Earlier this month venerable political scientist Robert Dahl passed away at the age of 98. His work has important implications for the study of nationalism in the contemporary world – with the Ukrainian crisis being a case in point.

Best known for his later writings on American democracy, one of his most important works is a study in comparative politics, Polyarchy, published back in 1971. Based on a systematic study of the 34 democracies he identified operating in the world at that time (out of 104 countries examined), Dahl looked for the conditions which seemed conducive to the flourishing of democracy.

Dahl was defending a pluralist interpretation of democracy – the idea that democracy is best served if there are several fairly well-organized blocs of interests competing for power, with no single group guaranteed a majority in every election. Ideally, there would be several dimensions of cleavage (ethnicity, religion, region, class) which would be “cross-cutting,” so that someone might belong to one group on one issue, and another on a second.

As a concept “pluralism” peaked in the 1950s and fell out of favor in the 1960s, under attacks from the Left who argued that it failed to notice the systematic exclusion of some groups from the political process, such as blacks and women. The Right were not interested in pluralism since they focused on individual rights. Dahl came up with the new term “polyarchy” as a replacement for “pluralism,” but it never caught on.

In his 1971 work, Dahl was rather doubtful about whether ethnic pluralism was good for democracy. He divided the countries into four groups based on their degree of ethnic fragmentation, and he found that 58% of the 26 most ethnically homogeneous countries were democratic, compared to just 18% of the 33 most ethnically diverse countries. Dahl concluded that “a competitive political system is less likely in countries with a considerable measure of socio-cultural pluralism.” (p. 111)

There were nevertheless six countries in that most diverse group that were functioning polyarchies as of 1970 (Sierra Leone, Ceylon, Malaya, India, Canada and Switzerland); and three in the next group of fairly high ethnic pluralism countries (Belgium, Lebanon and the Philippines). Looking back 44 years later, we see that out of that short list of 9 ethnically diverse polyarchies, three collapsed in war (Sierra Leone, Lebanon and Ceylon); Belgium is frozen in political deadlock; the Philippines battles a persistent Muslim insurrection; and even Canada narrowly escaped the secession of Quebec in 1995.

The implication is clear: ethnic pluralism makes it hard, but not impossible, for democracy to thrive.

Empirical political scientists have yet to reach a consensus on whether the data shows a consistent relationship between ethnic fragmentation and democracy. As Mark Beissinger points out, the data on ethnic pluralism is inconsistent and it is hard to isolate the impact of ethnicity from other variables such as poverty. In a 2012 study Wolfgang Merkel and Brigitte Weiffen found that ethnic fragmentation was low in established democracies (0.28) but twice as high in failed democratizers (0.58) and authoritarian regimes (0.49). (Table 3) They conclude that “most facets of heterogeneity do not hinder democratic transition[but] most of them complicate democratic consolidation.” That is, ethnic pluralism may weaken autocratic regimes, but in the longer term may make it more difficult for a functioning democracy to emerge.

We see this sadly illustrated by the case of Ukraine. Ukraine as is well known is a country divided between a Europe-oriented West and a Russia-oriented East, with an ambiguous tranche running through the middle of the country. This ethnic pluralism seems to have aided the emergence of a quasi-democratic, parliamentary system of government in the 1990s – certainly a system that was more democratic than that in neighboring Russia. Lucan Way described this as “pluralism by default,” referring to the case of Moldova – which like Ukraine was split between two language groups, a Russophone minority and a Romanian-speaking majority.

However, short-run gains seem to come at the expense of long-term stability. The divided electorate meant that neither side could achieve a commanding majority in free elections, so power oscillated between the Westerners (who won the re-run election in 2004) and the Easterners (who won in 2010). Dahl warned back in 1970 that a binary division seems more unstable than having three or more groups, since this allows for coalition bargaining.

Looking at the rest of the post-soviet space, we see that democracy did well in the Baltic states, which were quite ethnically homogeneous. Lithuania was demographically homogeneous, while Estonia and Latvia were engineered to be homogenous by denying citizenship to the non-Estonians and Latvians who made up half their population. But it did badly in some other ethnically homogenous countries, such as Armenia and Azerbaijan. In the former Yugoslavia, democracy has been established in post-Yugo republics that were ethnically homogenous, but not in Bosnia, where the three rival ethnic groups have not been able to trust each other sufficiently to engage in the sort of bargaining that makes pluralist democracy work.

Advocates of democracy promotion tended to ignore ethnicity while assuming that a common national identity will magically appear once democratic institutions are installed. From Bosnia to Iraq, the evidence suggests that democratizers ignore national identity at their peril.