1

Paper

Prior distributions for Bayesian data analysis in political science Gelman, Andrew Uploaded 02-25-2009 Keywords Bayesian inference hierarchical models mixture models prior information Abstract Prior information is often what makes Bayesian inference work. In the political science examples of which we are aware aware, information needs to come in, whether as regression predictors or regularization (that is, prior distributions) on parameters. We illustrate with a few examples from our own research.

2

Paper

The Power to Propose: A Natural Experiment in Politics Loewen, Peter Koop, Royce Fowler, James Uploaded 07-22-2009 Abstract In the study of democracy, an enduring question is whether citizens pay attention to what lawmakers do. Legislators frequently propose new laws, but observational studies cannot elucidate the effect such proposals have on citizen reactions to specific lawmakers, since any effects on electoral outcomes are confounded by unobserved individual differences in legislative and political skill. Here, we take advantage of a unique natural experiment in the Canadian House of Commons that allows us to estimate how the power to propose legislation affects elections. In the two most recent parliaments, the right of non-cabinet members to propose has been assigned by lottery. Comparing outcomes between those who were granted the right to propose and those who were not, we show that incumbents of the governing party enjoy a three and a half percentage point bonus in the electoral vote count following the allowed introduction of a single piece of legislation. This effect translates to a nine percent increase in the probability of winning the election. We also show that the causal effect does not result from media exposure or deterred entry of quality challengers who might otherwise have opposed the incumbent. Instead, government MPs who pass legislation receive more campaign donations, and money is associated with higher vote totals. These results are the first ever to show that what politicians do as lawmakers has a causal effect on the electorate.

3

Paper

Statistical Inference After Model Selection Berk, Richard Brown, Lawrence Zhao, Linda Uploaded 04-29-2009 Keywords Statistical Inference Model Selection Abstract Conventional statistical inference requires that a model of how the data were generated be known before the data are analyzed. Yet in criminology, and in the social sciences more broadly, a variety of model selection procedures are routinely undertaken followed by statistical tests and confidence intervals computed for a "final" model. In this paper, we examine such practices and show how they are typically misguided. The parameters being estimated are no longer well defined, and post-model-selection sampling distributions are mixtures with properties that are very different from what is conventionally assumed. Confidence intervals and statistical tests do not perform as they should. We examine in some detail the specific mechanisms responsible. We also offer some suggestions for better practice.

4

Paper

A Comparison of the Small-Sample Properties of Several Estimators for Spatial-Lag Count Models Hays, Jude Franzese, Robert Uploaded 07-22-2009 Keywords Interdependence Spatial Econometrics Spatial-Lag Models Count Data Poisson Nonlinear Least-Squares GMM Estimation Abstract Political scientists frequently encounter and analyze spatially interdependent count data. Applications include counts of coups in African countries, of state participation in militarized interstate disputes, and of bills sponsored by members of Congress, to name just a few. The extant empirical models for spatially interdependent counts and their corresponding estimators are, unfortunately, dauntingly complex, computationally costly, or both. They also generally tend 1) to treat spatial dependence as nuisance, 2) to stress spatial-error or spatial-heterogeneity models over spatial-lag models, and 3) to treat all observed spatial association as arising by one undifferentiated source. Prominent examples include the Winsorized count model of Kaiser and Cressie (1997) and Griffith�s spatially-filtered Poisson model (2002, 2003). Given the available options, the default approaches in most applied political-science research are to either to ignore spatial interdependence in count variables or to use spatially-lagged observed-counts as exogenous regressors, either of which leads to inconsistent estimates of causal relationships. We develop alternative nonlinear least-squares and method-of-moments estimators for the spatial-lag Poisson model that are consistent. We evaluate by Monte Carlo simulation the small sample performance of these relatively simple estimators against the naiive alternatives of current practice. Our results indicate substantial consistency improvements against minimal complexity and computational costs. We illustrate the model and estimators with an analysis of terrorist incidents around the world.

5

Paper

Modeling Dynamics in Time-Series-Cross-Section Political Economy Data Beck, Nathaniel Katz, Jonathan Uploaded 06-04-2009 Keywords dynamics TSCS political economy lagged dependent variable non-stationary Abstract This paper deals with a variety of dynamic issues in the analysis of time-series-cross-section (TSCS) data. While the issues raised are more general, we focus on applications to political economy. We begin with a discussion of specification and lay out the theoretical differences implied by the various types of time series models that can be estimated. It is shown that there is nothing pernicious in using a lagged dependent variable and that all dynamic models either implicitly or explicitly have such a variable; the differences between the models relate to assumptions about the speeds of adjustment of measured and unmeasured variables. When adjustment is quick it is hard to differentiate between the various models; with slower speeds of adjustment the various models make sufficiently different predictions that they can be tested against each other. As the speed of adjustment gets slower and slower, specification (and estimation) gets more and more tricky. We then turn to a discussion of estimation. It is noted that models with both a lagged dependent variable and serially correlated errors can easily be estimated; it is only OLS that is inconsistent in this situation. We then show, via Monte Carlo analysis shows that for typical TSCS data that fixed effects with a lagged dependent variable performs about as well as the much more complicated Kiviet estimator, and better than the Anderson-Hsiao estimator (both designed for panels).

6

Paper

Party Polarization in Congress: A Social Networks Approach Waugh, Andrew Pei, Liuyi Fowler, James Mucha, Peter Porter, Mason Uploaded 07-23-2009 Abstract We use the network science concept of modularity to measure polarization in the United States Congress. As a measure of the relationship between intra-community and extra-community ties, modularity provides a conceptually-clear measure of polarization that directly reveals both the number of relevant groups and the strength of their divisions. Moreover, unlike measures based on spatial models, modularity does not require predefined assumptions about the number of coalitions or parties, the shape of legislator utilities, or the structure of the party system. Importantly, modularity can be used to measure polarization across all Congresses, including those without a clear party divide, thereby permitting the investigation of partisan polarization across a broader range of historical contexts. Using this novel measure of polarization, we show that party influence on Congressional communities varies widely over time, especially in the Senate. We compare modularity to extant polarization measures, noting that existing methods underestimate polarization in periods in which party structures are weak, leading to artificial exaggerations of the extremeness of the recent rise in polarization. We show that modularity is a significant predictor of future majority party changes in the House and Senate and that turnover is more prevalent at medium levels of modularity. We utilize two individual-level variables, which we call "divisiveness" and "solidarity," from modularity and show that they are significant predictors of reelection success for individual House members, helping to explain why partially-polarized Congresses are less stable. Our results suggest that modularity can serve as an early-warning signal of changing group dynamics, which are reflected only later by changes in formal party labels.

7

Paper

Measuring the Effects of Voter Confidence on Political Participation Levin, Ines Alvarez, R. Michael Uploaded 06-22-2009 Keywords voter confidence turnout participation mexico matching causal effects Abstract In this paper we study the causal effect of voter confidence on participation decisions in the 2006 Mexican Election. Previous research has shown that voter confidence was a relevant factor in explaining participation during the years of the PRI hegemony. An open question is whether this relationship is still significant after the democratic transition taking place in the years 1997-2000. Moreover, in the previous literature, this problem was studied in a regression framework. In this article we argue that, since voter confidence and participation decisions are affected by similar covariates, a regression approach may lead to results which are too model dependent, and do not account for the heterogeneity of effects across voters. To solve this problem, we use matching methods, and find that voter confidence has considerable effects on participation decisions, but substantially different in magnitude from those found using the usual regression approach.

8

Paper

Competing Solutions to the Principal-Agent Model Haptonstahl, Stephen Uploaded 07-23-2009 Keywords bargaining principal agent risk aversion fairness quantal response equilibrium strategic statistical model random utility experiment Abstract Principal-Agent (PA) theory has been used for over three decades to model the relationship between an information-advantaged Agent and a Principal able to issue a contract ultimatum. For its common implementation as a game, the subgame-perfect Nash equilibrium is reasonably simple but generally wrong in predicting experimental or observational data. This paper implements PA theory theoretically and statistically as two kinds of strategic statistical model, then develops methods for testing competing behavioral hypotheses. I show that subgame-perfect Nash equilibrium, risk aversion/affinity, distributive justice/fairness theories, agent error, and random utility can be observationally distinct and how they might be distinguished statistically.

9

Paper

Agents and Outliers: Testing Organization with Committee Preference Expression Fortunato, David Uploaded 06-22-2009 Keywords ideal point estimation legislative organization theories of law making Abstract This paper offers a test of the three dominant schools of thought on the organization of the U.S. House of Representatives by revisiting the old question, Are committees composed of preference outliers? This study takes a new approach to the outlier question by explicitly assuming that the distribution of preferences among committee members varies from that of their colleagues on the floor. By making this assumption I free myself of the obligation of measuring ideology and focus instead on gauging the degree to which the committee crafted agenda allows these preference differences to be expressed --- or the degree to which committees are allowed to high jack the policy making process. Evaluating the latitude that committees take in setting the agenda allows me to assess not only the degree to which committee agents shirk from their principal but also the ability of the three dominant schools of thought on the organization of the U.S. House to predict legislative behavior. By generating jurisdiction specific estimates of agenda manipulation I find strong support for party dominated models of organization with a hierarchical ordering of agency in committee members and evidence for more outlier committees than previous research.

10

Paper

An Ecological Item-Response Model for Multiple Subsets of Respondents with Application to the European Court of Justice Malecki, Michael Uploaded 07-30-2009 Keywords ideal point estimation item response judicial politics ecological inference hierarchical model Abstract The European Court of Justice (ECJ) has fostered the development of a common European legal order, and in doing so, has asserted itself and its supremacy more, and more successfully, than any other international court. It has maintained features of international courts such as its composition of one judge per member state, while employing other tools of national high courts such as en banc decisions and organization into chambers, that together hide internal dissent and shield the ECJ from direct monitoring or curbing by the member states. The same shield has frustrated efforts to quantify the court's responsiveness to member states, with limited evidence that the ECJ yields to some member-state interest some of the time, but nonetheless has advanced integration beyond national governments' wishes. This equivocation arises at least in part from a failure to include relevant information about the court's composition and organization. In fact, the six-year renewable terms of judges, their previous qualifications and affiliations, and the internal organization into chambers all provide prior information that can and should be incorporated into a more complete model of judicial behavior. I develop an extension of the well-studied item-response model to infer judges' preferences, using the structured ecological data from the cases they heard and relevant prior information about judges and the national governments that appoint them as well as information about cases. I offer new, more rigorous tests for existing theoretical hypotheses about the ECJ's deference to certain actors and preference for integration. The model is applicable to other settings of structured ecological data. Many other national and international courts hear cases in subset chambers, and relevant prior information should be included rather than ignored in models of judicial behavior.

11

Paper

Estimation in Dirichlet Random Effects Models Kyung, Minjung Gill, Jeff Casella, George Uploaded 04-28-2009 Keywords generalized linear mixed model Dirichlet process random effects model precision parameter likelihood Gibbs sampling importance sampling probit mixed Dirichlet random effects model Abstract We develop a new Gibbs sampler for a linear mixed model with a Dirichlet process random effect term, which is easily extended to a generalized linear mixed model with a probit link function. Our Gibbs sampler exploits the properties of the multinomial and Dirichlet distribution, and is shown to be an improvement, in terms of operator norm and efficiency, over other commonly used MCMC algorithms. We also investigate methods for the estimation of the precision parameter of the Dirichlet process, finding that maximum likelihood may not be desirable, but a posterior mode is a reasonable approach. Examples are given to show how these models perform on real data. Our results complement both the theoretical basis of the Dirichlet process nonparametric prior and the computational work that has been done to date. Forthcoming: Annals of Statistics.

12

Paper

Balancing Competing Demands: Position-Taking and Election Proximity in the European Parliament Lindstaedt, Rene Slapin, Jonathan Vander Wielen, Ryan Uploaded 07-31-2009 Keywords Legislative Politics European Parliament Comparative Politics Bayesian IRT Parties Formal Theory Abstract Parties value unity, yet, members of parliament face competing demands, giving them incentives to deviate from the party. For members of the European Parliament (MEPs), these competing demands are national party and European party group pressures. Here, we look at how MEPs respond to those competing demands. We examine ideological shifts within a single parliamentary term to assess how European Parliament (EP) election proximity affects party group cohesion. Our formal model of legislative behavior with multiple principals yields the following hypothesis: When EP elections are proximate, national party delegations shift toward national party positions, thus weakening EP party group cohesion. For our empirical test, we analyze roll call data from the fifth EP (1999-2004) using Bayesian item response models. We find significant movement among national party delegations as EP elections approach, which is consistent with our theoretical model, but surprising given the existing literature on EP elections as second-order contests.

13

Paper

We're Not Lost, But How Did We Get Here? Jackson, John Uploaded 07-07-2009 Keywords Society for Political Methodology history Abstract This narrative recounts the beginning and early years of the Society for Political Methodology and what an initially small group of young, naive and energetic scholars did to and for Political Science.

14

Paper

The MAOA Gene Predicts Credit Card Debt De Neve, Jan-Emmanuel Fowler, James Uploaded 08-18-2009 Abstract This article presents the first evidence of a specific gene predicting real world economic behavior. Using data from the National Longitudinal Study of Adolescent Health, we show that individuals with a polymorphism of the MAOA gene that has lower transcriptional efficiency are significantly more likely to report having credit card debt. Having one or both MAOA alleles of the low efficiency type raises the average likelihood of having credit card debt by 7.8% and 15.9% respectively. About half of our population has one or both MAOA alleles of the low type. Prior research has linked this genetic variation to lack of conscientiousness, impulsivity, and addictive behavior.

15

Paper

We're Not Lost, But How Did We Get Here? Appended Documents Jackson, John Uploaded 07-07-2009 Keywords Society for Political Methodology history Abstract These are various documents to be appended to the narrative on the beginning of the Society for Political Methodology.

16

Paper

Quantitative Discovery from Qualitative Information: A General-Purpose Document Clustering Methodology King, Gary Grimmer, Justin Uploaded 07-19-2009 Keywords unsupervised learning discovery content analysis Abstract Many people attempt to discover useful information by reading large quantities of unstructured text, but because of known human limitations even experts are ill-suited to succeed at this task. This difficulty has inspired the creation of numerous automated cluster analysis methods to aid discovery. We address two problems that plague this literature. First, the optimal use of any one of these methods requires that it be applied only to a specific substantive area, but the best area for each method is rarely discussed and usually unknowable ex ante. We tackle this problem with mathematical, statistical, and visualization tools that define a search space built from the solutions to all previously proposed cluster analysis methods (and any qualitative approaches one has time to include) and enable a user to explore it and quickly identify useful information. Second, in part because of the nature of unsupervised learning problems, cluster analysis methods are not routinely evaluated in ways that make them vulnerable to being proven suboptimal or less than useful in specific data types. We therefore propose new experimental designs for evaluating these methods. With such evaluation designs, we demonstrate that our computer-assisted approach facilitates more efficient and insightful discovery of useful information than either expert human coders using qualitative or quantitative approaches or existing automated methods. We (will) make available an easy-to-use software package that implements all our suggestions.

17

Paper

The Importance of Fully Testing Conditional Theories Positing Interaction Golder, Matt Berry, William Milton, Daniel Uploaded 07-10-2009 Abstract In recent years, it has become common for political scientists to present marginal effect plots when interpreting results from interactive models. This has led to a dramatic improvement in the quality of research testing conditional theories. The typical practice is to (i) view one of the variables expected to interact, say Z, as the conditioning variable, (ii) offer a hypothesis about how the marginal effect of the other variable, X, is conditional on the value of Z, and (iii) construct a plot of the relationship between Z and the estimated marginal effect of X. All interactions are symmetric, though; when the effect of X is conditional on Z, the effect of Z must be conditional on X. In this paper, we illustrate that the failure of scholars to provide a second hypothesis about how the marginal effect of Z is conditional on the value of X, together with the corresponding marginal effect plot, means that scholars often subject their conditional theories to substantially weaker empirical tests than their data allow. The result is that much of the existing literature either understates or, more worryingly, overstates the empirical support for the conditional theories that political scientists have posited.

18

Paper

Bayesian statistical decision theory and a critical test for substantive significance Esarey, Justin Uploaded 09-09-2009 Keywords inference t-test substantive significance Bayesian Abstract I introduce a new critical test statistic, c*, that uses Bayesian statistical decision theory to help an analyst determine whether quantitative evidence supports the existence of a substantively meaningful relationship. Bayesian statistical decision theory takes a rational choice perspective toward evidence, allowing researchers to ask whether it makes sense to believe in the existence of a statistical relationship given how they value the consequences of correct and incorrect decisions. If a relationship of size c* is not important enough to influence future research and policy advice, then the evidence does not support the existence of a substantively significant effect. A replication of findings from the American Journal of Political Science and Journal of Politics illustrates that statistical significance at conventional levels is neither necessary nor sufficient to accept a hypothesis of substantive significance using c*. I also make software packages available for Stata and R that allow political scientists to easily use c* for inference in their own research.

19

Paper

Spike and Slab Prior Distributions for Simultaneous Bayesian Hypothesis Testing, Model Selection, and Prediction, of Nonlinear Outcomes Pang, Xun Gill, Jeff Uploaded 07-13-2009 Keywords Spike and Slab Prior Hypothesis Testing Bayesian Model Selection Bayesian Model Averaging Adaptive Rejection Sampling Generalized Linear Model Abstract A small body of literature has used the spike and slab prior specification for model selection with strictly linear outcomes. In this setup a two-component mixture distribution is stipulated for coefficients of interest with one part centered at zero with very high precision (the spike) and the other as a distribution diffusely centered at the research hypothesis (the slab). With the selective shrinkage, this setup incorporates the zero coefficient contingency directly into the modeling process to produce posterior probabilities for hypothesized outcomes. We extend the model to qualitative responses by designing a hierarchy of forms over both the parameter and model spaces to achieve variable selection, model averaging, and individual coefficient hypothesis testing. To overcome the technical challenges in estimating the marginal posterior distributions possibly with a dramatic ratio of density heights of the spike to the slab, we develop a hybrid Gibbs sampling algorithm using an adaptive rejection approach for various discrete outcome models, including dichotomous, polychotomous, and count responses. The performance of the models and methods are assessed with both Monte Carlo experiments and empirical applications in political science.

20

Paper

Characterizing the variance improvement in linear Dirichlet random effects models Kyung, Minjung Gill, Jeff Casella, George Uploaded 09-11-2009 Keywords Dirichlet processes mixture models Bayesian nonparametrics Abstract An alternative to the classical mixed model with normal random effects is to use a Dirichlet process to model the random effects. Such models have proven useful in practice, and we have observed a noticeable variance reduction, in the estimation of the fixed effects, when the Dirichlet process is used instead of the normal. In this paper we formalize this notion, and give a theoretical justification for the expected variance reduction. We show that for almost all data vectors, the posterior variance from the Dirichlet random effects model is smaller than that form the normal random effects model. Forthcoming: Statistics and Probability Letters

21

Paper

Regression Adjustments to Experimental Data: Do David Freedmanâ??s Concerns Apply to Political Science? Green, Donald Uploaded 07-15-2009 Keywords Experiments Regression Covariates Analysis of Covariance Abstract Abstract: One of David Freedman's important legacies was to raise awareness of the assumptions that underlie everyday statistical practice, such as regression analysis. His recent papers (Freedman 2008a, 2008b) offer stern warnings to those who offer regression analysis as an appropriate way to analyze experimental results. In particular, Freedman demonstrates that including pre-treatment covariates as controls leads to bias in finite samples and inaccurate standard errors. Freedman advises researchers against using regression adjustments for experiments involving fewer than 500 observations (2008a, p.191), a recommendation that has gained increasing attention and acceptance among social scientists. This paper argues that the ever-cautious Freedman was probably too cautious in his recommendations. After explicating the special features of Freedman's model, I use a combination of simulated and actual examples to show that as a practical matter the biases that Freedman pointed out tend to be negligible for N > 20. Pathological cases that could generate biases for larger experiments involve extreme outliers that would be readily detected through visual inspection.

22

Paper

When Mayors Matter: Estimating the Impact of Mayoral Partisanship on City Policy Gerber, Elisabeth Hopkins, Daniel Uploaded 09-18-2009 Keywords Regression discontinuity design partisanship urban fiscal policy Abstract U.S. cities are limited in their ability to set policy. Can these constraints mute the impact of mayorsâ?? partisanship on policy outcomes? We hypothesize that mayoral discretion--and thus partisanshipâ??s influence--will be more pronounced in policy areas where there is the less shared authority between local, state, and federal governments. To test this hypothesis, we create a novel data set combining U.S. mayoral election returns from 1990 to 2006 with urban fiscal data. Using regression discontinuity design, we find that cities that elect a Democratic mayor spend less on public safety, a policy area where local discretion is high, than otherwise similar cities that elect a Republican or Independent. We find no differences on tax policy, social policy, and other areas that are characterized by significant overlapping authority. These results have important implications for political accountability: mayors may not be able to influence the full range of policies that are nominally local responsibilities.

23

Paper

An Observational Study of Ballot Initiatives and State Outcomes Keele, Luke Uploaded 07-17-2009 Keywords causal inference matching ballot initiatives voter turnout difference-in-differences Abstract It has long been understood that the presence of the ballot initiative process leads to different outcomes among states. In general, extant research has found that the presence of ballot initiatives tends to increase voter turnout and depress state revenues and expenditures. I reconsider this possibility and demonstrate that past findings are an artifact of incorrect research design. Failure to account for differences in states often leads to a confounding association between ballot initiatives and voter turnout and fiscal policy. Here, I conduct an observational study based on a counterfactual model of inference to analyze the effects of ballot initiatives. The resulting research design leads to two analyses. First, I utilize the synthetic case control method, which allows me to compare over time outcomes in states with initiatives to states without initiatives while accounting for pretreatment baseline differences across states. Second, I use matching to assess voter turnout differences across metro areas along state boundaries with and without ballot initiatives. In both analyses, I find that ballot initiatives rarely have spillover effects on voter turnout and state fiscal policy.

24

Paper

From Nature to the Lab: The Methodology of Experimental Political Science and the Study of Causality Morton, Rebecca Williams, Kenneth Uploaded 09-18-2009 Keywords experiments causality Abstract In this manuscript we review the methodology of experimental political science and the study of causality.

25

Paper

How Prediction Markets can Save Event Studies Snowberg, Erik Wolfers, Justin Zitzewitz, Eric Uploaded 01-16-2009 Abstract Abstract Event studies have been used to address a variety of political questions -- from the economic effects of party control of government to the importance of complex rules in congressional committees. However, the results of event studies are notoriously sensitive to both choices made by researchers and external events. Specifically, event studies will generally produce different results depending on three interrelated things: which event window is chosen, the prior probability assigned to an event at the beginning of the event window, and the presence or absence of other events during the event window. In this paper we show how each of these may bias the results of event studies, and how prediction markets can mitigate these biases.

26

Paper

Identification, Inference, and Sensitivity Analysis for Causal Mediation Effects Imai, Kosuke Keele, Luke Yamamoto, Teppei Uploaded 07-20-2009 Keywords causal inference causal mediation analysis direct and indirect eects linear structural equation models sequential ignorability unmeasured confounders Abstract Causal mediation analysis is routinely conducted by applied researchers in a variety of disciplines including epidemiology, political science, psychology, and sociology. The goal of such an analysis is to investigate alternative causal mechanisms by examining the roles of intermediate variables that lie in the causal path between the treatment and outcome variables. In this paper, we first prove that under a particular version of sequential ignorability assumption, the average causal mediation effect (ACME) is nonparametrically identified. We compare our identifying assumption with those proposed in the literature. Some practical implications of our identification result are also discussed. In particular, the popular estimator based on the linear structural equation model (LSEM) can be interpreted as an ACME estimator if the linearity and no-interaction assumptions are satisfied in addition to the proposed assumption. We show that this assumption can easily be relaxed within the framework of LSEM. Second, we consider a simple nonparametric estimator of the ACME in order to relax distributional and functional form assumptions. We also discuss a more general nonparametric approach. Third, we propose a new sensitivity analysis that can be easily implemented by applied researchers within the standard LSEM framework. Like the existing identifying assumptions, the proposed assumption may be too strong in many applied settings. Thus, sensitivity analysis is essential in order to examine the robustness of empirical findings to the possible existence of an unmeasured confounder. Finally, we apply the proposed methods to a randomized experiment from political psychology.

27

Paper

Why Process Matters for Causal Inference Glynn, Adam Quinn, Kevin Uploaded 07-13-2009 Keywords causal effect process post-treatment mechanism mediation potential outcomes Abstract It is often assumed that the only way to assess the causal effects of an explanatory variable on an outcome variable is to compare the outcomes from units with differing values of the explanatory variable. In this paper, we provide a formal account of how within-unit causal process information (i.e., knowledge of the causal chain linking an explanatory variable to an outcome variable) can be used to make certain types of causal inferences without comparing outcomes from units with differing values of the explanatory variable. The methods discussed in this paper allow causal researchers to make full use of causal information that many had heretofore ignored. At the same time, because these methods are embedded in a Bayesian potential outcomes causal model, researchers are held to high standards of transparency and logical consistency. We illustrate these methods with an application to the effects of election day registration on African American turnout. This analysis shows that previous regression or matching estimates for these effects are likely overstated.

28

Paper

The 2008 Presidential Primaries through the Lens of Prediction Markets Malhotra, Neil Snowberg, Erik Uploaded 01-16-2009 Abstract Abstract To explore the influence of primary results during the 2008 nomination process we leverage a previously unused methodology --- the analysis of prediction market contracts. The unique structure of prediction markets allows us to address two unexplored questions. First, we analyze whether primary results affect candidates' chances in the general election, as candidates who take strong positions during the nomination contest may be unable to easily appeal to centrist voters in the general election. We also assess whether states with early primaries, such as Iowa and New Hampshire, have a disproportionate effect on the nominating process. We show that the length of the primary season has a minimal impact of the electability of candidates in the general election, and that some states have a disproportionate impact on the nominating process. However, the states that have the largest impact are not necessarily New Hampshire and Iowa, the states that have often been assumed to be the most influential because of their early position on the primary calendar.

29

Paper

A General Approach to Causal Mediation Analysis Imai, Kosuke Keele, Luke Tingley, Dustin Uploaded 07-20-2009 Keywords causal inference causal mechanisms sensitivity analysis sequential ignorability structural equation modeling unobserved confounder Abstract In a highly influential paper, Baron and Kenny (1986) proposed a statistical procedure to conduct a causal mediation analysis and identify possible causal mechanisms. This procedure has been widely used across many branches of the social and medical sciences and especially in psychology and epidemiology. However, one major limitation of this approach is that it is based on a set of linear regressions and cannot be easily extended to more complex situations that are frequently encountered in applied research. In this paper, we propose an approach that generalizes the Baron-Kenny procedure. Our method can accommodate linear and nonlinear relationships, parametric and nonparametric models, continuous and discrete mediators, and various types of outcome variables. We also provide a formal statistical justification for the proposed generalization of the Baron-Kenny procedure by placing causal mediation analysis within the widely-accepted counterfactual framework of causal inference. Finally, we develop a set of sensitivity analyses that allow applied researchers to quantify the robustness of their empirical conclusions. Such sensitivity analysis is important because as we show the Baron-Kenny procedure and our generalization of it rest on a strong and untestable assumption even in randomized experiments. We illustrate the proposed methods by applying them to a randomized field experiment, the Job Search Intervention Study (JOBS II). We also offer easy-to-use software that implements all of our proposed methods.

30

Paper

Tobler's Law, Urbanization, and Electoral Bias: Why Compact, Contiguous Districts are Bad for the Democrats Chen, Jowei Rodden, Jonathan Uploaded 11-11-2009 Keywords elections voting party competition legislative districting simulations electoral geography spatial autocorrelation Abstract When one of the major parties in the United States wins a substantially larger share of the seats than its vote share would seem to warrant, the conventional explanation lies in manipulation of maps by the party that controls the redistricting process. Yet this paper uses a unique data set from Florida to demonstrate a common mechanism through which substantial partisan bias can emerge purely from residential patterns. When partisan preferences are spatially dependent and partisanship is highly correlated with population density, any districting scheme that generates relatively compact, contiguous districts will tend to produce bias against the urban party. In order to demonstrate this empirically, we apply automated districting algorithms driven solely by compactness and contiguity parameters, building winner-take-all districts out of the precinct-level results of the tied Florida presidential election of 2000. The simulation results demonstrate that with 50 percent of the votes statewide, the Republicans can expect to win around 59 percent of the seats without any "intentional" gerrymandering. This occurs because urban districts tend to be homogeneous and Democratic while suburban and rural districts tend to be moderately Republican. Thus in Florida and other states where Democrats are highly concentrated in cities, the seemingly apolitical practice of requiring compact, contiguous districts will produce systematic pro-Republican electoral bias.

31

Paper

The Split Population Logit (SPopLogit): Modeling Measurement Bias in Binary Data Beger, Andreas DeMeritt, Jacqueline Moore, Will Hwang, Wonjae Uploaded 02-03-2009 Keywords binary limited dependent variables measurement bias unobservability Abstract This study describes a split population logit model that can be useful to researchers who are modeling a binary dependent variable that is measured with a biased instrument. To motivate the study we identify two common, yet widely unrecognized, circumstances in which political scientists are likely to study dichotomous variables that have been measured with bias. In one such setting (e.g., surveys) the strategic interests of actors will lead them to misrepresent an attitude or behavior. In another such setting (e.g., content analysis of events) researchers' instruments are unable to distinguish between the absence of a characteristic or event and missing data. We briefly argue that "unobservability," "zero-inflated," and other models form a single class of models that allow researchers to model the bias in operational instruments, and thus not only correct bias in statistical inference but, more importantly, produce theoretical accounts of the bias and then test the hypotheses that those accounts imply. We derive the likelihood function for the split population logit model, describe the properties of its MLEs, present the results from a Monte Carlo study, and briefly describe code that researchers can use to implement the model in the Stata statistical package.

32

Paper

Causal Mediation Analysis in R Imai, Kosuke Keele, Luke Tingley, Dustin Yamamoto, Teppei Uploaded 07-20-2009 Abstract Causal mediation analysis is widely used across many disciplines to investigate possible causal mechanisms. Such an analysis allows researchers to explore causal pathways, going beyond the estimation of simple causal effects. Recently, Imai, Keele and Yamamoto (2008) and Imai, Keele, and Tingley (2009) developed general algorithms to estimate causal mediation effects with the variety of data types that are often encountered in practice. The new algorithms can estimate causal mediation effects for linear and nonlinear relationships, with parametric and nonparametric models, with continuous and discrete mediators, and various types of outcome variables. In this paper, we show how to implement these algorithms in the statistical computing language R. Our easy-to-use software, mediation, takes advantage of the object-oriented programming nature of the R language and allows researchers to estimate causal mediation effects in a straightforward manner. Finally, mediation also implements sensitivity analyses which can be used to formally assess the robustness of findings to the potential violations of the key identifying assumption. After describing the basic structure of the software, we illustrate its use with several empirical examples.

33

Paper

Answer Key (Odd Numbers): Esssential Mathematics for Political and Social Research Gill, Jeff Uploaded 05-02-2009 Keywords answer key mathematics for political science trigonometry calculus matrix algebra probability and set theory Markov chains Abstract Hopefully this post is not too self-promotional. I get lots of requests for the answer key to ESSENTIAL MATHEMATICS FOR POLITICAL AND SOCIAL RESEARCH (Cambridge University Press 2006), which the press has restricted to teaching faculty. Recently, however, I have recently received permission to make generally available the answers worked-out in detail for odd numbered problems. This file provides these.

34

Paper

Sampling Schemes for Generalized Linear Dirichlet Random Effects Models Kyung, Minjung Gill, Jeff Casella, George Uploaded 02-18-2009 Keywords generalized linear mixed Dirchlet model Markov chain Monte Carlo Dirichlet process priors for random effects precision parameters Scottish Social Attitudes Survey terrorism targeting Abstract We evaluate MCMC sampling schemes for a variety of link functions in generalized linear models with Dirichlet random effects. We find that models using Dirichlet process priors for the random effects tend to capture information in the data in a nonparametric fashion. In fitting the the Dirichlet process, dealing with the precision parameter has troubled model specifications in the past. Here we find that incorporating this parameter into the MCMC sampling scheme is not only computationally feasible, but also results in a more robust set of estimates, in that they are marginalized-over rather than conditioned-upon. Applications are provided with social science problems in areas where the data can be difficult to model. In all, we find that these models provide superior Bayesian posterior results in theory, simulation, and application.

35

Paper

Joint Modeling of Dynamic and Cross-Sectional Heterogeneity: Introducing Hidden Markov Panel Models Park, Jong Hee Uploaded 07-14-2009 Keywords Bayesian statistics Fixed-effects Hidden Markov models Markov chain Monte Carlo methods Random-effects Reversible jump Markov chain Monte Carlo Abstract Researchers working with panel data sets often face situations where changes in unobserved factors have produced changes in the cross-sectional heterogeneity across time periods. Unfortunately, conventional statistical methods for panel data are based on the assumption that the unobserved cross-sectional heterogeneity is time constant. In this paper, I introduce statistical methods to diagnose and model changes in the unobserved heterogeneity. First, I develop three combinations of a hidden Markov model with panel data models using the Bayesian framework; (1) a baseline hidden Markov panel model with varying fixed effects and varying random effects; (2) a hidden Markov panel model with varying fixed effects; and (3) a hidden Markov panel model with varying intercepts. Second, I present model selection methods to diagnose the dynamic heterogeneity using the marginal likelihood method and the reversible jump Markov chain Monte Carlo method. I illustrate the utility of these methods using two important ongoing political economy debates; the relationship between income inequality and economic growth and the effect of institutions on income inequality.

36

Paper

Language Access and Initiative Outcomes: Did the Voting Rights Act Influence Support for Bilingual Education? Uploaded 12-17-2009 Keywords regression discontinuity design multilevel modeling immigrant political incorporation language access elections Voting Rights Act Abstract This paper investigates one tool designed to enfranchise immigrants: foreign-language election materials. Specifically, it estimates the impact of Spanish-language assistance provided under Section 203 of the Voting Rights Act. Focusing on a California initiative on bilingual education, it tests how Spanish-language materials influenced turnout and election outcomes in Latino neighborhoods. It also considers the possibility of an anti-Spanish backlash in non-Hispanic white neighborhoods. Empirically, the analysis couples a regression discontinuity design with multilevel modeling to isolate the impact of Section 203. The analysis finds that Spanish-language assistance increased turnout and reduced support for ending bilingual education in Latino neighborhoods with many Spanish speakers. It finds hints of backlash among non-Hispanic white precincts, but not with the same certainty. The turnout finding gains additional support from multilevel regression discontinuity analyses of 2004 Latino voter turnout nationwide. For Latino citizens who speak little English, the availability of Spanish ballots increases turnout and influences election outcomes as well.