close× Call Us
Title Explaining Declines in Clientelism through a Virtual-World Experimental Design
Post date 05/13/2015
C1 Background and Explanation of Rationale "How is Brazil at the turn of the 21st century like Great Britain in the 19th? What does the Roman Republic have in common with the U.S. at the turn of the 20th? In all of these places, clientelism—the exchange of money, goods, or in-kind benefits for votes between politicians and individual voters —was an important and prominent way of winning votes. Just like Americans and Britons, however, Brazilians do not vote the way they used to. Why do people who used to vote clientelistically stop voting on that basis? There is a tremendous amount of temporal and cross-sectional variation in the prevalence of clientelist voting and it largely reminds unexplained. The decline of clientelist voting is one of the more puzzling and persistent phenomena in the fields of democratization, political modernization, and political development, and it is my research question. I plan to use the Fulbright-Hays to gather qualitative data before, during, and after the 2016 mayoral and city council elections in Brazil so as to help answer this question in my dissertation. I am building a framework for analyzing and explaining the importance of clientelism as a determinant of vote choice. My primary unit of analysis is the individual voter and I use a multi-method approach with qualitative, cross-national survey, and experimental data. My argument is that increases in economic security lead to declines in clientelist voting, but only in the presence of certain systemic and individual-level factors that allow voters to easily identify non-clientelist candidates and distinguish them from one another. Most political science research on vote choice tends to be U.S.-centric and avoid even talking about clientelist voting, whereas the majority of clientelism research tends to overlook that it can coexist with other forms of voting. I take a synthetic approach, operating under the belief that clientelism needs to be treated like—and measured alongside—other common determinants of vote choice, such as policy, retrospective judgment, and candidate personality. Clientelist voting’s importance varies by country, voter, and election; it is one explanatory variable among many. My dissertation will provide an integrated framework within which I can estimate the importance of various determinants of vote choice at once. I will use this framework to explain why clientelist voting decreases when it does. To answer this question, I will conduct a within-video-game experiment in which I embed virtual elections featuring clientelist and non-clientelist candidates within an immersive virtual world and manipulate in-game variables such as income to determine whether players experimental treatments of income—impossible to do in the real world and impractical in a laboratory environment—alone affect clientelist voting. I will also manipulate contextual variables such as numbers of candidates. I argue that increases in economic security are the main force behind declines in clientelist voting, not the sole determinant. Increases in economic security are not alone sufficient to cause declines in clientelist voting; they are only a necessary condition. Without sufficient economic security, voters care primarily about basic goods. For those without money, food, or access to the state, providing them can win their votes. Other concerns, such as policy or candidate personality, tend to become ancillary when confronted with needs that are so immediate. Politicians see this opportunity, and if they can offer these things in exchange for votes, they do. When these basic needs are met, however, voters begin to want other things, such as self-expression, political participation, and the ability to influence policy. Yet while I agree with extant research that a meaningful decrease in poverty is a necessary condition for a decline in clientelist voting, I deviate from it by claiming that decreases in poverty alone do not constitute a sufficient condition. This is because a decline in clientelist voting requires the presence of another factor: a viable alternative to clientelist politicians. Even if not particularly satisfied with a status quo of clientelist candidates, it is difficult for a voter to desire a non-clientelist candidate if such an option is not realistic. The demand side of the voting equation, in other words, is inextricably linked with the question of supply. Supply, in turn, depends upon a variety of factors at both the individual and the macro level. An individual voter must also the political knowledge and interest necessary to identify alternative candidates and distinguish them from one another. Some voters will have neither the willingness nor the capacity to carry out these functions even in elections with two candidates, whereas some (if not many) will be able to do so in races with thousands. The macro level also matters because it helps determine how many candidates participate in each race. If one cannot compare the ideological positions, proposals, and personal characteristics of each candidate with one another, she will not deviate from a status quo of clientelist voting. This is a tougher test than it might appear. Electoral systems, for instance, matter because voters only have the willingness and ability to absorb a certain amount of information about political candidates. There is a tangible difference between a voter’s ability to understand the policy differences between two candidates and her ability to do so for a thousand."
C2 What are the hypotheses to be tested? "H1: As an individual's income increases, her likelihood of voting clientelistically decreases. H2: As the number of candidates for an office decreases, the likelihood of an individual voter voting clientelistically decreases. H3: As an individual's income increases and the number of candidates for an office decreases, her likelihood of voting clientelistically decreases (interactive effect)"
C3 How will these hypotheses be tested? * "These hypotheses will be tested with data collected from my interactive game environment, in which I will hold elections between non-clientelist and clientelist candidates in which the user must vote. After having held one election within my game environment, I will randomly assign a treatment (with an explanation within my storyline) to a certain population of players in which they receive a sizable boost of in-game currency. I will then hold another election and measure how many less players voted for a clientelist candidate. I will also randomly assign a second treatment at the beginning of the game in regards to the number of candidates the user must choose from in the election. One group will have 2 candidates for the post, one will have 4, and another will have 8. As a result, I expect that the income treatment will only have a statistically significant effect on vote decision in the games with 2 and 4 candidates, and not in the games with 8 candidates. I will use t-tests to measure the significance of my effects, as well as logistic regressions with various control variables (to measure robustness)."
C4 Country
C5 Scale (# of Units) not provided by authors
C6 Was a power analysis conducted prior to data collection? Data collection has not yet commenced
C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval? No
C8 IRB Number not provided by authors
C9 Date of IRB Approval not provided by authors
C10 Will the intervention be implemented by the researcher or a third party? Researchers
C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research? No
C12 If relevant, is there an advance agreement with the implementation group that all results can be published? No
C13 JEL Classification(s) not provided by authors