Using group-based trajectory modeling in conjunction with propensity scores to improve balance

This paper builds upon two prior papers by Haviland and Nagin (Psychometrika 70:1–22, 2005) and Haviland, Nagin, and Rosenbaum (Working paper, 2006) that attempt to bring the key attributes of an experiment to the analysis of non-experimental longitudinal data. Using a case study of the facilitation effect of gang membership on violence, it systematically examines the contribution of group-based trajectory modeling to the achievement of covariate balance in observational data. In this case study, inclusion of the posterior probabilities of group membership (PPGM), from a model on the pre-treatment measures of the outcome variable, created closer balance on these key covariates than did analyses that did not include them. Still closer balance was obtained on these key covariates by stratifying the analysis by trajectory group. This stratification was achieved by fitting separate propensity score models and matching gang joiners to gang abstainers within trajectory group. In addition, we demonstrated that further balance could be obtained on additional covariates by including PPGM from a model on pre-treatment longitudinal data of these covariates. While this case study is only one empirical example, we believe that it provides useful empirical evidence on the value of performing within trajectory group causal inference in observational longitudinal data and on the use of the PPGM in achieving balance in propensity score-based causal inference.

Policy options for combatting tax noncompliance

The Internal Revenue Service estimates that the “tax gap,” the difference between taxes owed and taxes voluntarily paid, is approaching 100 billion. This article examines seven categories of enforcement instruments for combatting noncompliance. The discussion has two objectives: (1) to summarize what is known about the effectiveness of each type of instrument in reducing noncompliance, and (2) to delineate criteria for evaluating the desirability of enforcement interventions of each type. Suggested criteria differ across classes of instruments, but several themes are repeatedly emphasized. One is that a relevant dimension of effectiveness is equity; a second is that private costs matter, particularly when incurred by the already compliant and independent third parties.     

Comparing the effects of community service and short-term imprisonment on recidivism: a matched samples approach

This study uses longitudinal official record data on adult offenders in The Netherlands (n = 4,246) to compare recidivism after community service to that after short-term imprisonment. To account for possible bias due to selection of offenders into these types of sanctions, we control for a large set of confounding variables using a combined method of ‘matching by variable’ and ‘propensity score matching’. Our findings demonstrate that offenders recidivate significantly less after having performed community service compared to after having been imprisoned. This finding holds for both the short- and long-term. Furthermore, using the Rosenbaum bounds method, we show that the results are robust for hidden bias.     

Response to “Methodological Sensitivities to Latent Class Analysis of Long-Term Criminal Trajectories”

This response to Methodological Sensitivities to Latent Class Analysis of Long-Term Criminal Trajectories is intended to make three key points: (1) the data sensitivities that are explored in that paper are examples of two generic data problems that transcend the specific methodology used to analyze longitudinal data, (2) while real sensitivities are identified, the overall stability of results is also noteworthy, and (3) the groups must be understood to be approximations.     

Statistical Sampling in Tax Audits

The courts, with some important qualifications, have been reluctant to uphold tax assessments based on a review of only a sample of all transactions. In this article we argue that audit assessments based on appropriately drawn and analyzed statistical samples do not suffer from the defects that the courts have correctly concluded mar assessments based on nonstatistical samples. We do, however, argue that because of the inherent imprecision of assessments based on a less-than-complete review of all records, the calculation of the assessment should include a factor to take into account the risk that the taxpayer has been overassessed. We suggest an assessment rule that does just this and also recommend guidelines for the use of statistical sampling in tax audits.     

Population Heterogeneity and State Dependence: State of the Evidence and Directions for Future Research

Criminological research has consistently uncovered a positive correlationbetween past and current criminal behavior. Continuity in offending overtime can be attributed to at least two processes—populationheterogeneity and state dependence. A population heterogeneity processattributes stability in offending over time to differences in ananti-social characteristic (self-control, impulsivity, psychopathicpersonality) across persons that is established early in life andtime-stable thereafter. An implication of a population heterogeneityexplanation for continuity in offending over time is that the anti-socialcharacteristic is likely to have reverberations throughout life, takingmany manifestations later in life (unemployment, drug addiction, maritalinstability). Any observed correlation between these later life events andcriminality, therefore, is spurious rather than causal, due to the factthat they are all the effects of a common cause. A state dependenceexplanation, in contrast, attributes observed stability in criminaloffending to a process of contagion. That is, criminal behavior has agenuine causal effect on subsequent criminality by eroding constraintsand strengthening incentives to crime. The implication of a statedependence process is that criminal conduct may be influenced by laterlife events. In this paper, we draw a connection between populationheterogeneity and state dependence processes and extant criminologicaltheory. We also review the literature that has investigated theplausibility of these two processes. Finally, concluding that weknow very little about them we make recommendations for future researchon population heterogeneity and state dependence.