Knowledge that Transforms

We exploit increases in postpartum length of stay generated by legislative changes in the late 1990s to identify the impact of greater hospital care on the health of newborns. Using all births in California over the 1995–2000 period, two-stage least-square estimates show that increased treatment intensity had a modest impact on readmission probabilities for the average newborn. Allowing the treatment effect to vary by two objective measures of medical need demonstrates that the law had large impacts for those with the greatest likelihood of a readmission. The results suggest that the returns to average and marginal patients vary considerably in this context.

This paper presents an alternative method to derive the limiting distribution of residual-based statistics. Our method does not impose an explicit assumption of (asymptotic) smoothness of the statistic of interest with respect to the model's parameters and thus is especially useful in cases where such smoothness is difficult to establish. Instead, we use a locally uniform convergence in distribution condition, which is automatically satisfied by residual-based specification test statistics. To illustrate, we derive the limiting distribution of a new functional form specification test for discrete choice models, as well as a runs-based tests for conditional symmetry in dynamic volatility models.

We derive Lagrange multiplier and likelihood ratio specification tests for the null hypotheses of multivariate normal and Student-t innovations using the generalized hyperbolic distribution as our alternative hypothesis. We decompose the corresponding Lagrange multiplier-type tests into skewness and kurtosis components. We also obtain more powerful one-sided Kuhn-Tucker versions that are equivalent to the likelihood ratio test, whose asymptotic distribution we provide. Finally, we conduct detailed Monte Carlo exercises to study the size and power properties of our proposed tests in finite samples.

Governments provide public information to reduce information imperfections. Do households rely on public signals to inform themselves about market conditions? To identify the importance of public information in households' price expectations, we take advantage of a unique natural experiment in Ecuador where the published inflation rate had been different from the correct rate over a period of fourteen months due to a software error. We find that the public signal about prices plays an important role in households' price expectations, and the change in price expectations affects their savings choices. The effect is stronger for better-educated and older people.

Nonparametric estimators of treatment effects are often applied in settings where clustering may be important. We provide a general methodology for consistently estimating the variance of a large class of nonparametric estimators, including the simple matching estimator, in the presence of clustering. Software for implementing our variance estimator is available in Stata.

There is controversy about whether geography matters mainly because of its contemporaneous impact on economic outcomes or because of its interaction with historical events. Looking at terrain ruggedness, we are able to estimate the importance of these two channels. Because rugged terrain hinders trade and most productive activities, it has a negative direct effect on income. However, in Africa rugged terrain afforded protection to those being raided during the slave trades. Since the slave trades retarded subsequent economic development, in Africa ruggedness has also had a historical indirect positive effect on income. Studying all countries worldwide, we find that both effects are significant statistically and that for Africa the indirect positive effect dominates the direct negative effect. Looking within Africa, we also provide evidence that the indirect effect operates through the slave trades.

I exploit the variation in demographic change across the United States to estimate the relationship between the age distribution in the population and the magnitude of cyclical output volatility. According to panel regression estimates, the relative supply of young workers, or youth share, has a statistically significant effect on the volatility of state-by-state GDP. Moreover, changes to the age distribution can account for up to 58% of the recent reduction in business cycle fluctuations, indicating a critical link between the youth share and output volatility.

We examine the impact of the Internet on the leading American recreation activity: watching television. We run a panel regression using television viewing, Internet penetration, and socioeconomic variables for a large number of American cities starting before the birth of the Web. We find that the Internet's effect on television viewing varies by age group, reducing it by a moderate amount for the youngest Americans but having no impact on the viewing of the oldest Americans. We hypothesize that the overall effect is likely to increase over time as older age groups have more experience with the Internet's recreational opportunities.

Since 1970, the fraction of mixed-race black-white births has increased nearly ninefold. This paper describes basic facts about the behaviors and outcomes of black-white mixed-race individuals. Unsurprisingly, on a host of background and achievement characteristics, as well as adult outcomes, mixed-race individuals fall in between whites and blacks. When it comes to engaging in risky and antisocial adolescent behavior, however, mixed-race adolescents are stark outliers compared to both blacks and whites. We argue that these behavioral patterns are most consistent with a two-sector Roy model, in which mixed-race adolescents, not having a predetermined peer group, engage in more risky behaviors in order to be accepted.

Expert agents, such as lawyers, play a prominent role in conflict resolution, yet little is known about how they affect outcomes. We construct a model that permits us to estimate the influence of agents and test whether the parties in a dispute face prisoner's dilemma incentives. Using eighteen years of final-offer arbitration data from New Jersey, we find the parties do significantly better when they retain agents and that the parties learn about this benefit over time. However, we also find that the gain to using an agent is fully offset when the opposing party also hires an agent. Since agents are costly, this noncooperative equilibrium is Pareto inferior.