enough, perhaps even experience a revolution. This mechanism can also explain the persistence of nondemocratic regimes in the highly unequal countries of Latin America, such as El Salvador and Paraguay. This account, then, suggests that democracy has the best chance to emerge in societies with middle levels of inequality. Here, the citizens are not totally satisfied with the existing system, and the elites are not so averse to democracy that they resort to repression to prevent it. This is the situation we find in Britain and Argentina in the late nineteenth and early twentieth centuries.
6.5.2 Consolidation
Inequality also critically influences the propensity of a democracy to consolidate. Because the main threat against democracy comes from its redistributive nature, the greater redistribution away from the elites the more likely they are to find it in their interest to mount a coup against it. Therefore, greater inequality is likely to destabilize democracy because, as observed previously, the burden of democracy on the elites is increasing in the income gap between them and the citizens.
This comparative static result with respect to inequality offers a potential explanation for why democracy may have been more difficult to consolidate in Latin America than in Western Europe. Latin American societies are considerably more unequal and, therefore, suffer more from distributional conflict between the elites and the citizens. Our framework predicts that in highly unequal societies, democratic policies should be highly redistributive but then abruptly come to an end with a coup that reverts back to much less redistributive policies. This pattern is reminiscent of the oscillations of many Latin American countries between the highly redistributive but unsustainable populist policies of short-lived democracies and the fiscally more conservative approach of subsequent nondemocratic regimes. Tellingly, Kaufman and Stallings (1991, p. 27) also emphasize a close connection between unconsolidated democracy and populist redistribution:
... established democracies (Venezuela, Colombia and Costa Rica in our study) were also associated with orthodox macro policies.... It was the transitional democracies (Peru, Argentina and Brazil) that followed populist policies.
Combining the effects of inequality on democratization and coups, we can see that equal societies never democratize in the first place. This helps to account for Singapore’s path of political development. Higher but still relatively low levels of inter-group inequality lead societies to democratize and, once created, democracy is consolidated because it is not so costly for the elites that a coup is desirable. This may capture Britain’s path of political development. Even higher levels of inequality still lead to democratization, but democracy does not consolidate because coups are attractive. As a result, the outcome is unconsolidated democracy, which is the path that Argentina followed in the twentieth century. Finally, at the highest levels of inequality, democracy is so threatening for the elites that they use repression to avoid it, a situation that characterized South Africa until 1994.6
6.6 The Middle Class
6.6.1 Democratization
Perhaps the most famous treatise on the origins of democracy is Moore’s (1966) Social Origins of Dictatorship and Democracy. Our work owes a natural intellectual debt to Moore, especially because we paraphrased his title. In our theory, the major factor that distinguishes democracy from nondemocracy is the greater political equality of democracies; so far, we have only distinguished between two groups: the elites and the citizens. This was mainly for simplification (again, an application of Occam’s razor). Nevertheless, in many circumstances, a third group between the elite and the great mass of citizens may be of significance. In general, this group could be identified in different ways but, following the emphasis of many scholars, it is useful to think of this group as the middle class forming a distinct political actor. When the middle class is brought into our framework, we obtain a range of interesting results, some of them vindicating the emphasis that Moore and other scholars placed on the middle class.
The first role that the middle class can play in the emergence of democracy is as the driver of the process. Recall that in our framework, democracy emerges in response to a serious revolutionary threat or significant social unrest. The middle class can be the driver in this process by playing a key role in the revolutionary movement or by fueling and maintaining it. Almost all revolutionary movements were led by middle-class actors and, more important, a number of the major challenges to the existing regime; for example, the uprisings that helped induce the First Reform Act in Britain or those during the Paris Commune in France or the revolts of the Radical Party in Argentina were largely middle-class movements (see O’Donnell and Schmitter 1986, pp. 50-2, on the crucial role of the middle class in contemporary democratizations). Therefore, the middle class - by virtue of its more comfortable economic situation and the greater education of its members - can be a critical catalyst in the process toward democracy. This might also explain why many of the early moves toward democracy in Europe were only partial. If the middle class is the key actor, it may be sufficient for the elites to co-opt the middle class rather than concede a comprehensive democracy to all those who are excluded from the political system. The resulting picture resembles the gradual move toward democracy experienced in much of Western Europe: first, the middle classes are included in the political process and then the franchise is extended to the mass of citizens.
Perhaps the more important role of the middle class is that of a buffer in the conflict between the elites and the citizens. Recall that when the elites expect democracy to adopt policies highly unfavorable to them, they prefer repression to democratization. The presence of a large and relatively affluent middle class ensures that they play an important role in democratic politics and, because they are more prosperous than the citizens, they will typically support policies much closer to those that the elites prefer. Therefore, by limiting the amount of policy change induced by democracy, a large and affluent middle class may act like a buffer between the elites and the citizens in democracy. It does this by simultaneously making democratization more attractive for the elites than repression and changing policy enough that the citizens are content not to revolt.
The role of the middle class in the transition to democracy might help us understand the contrast between the political histories of Costa Rica and Colombia on the one hand and Guatemala, El Salvador, and Nicaragua on the other. Despite many similarities in their colonial histories and economic structures, these five countries have had very
different political trajectories (Paige 1997; Nugent and Robinson 2002). Costa Rica and Colombia have become stable albeit restricted democracies since the middle of the nineteenth century and successfully made the transition to effective universal suffrage in 1948 and 1936, respectively. Guatemala, El Salvador, and Nicaragua, on the other hand, were dominated by dictators in the nineteenth century and initial moves toward democracy - for example, in El Salvador in the late 1920s and in Guatemala between 1945 and 1954 - were snuffed out by coups and repression. These three societies made the transition to democracy very late. One important difference among these countries is that there is a relatively large and affluent middle class, especially smallholder coffee producers, in Costa Rica and Colombia but not in the other three. Perhaps as a consequence, democratic politics, once installed, has been much more conflict-ridden in Guatemala, El Salvador, and Nicaragua than in Costa Rica and Colombia.
6.6.2 Consolidation
The middle class may play an important role in consolidating democracy by limiting redistribution. A society with a large and affluent middle class will engage only in limited redistribution away from the elites toward the citizens and, therefore, provide a much smaller threat to the interests of the elites. This might be useful in understanding why many Western European and some Latin American societies, like Costa Rica and Colombia, with comparatively large middle classes have also had relatively stable democracies, whereas El Salvador and Guatemala, which lack such a middle-class buffer, have had difficulty consolidating democracy.
6.7 Globalization
There is no doubt that there are stronger economic links between nations today than forty years ago. Countries are more closely linked internationally today, with economic organizations such as the European Union, NAFTA, Mercosur, and Asean; there are much larger volumes of goods and services being traded, and much larger cross-border financial transactions. Do these major economic and political changes have implications for the circumstances under which democracy will arise and consolidate?
6.7.1 Democratization
Globalization might contribute to democratization in a number of distinct ways. First, international financial integration means that capital owners, the elites, can more easily take their money out of a given country. This makes it more difficult to tax the elites and reduces the extent to which democracy can pursue populist and highly majoritarian policies. International financial integration, therefore, makes the elites feel more secure about democratic politics and discourages them from using repression to prevent a transition from nondemocracy to democracy.
Second, international trade affects factor prices and, via this channel, modifies redistributive politics. Countries differ in their factor endowments, and the relative abundance of factors of production determines patterns of specialization and the impact of trade on relative prices. One implication of increased international trade is an increase in the rewards to the relatively abundant factor in each country. In the case of less developed nations-which are typically those still in nondemocracy today and, therefore, the main candidates for democratization - this means an increase in the rewards to labor. Intuitively, before the advent of significant trade flows, less developed countries had an excess of labor and a shortage of capital, depressing the rewards to labor and increasing those to capital. Trade opening will pull these rewards toward those prevailing in the rest of the world, thus increasing the rewards to labor and potentially reducing the return to capital. Trade opening will, therefore, reduce the gap between the incomes of labor and capital, thus changing the extent of inequality between capital owners and labor owners.
The specific implications of our framework depend on three things: (1) the nature of relative factor abundance; (2) the nature of political identities; and (3) where a country is on the inverted U-shaped relationship between inter-group inequality and democratization. Imagine that nondemocratic countries are labor abundant, political conflict is between rich capital-owning elites and poor labor-owning citizens, and inequality is sufficiently high that the elites use repression to stay in power. In this case, increased trade integration will reduce the extent of inequality between the elites and the citizens and will make democracy less redistributive. Because democracy will then be less threatening to the elites, they will be less inclined to use repression to avoid democracy. In such circumstances, globalization promotes democracy. Nevertheless, our framework does not imply that the impact of globalization on factor prices always promotes democracy. Let’s continue to postulate that conflict is between the rich and the poor and that we are on the part of the inverted U-shaped relationship where the rich use repression to stay in power. Now consider Latin American countries such as Argentina, Brazil, and Uruguay in the late nineteenth century. In these countries, the elites owned a lot of land and they were also land abundant. As predicted by the theory of international trade, pre-First World War globalization led to large increases in returns to land (O’Rourke and Williamson 1999). In our framework, this increases inter-group inequality and makes the elites less likely to democratize. It also increases the proportion of elite wealth invested in land, another factor that we suggest makes democracy more threatening to the elites. By the converse of these arguments, in this case, globalization would impede democratization (as long as we are on the part of the inverted U-shaped relationship where inequality discourages democratization, as assumed previously).
Third, increased international trade also means that disruption of economic activity may become more costly for many less developed nations that are now integrated into the world economy and, therefore, repression may now be much more costly for the elites, again favoring democracy.
Finally, increased political integration and the end of the Cold War (if not hijacked by the war against terrorism) might imply that countries that repress their citizens can perhaps expect stronger sanctions and reactions from the democratic world. This effectively increases the costs of repression, promoting democracy. This might be especially important because a number of nondemocratic regimes in the Cold War Era, such as Mobutu’s disastrous dictatorship in Zaire, were kept alive by the explicit or implicit support of the international community.
6.7.2 Consolidation
Just as globalization can induce democratization, so it can aid democratic consolidation. Indeed, all of the mechanisms listed that link increased globalization to democratization also imply that coups will be less likely. This is either because coups become more costly in a more integrated world or because globalization implies that democracy is less threatening to the elites.
7. Political Identities and the Nature of Conflict
Most of the comparative static results discussed so far do not depend on the identity of the elite; they apply even in societies where the nature of political conflict is not along class lines. In South Africa, race may be more salient, although race and socioeconomic class overlap to a large extent. In Rwanda, it may be more plausible to think of groups forming along the lines of ethnicity: Hutu or Tutsi. In Mauritius, political conflict has been between people of East Indian descent and a heterogeneous coalition of others, some of whom are rich (i.e., the white sugar planters and Chinese business elites) and some very poor (i.e., mostly the descendents of African slaves). In the latter case, there is no simple overlap between ethnicity or race and class (Bowman 1991).
As long as one accepts the premise that the interests of individuals are partly about economic outcomes, our basic analysis remains unaltered. Consider our ideas about political institutions. Here, we showed that if political institutions were such as to limit the type of policies that could occur in democracy, they tended to induce consolidated democracy. This result applies even in Mauritius. If institutions limit democracies, then they limit what the East Indian majority can do to the Creole minority. Hence, they reduce the incentive of a Creole dictatorship to repress democracy and, once democracy has been created, they make coups less attractive - exactly as in our previous analysis.
Next, consider the ideas we developed about the connection between the composition of the wealth of the elite and democratization or coups. These ideas apply immediately in this case. Even when politics is East Indian against Creole, as the economy develops and capital becomes more important than land, repression and coups become more costly and (pro-East Indian) democracy becomes less redistributive. As in our baseline analysis, this tends to create a consolidated democracy, even in Mauritius. Interestingly, Mauritius has been a consolidated democracy since independence, and this process of consolidation has taken place in the context of the radically declining importance of land, the rapid development of industry, and the expansion in the importance of human capital.
The nature of political identities may undoubtedly influence the form of collective choice under democracy, which ties our analysis to several important traditions in political science. For example, contrast a society where political identities and cleavages are on the basis of class with one where there are many cross-cutting cleavages or race, ethnicity, religion, or region. The pluralist model of democracy is one in which society is indeed divided into many different groups. A standard claim about a pluralistically based society is that it generates less income redistribution and smaller welfare states because the many different cleavages stop a broad coalition in favor of redistribution emerging. In consequence, for instance, pluralistic societies do not have strong socialist parties (Lipset and Marks 2000). If this is the case, then our theory suggests that such societies would be more likely to have consolidated democracy because elites would have little to fear from majority rule. This helps explain the longevity and stability of democracy in the United States, often thought to be the epitome of a pluralistic society.