1

Paper

Pooling Disparate Observations Bartels, Larry M. Uploaded 01-01-1995 Keywords induction statistical inference Bayesian statistics econometrics observations F-test pooling fractional pooling Abstract Data analysts frequently face difficult choices about whether to pool disparate observations in their statistical analyses. I explore the inferential ramifications of such choices, and propose a new technique, dubbed "fractional pooling," which provides a simple way to incorporate prior beliefs about the theoretical relevance of disparate observations. The technique is easy to implement and has a plausible rationale in Bayesian statistical theory. I illustrate the potential utility of fractional pooling by applying the technique to political data originally analyzed by Ashenfelter (1994), Powell (1982), and Alesina et al. (1993). These examples demonstrate that conventional approaches to analyzing disparate observations can be seriously misleading, and that the approach proposed here can enrich our understanding of the inferential implications of unavoidably subjective judgments about the theoretical relevance of available data.

2

Paper

Correlated Disturbances in Discrete Choice Models:A Comparison of Multinomial Probit Models Alvarez, R. Michael Nagler, Jonathan Uploaded 01-01-1995 Keywords econometrics logit multinomial probit gev discrete-choice monte-carlo Abstract Correlated Disturbances in Discrete Choice Models: A Comparison of Multinomial Probit Models and Logit Models In political science, there are many cases where individuals make discrete choices from more than two alternatives. This paper uses Monte Carlo analysis to examine several questions about one class of discrete choice models --- those involving both alternative-specific and individual-specific variables on the right-hand side --- and demonstrates several findings. First, the use of estimation techniques assuming uncorrelated disturbances across alternatives in discrete choice models can lead to significantly biased parameter estimates. This point is tempered by the observation that probability estimates based on the full choice set generated from such estimates are not likely to be biased enough to lead to incorrect inferences. However, attempts to infer the impact of altering the choice set -- such as by removing one of the alternatives -- will be less successful. Second, the Generalized Extreme Value (GEV) model is extremely unreliable when the pattern of correlation among the disturbances is not as restricted as the GEV model assumes. GEV estimates may suggest grouping among the choices that is in fact not present in the data. Third, in samples the size of many typical political science applications -- 1000 observations -- Multinomial Probit (MNP) is capable of recovering precise estimates of the parameters of the systemic component of the model, though MNP is not likely to generate precise estimates of the relationship among the disturbances in samples of this size. Paradoxically, MNP's primary benefit is its ability to uncover relationships among alternatives and to correctly estimate the affect of removing an alternative from the choice set. Thus this paper suggests the increased use of MNP by political scientists examining discrete choice problems when the central question of interest is the effect of removing an alternative from the choice set. We demonstrate that for other questions, models positing independent disturbances may be `close enough.'

3

Paper

Nuisance vs. Substance: Specifying and Estimating Time-Series--Cross-Section Model Beck, Nathaniel Katz, Jonathan Uploaded 01-01-1995 Keywords Econometrics Time-series--cross-section GLS FGLS EGLS Parks Robust standard errors GLS--ARMA Abstract In a previous article we showed that ordinary least squares with panel corrected standard errors is superior to the Parks generalized least squares approach to the estimation of time-series--cross-section models. In this article we compare our proposed method to another leading technique, Kmenta's ``cross-sectionally heteroskedastic and timewise autocorrelated'' model. This estimator uses generalized least squares to correct for both panel heteroskedasticity and temporally correlated errors. We argue that it is best to model dynamics via a lagged dependent variable, rather than via serially correlated errors. The lagged dependent variable approach makes it easier for researchers to examine dynamics and allows for natural generalizations in a manner that the serially correlated errors approach does not. We also show that the generalized least squares correction for panel heteroskedasticity is, in general, no improvement over ordinary least squares and is, in the presence of parameter heterogeneity, inferior to it. In the conclusion we present a unified method for analyzing time-series--cross-section data.

4

Paper

Electoral Reform and Legislative Structure: The Effects of Australian Ballot Laws on House Committee Tenure Katz, Jonathan Sala, Brian R. Uploaded 01-01-1995 Keywords congress personal vote australian ballot duration model Abstract Most scholars agree that members of Congress are strongly motivated by their desire for reelection. This assumption implies that MCs adopt institutions, rules and norms of behavior in part to serve their electoral interests. Direct tests of the electoral connection are rare, however, because significant, exogenous changes in the electoral environment are difficult to identify. In this paper, we develop and test an electoral rationale for the norm of committee tenure, in which returning MCs typically retain their same assignments. We examine tenure patterns before and after a major, exogenous change in the electoral system -- the states' rapid adoption of Australian Ballot laws in the early 1890s. The ballot changes, we argue, induced new ``personal vote'' electoral incentives, which contributed to the adoption of ``modern'' Congressional institutions such as ``property rights'' to committee assignments. We demonstrate that there was a marked increase in assignment stability after 1892, when a majority of states had put the new ballot laws into force -- earlier than previous studies have suggested.

5

Paper

Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election Alvarez, R. Michael Nagler, Jonathan Uploaded 01-01-1995 Keywords Elections Campaigns Perot Multinomial Probit Economic Voting Angry Voters Abstract Theory: Theories of presidential elections (economic voting and spatial issue and ideology models), combined with the popular explanation of "angry voting", are used to account for voter choice in the 1992 Presidential Election. Hypotheses: Voter choice in this three-candidate race is a function of economic perceptions, issue and ideological positions of voters and candidates, or ``voter anger.'' Methods: Multinomial probit analysis of 1992 National Election Studies data including individual-specific and alternative-specific variables. Simulations based on counterfactual scenarios of ideological positions of the candidates and of voter perceptions of the economy. Results: The economy was the dominant factor in accounting for voter decisions in 1992, and Clinton, not Perot, was the beneficiary of economic discontent. While issues (mainly abortion) and ideology did play some role, Clinton was not perceived by the electorate as a ``New Democrat.'' We find little support for the hypothesis of ``angry voting.'' Last, Perot took more votes from Bush than from Clinton.

6

Paper

Strategic Position-Taking and the Timing of Voting Decisions in Congress Box-Steffensmeier, Janet M. Zorn, Christopher Arnold, Laura W. Uploaded 01-01-1995 Keywords timing Congressional voting duration models proportional hazards survival rate Abstract Voting behavior is intimately linked with many of the most prominent questions of concern to students of legislatures, including the strength of legislative parties and factions, the parameters of individual decision making, and the nature of representation (Collie 1985). One critical element of voting in legislatures is the timing of various choices legislators make. The study of strategic position taking and the timing of voting decisions is important for three major reasons: it adds information about the context and sequence of decision making; the analysis more closely approximates members' strategic considerations; and finally, in contrast to most of the literature on legislative roll call voting, the process is examined rather than strictly the result. Yet, despite the importance of position taking and timing, no one has examined comprehensively this crucial aspect of timing. Research on the timing of voting decisions provides insight into theoretical questions regarding the strategic behavior of legislators, institutional constraints on member behavior, and strategies of interest group influence. The project examines the vote to ratify the North American Free Trade Agreement, which has been called ". . . the most important vote on Capital Hill since the Berlin Wall came down" (Frenzel 1994, 3).

7

Paper

Unit Roots and Causal Inference in Political Science Freeman, John R. Williams, John T. Houser, Daniel Kellstedt, Paul Uploaded 01-01-1995 Keywords Time series unit roots VAR FM-VAR Abstract In the 1980s political scientists were introduced to vector autoregression (Sims, 1980). In the years that followed, they used this method to evaluate competing theories (Goldstein and Freeman, 1990, 199l; Freeman and Alt, 1994; Williams, 1990) and to test the validity of the restrictions in their regression models (MacKuen, Erikson, and Stimson, 1992). In the process, important empirical anomalies came to light. At about this same time, econometricians identified and began to evaluate the problems which unit roots and cointegration produced in vector autoregression and related time series methods. These problems had to do with nothing less than the validity of Granger causality tests and other inferential tools which are the heart of the approach. This research was important because econometricians had discovered years before that many economic time series are first-order integrated (Nelson and Plosser, 1982). Studying the trend properties of economic time series therefore is considered essential in time series econometrics. Recently political scientists (Ostrom and Smith, 1993; Durr, 1993) have argued that certain political time series contain unit roots as well. Yet, to date, no political scientist has made any such demonstration, let alone explained what should be done to put our results on sounder footings if, in fact, our level VARs are faulty. This is the purpose of this paper. In it, we explain the problems which unit roots and cointegration produce in level VARs--why it is so important to take into account the trend properties of one's data. We then review several approaches to solving these problems. One of these approaches, Phillips's (1995) Fully Modified Vector Autoregression (FM-VAR) is singled out for closer study. The theoretical nature of FM-VAR is briefly explained and some practical difficulties in implementing the associated estimation techniques and hypothesis tests are discussed. Finally, the usefulness of FM-VAR is explored in several analyses which parallel the main uses of level VARs mentioned above. These are a stylized Monte Carlo analysis; a reanalysis of Freeman's (1983) study of arms races; a retest of the specifications of MacKuen, Erikson, and Stimson's (1992) model of approval; and a reexamination of the exogeneity-of-vote intentions anomaly in Freeman, Williams and Lin's (1989) study of British government spending.

8

Paper

Do Majority-Minority Districts Maximize Black Representation in Congress Epstein, David O'Halloran, Sharyn Cameron, Charles Uploaded 01-01-1995 Keywords districting voting rights act minority representation electoral systems semi-parametric estimation Abstract This paper investigates the question of whether or not concentrated minority districts, which increase the probability that minorities are elected to office but decrease minority influence elsewhere, maximize overall black representation in Congress. We address this question in a three-step process: we first estimate representation equations that link constituency preferences to the actions of their representative; then electoral equations that link constituency characteristics to the type of representative elected; and finally combine these two effects to simulate the districting strategies that maximize substantive minority representation. We find that outside of the South, dividing minority voters equally across districts maximizes representation, while in the South the optimal scheme creates concentrated districts on the order of 47% black voting age population. We also conclude that minority candidates have substantial chances of being elected from districts with less than 50% minority voters, and that in the face of a national Republican tide, optimal districting schemes will concentrate minority voters less, rather than more.

9

Paper

The Spatial Model and Specification of Choice Models Alvarez, R. Michael Nagler, Jonathan Uploaded 01-01-1995 Keywords Multinomial Logit Spatial Model Conditional Logit Elections Abstract The spatial model has been in use in political science for close to 30 years, and in that period it has achieved a place of prime importance as our paradigm of the process of candidate-choice used by voters. For much of this time political scientists have estimated models of candidate-choice using binary logit or probit, even in cases where there were more than two choices facing voters. Recently discrete choices models beyond binary logit and probit have been making their way into use in political science with increasing frequency. The properties of these models, and their relationship to the spatial model, are frequently misunderstood. This paper demonstrates four essential points. First, the popular multinomial logit model is in fact equivalent to running a series of binary logit models. It involves nothing more than pairwise comparisons of the choices. Second, despite containing no information about the choices, the multinomial logit model provides reduced form estimates of the effect of characteristics of choices that are equivalent to the estimates of such effects provided by the conditional logit model - which does utilize information about the characteristics of the choices. Third, the multinomial logit model cannot offer any inferences as to effects of changing the characteristics of the choices, or introducing additional choices; whereas the conditional logit model can offer such inferences. Fourth, the classic spatial model has a flaw in multi-candidate settings that has been overlooked, with more than two candidates the spatial model explicitly contradicts an aspect of voter behaviour widely believed to be prevalent: the tendency of voters to view certain candidates as `similar' alternatives, and thus for the presence of additional candidates to effect asymettrically the probability of existing candidates being chosen.

10

Paper

The Robustness of Normal-theory LISREL Models: Tests Using a New Optimizer, the Bootstrap, and Sampling Experiments, with Applications Mebane, Walter R. Sekhon, Jasjeet Wells, Martin T. Uploaded 01-01-1995 Keywords statistics estimation covariance structures linear structural relations LISREL bootstrap confidence intervals BCa specification tests goodness-of-fit hypothesis tests optimization evolutionary programming genetic algorithms monte carlo sampling experiment Abstract Asymptotic results from theoretical statistics show that the linear structural relations (LISREL) covariance structure model is robust to many kinds of departures from multivariate normality in the observed data. But close examination of the statistical theory suggests that the kinds of hypotheses about alternative models that are most often of interest in political science research are not covered by the nice robustness results. The typical size of political science data samples also raises questions about the applicability of the asymptotic normal theory. We present results from a Monte Carlo sampling experiment and from analysis of two real data sets both to illustrate the robustness results and to demonstrate how it is unwise to rely on them in substantive political science research. We propose new methods using the bootstrap to assess more accurately the distributions of parameter estimates and test statistics for the LISREL model. To implement the bootstrap we use optimization software two of us have developed, incorporating the quasi-Newton BFGS method in an evolutionary programming algorithm. We describe methods for drawing inferences about LISREL models that are much more reliable than the asymptotic normal-theory techniques. The methods we propose are implemented using the new software we have developed. Our bootstrap and optimization methods allow model assessment and model selection to use well understood statistical principles such as classical hypothesis testing.

11

Paper

Re-thinking Equilibrium Presidential Approval: Markov-Switching Error Correction Jackman, Simon Uploaded 01-01-1995 Keywords (none submitted) Abstract I present a re-working of an error-correction model of presidential approval. In large measure this is driven by an attempt to make better sense of what appears to be an interesting period in American politics---the phenomenal rise and fall in approval for George Bush. The innovation I present is to allow the weight of the determinants of presidential approval to themselves change over time. I do this by estimating a switching-regime time-series model of aggregate presidential approval, and here I fit two distinct statistical regimes. There are good reasons why we would a model of presidential approval to do this. For one thing, the time series here span a long slew of technological changes in the relationship between the public, the economy, and political leaders:~it seems reasonable that even in the aggregate, ``different things will matter at different times'' in assessing a president. Furthermore (though this point is not pursued in the paper), if there is a fair amount of heterogeneity both in the content and the dynamics of information-flows in the American political economy, then a one-regime statistical model will not characterize the aggregate-level process as well as a (non-linear) switching-regime model. The two-regime model I present does appear to be a better characterization than a single-state model. I find presidential approval to typically track inflation and, to a lesser extent, movements in the stock market, subject to the large shocks associated with a rally event, which dissipate slowly. Occasionally though, and often hot on the heels of a rally event, presidential approval slips into another regime, in which unemployment and economic expectations are dominant, and in which approval quickly reverts to levels in line with those economic indicators. Almost as quickly, approval returns to its former slowly-correcting state. This characterization is quite different to previous models of presidential approval. Importantly, my findings suggest that the mixture between perception and reality brought to bear in assessing a president is not at all constant. Different parts of reality are deemed more relevant than at others. Typically, the piece of the macro-economy relevant to presidential approval is inflation. After this relationship has been disturbed by the disequilibriating shock of a rally event, unemployment and a ``perception'' variable---expectations about business conditions---enter the fray. Also novel here is the finding that economic perceptions (i.e., economic expectations, a subjective variable) operate to ratchet presidential approval back in line with economic fundamentals. Though of course, given that economic expectations are almost always overly optimistic, it would seem that if anything, the post-rally slide in presidential approval would be faster if we possessed perfect economic foresight. What is interesting though is that economic ``perceptions'' (as opposed to economic ``reality'') are not part of the typical set of determinants of presidential approval. This will be news to students of public opinion. Error-laden, subjective information about the future course of the macro-economy comes into play when approval has strayed away from levels we would expect on the basis of inflation. Furthermore, I find that the switching between statistical regimes has become more volatile in recent times. In particular, George Bush's presidency is marked by quite abrupt switching between the two regimes, suggesting that the mix of considerations used to assess Bush was rapidly changing.