A Caltech Library Service

Three Essays in the Dynamics of Political Behavior


Kim, Seo-young Silvia (2020) Three Essays in the Dynamics of Political Behavior. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/jhvx-cb34.


In this thesis, I empirically assess the dynamics of political behavior. More specifically, I analyze what creates — or does not create — change in political participation, such as voting in elections and contributing to campaigns. Through this, I intend to show that paying close attention to dynamics can help answer fundamental questions of political behavior and offer important insights for real-life policies.

In Chapter 1, I focus on how non-political life events and election administration policy impact voter turnout. I analyze (1) the effect of moving on turnout over time and (2) how an election administration policy helps with the recovery of lowered turnout by lowering the re-registration burden of movers.

Moving depresses turnout by imposing various costs on voters. However, movers eventually settle down, and such detrimental effects can disappear over time. I analyze these dynamics using United States Postal Services (USPS) data and detailed voter panel data from Orange County, California. Using a generalized additive model, I show that previously registered voters who move close to the election are significantly less likely to vote (at most -16.2 percentage points), and it takes at least six months on average for turnout to recover. This dip-and-recovery is not observed for within-precinct moves, suggesting that costs of moving matter only when the voter's environment has changed much. I then evaluate an election administration policy that resolves their re-registration burden. This policy proactively tracks movers, updates their registration records for them, and notifies them by mailings. Using a natural experiment, I find that this policy is effective in boosting turnout (+5.9 percentage points). This success of a simple, pre-existing, and non-partisan safety net is promising, and I conclude by discussing policy implications.

Chapter 2 (published at Election Law Journal, doi: 10.1089/elj.2019.0593, coauthored with R. Michael Alvarez and Jonathan N. Katz) shows how the participation dynamics of political participation differ between two distinct classes of donors---hidden and visible (from data), based on their amount contributed. In campaign finance we find that there is something about the data generating process that is often overlooked, but which affects the interpretation of data greatly. This precedes Chapter 3 as it provides some important intuitions as to how the data should be filtered, wrangled, and interpreted for usage.

More specifically, inferences about individual campaign contributors are limited by how the Federal Election Commission (FEC) collects and reports data. Only transactions that exceed a cycle-to-date total of \$200 are individually disclosed, so that contribution histories of many donors are unobserved. We contrast visible donors and "hidden donors," or small donors who are invisible due to censoring and routinely ignored in existing research. I use the Sanders presidential campaign in 2016, whose unique campaign structure received money only through an intermediary (or conduit) committee. These are governed by stricter disclosure statutes, allowing us to study donors who are normally hidden. For the Sanders campaign, there were seven hidden donors for every visible donor, and altogether, hidden donors were responsible for 33.8\% of Sanders' campaign funds. We show that hidden donors start giving relatively later, with contributions concentrated around early primaries. We suggest that as presidential campaign strategies change towards wooing smaller donors, more research on what motivates them is necessary.

In Chapter 3, I focus on how events in the election cycle affect political behavior — this time, campaign contributions. I show how the aggregate behavior of campaign contributors is not affected as a function of election cycle dynamics and events.

Using the 2016 campaign finance data from the FEC as a daily time-series, I test the hypothesis that if presidential donors are either instrumental or momentum-driven, they will be responsive to events that reveal new information about candidate viability, such as early victories or unexpected upsets in primaries. I employ the sequential segmentation spline method to detect structural breaks while providing smooth estimates between the jumps. I find that on the national level, daily aggregates for any candidate is a slow-moving, smooth process, without any particular critical events. Even when data is disaggregated by state, events expected to create shocks hardly ever do, such as the Iowa caucus or the New Hampshire primary. This is also observed for a preliminary analysis of the 2020 contribution data. I conclude that campaign contributing is, in aggregate, a smooth process, and that donors are neither uniformly instrumental nor momentum-driven.

In all these chapters, my methodological contribution is in taking advantage of extremely large administrative datasets and harnessing the power of the large sample size with nonparametric and semiparametric methods. The rich world of nonparametric and semiparametric methods remains largely untapped by political science studies. I hope to show through this thesis that they can answer new questions, answer old questions in new ways, and provide strong insight that the default linearity model cannot provide.

Item Type:Thesis (Dissertation (Ph.D.))
Subject Keywords:Social sciences; political science; political behavior; political participation; political dynamics; turnout; elections; election administration; campaign finance; campaign contributions
Degree Grantor:California Institute of Technology
Division:Humanities and Social Sciences
Major Option:Social Science
Minor Option:Political Science
Thesis Availability:Public (worldwide access)
Research Advisor(s):
  • Katz, Jonathan N. (advisor)
  • Alvarez, R. Michael (co-advisor)
Thesis Committee:
  • Alvarez, R. Michael (chair)
  • Katz, Jonathan N.
  • Gibilisco, Michael B.
  • Shum, Matthew S.
Defense Date:13 May 2020
Non-Caltech Author Email:sskim.research (AT)
Funding AgencyGrant Number
John Randolph Haynes and Dora Haynes FoundationUNSPECIFIED
Lance Davis FellowshipUNSPECIFIED
Record Number:CaltechTHESIS:05182020-162259898
Persistent URL:
Related URLs:
URLURL TypeDescription to publication of Chapter 2.
Kim, Seo-young Silvia0000-0002-8801-9210
Default Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:13710
Deposited By: Seo Young Kim
Deposited On:01 Jun 2020 21:51
Last Modified:07 Dec 2020 16:38

Thesis Files

PDF - Final Version
See Usage Policy.


Repository Staff Only: item control page