Search results

The prevalence and stability of marriage has declined in the United States as the economic lives of men and women have converged. Family change has not been uniform, however, and the widening gaps in marital status, relationship stability, and childbearing between socioeconomic groups raise concerns about child well-being in poor families and future inequality. This paper uses data from a recent cohort of young adults – Wave IV of the National Longitudinal Study of Adolescent Health – to investigate whether disparities in cognitive ability and non-cognitive skills contribute to this gap. Blinder–Oaxaca decompositions of differences in key family outcomes across education groups show that, though individual non-cognitive traits are significantly associated with union status, relationship instability, and single motherhood, they collectively make no significant contribution to the explanation of educational gaps for almost all of these outcomes. Measured skills can explain as much as 25 percent of differences in these outcomes by family background (measured by mother’s education), but this effect disappears when own education is added to the model. Both cognitive and non-cognitive skills are strongly predictive of educational attainment but, conditional on education, explain very little of the socioeconomic gaps in family outcomes for young adults.

Standard estimation of ARMA models in which the AR and MA roots nearly cancel, so that individual coefficients are only weakly identified, often produces inferential ranges for individual coefficients that give a spurious appearance of accuracy. We remedy this problem with a model that uses a simple mixture prior. The posterior mixing probability is derived using Bayesian methods, but we show that the method works well in both Bayesian and frequentist setups. In particular, we show that our mixture procedure weights standard results heavily when given data from a well-identified ARMA model (which does not exhibit near root cancellation) and weights heavily an uninformative inferential region when given data from a weakly-identified ARMA model (with near root cancellation). When our procedure is applied to a well-identified process the investigator gets the “usual results,” so there is no important statistical cost to using our procedure. On the other hand, when our procedure is applied to a weakly identified process, the investigator learns that the data tell us little about the parameters – and is thus protected against making spurious inferences. We recommend that mixture models be computed routinely when inference about ARMA coefficients is of interest.

The economics literature on gender has expanded considerably in recent years, fueled in part by new sources of data, including from experimental studies of gender differences in preferences and other traits. At the same time, economists have been developing more realistic models of psychological and social influences on individual choices and the evolution of culture and social norms. Despite these innovations, much of the economics of gender has been left behind, and still employs a reductive framing in which gender gaps in economic outcomes are either due to discrimination or to “choice.” I suggest here that the persistence of this approach is due to several distinctive economic habits of mind – strong priors driven by market bias and gender essentialism, a perspective that views the default economic agent as male, and an oft-noted tendency to avoid complex problems in favor of those that can be modeled simply. I also suggest some paths forward.

In this paper, we expand Kleibergen and Zivot's (2003) Bayesian two-stage (B2S) model by allowing for unequal variances. Our choice for modeling heteroscedasticity is a fully Bayesian parametric approach. As an application, we present a cross-country Cobb–Douglas production function estimation.

This chapter reviews developments to improve on the poor performance of the standard GMM estimator for highly autoregressive panel series. It considers the use of the ‘system’ GMM estimator that relies on relatively mild restrictions on the initial condition process. This system GMM estimator encompasses the GMM estimator based on the non-linear moment conditions available in the dynamic error components model and has substantial asymptotic efficiency gains. Simulations, that include weakly exogenous covariates, find large finite sample biases and very low precision for the standard first differenced estimator. The use of the system GMM estimator not only greatly improves the precision but also greatly reduces the finite sample bias. An application to panel production function data for the U.S. is provided and confirms these theoretical and experimental findings.