F12(b) - Workshop 12: Voter Decision Making
Date: Jun 5 | Time: 02:00pm to 03:30pm | Location: SWING 107
Chair/Président/Présidente : Eline de Rooij (Simon Fraser University)
Discussant/Commentateur/Commentatrice : Eline de Rooij (Simon Fraser University)
Voter Rationality in the Shadow of Foreign Aid: Syeda ShahBano Ijaz (University of California San Diego)
Abstract: While driving down the I-5 in Southern California, I recently noticed a road sign that read, “your tax dollars are being spent here”. This form of accountability that draws a clear linkage between taxation and tangible public goods is starkly missing in developing countries: in part because there are not many tax dollars to go around and in part because voters may not reward the construction of six-lane highways. The median voter in economically and politically fragile countries has low income per capita, faces uncertainty about livelihood and pays either zero or minimal taxes. What, for such voters, is good governance? Do these voters shun aid dependency and clientelism, and reward investment in long run public goods projects? I argue that there is a pressing need to reconceptualize the standards of good governance in aid dependent economies as voters face uncertainty and time inconsistency. Using a survey experiment on a sample of undergraduate students in Pakistan, I develop a theory of voter preferences regarding public goods and political candidates. I find that when forced to choose, voters almost always prefer cash grant programs over tangible public goods. I also find a weak preference for political candidates who have a higher ability to secure foreign aid for their constituency. While voters prefer projects to be funded through the Government of Pakistan rather than a foreign aid agency, my overall results confirm that voters’ preferences are aligned with higher uncertainty and time inconsistency issues.
Experiments on Party Cue Effects: Does Party Identification Account for Treatment Effect Heterogeneity?: Eric Guntermann (University of California, Berkeley)
Abstract: In recent years, scholars have published a large number of experimental studies showing that, when citizens read policy positions and candidate endorsements from parties, they react by supporting the position or candidate of the party with which they identify. However, no one ever has ever shown that party identification accounts for differences in responses to party cues across respondents. I propose that positive and negative attitudes towards parties may play a role in accounting for the different ways people respond to information from parties. I reanalyze data from all experimental studies whose data are available and test whether party identification or party attitudes best moderate the receipt of party cue treatments. I expect to find heterogeneity in treatment effects that party identification fails to account for but that can be explained by party attitudes.
Do Voters Infer Positional Proximity to Candidates Based on Shared Local Ties?: Julia Schulte-Cloos (European University Institute), Paul Cornelius Bauer (Mannheim Centre for European Social Research)
Abstract: Does candidates' localness help them to attract votes? Integrating theories on the personal vote with approaches of informational shortcuts, this study posits that candidates fare better in electoral contexts in which they have local ties. We first test this proposition by relying on a novel dataset of all German candidates (N = 9306) that ran for office as direct candidates in the two most recent elections in Germany (2013, 2017). Exploiting the nature of the German two-tier electoral system and drawing on within electoral district and within candidate comparisons, we find support that candidates' localness substantively affects their vote share. This finding holds even in those districts where they have no chance of winning the single-member districts and voters hence cannot expect redistributional benefits from candidates. Second, we implement a survey experimental design as to assess whether candidates' localness serves as a heuristic for voters to infer positional proximity. We present respondents with a fictive ballot paper, thereby implicitly making the information on the localness of a candidate salient. We randomly assign a local candidate from the respective municipality of individuals (treatment condition) or no local candidate (control condition) to survey respondents by using information on their geolocation. The results show that voters are (a) more likely to participate in the election and (b) feel positionally closer to the local candidates. The findings bear crucial importance in advancing our understanding of the mechanisms accounting for the well-documented "friends-and-neighbour" effect.
Political Parties, Pressure Groups, and Group Pressures: A Search for Democracy’s Missing Demand Function: Mark Pickup (Simon Fraser University), Eric Groenendyk (University of Memphis), Erik O. Kimbrough (Chapman University)
Abstract: In the classical theory of democracy, interest groups organize to demand policy and parties organize to supply it. A common concern in identity-based theories of democracy is that a combination of collective action problems and strong partisan identities weakens the ability of interest groups to organize demand for policy and thus gives the parties power to control both the supply of and demand for policy. We suggest that membership in interest groups can also be a source of identity, and that the same kind of identity-driven motivations that encourage partisan behavior could potentially provide a counterbalance, restoring interest groups to their classical role. We design a series of experiments to test 1) whether partisan motivations can shape behavior, potentially driving individuals to act against their material interests; 2) whether the impact of partisan motivation is amplified in the presence of collective action problems; 3) whether competition between interest groups can offset some or all of the impact of partisan motivations; and 4) whether this effect is more pronounced for pre-existing identities than for minimal group identities.