Assessing Stability and Change in Criminal Offending: A Comparison of Random Effects, Semiparametric, and Fixed Effects Modeling Strategies

An important theoretical problem for criminologists is an explanation forthe robust positive correlation between prior and future criminaloffending. Nagin and Paternoster (1991) have suggested that the correlationcould be due to time-stable population differences in the underlyingproneness to commit crimes (population heterogeneity) and/or thecriminogenic effect that crime has on social bonds, conventionalattachments, and the like (state dependence). Because of data andmeasurement limitations, the disentangling of population heterogeneityand state dependence requires that researchers control for unmeasuredpersistent heterogeneity. Frequently, random effects probit models havebeen employed, which, while user-friendly, make a strong parametricassumption that the unobserved heterogeneity in the population follows anormal distribution. Although semiparametric alternatives to the randomeffects probit model have recently appeared in the literature to avoid thisproblem, in this paper we return to reconsider the fully parametric model. Viasimulation evidence, we first show that the random effects probit modelproduces biased estimates as the departure of heterogeneity from normalitybecomes more substantial. Using the 1958 Philadelphia cohort data, we thencompare the results from a random effects probit model with a semiparametricprobit model and a fixed effects logit model that makes no assumptions aboutthe distribution of unobserved heterogeneity. We found that with this dataset all three models converged on the same substantive result—evenafter controlling for unobserved persistent heterogeneity, with models thattreat the unobserved heterogeneity very differently, prior conduct had apronounced effect on subsequent offending. These results are inconsistentwith a model that attributes all of the positive correlation between priorand future offending to differences in criminal propensity. Sinceresearchers will often be completely blind with respect to the tenabilityof the normality assumption, we conclude that different estimationstrategies should be brought to bear on the data.     

Harvesting experience: A reappraisal of the U.S. Climate Change Action Plan

For mitigating climate change and adapting to whatever impacts we cannot avoid, there are no politically feasible alternatives to improvements in the U.S. Climate Change Action Plan at this time or for the foreseeable future. Yet improvements in the Action Plan have been obstructed by the diversion of attention and other resources to negotiating a binding international agreement, to developing a predictive understanding of global change, and to documenting the failure of the Action Plan to meet its short-term goal for the reduction of aggregate greenhouse gas emissions. Continuous improvements depend upon reallocating attention and other resources to the Action Plan, and more specifically, to the many small-scale policies that have already succeeded by climate change and no regrets criteria under the Action Plan. Sustaining the effort over the long term depends on harvesting experience from these small-scale successes for diffusion and adaptation elsewhere on a voluntary basis.     

Specificity versus representativeness: methodological approaches to the study of Mexico-U.S. migration

"This article compares two sources of data on Mexico-U.S. migration, based on radically different methodologies: the Mexican Migration Project (Promig) and the National Survey of Population Dynamics (Enadid).... This comparative study shows that a micro-social design drawing on multiple community samples, such as Promig, can solve the methodological conflict between specificity and representativeness.... The authors' research also highlights the problem of selectivity and specificity entailed by traditional surveys such as Enadid as a result of restricting their samples to international residents in Mexico and attempting to explain such a complex, socioeconomic process using a limited number of variables." (EXCERPT)     

The regional dimension of Mexican migration to the United States

"Although Mexican migration to the United States dates back over a century, reliable statistics documenting its geographical distribution have only recently become available. This new information, together with the creation of a Geographical Data System and the recent development of multilevel analysis, have allowed one to examine the regional context of Mexican migration to the U.S.... This paper draws on two new tools for context analysis and examines how they can be used to study international migration: first, the Geographical Data systems, which measure physical factors (aridity, isolation, land use, environmental degradation), together with socioeconomic statistics and familial organization and reproduction." (EXCERPT)     

The Benefit Equity Principle and Willingness to Pay for City Services

In 1991 and 1992, the City of Eugene, Oregon engaged in a citizen participation process to determine public support for a variety of spending and budget balancing alternatives. This process, called Eugene Decisions, employed an innovative combination of public forums, budget balancing exercises and representative surveys. Two of the surveys, which included questions about individual service use, are analyzed here in order to test whether the use of city services is related to willingness to pay taxes and fees. The concept of benefit equity is supported if service use is positively associated with willingness to pay.