Working Papers
2004
30 records found
Shaken, Not Stirred: Evidence on Ballot Order Effects from the California Alphabet Lottery, 1978 - 2002
Ho, Daniel E., Imai, Kosuke
Submitted: 2004-02-02
Keywords: ballots, elections, causal inference, natural experiment, randomization, fisher test, partisan cue
Abstract: (click to show/hide) We analyze a natural experiment to answer the longstanding question of whether the name order of candidates on ballots affects election outcomes. Since 1975, California law has mandated randomizing the ballot order with a lottery, where alphabet letters would be shaken vigorously and selected from a container. Previous studies, relying overwhelmingly on non-randomized data, have yielded conflicting results about whether ballot order effects even exist. Using improved statistical methods, our analysis of statewide elections from 1978 to 2002 reveals that in general elections ballot order has a significant impact only on minor party candidates and candidates for nonpartisan offices. In primaries, however, being listed first benefits everyone. In fact, ballot order might have changed the winner in roughly nine percent of all primary races examined. These results are largely consistent with a theory of partisan cuing. We propose that all electoral jurisdictions randomize ballot order to minimize ballot effects.
Ho, Daniel E., Imai, Kosuke
Submitted: 2004-02-02
Keywords: ballots, elections, causal inference, natural experiment, randomization, fisher test, partisan cue
Abstract: (click to show/hide) We analyze a natural experiment to answer the longstanding question of whether the name order of candidates on ballots affects election outcomes. Since 1975, California law has mandated randomizing the ballot order with a lottery, where alphabet letters would be shaken vigorously and selected from a container. Previous studies, relying overwhelmingly on non-randomized data, have yielded conflicting results about whether ballot order effects even exist. Using improved statistical methods, our analysis of statewide elections from 1978 to 2002 reveals that in general elections ballot order has a significant impact only on minor party candidates and candidates for nonpartisan offices. In primaries, however, being listed first benefits everyone. In fact, ballot order might have changed the winner in roughly nine percent of all primary races examined. These results are largely consistent with a theory of partisan cuing. We propose that all electoral jurisdictions randomize ballot order to minimize ballot effects.
The Etiology of Public Support for the Designated Hitter Rule
Zorn, Christopher, Gill, Jeff
Submitted: 2004-03-21
Keywords: baseball, designated hitter, public opinion, ideology, selection model
Abstract: (click to show/hide) Since its introduction in 1973, major league baseball’s designated hitter (DH) rule has been the subject of continuing controversy. Here, we investigate the political and socio–demographic determinants of public opinion towards baseball’s DH rule, using data from a nationwide poll conducted during September, 1997. Our findings suggest that, while both self–proclaimed Democrats and Republicans are more likely to follow baseball than are political independents, it is Democrats, not Republicans, who tend to favor the DH. In addition, older respondents were more likely to oppose the rule, while respondents from the Midwest tended to favor it.
Zorn, Christopher, Gill, Jeff
Submitted: 2004-03-21
Keywords: baseball, designated hitter, public opinion, ideology, selection model
Abstract: (click to show/hide) Since its introduction in 1973, major league baseball’s designated hitter (DH) rule has been the subject of continuing controversy. Here, we investigate the political and socio–demographic determinants of public opinion towards baseball’s DH rule, using data from a nationwide poll conducted during September, 1997. Our findings suggest that, while both self–proclaimed Democrats and Republicans are more likely to follow baseball than are political independents, it is Democrats, not Republicans, who tend to favor the DH. In addition, older respondents were more likely to oppose the rule, while respondents from the Midwest tended to favor it.
Reconsidering Tests for Ambivalence in Political Choice Survey Data
Glasgow, Garrett
Submitted: 2004-03-21
Keywords: ambivalence, heteroskedastic discrete choice
Abstract: (click to show/hide) The concept of ambivalence challenges the assumption that individuals combine their positive and negative attitudes towards objects in their choice set into unidimensional attitudes, instead maintaining that individuals can simultaneously hold conflicting attitudes. Unfortunately, most tests for ambivalence in political choice survey data are inconclusive. In particular, the empirical results of these tests could also be explained by a choice model with unidimensional attitudes. There are two related reasons for this. First, individuals who appear to be close to neutrality or indifference in a choice model with unidimensional attitudes are expected to have observed choice behavior identical to that expected from ambivalent individuals. Second, the measures of ambivalence developed and used in survey-based studies of ambivalence in political choice are closely related to measures of neutrality or indifference in a unidimensional attitude choice model. Taken together, these two observations point out the need to reconsider our empirical tests of ambivalence if we wish to determine if and how ambivalence influences individual political choice behavior.
Glasgow, Garrett
Submitted: 2004-03-21
Keywords: ambivalence, heteroskedastic discrete choice
Abstract: (click to show/hide) The concept of ambivalence challenges the assumption that individuals combine their positive and negative attitudes towards objects in their choice set into unidimensional attitudes, instead maintaining that individuals can simultaneously hold conflicting attitudes. Unfortunately, most tests for ambivalence in political choice survey data are inconclusive. In particular, the empirical results of these tests could also be explained by a choice model with unidimensional attitudes. There are two related reasons for this. First, individuals who appear to be close to neutrality or indifference in a choice model with unidimensional attitudes are expected to have observed choice behavior identical to that expected from ambivalent individuals. Second, the measures of ambivalence developed and used in survey-based studies of ambivalence in political choice are closely related to measures of neutrality or indifference in a unidimensional attitude choice model. Taken together, these two observations point out the need to reconsider our empirical tests of ambivalence if we wish to determine if and how ambivalence influences individual political choice behavior.
Prior Distributions for Variance Parameters in Hierarchical Models
Gelman, Andrew
Submitted: 2004-03-28
Keywords: Bayesian inference, hierarchical model, multilevel model, noninformative prior distribution, weakly informative prior distribution
Abstract: (click to show/hide) Various noninformative prior distributions have been suggested for scale parameters in hierarchical models. We construct a new folded-noncentral-$t$ family of conditionally conjugate priors for hierarchical standard deviation parameters, and then consider noninformative and weakly informative priors in this family. We use an example to illustrate serious problems with the inverse-gamma family of "noninformative" prior distributions. We suggest instead to use a uniform prior on the hierarchical standard deviation, using the half-$t$ family when the number of groups is small and in other settings where a weakly informative prior is desired.
Gelman, Andrew
Submitted: 2004-03-28
Keywords: Bayesian inference, hierarchical model, multilevel model, noninformative prior distribution, weakly informative prior distribution
Abstract: (click to show/hide) Various noninformative prior distributions have been suggested for scale parameters in hierarchical models. We construct a new folded-noncentral-$t$ family of conditionally conjugate priors for hierarchical standard deviation parameters, and then consider noninformative and weakly informative priors in this family. We use an example to illustrate serious problems with the inverse-gamma family of "noninformative" prior distributions. We suggest instead to use a uniform prior on the hierarchical standard deviation, using the half-$t$ family when the number of groups is small and in other settings where a weakly informative prior is desired.
Bayesian Measures of Explained Variance and Pooling in Multilevel (Hierarchical) Models
Gelman, Andrew, Pardoe, Iain
Submitted: 2004-04-16
Keywords: adjusted R-squared, Bayesian inference, hierarchical model, multilevel regression, partial pooling, shrinkage
Abstract: (click to show/hide) Explained variance (R2) is a familiar summary of the fit of a linear regression and has been generalized in various ways to multilevel (hierarchical) models. The multilevel models we consider in this paper are characterized by hierarchical data structures in which individuals are grouped into units (which themselves might be further grouped into larger units), and there are variables measured on individuals and each grouping unit. The models are based on regression relationships at different levels, with the first level corresponding to the individual data, and subsequent levels corresponding to between-group regressions of individual predictor effects on grouping unit variables. We present an approach to defining R2 at each level of the multilevel model, rather than attempting to create a single summary measure of fit. Our method is based on comparing variances in a single fitted model rather than comparing to a null model. In simple regression, our measure generalizes the classical adjusted R2. We also discuss a related variance comparison to summarize the degree to which estimates at each level of the model are pooled together based on the level-specific regression relationship, rather than estimated separately. This pooling factor is related to the concept of shrinkage in simple hierarchical models. We illustrate the methods on a dataset of radon in houses within counties using a series of models ranging from a simple linear regression model to a multilevel varying-intercept, varying-slope model.
Gelman, Andrew, Pardoe, Iain
Submitted: 2004-04-16
Keywords: adjusted R-squared, Bayesian inference, hierarchical model, multilevel regression, partial pooling, shrinkage
Abstract: (click to show/hide) Explained variance (R2) is a familiar summary of the fit of a linear regression and has been generalized in various ways to multilevel (hierarchical) models. The multilevel models we consider in this paper are characterized by hierarchical data structures in which individuals are grouped into units (which themselves might be further grouped into larger units), and there are variables measured on individuals and each grouping unit. The models are based on regression relationships at different levels, with the first level corresponding to the individual data, and subsequent levels corresponding to between-group regressions of individual predictor effects on grouping unit variables. We present an approach to defining R2 at each level of the multilevel model, rather than attempting to create a single summary measure of fit. Our method is based on comparing variances in a single fitted model rather than comparing to a null model. In simple regression, our measure generalizes the classical adjusted R2. We also discuss a related variance comparison to summarize the degree to which estimates at each level of the model are pooled together based on the level-specific regression relationship, rather than estimated separately. This pooling factor is related to the concept of shrinkage in simple hierarchical models. We illustrate the methods on a dataset of radon in houses within counties using a series of models ranging from a simple linear regression model to a multilevel varying-intercept, varying-slope model.
Treatment effects in before-after data
Gelman, Andrew
Submitted: 2004-04-27
Keywords: correlation, experiments, interactions, hierarchical models, observational studies, variance components
Abstract: (click to show/hide) In experiments and observations with before-after data, the correlation between "before" and "after" measurements is typically higher among the controls than among the treated units, violating the usual assumptions of equal variance and a constant treatment effect. We illustrate with three applied examples and then discuss models that could be used to fit this phenomenon, which we argue is related to the
Gelman, Andrew
Submitted: 2004-04-27
Keywords: correlation, experiments, interactions, hierarchical models, observational studies, variance components
Abstract: (click to show/hide) In experiments and observations with before-after data, the correlation between "before" and "after" measurements is typically higher among the controls than among the treated units, violating the usual assumptions of equal variance and a constant treatment effect. We illustrate with three applied examples and then discuss models that could be used to fit this phenomenon, which we argue is related to the
The Most Liberal Senator: Analyzing and Interpreting Congressional Roll Calls
Clinton, Joshua, Jackman, Simon, Rivers, Doug
Submitted: 2004-05-12
Keywords: ideal points, roll call voting, 2004 presidential election
Abstract: (click to show/hide) The non-partisan National Journal recently declared Senator John Kerry to be the "top liberal" in the Senate based on analysis of 62 roll calls in 2003. Although widely reported in the media (and the subject of a debate among the Democratic presidential candidates), we argue that this characterization of Kerry is misleading in at least two respects. First, when we account for the "margin of error: in the voting scores -- which is considerable for Kerry given that he missed 60% of the National Journal's key votes while campaigning -- we discover that the probability that Kerry is the "top liberal" is only .30, and that we cannot reject the conclusion that Kerry is only the 20th most liberal senator. Second, we compare the position of the President Bush on these key votes; including the President's announced positions on these votes reveals the President to be just as conservative as Kerry is liberal (i.e., both candidates are extreme relative to the 108th Senate). A similar conclusion holds when we replicate the analysis using all votes cast in the 107th Senate. A more comprehensive analysis than that undertaken by National Journal (including an accounting of the margins of error in voting scores) shows although Kerry belongs to the most liberal quintile of the contemporary Senate, Bush belongs to the most conservative quintile.
Clinton, Joshua, Jackman, Simon, Rivers, Doug
Submitted: 2004-05-12
Keywords: ideal points, roll call voting, 2004 presidential election
Abstract: (click to show/hide) The non-partisan National Journal recently declared Senator John Kerry to be the "top liberal" in the Senate based on analysis of 62 roll calls in 2003. Although widely reported in the media (and the subject of a debate among the Democratic presidential candidates), we argue that this characterization of Kerry is misleading in at least two respects. First, when we account for the "margin of error: in the voting scores -- which is considerable for Kerry given that he missed 60% of the National Journal's key votes while campaigning -- we discover that the probability that Kerry is the "top liberal" is only .30, and that we cannot reject the conclusion that Kerry is only the 20th most liberal senator. Second, we compare the position of the President Bush on these key votes; including the President's announced positions on these votes reveals the President to be just as conservative as Kerry is liberal (i.e., both candidates are extreme relative to the 108th Senate). A similar conclusion holds when we replicate the analysis using all votes cast in the 107th Senate. A more comprehensive analysis than that undertaken by National Journal (including an accounting of the margins of error in voting scores) shows although Kerry belongs to the most liberal quintile of the contemporary Senate, Bush belongs to the most conservative quintile.
What Divides Us?: The Image and Organization of Political Science
Grant, Tobin J.
Submitted: 2004-05-25
Keywords: political science, multidimensional scaling, sociology of science
Abstract: (click to show/hide) The dominant image of political science presented in the literature is one of a field divided along a methodological dimension. According to this hard-soft model, political scientists vary primarily in their use of "hard" scientific methods and approaches to politics and "soft" humanistic means of understanding politics. I test whether this image accurately depicts the organization of political science. I multidimensional scalings of political science based on the organization of political scientists within APSA and the organization of the APSA annual meeting. These models locate the primary division within the discipline as one based on the types of political phenomena examined, with international relations studying state actors standing at one end of a continuum and American politics scholars, particularly those studying local politics, at the other end. The methodological differences are important but are secondary to this first dimension. This model is consistent for the organization of both political scientists and their scholarship.
Grant, Tobin J.
Submitted: 2004-05-25
Keywords: political science, multidimensional scaling, sociology of science
Abstract: (click to show/hide) The dominant image of political science presented in the literature is one of a field divided along a methodological dimension. According to this hard-soft model, political scientists vary primarily in their use of "hard" scientific methods and approaches to politics and "soft" humanistic means of understanding politics. I test whether this image accurately depicts the organization of political science. I multidimensional scalings of political science based on the organization of political scientists within APSA and the organization of the APSA annual meeting. These models locate the primary division within the discipline as one based on the types of political phenomena examined, with international relations studying state actors standing at one end of a continuum and American politics scholars, particularly those studying local politics, at the other end. The methodological differences are important but are secondary to this first dimension. This model is consistent for the organization of both political scientists and their scholarship.
Practical Issues in Implementing and Understanding Bayesian Ideal Point Estimation
Bafumi, Joseph, Gelman, Andrew, Park, David K., Kaplan, Noah
Submitted: 2004-06-11
Keywords: Ideal points, Bayesian, Logistic regression, Rasch model
Abstract: (click to show/hide) In recent years, logistic regression (Rasch) models have been used in political science for estimating ideal points of legislators and Supreme Court justices. These models present estimation and identifiability challenges, such as improper variance estimates, scale and translation invariance, reflection invariance, and issues with outliers. We resolve these issues using Bayesian hierarchical modeling, linear transformations, informative regression predictors, and explicit modeling for outliers. In addition, we explore new ways to usefully display inferences and check model fit.
Bafumi, Joseph, Gelman, Andrew, Park, David K., Kaplan, Noah
Submitted: 2004-06-11
Keywords: Ideal points, Bayesian, Logistic regression, Rasch model
Abstract: (click to show/hide) In recent years, logistic regression (Rasch) models have been used in political science for estimating ideal points of legislators and Supreme Court justices. These models present estimation and identifiability challenges, such as improper variance estimates, scale and translation invariance, reflection invariance, and issues with outliers. We resolve these issues using Bayesian hierarchical modeling, linear transformations, informative regression predictors, and explicit modeling for outliers. In addition, we explore new ways to usefully display inferences and check model fit.
Can Voting Reduce Welfare? Evidence from the US Telecommunications Sector
Falaschetti, Dino
Submitted: 2004-06-15
Keywords: Electoral Institutions, Voter Turnout, Capture Theory, Regulatory Commitment, Telecommunications Policy, Economic Welfare
Abstract: (click to show/hide) Voter turnout is popularly cited as reflecting a polity's health. The ease with which electoral members influence policy can, however, constrain an economy's productive capacity. For example, while influential electorates might carefully monitor political agents, they might also "capture" them. In the latter case, electorates transfer producer surplus to consumers at the expense of social welfare - i.e., a "healthy" polity's economy rests at an inferior equilibrium. I develop evidence that the US telecommunications sector may have realized such an outcome. This evidence is remarkably difficult to dismiss as an artifact of endogeneity bias, and appears important for several audiences. For example, the normative regulation literature calls for constraints on producers' market power, while the institutions and commitment literature calls for checks on political agents' opportunism. Evidence that I develop here suggests that, unbound by similar constraints, electoral principals might effectively control their political agents while significantly retarding their economic agents' productive incentives.
Falaschetti, Dino
Submitted: 2004-06-15
Keywords: Electoral Institutions, Voter Turnout, Capture Theory, Regulatory Commitment, Telecommunications Policy, Economic Welfare
Abstract: (click to show/hide) Voter turnout is popularly cited as reflecting a polity's health. The ease with which electoral members influence policy can, however, constrain an economy's productive capacity. For example, while influential electorates might carefully monitor political agents, they might also "capture" them. In the latter case, electorates transfer producer surplus to consumers at the expense of social welfare - i.e., a "healthy" polity's economy rests at an inferior equilibrium. I develop evidence that the US telecommunications sector may have realized such an outcome. This evidence is remarkably difficult to dismiss as an artifact of endogeneity bias, and appears important for several audiences. For example, the normative regulation literature calls for constraints on producers' market power, while the institutions and commitment literature calls for checks on political agents' opportunism. Evidence that I develop here suggests that, unbound by similar constraints, electoral principals might effectively control their political agents while significantly retarding their economic agents' productive incentives.
Parameterization and Bayesian Modeling
Gelman, Andrew
Submitted: 2004-06-15
Keywords: censored data, data augmentation, Gibbs sampler, hierarchical model, missing data imputation, parameter expansion, prior distribution, truncated data
Abstract: (click to show/hide) Progress in statistical computation often leads to advances in statistical modeling. For example, it is surprisingly common that an existing model is reparameterized, solely for computational purposes, but then this new configuration motivates a new family of models that is useful in applied statistics. One reason this phenomenon may not have been noticed in statistics is that reparameterizations do not change the likelihood. In a Bayesian framework, however, a transformation of parameters typically suggests a new family of prior distributions. We discuss examples in censored and truncated data, mixture modeling, multivariate imputation, stochastic processes, and multilevel models.
Gelman, Andrew
Submitted: 2004-06-15
Keywords: censored data, data augmentation, Gibbs sampler, hierarchical model, missing data imputation, parameter expansion, prior distribution, truncated data
Abstract: (click to show/hide) Progress in statistical computation often leads to advances in statistical modeling. For example, it is surprisingly common that an existing model is reparameterized, solely for computational purposes, but then this new configuration motivates a new family of models that is useful in applied statistics. One reason this phenomenon may not have been noticed in statistics is that reparameterizations do not change the likelihood. In a Bayesian framework, however, a transformation of parameters typically suggests a new family of prior distributions. We discuss examples in censored and truncated data, mixture modeling, multivariate imputation, stochastic processes, and multilevel models.
Macro vs. Micro-Level Perspectives on Economic Voting: Is the Micro-Level Evidence Endogenously Induced?
Erikson, Robert S.
Submitted: 2004-07-10
Keywords: economic voting, vote choice
Abstract: (click to show/hide) Many of the findings regarding economic voting derive from the micro-level analyses of survey data, in which respondents' survey evaluations of the economy are shown to predict the vote. This paper investigates the causal nature of this relationship and argues that cross-sectional consistency between economic evaluations and vote choice is mainly if not entirely due to vote choice influencing the survey response. Moreover, the evidence suggest that apart from this endogenously induced partisan bias, almost all of the cross-sectional variation in survey evaluations of the economy is random noise rather than actual beliefs about economic conditions In surveys, the mean evaluations reflect the economic signal that predicts the aggregate vote. Following Kramer (1983), economic voting is best studied at the macro-level rather than the micro-level.
Erikson, Robert S.
Submitted: 2004-07-10
Keywords: economic voting, vote choice
Abstract: (click to show/hide) Many of the findings regarding economic voting derive from the micro-level analyses of survey data, in which respondents' survey evaluations of the economy are shown to predict the vote. This paper investigates the causal nature of this relationship and argues that cross-sectional consistency between economic evaluations and vote choice is mainly if not entirely due to vote choice influencing the survey response. Moreover, the evidence suggest that apart from this endogenously induced partisan bias, almost all of the cross-sectional variation in survey evaluations of the economy is random noise rather than actual beliefs about economic conditions In surveys, the mean evaluations reflect the economic signal that predicts the aggregate vote. Following Kramer (1983), economic voting is best studied at the macro-level rather than the micro-level.
Cuing and Coordination in American Elections
Mebane, Walter R.
Submitted: 2004-07-16
Keywords: evolutionary game, GEV, political behavior, strategic coordination, policy moderation
Abstract: (click to show/hide) I use evolutionary game models based on pure imitation to reexamine recent findings that strategic coordination characterizes the American electorate. Imitation means that voters who are dissatisfied with their strategy adopt the strategy of the first voter they encounter who is similar to them. In the replicator dynamics such imitation implies, everyone ultimately uses the coordinating strategy, but I study what happens over time spans that are relevant for voters. I consider three evolutionary models, including two that involve partisan cuing. Simulations using National Election Studies data from presidential years 1976-96 suggest that many voters use an unconditional strategy, usually a strategy of voting a straight ticket matching their party identification. I then estimate a choice model that incorporates an approximation to the evolutionary dynamics. The results support partisan cuing and confirm that most voters vote unconditionally. The estimates also support previous findings regarding policy moderation and institutional balancing.
Mebane, Walter R.
Submitted: 2004-07-16
Keywords: evolutionary game, GEV, political behavior, strategic coordination, policy moderation
Abstract: (click to show/hide) I use evolutionary game models based on pure imitation to reexamine recent findings that strategic coordination characterizes the American electorate. Imitation means that voters who are dissatisfied with their strategy adopt the strategy of the first voter they encounter who is similar to them. In the replicator dynamics such imitation implies, everyone ultimately uses the coordinating strategy, but I study what happens over time spans that are relevant for voters. I consider three evolutionary models, including two that involve partisan cuing. Simulations using National Election Studies data from presidential years 1976-96 suggest that many voters use an unconditional strategy, usually a strategy of voting a straight ticket matching their party identification. I then estimate a choice model that incorporates an approximation to the evolutionary dynamics. The results support partisan cuing and confirm that most voters vote unconditionally. The estimates also support previous findings regarding policy moderation and institutional balancing.
Unifying Political Metrology: A Probilistic Model of Measurement
Grant, J. Tobin
Submitted: 2004-07-21
Keywords: Measurement
Abstract: (click to show/hide) Political science needs an improved metrology, which includes both measurement theory and applied assessments of measurement procedures. I discuss central metrological concepts and their application to political science. I present a probilistic model of measurement that is grounded in well-established measurement theory. The model incorporates recent work in metrology that emphasizes the uncertainty of all measurements. This model has implications for political science measures, including the criteria used to evaluate measurements, the role of qualitative measurements, and the tasks needed to improve measurements. I conclude with a discussion of how political science can improve its metrology.
Grant, J. Tobin
Submitted: 2004-07-21
Keywords: Measurement
Abstract: (click to show/hide) Political science needs an improved metrology, which includes both measurement theory and applied assessments of measurement procedures. I discuss central metrological concepts and their application to political science. I present a probilistic model of measurement that is grounded in well-established measurement theory. The model incorporates recent work in metrology that emphasizes the uncertainty of all measurements. This model has implications for political science measures, including the criteria used to evaluate measurements, the role of qualitative measurements, and the tasks needed to improve measurements. I conclude with a discussion of how political science can improve its metrology.
Parametric and Nonparametric Bayesian Models for Ecological Inference in 2 x 2 Tables
Imai, Kosuke, Lu, Ying
Submitted: 2004-07-21
Keywords: Aggregate data, Data augmentation, Density estimation, Dirichlet process prior, Normal mixtures, Racial voting
Abstract: (click to show/hide) The ecological inference problem arises when making inferences about individual behavior from aggregate data. Such a situation is frequently encountered in the social sciences and epidemiology. In this article, we propose a Bayesian approach based on data augmentation. We formulate ecological inference in $2 times 2$ tables as a missing data problem where only the weighted average of two unknown variables is observed. This framework directly incorporates the deterministic bounds, which contain all information available from the data, and allow researchers to incorporate the individual-level data whenever available. Within this general framework, we first develop a parametric model. We show that through the use of an $EM$ algorithm, the model can formally quantify the effect of missing information on parameter estimation. This is an important diagnostic for evaluating the degree of aggregation effects. Next, we introduce a nonparametric Bayesian model using a Dirichlet process prior to relax the distributional assumption of the parametric model. Through simulations and an empirical application, we evaluate the relative performance of our models and other existing methods. We show that in many realistic scenarios, aggregation effects are so severe that more than half of the information is lost, yielding estimates with little precision. We also find that our nonparametric model generally outperforms parametric models. C-code, along with an R interface, is publicly available for implementing our Markov chain Monte Carlo algorithms to fit the proposed models.
Imai, Kosuke, Lu, Ying
Submitted: 2004-07-21
Keywords: Aggregate data, Data augmentation, Density estimation, Dirichlet process prior, Normal mixtures, Racial voting
Abstract: (click to show/hide) The ecological inference problem arises when making inferences about individual behavior from aggregate data. Such a situation is frequently encountered in the social sciences and epidemiology. In this article, we propose a Bayesian approach based on data augmentation. We formulate ecological inference in $2 times 2$ tables as a missing data problem where only the weighted average of two unknown variables is observed. This framework directly incorporates the deterministic bounds, which contain all information available from the data, and allow researchers to incorporate the individual-level data whenever available. Within this general framework, we first develop a parametric model. We show that through the use of an $EM$ algorithm, the model can formally quantify the effect of missing information on parameter estimation. This is an important diagnostic for evaluating the degree of aggregation effects. Next, we introduce a nonparametric Bayesian model using a Dirichlet process prior to relax the distributional assumption of the parametric model. Through simulations and an empirical application, we evaluate the relative performance of our models and other existing methods. We show that in many realistic scenarios, aggregation effects are so severe that more than half of the information is lost, yielding estimates with little precision. We also find that our nonparametric model generally outperforms parametric models. C-code, along with an R interface, is publicly available for implementing our Markov chain Monte Carlo algorithms to fit the proposed models.
Randomization Inference with Natural Experiments: An Analysis of Ballot Effects in the 2003 California Recall Election
Imai, Kosuke, Ho, Daniel
Submitted: 2004-07-21
Keywords: casual inference, Fisher's exact test, inversion, political science, voting behavior, elections
Abstract: (click to show/hide) Since the 2000 U.S. Presidential election, social scientists have rediscovered a long tradition of research that investigates the effects of ballot format on voting. Using a new dataset collected by the New York Times, we investigate the causal effects of being listed on the first ballot page in the 2003 California gubernatorial recall election. California law mandates a complex randomization procedure of ballot order that approximates a classical randomized experiment in real world settings. The recall election also poses particular statistical challenges with an unprecedented 135 candidates running for the office. We apply (nonparametric) randomization inference based on Fisher's exact test, which incorporates the complex randomization procedure and yields accurate confidence intervals. Conventional asymptotic model-based inferences are found to be highly sensitive to assumptions and model specification. Randomization inference suggests that roughly half of the candidates gained more votes when listed on the first page of ballot.
Imai, Kosuke, Ho, Daniel
Submitted: 2004-07-21
Keywords: casual inference, Fisher's exact test, inversion, political science, voting behavior, elections
Abstract: (click to show/hide) Since the 2000 U.S. Presidential election, social scientists have rediscovered a long tradition of research that investigates the effects of ballot format on voting. Using a new dataset collected by the New York Times, we investigate the causal effects of being listed on the first ballot page in the 2003 California gubernatorial recall election. California law mandates a complex randomization procedure of ballot order that approximates a classical randomized experiment in real world settings. The recall election also poses particular statistical challenges with an unprecedented 135 candidates running for the office. We apply (nonparametric) randomization inference based on Fisher's exact test, which incorporates the complex randomization procedure and yields accurate confidence intervals. Conventional asymptotic model-based inferences are found to be highly sensitive to assumptions and model specification. Randomization inference suggests that roughly half of the candidates gained more votes when listed on the first page of ballot.
The Estimation of Time-Invariant Variables in Panel Analyses with Unit Fixed Effects
Plümper, Thomas, Troeger, Vera E.
Submitted: 2004-07-23
Keywords: Time Invariant Variables, Unit effects, Monte Carlo, Hausman-Taylor
Abstract: (click to show/hide) This paper analyzes the estimation of time-invariant variables in panel data models with unit-effects. We compare three procedures that have frequently been employed in comparative politics, namely pooled-OLS, random effects and the Hausman-Taylor model, to a vector decomposition procedure that allows estimating time-invariant variables in an augmented fixed effects approach. The procedure we suggest consists of three stages: the first stage runs a fixed-effects model without time-invariant variables, the second stage decomposes the unit-effects vector into a part explained by the time-invariant variables and an error term, and the third stage re-estimates the first stage by pooled-OLS including the time invariant variables plus the error term of stage 2. We use Monte Carlo simulations to demonstrate that this method works better than its alternatives in estimating typical models in comparative politics. Specifically, the unit fixed effects vector decomposition technique performs better than both pooled OLS and random effects in the estimation of time-invariant variables correlated with the unit effects and better than Hausman-Taylor in estimating the time-invariant variables correlated with the unit effects. Finally, we re-analyze recent work by Huber and Stephens (2001) as well as by Beramendi and Cusack (2004). These analyses seek to cope with the problem of time-invariant variables in panel data.
Plümper, Thomas, Troeger, Vera E.
Submitted: 2004-07-23
Keywords: Time Invariant Variables, Unit effects, Monte Carlo, Hausman-Taylor
Abstract: (click to show/hide) This paper analyzes the estimation of time-invariant variables in panel data models with unit-effects. We compare three procedures that have frequently been employed in comparative politics, namely pooled-OLS, random effects and the Hausman-Taylor model, to a vector decomposition procedure that allows estimating time-invariant variables in an augmented fixed effects approach. The procedure we suggest consists of three stages: the first stage runs a fixed-effects model without time-invariant variables, the second stage decomposes the unit-effects vector into a part explained by the time-invariant variables and an error term, and the third stage re-estimates the first stage by pooled-OLS including the time invariant variables plus the error term of stage 2. We use Monte Carlo simulations to demonstrate that this method works better than its alternatives in estimating typical models in comparative politics. Specifically, the unit fixed effects vector decomposition technique performs better than both pooled OLS and random effects in the estimation of time-invariant variables correlated with the unit effects and better than Hausman-Taylor in estimating the time-invariant variables correlated with the unit effects. Finally, we re-analyze recent work by Huber and Stephens (2001) as well as by Beramendi and Cusack (2004). These analyses seek to cope with the problem of time-invariant variables in panel data.
Time-Series--Cross-Section Issues: Dynamics, 2004
Beck, Nathaniel, Katz, Jonathan
Submitted: 2004-07-24
Keywords: Time-series--cross-section data, lagged dependent variables, Nickell bias, specification, integration
Abstract: (click to show/hide) This paper deals with a variety of dynamic issues in the analysis of time-series--cross-section (TSCS) data raised by recent papers; it also more briefly treats some cross-sectional issues. 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). It is also shown that there is nothing pernicious in using a lagged dependent variable, and 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 models, and analysts may choose on grounds of convenience (assuming that the model passes standard econometric tests). When adjustment is slow it may be the case that the data are integrated, which means that no method developed for the stationary case is appropriate. At the cross-sectional level, it is argued that the critical issue is assessing heterogeneity; a variety of strategies for this assessment are discussed.
Beck, Nathaniel, Katz, Jonathan
Submitted: 2004-07-24
Keywords: Time-series--cross-section data, lagged dependent variables, Nickell bias, specification, integration
Abstract: (click to show/hide) This paper deals with a variety of dynamic issues in the analysis of time-series--cross-section (TSCS) data raised by recent papers; it also more briefly treats some cross-sectional issues. 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). It is also shown that there is nothing pernicious in using a lagged dependent variable, and 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 models, and analysts may choose on grounds of convenience (assuming that the model passes standard econometric tests). When adjustment is slow it may be the case that the data are integrated, which means that no method developed for the stationary case is appropriate. At the cross-sectional level, it is argued that the critical issue is assessing heterogeneity; a variety of strategies for this assessment are discussed.
Empirical Modeling Strategies for Spatial Interdependence: Omitted-Variable vs. Simultaneity Biases
Hays, Jude, Franzese, Robert
Submitted: 2004-07-24
Keywords: Spatial Lag Models, Diffusion, Omitted Variable Bias, Simultaneity Bias
Abstract: (click to show/hide) Scholars recognize that time-series-cross-section data typically correlate across time and space, yet they tend to model temporal dependence directly while addressing spatial interdependence solely as nuisance to be “corrected” (FGLS) or to which to be “robust” (PCSE). We demonstrate that directly modeling spatial interdependence is methodologically superior, offering efficiency gains and generally helping avoid biased estimates even of “non-spatial” effects. We first specify empirical models representing two modern approaches to comparative and international political economy: (context-conditional) open-economy comparative political-economy (i.e., common stimuli, varying responses) and international political-economy, which implies interdependence (plus closed-economy and orthogonal-open-economy predecessors). Then we evaluate four estimators—non-spatial OLS, spatial OLS, spatial 2SLS-IV, and spatial ML—for analyzing such models in spatially interdependent data. Non-spatial OLS suffers from potentially severe omitted-variable bias, tending to inflate estimates of common-stimuli effects especially. Spatial OLS, which specifies interdependence directly via spatial lags, dramatically improves estimates but suffers a simultaneity bias, which can be appreciable under strong interdependence. Spatial 2SLS-IV, which instruments for spatial lags of dependent variables with spatial lags of independent variables, yields unbiased and reasonably efficient estimates of both common-stimuli and diffusion effects, when its conditions hold: large samples and fully exogenous instruments. A tradeoff thus arises in practice between biased-but-efficient spatial OLS and consistent- (or, at least, less-biased-) but-inefficient spatial 2SLS-IV. Spatial ML produces good estimates of non-spatial effects under all conditions but is computationally demanding and tends to underestimate the strength of interdependence, appreciably so in small-N samples and when the true diffusion-strength is modest. We also explore the standard-error estimates from these four procedures, finding sizable inaccuracies by each estimator under differing conditions, and PCSE’s do not necessarily reduce these inaccuracies. By an accuracy-of-reported-standard-errors criterion, 2SLS-IV seems to dominate. Finally, we explore the spatial 2SLS-IV estimator under varying patterns of interdependence and endogeneity, finding that its estimates of diffusion strength suffer only when a condition we call cross-spatial endogeneity, wherein dependent variables (y’s) in some units cause explanatory variables (x’s) in others, prevails.
Hays, Jude, Franzese, Robert
Submitted: 2004-07-24
Keywords: Spatial Lag Models, Diffusion, Omitted Variable Bias, Simultaneity Bias
Abstract: (click to show/hide) Scholars recognize that time-series-cross-section data typically correlate across time and space, yet they tend to model temporal dependence directly while addressing spatial interdependence solely as nuisance to be “corrected” (FGLS) or to which to be “robust” (PCSE). We demonstrate that directly modeling spatial interdependence is methodologically superior, offering efficiency gains and generally helping avoid biased estimates even of “non-spatial” effects. We first specify empirical models representing two modern approaches to comparative and international political economy: (context-conditional) open-economy comparative political-economy (i.e., common stimuli, varying responses) and international political-economy, which implies interdependence (plus closed-economy and orthogonal-open-economy predecessors). Then we evaluate four estimators—non-spatial OLS, spatial OLS, spatial 2SLS-IV, and spatial ML—for analyzing such models in spatially interdependent data. Non-spatial OLS suffers from potentially severe omitted-variable bias, tending to inflate estimates of common-stimuli effects especially. Spatial OLS, which specifies interdependence directly via spatial lags, dramatically improves estimates but suffers a simultaneity bias, which can be appreciable under strong interdependence. Spatial 2SLS-IV, which instruments for spatial lags of dependent variables with spatial lags of independent variables, yields unbiased and reasonably efficient estimates of both common-stimuli and diffusion effects, when its conditions hold: large samples and fully exogenous instruments. A tradeoff thus arises in practice between biased-but-efficient spatial OLS and consistent- (or, at least, less-biased-) but-inefficient spatial 2SLS-IV. Spatial ML produces good estimates of non-spatial effects under all conditions but is computationally demanding and tends to underestimate the strength of interdependence, appreciably so in small-N samples and when the true diffusion-strength is modest. We also explore the standard-error estimates from these four procedures, finding sizable inaccuracies by each estimator under differing conditions, and PCSE’s do not necessarily reduce these inaccuracies. By an accuracy-of-reported-standard-errors criterion, 2SLS-IV seems to dominate. Finally, we explore the spatial 2SLS-IV estimator under varying patterns of interdependence and endogeneity, finding that its estimates of diffusion strength suffer only when a condition we call cross-spatial endogeneity, wherein dependent variables (y’s) in some units cause explanatory variables (x’s) in others, prevails.
The Varying Role of Voter Information across Democratic Societies
Sekhon, Jasjeet
Submitted: 2004-07-26
Keywords: Voter Information, Elections, Causal Inference, Matching, Propensity Score Matching, Robust Estimation, Democratization, Survey Data
Abstract: (click to show/hide) Using new robust matching methods for making causal inferences from survey data, I demonstrate that there are profound differences between how voters behave in mature democracies versus how they behave in new ones. The problems of voter ignorance and inattentiveness are not as serious in mature democracies as many analysts have suggested but are of grave concern in new democracies. Citizens in mature democracies are able to accomplish something that citizens in fledgling democracies are not: inattentive and poorly informed citizens are able to vote like their better informed compatriots and hence need to pay little attention to political events such as election campaigns in order to vote as if they were attentive. The results from the U.S. (which rely on various National Election Studies) and Mexico (2000 Panel Study) are reported in detail. Results from other countries are briefly reported.
Sekhon, Jasjeet
Submitted: 2004-07-26
Keywords: Voter Information, Elections, Causal Inference, Matching, Propensity Score Matching, Robust Estimation, Democratization, Survey Data
Abstract: (click to show/hide) Using new robust matching methods for making causal inferences from survey data, I demonstrate that there are profound differences between how voters behave in mature democracies versus how they behave in new ones. The problems of voter ignorance and inattentiveness are not as serious in mature democracies as many analysts have suggested but are of grave concern in new democracies. Citizens in mature democracies are able to accomplish something that citizens in fledgling democracies are not: inattentive and poorly informed citizens are able to vote like their better informed compatriots and hence need to pay little attention to political events such as election campaigns in order to vote as if they were attentive. The results from the U.S. (which rely on various National Election Studies) and Mexico (2000 Panel Study) are reported in detail. Results from other countries are briefly reported.
Models of Intertemporal Choice
Wand, Jonathan
Submitted: 2004-07-26
Keywords: choice, extremal process, utility maximizing, dynamic, discrete, lagged dependent variable, panel
Abstract: (click to show/hide) In this paper, I consider the behavior of individuals making repeated choices over a finite set of discrete alternatives. Individuals are assumed to maximize utility each time they are faced with a choice, without affecting the utility or availability of future choices. I build on a class of models where serial correlation in choices is due to a process of learning over time about the merits of alternatives, rather than due to unobserved persistent effects. I provide new analytical results for characterizing transition probabilities between choices without imposing restrictions on how the systematic component of utilities may change over time.
Wand, Jonathan
Submitted: 2004-07-26
Keywords: choice, extremal process, utility maximizing, dynamic, discrete, lagged dependent variable, panel
Abstract: (click to show/hide) In this paper, I consider the behavior of individuals making repeated choices over a finite set of discrete alternatives. Individuals are assumed to maximize utility each time they are faced with a choice, without affecting the utility or availability of future choices. I build on a class of models where serial correlation in choices is due to a process of learning over time about the merits of alternatives, rather than due to unobserved persistent effects. I provide new analytical results for characterizing transition probabilities between choices without imposing restrictions on how the systematic component of utilities may change over time.
An Experimental Test of Proximity and Directional Voting
Paolino, Philip, Lacy, Dean
Submitted: 2004-07-27
Keywords: experiment, issue voting, directional
Abstract: (click to show/hide) Lewis and King (2000) discuss difficulties in evaluating the proximity hypothesis about issue voting versus the directional hypothesis. In this paper, we propose as a solution to this problem is asking individuals to evaluate candidates generated to represent certain issue positions experimentally. Such an approach controls candidates' positions, while holding other features constant, presents these fictitious candidates to randomly assigned subjects, and examines whether the relationship between subjects' evaluations of these candidates and their ideological beliefs is consistent with proximity or directional theory. Our results provide slightly more support for proximity theory, but our data are not entirely conclusive on this point.
Paolino, Philip, Lacy, Dean
Submitted: 2004-07-27
Keywords: experiment, issue voting, directional
Abstract: (click to show/hide) Lewis and King (2000) discuss difficulties in evaluating the proximity hypothesis about issue voting versus the directional hypothesis. In this paper, we propose as a solution to this problem is asking individuals to evaluate candidates generated to represent certain issue positions experimentally. Such an approach controls candidates' positions, while holding other features constant, presents these fictitious candidates to randomly assigned subjects, and examines whether the relationship between subjects' evaluations of these candidates and their ideological beliefs is consistent with proximity or directional theory. Our results provide slightly more support for proximity theory, but our data are not entirely conclusive on this point.
Monotone Comparative Statics in Models of Politics: A Method for Simplifying Analysis and Enhancing Empirical Content
Bueno de Mesquita, Ethan, Ashworth, Scott
Submitted: 2004-08-18
Keywords: game theory, formal theory, empirical implications of theoretical models, comparative statics,
Abstract: (click to show/hide) We elucidate a powerful yet simple method for deriving comparative statics conclusions for a wide variety of models: Monotone Comparative Statics (Milgrom and Shannon, 1994). Monotone comparative static methods allow researchers to extract robust, substantive empirical implications from formal models that can be tested using ordinal data and simple non-parametric tests. They also replace a diverse range of more technically di±cult mathematics (facilitating richer, more realistic models), a large set of assumptions that are hard to understand or justify substantively (highlighting the political intuitions underlying a model's results), and a complicated set of methods for extracting implications from models. We present an accessible introduction to the central monotone comparative statics results and a series of practical tools for using these techniques in applied models (with reference to original sources, when relevant). Throughout we demonstrate the techniques with examples drawn from political science.
Bueno de Mesquita, Ethan, Ashworth, Scott
Submitted: 2004-08-18
Keywords: game theory, formal theory, empirical implications of theoretical models, comparative statics,
Abstract: (click to show/hide) We elucidate a powerful yet simple method for deriving comparative statics conclusions for a wide variety of models: Monotone Comparative Statics (Milgrom and Shannon, 1994). Monotone comparative static methods allow researchers to extract robust, substantive empirical implications from formal models that can be tested using ordinal data and simple non-parametric tests. They also replace a diverse range of more technically di±cult mathematics (facilitating richer, more realistic models), a large set of assumptions that are hard to understand or justify substantively (highlighting the political intuitions underlying a model's results), and a complicated set of methods for extracting implications from models. We present an accessible introduction to the central monotone comparative statics results and a series of practical tools for using these techniques in applied models (with reference to original sources, when relevant). Throughout we demonstrate the techniques with examples drawn from political science.
Entropy optimization: computer implementation of the MaxEnt and MinxEnt principles
Veiga, Alvaro, Mattos, Rogerio
Submitted: 2004-09-01
Keywords: entropy optimization, Shannon's measure, Kullback's measure, ecological inference, MaxEnt, MinxEnt
Abstract: (click to show/hide) The entropy optimization principles MaxEnt of Jaynes (1957a,b) and MinxEnt of Kullback (1959) can be used in a variety of scientific fields. Among many possible applications, the principles are suitable to tackle the ecological inference problem that often shows up in social science research. Formally, both principles involve the constrained optimization of entropy measures that are intrinsically nonlinear functions of probabilities. Since each is a nonlinear programming problem, the solutions to both depend on iterative search algorithms. In addition, the constraints that probabilities are non–negative and sum up to one restrict in a particular way the solution space. The paper presents in detail a computer efficient implementation of those two principles in the linearly constrained case that makes a prior check for the existence of solution to the optimization problems. A description, made with the aid of two flowcharts, of an algorithm allows interested researchers to develop computer codes in practically any language. The authors also make available their own, easy–to–use codes written in MatLab.
Veiga, Alvaro, Mattos, Rogerio
Submitted: 2004-09-01
Keywords: entropy optimization, Shannon's measure, Kullback's measure, ecological inference, MaxEnt, MinxEnt
Abstract: (click to show/hide) The entropy optimization principles MaxEnt of Jaynes (1957a,b) and MinxEnt of Kullback (1959) can be used in a variety of scientific fields. Among many possible applications, the principles are suitable to tackle the ecological inference problem that often shows up in social science research. Formally, both principles involve the constrained optimization of entropy measures that are intrinsically nonlinear functions of probabilities. Since each is a nonlinear programming problem, the solutions to both depend on iterative search algorithms. In addition, the constraints that probabilities are non–negative and sum up to one restrict in a particular way the solution space. The paper presents in detail a computer efficient implementation of those two principles in the linearly constrained case that makes a prior check for the existence of solution to the optimization problems. A description, made with the aid of two flowcharts, of an algorithm allows interested researchers to develop computer codes in practically any language. The authors also make available their own, easy–to–use codes written in MatLab.
On estimates of split-ticket voting
Johnston, Ron, Gschwend, Thomas, Pattie, Charles
Submitted: 2004-09-20
Keywords: split-ticket, Ecological Inference, Entropy-Maximizing
Abstract: (click to show/hide) Kimball on split-ticket voting in the USA, suggesting that their estimates of the volume of such voting (derived using King's EI method) across Congressional Districts and States are unreliable. Using part of the Burden-Kimball data set, we report on a parallel set of estimates generated by a different procedure (EMax), which employs three rather than two sets of bounds. The results are extremely similar to Burden and Kimball's, providing strong circumstantial evidence for their conclusions regarding the impact of campaign spending and other influences on the volume of split-ticket voting.
Johnston, Ron, Gschwend, Thomas, Pattie, Charles
Submitted: 2004-09-20
Keywords: split-ticket, Ecological Inference, Entropy-Maximizing
Abstract: (click to show/hide) Kimball on split-ticket voting in the USA, suggesting that their estimates of the volume of such voting (derived using King's EI method) across Congressional Districts and States are unreliable. Using part of the Burden-Kimball data set, we report on a parallel set of estimates generated by a different procedure (EMax), which employs three rather than two sets of bounds. The results are extremely similar to Burden and Kimball's, providing strong circumstantial evidence for their conclusions regarding the impact of campaign spending and other influences on the volume of split-ticket voting.
Difficult Choices: An Evaluation of Heterogenous Choice Models
Keele, Luke, Park, David K.
Submitted: 2004-10-14
Keywords: probit, discrete choice, heteroskedasticity, heteroskedastic probit
Abstract: (click to show/hide) While the derivation and estimation of heterogeneous choice models appears straightforward, the properties of such models are not well understood. Using a series of Monte Carlo experiments, we focus on the properties of both heteroskedastic probit and heteroskedastic ordered probit models. We also test how robust these models are to both specification and measurement error. We find that estimates in heterogeneous choice models tend to be biased in all but ideal conditions, and can often lead to incorrect inferences.
Keele, Luke, Park, David K.
Submitted: 2004-10-14
Keywords: probit, discrete choice, heteroskedasticity, heteroskedastic probit
Abstract: (click to show/hide) While the derivation and estimation of heterogeneous choice models appears straightforward, the properties of such models are not well understood. Using a series of Monte Carlo experiments, we focus on the properties of both heteroskedastic probit and heteroskedastic ordered probit models. We also test how robust these models are to both specification and measurement error. We find that estimates in heterogeneous choice models tend to be biased in all but ideal conditions, and can often lead to incorrect inferences.
Taking Time Seriously: Dynamic Regression
Keele, Luke, De Boef, Suzanna
Submitted: 2004-10-14
Keywords: Time series, error correction models, lagged dependent variables, ADL models
Abstract: (click to show/hide) Dramatic change in the world around us has stimulated a wealth of interest in research questions about the dynamics of political processes. At the same time, we have seen increases in the number of time series data sets and the length of typical time series. While advances in time series methods have helped us to think about political change in important ways, too often published time series analysis displays shortcomings in three areas. First, analysts often estimate models without testing the restrictions implied by their specification. Second, applied researchers link the theoretical concept of equilibrium with the existence of cointegration and use of error correction models. Third, those estimating time series models have often done a poor job of interpreting their statistical results. The consequences, at best, are poor connections between theory and tests and thus a limited cumulation of knowledge. Often, the costs include biased results and incorrect inferences as well. Here, we outline techniques for the estimation of linear models with dynamic specification. In general, we recommend that analysts start with a combination of general dynamic models and test for restrictions before adopting a particular specification. Finally, we recommend that analysts make use of the wide array of information that can be gleaned from dynamic specifications. We illustrate this strategy with data Congressional approval and tax rates across OECD countries.
Keele, Luke, De Boef, Suzanna
Submitted: 2004-10-14
Keywords: Time series, error correction models, lagged dependent variables, ADL models
Abstract: (click to show/hide) Dramatic change in the world around us has stimulated a wealth of interest in research questions about the dynamics of political processes. At the same time, we have seen increases in the number of time series data sets and the length of typical time series. While advances in time series methods have helped us to think about political change in important ways, too often published time series analysis displays shortcomings in three areas. First, analysts often estimate models without testing the restrictions implied by their specification. Second, applied researchers link the theoretical concept of equilibrium with the existence of cointegration and use of error correction models. Third, those estimating time series models have often done a poor job of interpreting their statistical results. The consequences, at best, are poor connections between theory and tests and thus a limited cumulation of knowledge. Often, the costs include biased results and incorrect inferences as well. Here, we outline techniques for the estimation of linear models with dynamic specification. In general, we recommend that analysts start with a combination of general dynamic models and test for restrictions before adopting a particular specification. Finally, we recommend that analysts make use of the wide array of information that can be gleaned from dynamic specifications. We illustrate this strategy with data Congressional approval and tax rates across OECD countries.
Social Capital, Government Performance, and the Dynamics of
Keele, Luke
Submitted: 2004-10-14
Keywords: trust in government, social capital, time series, error correction models
Abstract: (click to show/hide) In extant research on trust in government, a tension has developed between whether the movement of trust over time is a function of political performance or political alienation. In performance based explanations trust responds to the economy, Congress and the President and should move frequently over time. Under theories of political alienation, government performance matters little as hostility toward both political leaders and the political process causes distrust of government, and is a direct threat to government legitimacy. Using aggregate data in a time series analysis of trust in government, I find that both political alienation, as measured by social capital and performance have important but differing effects on trust. Government performance has an immediate effect on trust while movement in social capital sets the long-term level of trust. The qualitative outcome is that trust embodies both performance and political alienation and is an important indicator of citizen satisfaction with government.
Keele, Luke
Submitted: 2004-10-14
Keywords: trust in government, social capital, time series, error correction models
Abstract: (click to show/hide) In extant research on trust in government, a tension has developed between whether the movement of trust over time is a function of political performance or political alienation. In performance based explanations trust responds to the economy, Congress and the President and should move frequently over time. Under theories of political alienation, government performance matters little as hostility toward both political leaders and the political process causes distrust of government, and is a direct threat to government legitimacy. Using aggregate data in a time series analysis of trust in government, I find that both political alienation, as measured by social capital and performance have important but differing effects on trust. Government performance has an immediate effect on trust while movement in social capital sets the long-term level of trust. The qualitative outcome is that trust embodies both performance and political alienation and is an important indicator of citizen satisfaction with government.
Analyzing the US Senate in 2003: Similarities, Networks, Clusters and Blocs
Jakulin, Aleks
Submitted: 2004-10-27
Keywords: roll call analysis, latent variable models, MCMC, information theory, clustering, visualization
Abstract: (click to show/hide) To analyze the roll calls in the US Senate in year 2003, we have employed the methods already used throughout the science community for analysis of genes, surveys and text. With information-theoretic measures we assess the association between pairs of senators based on the votes they cast. Furthermore, we can evaluate the influence of a voter by postulating a Shannon information channel between the outcome and a voter. The matrix of associations can be summarized using hierarchical clustering, multi-dimensional scaling and link analysis. With a discrete latent variable model we identify blocs of cohesive voters within the Senate, and contrast it with continuous ideal point methods. Under the bloc-voting model, the Senate can be interpreted as a weighted vote system, and we were able to estimate the empirical voting power of individual blocs through what-if analysis.
Jakulin, Aleks
Submitted: 2004-10-27
Keywords: roll call analysis, latent variable models, MCMC, information theory, clustering, visualization
Abstract: (click to show/hide) To analyze the roll calls in the US Senate in year 2003, we have employed the methods already used throughout the science community for analysis of genes, surveys and text. With information-theoretic measures we assess the association between pairs of senators based on the votes they cast. Furthermore, we can evaluate the influence of a voter by postulating a Shannon information channel between the outcome and a voter. The matrix of associations can be summarized using hierarchical clustering, multi-dimensional scaling and link analysis. With a discrete latent variable model we identify blocs of cohesive voters within the Senate, and contrast it with continuous ideal point methods. Under the bloc-voting model, the Senate can be interpreted as a weighted vote system, and we were able to estimate the empirical voting power of individual blocs through what-if analysis.
Voting cycles and institutional paradoxes: a model of partisan control and change in state politics
Brierly, Allen
Submitted: 2004-11-05
Keywords: EITM, election and voting cycles, measurement of political party competition, state elections
Abstract: (click to show/hide) This study applies a formal model of political competition to analyze partisan control and changes in partisan control of state government. The analysis is a straightforward application of both traditional theories of political parties and a social choice understanding of the role agenda setting plays in electoral competition. The models incorporate the traditional classification and estimation of party competition, while extending the more formal analysis of agenda setting to duopoly competition in a long-run electoral context. The findings synethesize a variety of recent and traditional hypotheses concerning state politics, governance, and elections. The results describe the extent and scope of divided government and compare the stability of unified versus divided partisan control. Theories of party change are also incorporated in the model to test the stability of partisan control and to classify different types of political competition. This study presents both a description and a discussion of the arguments for competition, linking the merits of increasing competition to the consequences of unstable party changes and divided partisan control.
Brierly, Allen
Submitted: 2004-11-05
Keywords: EITM, election and voting cycles, measurement of political party competition, state elections
Abstract: (click to show/hide) This study applies a formal model of political competition to analyze partisan control and changes in partisan control of state government. The analysis is a straightforward application of both traditional theories of political parties and a social choice understanding of the role agenda setting plays in electoral competition. The models incorporate the traditional classification and estimation of party competition, while extending the more formal analysis of agenda setting to duopoly competition in a long-run electoral context. The findings synethesize a variety of recent and traditional hypotheses concerning state politics, governance, and elections. The results describe the extent and scope of divided government and compare the stability of unified versus divided partisan control. Theories of party change are also incorporated in the model to test the stability of partisan control and to classify different types of political competition. This study presents both a description and a discussion of the arguments for competition, linking the merits of increasing competition to the consequences of unstable party changes and divided partisan control.
