Knowledge that Transforms

Policymakers often test expensive new programs on relatively small samples. Formally incorporating informative Bayesian priors into impact evaluation offers the promise to learn more from these experiments. We evaluate a Colombian program for 200 firms which aimed to increase exporting. Priors were elicited from academics, policymakers, and firms. Contrary to these priors, frequentist estimation cannot reject null effects in 2019, and finds some negative impacts in 2020. For binary outcomes like whether firms export, frequentist estimates are relatively precise, and Bayesian posterior intervals update to overlap almost completely with standard confidence intervals. For outcomes like increasing export variety, where the priors align with the data, the value of these priors is seen in posterior intervals that are considerably narrower than the confidence intervals. Finally, for noisy outcomes like export value, posterior intervals show almost no updating from priors, highlighting how uninformative the data are about such outcomes. Future policy experiments could use these posteriors as priors in a Bayesian or empirical Bayesian analysis.

India's Industrial Disputes Act (IDA) requires large manufacturing plants to pay substantial costs if they wish to shrink their workforce. Since the early 2000s, these large plants have dramatically increased their use of contract workers who are not subject to these regulatory constraints. Between 2000 and 2015, the contract labor share in non‐managerial employment nearly doubled at establishments with more than 100 workers (from 21 to 40 percentage points), while it only increased from 14 to 17 percentage points at establishments with less than 50 workers. Over the same period, the thickness of the right tail of the establishment size distribution in formal Indian manufacturing plants increased, the average product of labor at large plants declined, the job creation rate for large plants increased, and the probability that large plants introduced new products rose. We argue that these changes were caused by the increased adoption of contract labor. In a model of establishment growth subject to firing costs, we show that easing access to contract labor increased TFP in Indian manufacturing by 7.3% since the early 2000s, occurring all through a one‐time reduction in misallocation between large and small plants with negligible change in the long‐run growth rate.

We develop a stylized model that allows us to estimate a value-added measure for nursing homes ("SNFs") which accounts for patient selection both into and out of a SNF. We use the model, together with detailed data on the physical and mental health of about 6 million Medicare SNF patients between 2011 and 2016, to estimate the value added for about 14,000 distinct SNFs. We document substantial heterogeneity in value added. Nationwide, compared to a 10th percentile SNF, a 90th percentile SNF is able to discharge a patient at the same health level almost a week sooner, or one quarter of the median length of stay. Heterogeneity in value added within a market is almost as large as it is nationwide. Our results point to the potential for substantial gains through policies that encourage reallocation of patients to higher-quality SNFs within their market.

We study the effect of incarceration on wages, self-employment, and taxes and transfers in North Carolina and Ohio using two quasi-experimental research designs: discontinuities in sentencing guidelines and random assignment to judges. Across both states, incarceration generates short-term drops in economic activity while individuals remain in prison. As a result, a year-long sentence decreases cumulative earnings over five years by 13%. Beyond five years, however, there is no evidence of lower employment, wage earnings, or self-employment in either state, as well as among defendants with no prior incarceration history. These results suggest that upstream factors, such as other types of criminal justice interactions or pre-existing labor market detachment, are more likely to be the cause of low earnings among the previously incarcerated, who we estimate would earn just $5,000 per year on average if spared a prison sentence.

Based on the GARCH literature, Engle and Russell (1998) established consistency and asymptotic normality of the QMLE for the autoregressive conditional duration (ACD) model, assuming strict stationarity and ergodicity of the durations. Using novel arguments based on renewal process theory, we show that their results hold under the stronger requirement that durations have finite expectation. However, we demonstrate that this is not always the case under the assumption of stationary and ergodic durations. Specifically, we provide a counterexample where the MLE is asymptotically mixed normal and converges at a rate significantly slower than usual. The main difference between ACD and GARCH asymptotics is that the former must account for the number of durations in a given time span being random. As a by‐product, we present a new lemma which can be applied to analyze asymptotic properties of extremum estimators when the number of observations is random.

How do collateral requirements impact consumer borrowing behavior? Using administrative loan application and performance data from the U.S. Federal Disaster Loan Program, we exploit a loan amount threshold above which households must post their residence as collateral. Our bunching estimates suggest that the median borrower is willing to give up 40% of their loan amount to avoid posting collateral. Exploiting time variation in the threshold, we estimate collateral causally reduces default rates by 36%. Finally, we structurally estimate households' attachment to their homes, net of any equity, and find a median value of $11,000. Attachment creates a wedge between lender and borrower valuation of collateral of 15%. Our results explain high perceived default costs in the mortgage market, and document the importance of collateral for reducing moral hazard in consumer credit markets.

Milgrom and Shannon (1994) provide necessary and sufficient conditions on parameterized optimization problems for their solution sets to be globally monotone in the parameter. We establish that their conditions may be significantly relaxed when focusing on discrete, binary comparisons between solution sets. Such binary comparisons are ubiquitous in economics and may involve comparing the same decision maker across two distinct regimes or two distinct decision makers with related objectives (e.g., a monopolist firm versus a social planner). While the single‐crossing property remains prominent in the theory, quasisupermodularity of the objective functions of interest is not needed. Our approach relies upon a novel method of embedding a new optimization problem with a quasisupermodular objective function “between” the two original problems of interest. In smooth problems, sufficient conditions for our new assumptions may be verified by elementary differential comparisons, making them well suited for applied work. We illustrate the relevance of this novel approach with several economic applications.

We propose a theory of how market power affects wage inequality. We ask how goods and labor market power jointly determine the level of wages, the skill premium, and wage inequality. We then use detailed microdata from the U.S. Census Bureau between 1997 and 2016 to estimate the parameters of labor supply, technology, and the market structure. We find that a less competitive market structure lowers the average wage of high‐skilled workers by 11.3%, and of low‐skilled workers by 12.2%, contributes 8.1% to the rise in the skill premium, and accounts for 54.8% of the increase in between‐establishment wage variance.

To what extent is a worker's human capital growth affected by the quality of his coworkers? To answer this question, we develop and estimate a model in which the productivity and the human capital growth of an individual depend on the average human capital of his coworkers. The measured production function is supermodular: The marginal product of a more knowledgeable individual is increasing in the human capital of his coworkers. The measured human capital accumulation function is convex: An individual's human capital growth is increasing in coworkers' human capital only when paired with more knowledgeable coworkers, but independent of coworkers' human capital when paired with less knowledgeable coworkers. Learning from coworkers accounts for two thirds of the stock of human capital accumulated on the job. Technological changes that increase production supermodularity lead to labor market segregation and, by reducing the opportunities for low human capital workers to learn from better coworkers, lead to a decline in aggregate human capital and output.

An agent selectively samples attributes of a complex project so as to influence the decision of a principal. The players disagree about the weighting, or relevance, of attributes. The correlation across attributes is modeled through a Gaussian process, the covariance function of which captures pairwise attribute similarity. The key trade‐off in sampling is between the alignment of the players' posterior values for the project and the variability of the principal's decision. Under a natural property of the attribute correlation—the nearest‐attribute property (NAP)—each optimal attribute is relevant for some player and at most two optimal attributes are relevant for only one player. We derive comparative statics in the strength of attribute correlation and examine the robustness of our findings to violations of NAP for a tractable class of distance‐based covariances. The findings carry testable implications for attribute‐based product evaluation and strategic selection of pilot sites.