Source · aikyam school

Heterogeneous Preferences and the Efficacy of Public School Choice

Original source

Justine S. Hastings; Thomas J. Kane; Douglas O. Staiger

Heterogeneous Preferences and the Efficacy of Public School Choice

working paper

Search for this source ↗

Permission for School’s use is unestablished.

Used in these cards

Incorporate income-contingent priority tiers in school choice allocation algorithms to guarantee low-income access to low-poverty schools.

"In addition, the district gave a 'priority boost' to low-income students choosing to attend schools with low concentrations of low-income students. Hence, choices for top schools by students with underperforming home-schools would be given top priority."

Simulate demand elasticities separately across performance tiers rather than applying aggregate Herfindahl-Hirschman Indices when evaluating public market incentives.

"We find that demand-side pressure is largest for high-performing schools that serve very elastic families and minimal for low-performing schools that serve inelastic, disadvantaged families."

Calculate empirical Bayes posterior expectations of individual random coefficients to create un-attenuated interaction terms for secondary causal regressions.

"$\hat{\beta}_{i}^{S}$ is a posterior estimate of the weight each parent placed on school tests scores, calculated from our demand model using Bayes' rule... coefficient estimates for terms involving $\hat{\beta}_{i}^{S}$ are not attenuated by the usual measurement error bias because the measurement error $(\beta_{i}^{S}-\hat{\beta}_{i}^{S})$ is uncorrelated with the posterior."

Isolate spatial proximity preferences from endogenous location decisions by estimating demand models on subsamples undergoing exogenous assignment boundary shifts.

"To test for the potential effects of residential sorting on our estimates, we re-estimated our model for the subsample of students who were reassigned... The estimated parameters were qualitatively and quantitatively similar in the reassigned subsample compared to those for the full sample, suggesting that endogenous residential location is not a major source of bias in this data."

Use posterior estimates of individual choice parameters as instrumental variable interactions in randomized trial evaluations to uncover hidden treatment heterogeneity.

"Using the exploded-mixed-logit model to estimate the weight that each parent placed on test scores when choosing a school, we find that students of parents with weights at the 95th percentile experienced significant rises in End of Grade test scores of approximately 0.1 student-level standard deviations, while students of parents placing little value on academics actually experienced declines in achievement."

Distribute simplified, high-salience school performance scorecards directly to low-income households during choice enrollment windows to increase quality elasticity.

"Hastings and Weinstein (2008) find less heterogeneity, with most parents choosing schools with higher test scores, when they receive simplified information on school test scores."

Publish regression-adjusted value-added metrics alongside raw test scores in official school report cards to shift parental demand toward institutional productivity.

"The value-added measure in panel 4 does not fit the observed choice data well at all. This may not be surprising since this statistic is not available to parents and is not easy to calculate given the available average score statistics."

Isolate lottery instrumental variables strictly to oversubscribed marginal priority groups where admission probabilities are strictly bounded between 0 and 1.

"We exploit lottery random assignment by analyzing the subset of students choosing non-home-schools that were oversubscribed and limit our sample to students in marginal priority groups within those schools - priority groups for which lottery number alone determined initial admission."

Elicit ranked preferences in choice surveys to dynamically modify individual choice sets and separate systematic utility weights from extreme-value error terms.

"The multiple-ranked responses provided by each parent creates variation in the choice set by effectively removing the prior chosen school from the subsequent choice set. This choice-set variation allows us to estimate the distribution of values parents place on school characteristics from observed substitution patterns for each individual"

Incorporate random coefficients and demographic interaction terms in discrete choice models to measure preference heterogeneity across consumer subgroups.

"We estimate an exploded-mixed-logit model of demand for schools and find that higher-SES parents are more likely to choose higher-performing schools, while minority families must trade off preferences for high-performing schools against preferences for a predominantly minority school."

Structure busing zones and transportation subsidies explicitly around low-income housing blocks to neutralize distance disincentives when expanding school choice access.

"Focusing first on the means of the coefficients, parents of both elementary and middle school students in all four subsamples were less likely to choose a school far away, and more likely to choose a school that had high test scores, was their home-school, had a majority of students their own race, was their school in the prior year, or was a school in their choice zone (assuring school bus transportation)."

Design school choice policies with targeted information interventions and revised assignment zones to minimize racial composition trade-offs when selecting high-performing academic programs.

"This is particularly true for minority families who must sacrifice attending a predominantly minority school to select a high-test-score school. Our results imply that public school choice may widen rather than narrow the gap in achievement."

Incorporate interaction terms between user baseline ability and income when modeling demand elasticity for public services.

"The positive interactions with school score indicate that the average weight placed on school score increased with both neighborhood income and the student's baseline test score for parents of students in all four subgroups."

Evaluate preference parameters from initial policy implementation years to obtain un-strategic baseline demand estimates before gaming behaviors emerge among participants.

"We tested for the presence of strategic behavior in the first year of choice by exploiting the redrawing of school boundaries... We did not, however, find significant differences in the exploded-logit parameters (particularly for the coefficients relating to school test scores) across families experiencing positive and negative shocks to their guaranteed school"