In the analysis of causal effects in non-experimental studies, conditioning on observable covariates is one way to try to reduce unobserved confounder bias. However, a developing literature has shown that conditioning on certain covariates may increase bias, and the mechanisms underlying this phenomenon have not been fully explored. We add to the literature on bias-increasing covariates by first introducing a way to decompose omitted variable bias into three constituent parts: bias due to an unobserved confounder, bias due to excluding observed covariates, and bias due to amplification. This leads to two important findings. Although instruments have been the primary focus of the bias amplification literature to date, we identify the fact that the popular approach of adding group fixed effects can lead to bias amplification as well. This is an important finding because many practitioners think that fixed effects are a convenient way to account for any and all group-level confounding and are at worst harmless. The second finding introduces the concept of bias unmasking and shows how it can be even more insidious than bias amplification in some cases. After introducing these new results analytically, we use constructed observational placebo studies to illustrate bias amplification and bias unmasking with real data. Finally, we propose a way to add bias decomposition information to graphical displays for sensitivity analysis to help practitioners think through the potential for bias amplification and bias unmasking in actual applications.

Suppose that an experimenter has collected a sample as well as baseline information about the units in the sample. How should she allocate treatments to the units in this sample? We argue that the answer does not involve randomization if we think of experimental design as a statistical decision problem. If, for instance, the experimenter is interested in estimating the average treatment effect and evaluates an estimate in terms of the squared error, then she should minimize the expected mean squared error (MSE) through choice of a treatment assignment. We provide explicit expressions for the expected MSE that lead to easily implementable procedures for experimental design.

When facing small numbers of observations or rare events, political scientists often encounter separation, in which explanatory variables perfectly predict binary events or nonevents. In this situation, maximum likelihood provides implausible estimates and the researcher might want incorporate some form of prior information into the model. The most sophisticated research uses Jeffreys’ invariant prior to stabilize the estimates. While Jeffreys’ prior has the advantage of being automatic, I show that it often provides too much prior information, producing smaller point estimates and narrower confidence intervals than even highly skeptical priors. To help researchers assess the amount of information injected by the prior distribution, I introduce the concept of a partial prior distribution and develop the tools required to compute the partial prior distribution of quantities of interest, estimate the subsequent model, and summarize the results.

In this article, we evaluate the usefulness of Google Consumer Surveys (GCS) as a low-cost tool for doing rigorous social scientific work. We find that its relative strengths and weaknesses make it most useful to researchers who attempt to identify causality through randomization to treatment groups rather than selection on observables. This finding stems, in part, from the fact that the real cost advantage of GCS over other alternatives is limited to short surveys with a small number of questions. Based on our replication of four canonical social scientific experiments and one study of treatment heterogeneity, we find that the platform can be used effectively to achieve balance across treatment groups, explore treatment heterogeneity, include manipulation checks, and that the provided inferred demographics may be sufficiently sound for weighting and explorations of heterogeneity. Crucially, we successfully managed to replicate the usual directional finding in each experiment. Overall, GCS is likely to be a useful platform for survey experimentalists.

Existing approaches to measuring political disagreement from text data perform poorly except when applied to narrowly selected texts discussing the same issues and written in the same style. We demonstrate the first viable approach for estimating legislator-specific scores from the entire speech corpus of a legislature, while also producing extensive information about the evolution of speech polarization and politically loaded language. In the Irish Dáil, we show that the dominant dimension of speech variation is government–opposition, with ministers more extreme on this dimension than backbenchers, and a second dimension distinguishing between the establishment and anti-establishment opposition parties. In the U. S. Senate, we estimate a dimension that has moderate within-party correlations with scales based on roll-call votes and campaign donation patterns; however, we observe greater overlap across parties in speech positions than roll-call positions and partisan polarization in speeches varies more clearly in response to major political events.

Bowers, Fredrickson, and Panagopoulos (2013, Reasoning about interference between units: A general framework, Political Analysis 21(1):97–124; henceforth BFP) showed that one could use Fisher's randomization-based hypothesis testing framework to assess counterfactual causal models of treatment propagation and spillover across social networks. This research note improves the statistical inference presented in BFP (2013) by substituting a test statistic based on a sum of squared residuals and incorporating information about the fixed network for the simple Kolmogorov–Smirnov test statistic (Hollander 1999, section 5.4) they used. This note incrementally improves the application of BFP's “reasoning about interference” approach. We do not offer general results about test statistics for multi-parameter causal models on social networks here, but instead hope to stimulate further, and deeper, work on test statistics and sharp hypothesis testing.

Legislative redistricting alters the political and electoral context for some voters but not others, thus offering a potentially promising research design to study many questions of interest in political science. We apply this design to study the effect that descriptive representation has on co-ethnic political engagement, focusing on Hispanic participation following California's 2000 redistricting cycle. We show that when redistrictors draw legislative boundaries in California's 1990, 2000, and 2010 apportionment cycles, they systematically sort higher-participating Hispanic voters into majority-Hispanic (MH) jurisdictions represented by co-ethnic candidates, biasing subsequent comparisons of Hispanic participation across districts. Similar sorting occurs during redistricting in Florida and Texas, though here the pattern is reversed, with less-participating Hispanic voters redistricted to MH districts. Our study highlights important heterogeneity in redistricting largely unknown or underappreciated in previous research. Ignoring this selection problem could significantly bias estimates of the effect of Hispanic representation, either positively or negatively. After we correct for these biases using a hierarchical genetic matching algorithm, we find that, in California, being moved to a district with an Hispanic incumbent has little impact on Hispanic participation in our data.