Douglas Rivers

Selection Bias in Linear Regression, Logit and Probit Models

Missing data are common in observational studies due to self-selection of subjects. Missing data can bias estimates of linear regression and related models. The nature of selection bias and econometric methods for correcting it are described. The econometric approach relies upon a specification of the selection mechanism. We extend this approach to binary logit and probit models and provide a simple test for selection bias in these models. An analysis of candidate preference in the 1984 U.S. presidential election illustrates the technique.

The 2006 Cooperative Congressional Election Study

Abstract In 2006 Polimetrix, Inc. of Palo Alto, CA. fielded the Cooperative Congressional Election Study, the largest study of Congressional elections ever fielded in the US. The project was a joint venture of 38 universities and over 100 political scientists. In this paper, we detail the design and execution of the project, with special attention to the method by which the sample was generated. We show that the estimates from the Common Content of CCES outperform conventional estimates based on RDD phone surveys. We also argue that opt‐in panels, internet surveys, and cooperative ventures like CCES provide cost‐effective alternatives for social scientists under certain conditions. These types of surveys can provide reductions in RMSE over conventional methods when sample matching is used to ameliorate the biases that come with sampling from an opt‐in panel.

Institutions in International Relations: Understanding the Effects of the GATT and the WTO on World Trade

The General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO) have been touted as premier examples of international institutions, but few studies have offered empirical proof. This article comprehensively evaluates the effects of the GATT/WTO and other trade agreements since World War II. Our analysis is organized around two factors: institutional standing and institutional embeddedness. We show that many countries had rights and obligations, or institutional standing, in the GATT/WTO even though they were not formal members of the agreement. We also expand the analysis to include a range of other commercial agreements that were embedded with the GATT/WTO. Using data on dyadic trade since 1946, we demonstrate that the GATT/WTO substantially increased trade for countries with institutional standing, and that other embedded agreements had similarly positive effects. Moreover, our evidence suggests that international trade agreements have complemented, rather than undercut, each other.An earlier version of this article was presented at the 99th Annual Meeting of the American Political Science Association, Philadelphia, August 28–31, 2003. We thank Tim BA¼the, Joanne Gowa, Miles Kahler, Andrew Rose, Arthur Stein, Richard Steinberg, and seminar participants at Stanford University, the University of Chicago (PIPES), the University of California, Los Angeles, the University of California, San Diego, and the University of Virginia, for many helpful comments. We especially thank Claire Adida, Ashley Conner, Moonhawk Kim, Erin Krampetz, James Morrison, Mike Nardis, Natan Sachs, Rachel Rubinfeld, and Jessica Weeks for excellent research assistance. We are grateful for financial support from the National Science Foundation (CAREER grant SES-0548285 to Tomz), the Stanford Center for International Development, and the Vice Provost for Undergraduate Education at Stanford.

Heterogeneity in Models of Electoral Choice

Heterogeneity or the presence of a variety of decision rules in a population has usually been ignored in voting research. A method for handling heterogeneous preferences using rank order data is developed and applied to a simple issue-voting model. The estimated average effect of partisanship is substantially higher when the assumption of homogeneity is relaxed, though many self-identified partisans also use ideological criteria to evaluate candidates and many independents rely on partisan criteria.

The Analysis of Committee Power: An Application to Senate Voting on the Minimum Wage *

A widely noted empirical regularity of congressional behavior is that standing committees exert disproportionate influence on congressional choices. The observed phenomenon has a number of labels-committee influence, committee power, and (from the parent chambers perspective) deference to committees-and a large body of theoretical and empirical research has sought to determine when and why it exists. This paper takes as given only the weakest form the observation, namely, that committee power exists sometimes. It does not directly address the questions of when and why committee power exists, although much of the relevant literature is reviewed. Rather, it focuses on a prerequisite to the resolution of disputes about committee power. How can committee power be assessed empirically? Section I reviews three classes of explanations and identifies obstacles to convincing empirical tests of the accounts. Section II introduces an econometric approach for analyzing committee power. Section III applies the technique to a sequence of votes on minimum wage legislation in the Senate in 1977. Section IV extends the technique to multi-dimensional choice spaces. Section V is a discussion and summary.

Party Identification, Retrospective Voting, and Moderating Elections in a Federal System: West Germany, 1961-1989

The hypotheses of retrospective voting and moderating elections rationalize some empirical regularities in U.S. presidential and congressional elections that posed a challenge for the party identification hypothesis. Here, these hypotheses are applied to the German federal system that is characterized by staggered national and Land (provincial) elections. They are tested using data on real GNP growth at the national and Land levels, party vote shares in national and Land elections, party seat shares in national and Land parliaments, and the party composition of national and Land governments over the time period 1961-1989. Perhaps surprisingly, all three hypotheses—party identification, retrospective voting, and moderating elections—find empirical support when applied to the German federal system. Although these hypotheses were formulated with reference to U.S. political institutions, they travel well—bar some modifications that take into account special features of the German political system.

Experimental estimates of the impact of wage subsidies

The effects of a wage subsidy program on the duration of insured unemployment are investigated using experimental data. Participation in the experiment was voluntary and about one third of the subjects refused to take the subsidy voucher offered to them. Because subsidies appear to have stigmatic effects which tend to lower participation rates by high-skilled workers, experimental participants have longer average durations of unemployment than non-participants. However, correcting for self-selection, we find that wage subsidies can substantially increase a participants probability of reemployment. Subsidies are also compared to a search bonus proposal which is also cost effective, but, due to differences in participation patterns, has rather different effects.

The Mythical Swing Voter

Cross-sectional surveys conducted during the 2012 U.S. presidential campaign showed large swings in support for the Democratic and Republican candidates, especially before and after the first presidential debate. Using a unique (in terms of scale, frequency, and source) panel survey, we find that daily sample composition varied more in response to campaign events than did vote intentions. Multi-level regression and post-stratification (MRP) is used to correct for selection bias. Demographic post-stratification, similar to that used in most academic and media polls, is inadequate, but the addition of attitudinal variables (party identification, ideological self-placement, and past vote) appear to make selection ignorable in our data. We conclude that vote swings in 2012 were mostly sample artifacts and that real swings were quite small. While this account is at variance with most contemporaneous analyses, it better corresponds with our understanding of partisan polarization in modern American politics.