Knowledge that Transforms

The increasing popularity of regression discontinuity methods for causal inference in observational studies has led to a proliferation of different estimating strategies, most of which involve first fitting nonparametric regression models on both sides of a treatment assignment boundary and then reporting plug-in estimates for the effect of interest. In applications, however, it is often difficult to tune the nonparametric regressions in a way that is well calibrated for the specific target of inference; for example, the model with the best global in-sample fit may provide poor estimates of the discontinuity parameter, which depends on the regression function at boundary points. We propose an alternative method for estimation and statistical inference in regression discontinuity designs that uses numerical convex optimization to directly obtain the finite-sample-minimax linear estimator for the regression discontinuity parameter, subject to bounds on the second derivative of the conditional response function. Given a bound on the second derivative, our proposed method is fully data driven and provides uniform confidence intervals for the regression discontinuity parameter with both discrete and continuous running variables. The method also naturally extends to the case of multiple running variables.

Local projections (LP) is a popular methodology for the estimation of impulse responses (IR). Compared to the traditional VAR approach, LP allow for more flexible IR estimation by imposing weaker assumptions on the dynamics of the data. The nonparametric nature of LP comes at an efficiency cost, and in practice, the LP estimator may suffer from excessive variability. In this work, we propose an IR estimation methodology based on B-spline smoothing called smooth local projections (SLP). The SLP approach preserves the flexibility of standard LP, can substantially increase precision, and is straightforward to implement. A simulation study shows that SLP can deliver substantial gains in IR estimation over LP. We illustrate our technique by studying the effects of monetary shocks where we highlight how SLP can easily incorporate commonly employed structural identification strategies.

This paper provides direct evidence that learning about demand is an important driver of firms’ dynamics. We present a model of Bayesian learning in which firms are uncertain about idiosyncratic demand in each market and update their beliefs as noisy information arrives. Firms update their beliefs to a given demand shock more, the younger they are. We test and empirically confirm this prediction, using the structure of the model, together with exporter-level data, to identify demand shocks and the firms’ beliefs about future demand. Consistent with theory, we also find the learning process to be weakened in more uncertain environments.

This paper shows that since votes of members of the Federal Open Market Committee have been included in press statements, stock prices increase after the announcement when votes are unanimous but fall when dissent (which typically is due to preference for higher interest rates) occurs. This pattern started prior to the 2007–2008 financial crisis. The differences in stock market reaction between unanimity and dissent remain, even controlling for the stance of monetary policy and consecutive dissent. Statement semantics also do not seem to explain the documented effect. We find no differences between unanimity and dissent with respect to impact on market risk and Treasury securities.

This paper develops a criterion to assess equalization of opportunity that is consistent with theoretical views of equality of opportunity. We characterize inequality of opportunity as a situation where some groups in society enjoy an illegitimate advantage. In this context, equalization of opportunity requires that the extent of the illegitimate advantage enjoyed by the privileged groups falls. Robustness requires that this judgment be supported by the broadest class of individual preferences. We formalize this criterion in a decision-theoretic framework and derive an empirical condition for equalization of opportunity based on observed opportunity distributions. The criterion is used to assess the effectiveness of child care at equalizing opportunity among children, using quantile treatment effects estimates of a major child care reform in Norway. Overall, we find strong evidence supporting equalization of opportunity.

We evaluate the impact on youth crime of a welfare reform that tightened activation requirements for social assistance clients. The evaluation strategy exploits administrative individual data in combination with geographically differentiated implementation of the reform. We find that the reform reduced crime among teenage boys from economically disadvantaged families. Stronger reform effects on weekday versus weekend crime, reduced school dropout, and favorable long-run outcomes in terms of crime and educational attainment point to both incapacitation and human capital accumulation as key mechanisms. Despite lowered social assistance take-up, we uncover no indication that loss of income support pushed youth into crime.

In volume 94 of this REVIEW, Adams, Blackburn, and Cotti (ABC), using Fatal Accident Reporting System data from 1998 to 2006, find that a 10% increase in the minimum wage is associated with a 7% to 11% increase in alcohol-related fatal traffic accidents involving teen drivers. We find this result does not hold when the analysis period is expanded to include 1991 through 2013. In addition, auxiliary analyses provide no support for income-driven increases in alcohol consumption, the primary mechanism posited by ABC. Together, our results suggest that minimum wage increases are not a silent killer.

A new regression-based approach for the estimation of the tail index of heavy-tailed distributions with several important properties is introduced. First, it provides a bias reduction when compared to available regression-based methods; second, it is resilient to the choice of the tail length used for the estimation of the tail index; third, when the effect of the slowly varying function at infinity of the Pareto distribution vanishes slowly, it continues to perform satisfactorily; and fourth, it performs well under dependence of unknown form. An approach to compute the asymptotic variance under time dependence and conditional heteroskcedasticity is also provided.

This paper estimates the elasticity of substitution between capital and skill in manufacturing using immigration-induced variation in skill mix across U.S. counties between 1860 and 1930. We find that capital initially complemented both high- and low-skill labor (determined by literacy) and, unlike today, was more complementary with low-skill labor. Around 1890, capital increased its relative complementarity with high-skill labor. Simulations calibrated to our estimates imply the level of capital-skill complementarity after 1890 allowed the manufacturing sector to absorb the large wave of Eastern and Southern European immigrants with only a modest decline in less-skilled relative wages. This would not have been possible under the older production technology.

We model the correlations of brothers' earnings isolating the effect of fathers' earnings from additional residual influences shared between brothers. We separate the two effects by analysing sibling correlations and intergenerational correlations jointly within a unified framework. Our multi-person model of earnings dynamics distinguishes permanent from transitory shocks, allows for heterogeneous life cycle effects and nests previous models. Using data on the Danish population of father/first-son/second-son triplets, we corroborate the findings of studies that do not account for life cycle effects for those aged in their 30's, but find correlations twice as large at 25. The impact of intergenerational effects also varies over age, but is everywhere higher than found in previous studies ?by on average a factor of thirteen? and accounts for most of the sibling correlation. We provide evidence that lack of both life cycle effects and heterogeneous intergenerational transmission across families in previous studies explain the difference. When allowing for differential intergenerational transmission within families, we find mild evidence of stronger transmission to second sons.