View Full Document


Unformatted text preview:

Measuring ethnic voting Does proportional representation politicize ethnicity John D Huber March 24 2010 Abstract This paper develops a measure that can be used to compare ethnic voting levels across countries The measure examines the relationship between ethnicity and vote choice as it becomes easier to predict voting behavior in a country by knowing only voters ethnicity the ethnic voting measure increases The paper then uses data from 71 surveys to estimate ethnic voting levels in 45 countries and to examine the validity of the widely held assumption that proportional electoral laws lead to higher levels of ethnic voting The central finding is that all else equal proportional representation leads to less rather than more politicization of ethnicity By examining patterns of vote support across ethnic groups the paper suggests explanations for why PR often leads to less ethnic voting than is generally assumed I am grateful for research assistance from Tom Ogorzalek for research support from the National Science Foundation and for helpful comments from Kate Baldwin and Dawn Brancati I am also grateful to the Russell Sage Foundation where I was a visiting research scholar while writing this paper Professor Department of Political Science Columbia University and Visiting Scholar Russell Sage Foundation New York 1 Introduction Social scientists engage in robust debates about how to design democratic institutions for successful governance in ethnically divided societies and choosing the electoral law is widely held to be the most crucial decision that institutional engineers face There are sharp disagreements however about which electoral law is most appropriate Some scholars argue for proportional representation PR because it allows any group to have its own party thereby avoiding the frustration an individual would feel if his or her group is not representated e g Lijphart 1977 Lijphart 1999 Other scholars disagree arguing that the politicization of ethnicity occurring under PR is undesirable The goal instead should be to diffuse ethnicity by forcing parties to seek electoral coalitions that span different groups for example by adopting electoral rules that force vote pooling e g Horowitz 1985 and 1991 Despite deep disagreements about whether PR is a good idea in ethnically divided societies there is general agreement that it makes ethnic voting behavior more likely Since parties are easy to form under PR political elites can make appeals based on ethnicity and voters can choose parties that represent their groups So while scholars such as Lijphart and Horowitz disagree about whether facilitating ethnic voting behavior is a sensible thing to do they do not disagree that PR will politicize ethnicity in the electoral arena 1 Social scientists do not actually know however whether PR is associated with a heightened importance of ethnicity in voting behavior because there exists no measure that can be used to compare ethnic voting levels across countries Developing such a measure is important for adjudicating debates about how electoral institutions affect possibilities for stable governance in ethically divided societies And understanding the politicization of ethnicity in elections is important for other reasons as well Ethnic diversity is associated with a variety of governance issues in developed and developing democracies including lower levels of public goods provision higher corruption 1 See also Reilly and Reynolds 1999 Sisk and Reynolds 1998 and Tsbelis 1990 1 and slower economic growth Understanding the factors leading to the politicization of ethnicity can therefore shed light on the circumstances under which ethnic diversity causes governance problems in democracies This paper does not seek to adjudicate disputes about whether PR is a good choice in ethnically divided societies Instead the goal is to understand the degree to which ethnicity becomes politicized under different electoral laws which seems a crucial first step toward recommending particular institutional arrangements Specifically the paper has two objectives The first is to develop a measure of ethnic voting There are a variety of ways one could conceptualize ethnic voting including levels of support for ethnic parties the incidence of ethnic appeals by parties the propensity of ethnic group members to vote in the same way and the propensity of voters to support candidates from their own group to name several The measure developed here does not focus on ethnic parties candidates or appeals Instead it focuses on the relationship between ethnicity and vote choice As it becomes easier to predict voting behavior in a country by knowing only voters ethnicity the ethnic voting measure increases Focusing on the relationship between ethnicity and voting behavior is a sensible place to start in efforts to measure ethnic voting If ethnic appeals are strong or ethnic parties are present but voters do not let their ethnicity guide their vote it is doubtful that the politicization of ethnicity in the electoral arena exists to any meaningful extent And even if we do not observe explicit ethnic appeals by ethnic parties it is hard to deny the politicization of ethnicity if an individual s vote is accurately predicted by the individual s ethnicity While it is possible to measure ethnic voting at the group or party level the focus here is on measuring ethnic voting at the country level Without a country level measure it is difficult to study empirically the relationship between system level factors like the electoral law and the politicization of ethnicity And by focusing on a country level measure it is possible to take account of group size in measuring the politicization of ethnicity We should expect for example that if everyone from a particular ethnic group supported the same party and no one outside this 2 group supported the party then the degree to which this politicization of ethnicity is a problem in society should depend on the size of the group As a small group becomes larger the overall politicization of ethnicity in electoral politics should be said to increase Holding all else equal the measure proposed here increases as the size of groups becomes more equal The theoretical range of the measure is from 0 which occurs when the proportion of voters supporting each party is the same across groups to 1 which occurs when each group is represented by one party each group s party receives no support from outside the

Access the best Study Guides, Lecture Notes and Practice Exams

Loading Unlocking...

Join to view ethnic voting and access 3M+ class-specific study document.

We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view ethnic voting and access 3M+ class-specific study document.


By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?