Lack of available data and of comparable cases has made hypothesis testing in democratization a difficult endeavor. Fish attempts to unravel some of the debate among adherents of cultural, socioeconomic, critical-junctural, procedural, and volunteerist perspectives. Although single case studies can generate hypothesis, only large N studies can adjudicate between rival claims. To overcome these problems, therefore, "medium-N" studies attempt a structured focused comparison of a handful of cases. They have been successful in rejecting some hypothesis and providing some advances, but they still lack statistical rigor and still suffer from the "many variables and few cases" problem.

          Another approach is the use of longitudinal studies to expand the number of cases and test causal variables, but this may lead to the inclusion of widely dissimilar cases such as democracies from an earlier time when the term might have carried a significantly different meaning than present day standards. For these reasons, he proposes studying postcommunist regime change because the area is large and encompasses a wide diversity of social, cultural and religious conditions. However, they all shared a remarkable similar set of similarities, including the time period of transition (1989-1992), the character of antecedent regimes, and economies governed by plans and commands. From similar beginning points, the development trajectories diverged dramatically. Because of the intense investigation of these countries, leading to a great deal of data, they offer useful material for study. Disadvantages include the inability to consider variables relevant in other time periods, and the conclusions may lack applicability to other regions or other transitions. Because the number of cases is still small, the number of variables is also still limited.

          Despite the debate over how to measure democracy, Fish concludes that a rough consensus has begun to emerge. This has become embodied in a survey, published in1997, headed by Alexander Motyl and including a group of experts from Freedom House. The criteria are also published in Kratnycky 1997. On this scale, 1 is complete democratization and 7 the least. Fish says that this is like the ratings used by Freedom House, which evaluates democratization on a global scale, but does not say what any differences might be, except for different scaling to provide finer distinctions between countries.

          Of the major hypotheses examined by Fish, the first is the economic requisites of democratization. The modernization thesis claims that countries with a more advanced human development index would enjoy advantages in democratization. In order to test the modernization thesis, he uses the HDI instead of GNP/capita. All of these countries are engaged in a dual transition.

          Another thesis claims that those countries that begin the transition with private ownership and the beginnings of a market economy have advantages. However, there is no consensus on the direction of effect on the relation between economic reform and democratization. According to the "economic-reform-as-threat-to-democratization" theorists, the pain of economic reform delegitimizes rulers and the imperative of reform forces them to ignore public demands. [For this view he cites Przeworski 1991: 136-191; 1992: 54-57; Callaghy 1994: 135-140) In this view, democracy and capitalism are based on contradictory principles. Capitalism runs on inequality, but democracy requires full equality before the law and one-person, one-vote (Milliband 1992:119-122; Dahl 1992: 83-87; Lindblom 1977: 161-200, 268-275).

          The opposing view claims that 1) a segment of the population involved in production independent of state control is essential for democracy; 2) both capitalism and democracy depend on individual choice; 3) that the separation of economic from political power under capitalism allows the formation of competing groups that balance each other; and 4) capitalism channels the individual need for recognition and dignity to be channeled into productive directions.

          In order to measure economic reform he uses the World Bank data from the 1996 report From Plan to Market, from which he devises a score on a scale of 1 for low reform to 10 for high reform.

          Under the socio-cultural dimension, he uses religion as a proxy for the essence of culture. He justifies this from the speculation that religions that are hierarchical and based on corporate affiliations (collectivism) are less supportive of democracy than those that permit egalitarianism and individualism. Some speculate that Eastern Christianity is more hierarchical, mystical and collectivist and therefore less hospitable to democracy. Others claim that Islam is not supportive of democracy because Islamic thought denies the legitimacy of an independent political and public sphere, separate from God's dominion, and that it is antagonistic to individualism (extensive citation here). For those who defend Islam he cites Enayat 1982:120; Espositio and Voll 1996: 6-51). To test these hypotheses, he divides the post-Soviet space into three regions for the dominant religious tradition. For his final set of factors to be tested, he uses ethnic heterogeneity. This derives from the belief that ethnic divisions create difficulties for democracy because political parties have difficulty appealing across ethnic divisions, so they become representative of one type of identity rather than representing malleable interests. Fish goes so far as to say that the destruction of popular rule by interethnic conflict is one of the most salient features of our present time. He supports this by referring to the repeated reversals of democratization in multiethnic African countries and the promotion of democracy in Bosnia as farcical. He measures ethnic homogeneity as the percentage of the population represented by the largest group. He claims that this is roughly the inverse of a fractionalization score (Taylor and Hudson 1972: 214-217, 271-74).

          According to this method of operationalization, there is only a moderate relation between the HDI value and democratic reform (R² = 0.36). This does not contradict, but neither does it bolster the modernization theory. However, economic reform is highly correlated with democratization with an adjusted (R² = 0.73), which is statistically significant. Ethnic homogeneity fails to reach statistical significance. "Third, national culture may well influence democratization, though its effects are neither unequivocal nor fully comprehensible" (230). He draws this conclusion because the coefficient for western Christianity fails statistical significance, but the coefficient for Islam is substantively as well as statistically significant. Western Christianity might have an influence on democratization because western Christianity and economic reform are fairly highly correlated (r = -0.74). Because of this collinearity, it is difficult to separate the causal effects. However, " . . . the substantive and statistical significance and the robustness of the Muslim variable show that 'national culture' cannot be ignored as a possible influence on democratization" (231).

          Next Fish considers the question whether democratization might be driving economic reform, rather than the other way around, or whether it is an interactive reinforcing relationship. Fish concludes that the move toward the market is consistent with movement toward democracy.

          Because the predominant religious tradition is a long-established feature of societies there is no endogeneity problem, nor can religion and democratization result from any separate third factor. Therefore, the first decade of post-communism does provide evidence that Muslim countries face a disadvantage in democratization. Differences in socioeconomic development cannot explain the divergence. The third major conclusion of this paper is the negative finding that socioeconomic development is not strongly related to democratization and not statistically significant when other variables are included. "Relative poverty, therefore, is not shown to be a drag on democratization; nor is relative plenty shown to ensure more progress toward democracy" (233). Although none of the countries has pursued a pure liberal economic policy, broad trends toward marketization are traceable. The final section attempts to explain this unusual feature of post-communist states.

          In other parts of the world, people who champion economic neo-liberalism are less liberal in purely political affairs. For instance in the US or Germany, political liberals tend toward social-democratic policies and economic liberals are conservative or even authoritarian. In the communist world the combination of closed authoritarian polities with command-plan economies has meant that those who favor change tended to favor both democratization and rapid marketization. In other words, he is writing that the "elective affinity between liberalism (and anti-liberalisms) in post-communist countries is not strictly a matter of principles and ideology. It has roots in the structure of social interests" (235).

          Intellectuals in Russia, Ukraine, Romania and other countries had been debating and promoting the virtues of a modernizing, market-friendly form of developmental authoritarianism patterned on Chile or South Korea. However, the post-Communist states lacked any upper class that had power in the market but opposed political participation. Therefore, a technocratic bureaucratic elite motivated by capitalism and political exclusion could not emerge. Fish attributes the lack of inspiring developmental leaders to the meager number and influence of those who would be the constituents, backers and agents of such a program. Therefore, economic reform has not created pressures for authoritarianism. Economic policy in all the autocratic regions is characterized by massive rent-seeking and predation by rulers rather than liberalization.

          Fish admits that the ideological elective affinities and the political capacities of social groups and elites are not the total explanation. Because the state has a near monopoly on employment in communist countries citizens have little possibility to express dissent because they depend so deeply on the state for subsistence. Therefore, the status of economic reform holds a greater determining influence on the possibilities for democratization because reform releases the individual from such dependence on the state. Barrington Moore's phrase, "No bourgeois, no democracy" has been debated extensively and the confuting evidence has been fairly modest. In contrast to the absence of any middle class in the communist world, in countries such as Spain, Portugal, South Africa, Brazil, Argentina and Taiwan, a stratum of independent property owners exercised considerable political influence. A substantial stratum of the population works in the nonstate sector and has disposable income. This category of people did not exist ten years ago, and the appearance of this group required economic transformation.

          Finally, Fish explains the relations between economic liberalization and democratization: both have become linked in the minds of the elite and electorates with escape from external control or Russian hegemony. In contrast to many other regions of the world, especially Africa, where capitalism is linked with imperial exploitation, in the post-communist world it is linked with national liberation. Whereas in Africa politicians who promote structural reform face stiff popular resistance, in the post-communist world promoters of marketization are rarely tossed out of power. "'Because democracy and capitalism are highly compatible does not mean that the transitions to them are' (Diamond, 1995: 114). In fact, even many prominent theorists who regard democracy and capitalism as compatible in the long run do not regard the transition to them as necessarily compatible" (242).