Racial/ethnic differences in utilization of section 8 existing rental vouchers and certificates

Abstract

This study examines racial and ethnic differences in the success of Section 8 Existing Housing Program enrollees in using rental vouchers and certificates. The study uses data from a sample of large urban public housing authorities (PHAs) collected by Abt Associates for the Housing Voucher Demonstration.

The study finds that, excluding one site, average success rates for whites, blacks, and Hispanics are quite similar. However, minority success rates are significantly higher than nonminority success rates in some sites, and significantly lower in others. The racial and ethnic differences in success rates appear to be related to the proportion of enrollees in each program that were nonminority. Nonminority whites tend to be less successful than blacks or Hispanics in sites where most enrollees are minority, and more successful in sites where most enrollees are white. We believe this may reflect low enrollee success rates in searching for units outside the Section 8 submarket, as defined by existing recipients.     

Two-Step Hierarchical Estimation: Beyond Regression Analysis

e-mail: achen{at}princeton.edu Two-step estimators for hierarchical models can be constructedeven when neither stage is a conventional linear regressionmodel. For example, the first stage might consist of probitmodels, or duration models, or event count models. The secondstage might be a nonlinear regression specification. This notesketches some of the considerations that arise in ensuring thattwo-step estimators are consistent in such cases.     

WHAT WORKS BEST AND WHEN: ACCOUNTING FOR MULTIPLE SOURCES OF PURESELECTION BIAS IN PROGRAM EVALUATIONS

Most evaluations are still quasi‐experimental and most recent quasi‐experimental methodological research has focused on various types of propensity score matching to minimize conventional selection bias on observables. Although these methods create better‐matched treatment and comparison groups on observables, the issue of selection on unobservables still looms large. Thus, in the absence of being able to run randomized controlled trials (RCTs) or natural experiments, it is important to understand how well different regression‐based estimators perform in terms of minimizing pure selection bias, that is, selection on unobservables. We examine the relative magnitudes of three sources of pure selection bias: heterogeneous response bias, time‐invariant individual heterogeneity (fixed effects [FEs]), and intertemporal dependence (autoregressive process of order one [AR(1)]). Because the relative magnitude of each source of pure selection bias may vary in different policy contexts, it is important to understand how well different regression‐based estimators handle each source of selection bias. Expanding simulations that have their origins in the work of Heckman, LaLonde, and Smith ( 1999 ), we find that difference‐in‐differences (DID) using equidistant pre‐ and postperiods and FEs estimators are less biased and have smaller standard errors in estimating the Treatment on the Treated (TT) than other regression‐based estimators. Our data analysis using the Job Training Partnership Act (JTPA) program replicates our simulation findings in estimating the TT.     

Front‐Door Difference‐in‐Differences Estimators

We develop front‐door difference‐in‐differences estimators as an extension of front‐door estimators. Under one‐sided noncompliance, an exclusion restriction, and assumptions analogous to parallel trends assumptions, this extension allows identification when the front‐door criterion does not hold. Even if the assumptions are relaxed, we show that the front‐door and front‐door difference‐in‐differences estimators may be combined to form bounds. Finally, we show that under one‐sided noncompliance, these techniques do not require the use of control units. We illustrate these points with an application to a job training study and with an application to Florida's early in‐person voting program. For the job training study, we show that these techniques can recover an experimental benchmark. For the Florida program, we find some evidence that early in‐person voting had small positive effects on turnout in 2008. This provides a counterpoint to recent claims that early voting had a negative effect on turnout in 2008.     

Robust Estimation and Outlier Detection for Overdispersed Multinomial Models of Count Data

We develop a robust estimator—the hyperbolic tangent (tanh) estimator—for overdispersed multinomial regression models of count data. The tanh estimator provides accurate estimates and reliable inferences even when the specified model is not good for as much as half of the data. Seriously ill-fitted counts—outliers—are identified as part of the estimation. A Monte Carlo sampling experiment shows that the tanh estimator produces good results at practical sample sizes even when ten percent of the data are generated by a significantly different process. The experiment shows that, with contaminated data, estimation fails using four other estimators: the nonrobust maximum likelihood estimator, the additive logistic model and two SUR models. Using the tanh estimator to analyze data from Florida for the 2000 presidential election matches well-known features of the election that the other four estimators fail to capture. In an analysis of data from the 1993 Polish parliamentary election, the tanh estimator gives sharper inferences than does a previously proposed heteroskedastic SUR model .     

A Seemingly Unrelated Regression Model for Analyzing Multiparty Elections

This paper develops an estimator for models of election returnsin multiparty elections. It shares the same functional formasthe Katz–King estimator but is computationally simpler,can be used with any number of parties, and is based on moreconventional distributional assumptions. Small sample propertiesof the estimator are derived, which makes it particularly usefulin many of the applications where there are a relatively smallnumber of voting districts. The distributional assumptions arecontained in two elements. The first treats the observed votesas the outcomes resulting from sampling the voters in each district.The second stochastic element arises from the usual treatmentof the stochastic term in a regression model, namely, the inabilityof the included variables and the linear form to match the underlyingprocess perfectly. The model is then used to analyze the 1993Polish parliamentary elections. The results from this analysisare used to develop Monte Carlo experiments comparing severaldifferent yet feasible estimators. The conclusion is that anumber of accessible estimators, including the standard seeminglyunrelated regression model and the Beck–Katz model withpanel-corrected standard errors, are all good choices.     

Can Nonexperimental Estimates Replicate Estimates Based on Random Assignment in Evaluations of School Choice? A Within‐Study Comparison

The ability of nonexperimental estimators to match impact estimates derived from random assignment is examined using data from the evaluation of two interdistrict magnet schools. As in previous within‐study comparisons, nonexperimental estimates differ from estimates based on random assignment when nonexperimental estimators are implemented without pretreatment measures of academic performance. With comparison groups consisting of students drawn from the same districts or districts with similar student body characteristics as the districts where treatment group students reside, using pretreatment test scores reduces the bias in nonexperimental methods between 64 and 96 percent. Adding pretreatment test scores does not achieve as much bias reduction when the comparison group consists of students drawn from districts with different student body characteristics than the treatment group students’ districts. The results suggest that using pretreatment outcome measures and comparison groups that are geographically aligned with the treatment group greatly improves the performance of nonexperimental estimators.     

On Formal Theory and Statistical Methods: A Response to Carrubba, Yuen, and Zorn

e-mail: sign{at}mail.rochester.edu The recent article by Carrubba, Yuen, and Zorn (2007) (CYZ)attempts to relate the strategic random utility models in Signorino(1999, 2002, 2003) and in Signorino and Yilmaz (2003) to existinggame theory practice and to existing statistical techniques.It contributes to this literature by reminding us that comparativestatics analysis can be applied to the equilibria of these models.There are a number of claims in CYZ, however, that require clarification.In particular, the article's primary claim is that comparativestatics analysis, in combination with one of three proposedstatistical estimators, provides a simpler alternative to methodspreviously advocated. This claim (or combination of claims)is incorrect. When one examines the procedure CYZ recommends,it is no simpler for substantive researchers than anything previouslyrecommended. Moreover, none of the proposed estimators are new:they are exactly the same methods introduced in Signorino (1999,2003), in Signorino and Yilmaz (2003), in Signorino, Walker,and Bas (2002), and in Bas, Signorino, and Walker (2007).     

Estimating Dynamic Panel Data Models in Political Science

Panel data are a very valuable resource for finding empiricalsolutions to political science puzzles. Yet numerous publishedstudies in political science that use panel data to estimatemodels with dynamics have failed to take into account importantestimation issues, which calls into question the inferenceswe can make from these analyses. The failure to account explicitlyfor unobserved individual effects in dynamic panel data inducesbias and inconsistency in cross-sectional estimators. The purposeof this paper is to review dynamic panel data estimators thateliminate these problems. I first show how the problems withcross-sectional estimators arise in dynamic models for paneldata. I then show how to correct for these problems using generalizedmethod of moments estimators. Finally, I demonstrate the usefulnessof these methods with replications of analyses in the debateover the dynamics of party identification.     

Beyond the Unit Root Question: Uncertainty and Inference

A fundamental challenge facing applied time-series analysts is how to draw inferences about long-run relationships (LRR) when we are uncertain whether the data contain unit roots. Unit root tests are notoriously unreliable and often leave analysts uncertain, but popular extant methods hinge on correct classification. Webb, Linn, and Lebo (WLL; 2019) develop a framework for inference based on critical value bounds for hypothesis tests on the long-run multiplier (LRM) that eschews unit root tests and incorporates the uncertainty inherent in identifying the dynamic properties of the data into inferences about LRRs. We show how the WLL bounds procedure can be applied to any fully specified regression model to solve this fundamental challenge, extend the results of WLL by presenting a general set of critical value bounds to be used in applied work, and demonstrate the empirical relevance of the LRM bounds procedure in two applications.     

Subsidizing the affluent: the case of medical education.

State and federal funds are important sources of revenue for medical schools, and a strong case can be made for public support for these institutions. Although the federal role is more widely known, the states in fact provide the bulk of direct support for medical training. The nature of aid from the two sources differs in significant ways. Most federal aid supports research or patient care, but much of state aid goes to support unconditional tuition subsidies. The primary beneficiaries of these subsidies are relatively affluent nonminority students who are beginning lucrative careers in the medical field--careers that would be lucrative even if no subsidies were provided. Nor does it appear that general subsidies are needed to attract poor and minority applicants. While targeted loans and scholarships to individuals may be justified, general tuition subsidies are not.     

Equity and information: Information regulation,environmental justice,and risks from toxic chemicals

Decreases over time in pounds of industrial chemical emissions have led to concerns that nonminority, higher‐income communities have benefited disproportionately in reductions in risk. Toxic chemical release data, modeled for toxicity and dispersion in square kilometer units across 45 states, are used to test six sets of hypotheses of potential interactions between facilities and the communities surrounding them. The results provide mixed evidence that ethnicity explains chemical emission changes. Stronger support is found that changes in emission patterns are affected by a community's ability to process complex information and its capacity for collective action, and industrial production factors. Some support is found for the efficacy of state environmental policies that both do and do not fit the traditional regulatory model in encouraging risk reductions. © 2005 by the Association for Public Policy Analysis and Management     

Not Taking it on Faith: State and Religious Influences on European Court of Human Rights Judges in Freedom of Religion Cases

I analyze freedom of religion case law from the European Court of Human Rights (ECHR), arguing that the ECHR has been inconsistent and deferential to state governments. To account for this phenomenon, I consider two theoretical frameworks. First, I ask whether the religious affiliation of applicants matters when it comes to judges’ willingness to decide in their favor. Second, I consider whether ECHR judges are influenced by the preferences of the states that nominated them. Employing logistic regression analysis on an original dataset of 538 judge votes on 70 freedom of religion cases, I find that Muslim applicants are substantially less likely to receive favorable judge votes, and that judges are more likely to vote in favor of the states that nominated them. Additionally, I find that West European states are especially likely to receive favorable rulings, even when controlling for differences in state protections of religious liberty.     

Causal Complexity and Party Preference

Abstract.  Qualitative Comparative Analysis (QCA) overlaps logistic regression in explaining events, but challenges the latter's lack of accounting for causal complexity. QCA has only to a limited degree been applied to large-N studies or individuals as cases and has not incorporated the logic of probability. QCA and logistic regression are compared with respect to logic, procedure and outcome. Political orientations from five national surveys are adapted to the requirements of the two methods. The methods are demonstrated on explanations of individuals' party preferences. QCA and logistic regression converge and overlap in identifying degrees of causal complexity, in ascertaining model significance and in identifying antecedents to party preference. Results differ in degree, not in kind. A slightly more nuanced picture emerges using the QCA approach, whereas logistic regression delivers greater parsimony. Choice of method(s) is not arbitrary. QCA can easily be used on any large-N research problem. It should apply probability when appropriate.     

Change Agents or Cogs in the Machine? Female Managers and Unofficial Gender Equality in Federal Agencies

This article combines administrative personnel data with survey data to examine whether increases in female representation among the managerial ranks of federal agencies are associated with increases in unofficial gender equality in those agencies. Specifically, female representation at multiple levels of agencies’ personnel hierarchies are measured so as to examine whether these measures are associated with increases in lower-level female and male employees’ subjective experiences of power, status, and opportunity. The results indicate that female representation among the SES has salutary effects on General Schedule (GS) level 7–12 and General Schedule level 13–15 females. Additionally, the results suggest that female representation among GS 13–15 supervisors has salutary effects on GS 7–12 females. Finally, the results suggest that top-level female representation has negative effects on GS 7–12 males. These findings speak to the ongoing theoretical debate concerning female managers’ ability and willingness to bring about equality-inducing change in their organizations. They also speak to the question of whether a representative bureaucracy has unfavorable distributional consequences for members of nonminority groups.     

Understanding unequal turnout: Education and voting in comparative perspective

Well-educated citizens vote more frequently than the poorly educated in some countries, including the USA. However, in many countries, no such differences are observed. One classical explanation of the presence or absence of this inequality in voting is that the strength of left-wing forces sharpens or reduces it. An alternative explanation is that some institutional arrangements and contextual features disproportionately affect the voter participation of some individuals depending on their resources, thus shaping turnout inequality. These theories are tested using multilevel modeling with data from 28 advanced industrial democracies. Compulsory voting reduces inequalities because under this system quasi-universal turnout is achieved. In addition, the poorly educated vote more frequently when the voting procedure is easy and when there are few political parties, thus reducing turnout inequality. However, strong left-wing parties and trade unions are not associated with more equal turnout.     

Distinguishing territorial structure from electoral adventurism: The distinct sources of static and dynamic nationalization

Estimates of static nationalization do not always reflect stark qualitative differences between parties. We use a research design oriented around a comparison of sharply different parties—the unstable Democratic Left in Ecuador and the stable Broad Front in Uruguay—to develop the distinctiveness of static and dynamic nationalization. Snapshot measures that only consider a single election suggest that both parties are poorly statically nationalized; but we show that the former case is highly statically nationalized, and that the observed territorial differences arise because it is poorly dynamically nationalized. We adopt the linear mixed modeling approach to reduce the bias in extant estimators. The approach is also informative about the sources of variance in a party's territorial support: relatively stable district attributes account for static nationalization, while features unique to the electoral cycle account for dynamic nationalization. Substantively, our study alters conclusions about parties operating in highly unstable electoral contexts.     

Logical Inconsistency in EI-Based Second-Stage Regressions

The statistical procedure EI–R, in which point estimates produced by the King (1997) ecological inference technique are used as dependent variables in a linear regression, can be logically inconsistent insofar as the assumptions necessary to support EI–R's first stage (ecological inference via King's technique) can be incompatible with the assumptions supporting its second stage (linear regression). In light of this problem, we develop a specification test for logical consistency of EI–R and describe options available to a researcher who confronts test rejection. We then apply our test to the implementation of EI–R in Burden and Kimball's (1998) study of ticket splitting and find that this implementation is logically inconsistent. In correcting for this problem we show that Burden and Kimball's substantive results are artifacts of a self-contradictory statistical technique.     

Racial Diversity and Judicial Influence on Appellate Courts

This article evaluates the substantive consequences of judicial diversity on the U.S. Courts of Appeals. Due to the small percentage of racial minorities on the federal bench, the key question in evaluating these consequences is not whether minority judges vote differently from nonminority judges, but whether their presence on appellate courts influences their colleagues and affects case outcomes. Using matching methods, I show that black judges are significantly more likely than nonblack judges to support affirmative action programs. This individual‐level difference translates into a substantial causal effect of adding a black judge to an otherwise all‐nonblack panel. Randomly assigning a black counterjudge—a black judge sitting with two nonblack judges—to a three‐judge panel of the Courts of Appeals nearly ensures that the panel will vote in favor of an affirmative action program. These results have important implications for assessing the relationship between diversity and representation on federal courts.     

Switzerland's Future in Europe: Isolation or Constitutional Change

In order to move closer to the European Community and eventually to join it, Switzerland must reshape its political institutions. A major constitutional reform should provide for a real head of government, adapt direct democracy to the constraints of supra-nationality, and simplify the legislative process. In the ongoing debate in Switzerland, the question is not so much whether such reforms are necessary, but when they should occur. The author argues that the dominant doctrine of negotiating first with the EC and reforming later is not realistic. The so-called "double majority referendum," which is necessary for ratifying membership either in the European Economic Space or in the EC itself, is given as an illustration. This ratification procedure is not in accordance with the principle of "one man, one vote," and is clearly biased against Europe.