Knowledge that Transforms

This paper evaluates the presence of racial disparities in the issuance of consumer credit. Using a database of credit histories, I link location-based information on race with individual credit files. After controlling for place-specific factors such as housing vacancy rates and general population demographics, I find qualitatively large differences in the amount of credit offered to similarly qualified applicants living in black versus white areas. High data quality allows distinguishing between issuer-provision (supply) and utilization of credit (demand). Additional estimates using information on payday lending provide support for idea that issuers condition lending on location.

Despite their ubiquity, observational studies to infer the causal effect of a so-called immutable characteristic, such as race or sex, have struggled for coherence, given the unavailability of a manipulation analogous to a “treatment” in a randomized experiment and the danger of posttreatment bias. We demonstrate that a shift in focus from actual traits to perceptions of them can address both of these issues while facilitating articulation of other critical concepts, particularly the timing of treatment assignment. We illustrate concepts by discussing the designs of various studies of the role of race in trial court death penalty decisions.

Many researchers have addressed concerns of omitted variable bias in hedonic price models through the use of spatial fixed effects. We argue that this approach does not consider the biases introduced by effects that overlap the zone of capitalization for nonmarket goods. We show this bias can dominate the usual omitted variable bias using data on developer-provided open space. We control for multiscale capitalization and omitted variables simultaneously by extending the Hausman-Taylor estimator where the panel component of the estimator is defined over repeated transactions within neighborhoods rather than the traditional definition following a unique cross-sectional unit through time.

We analyze the effect of survey design on reported job satisfaction by exploiting two quasi-experiments in the British Household Panel Survey: a change in question design and parallel use of different interview modes. We show that apparently minor differences in survey design lead to substantial biases in econometric results, particularly on gender differences. The common empirical finding that women care less about wages and prefer to work fewer hours than men appears largely an artifact of survey design rather than a true behavioral difference.

We take advantage of a natural experiment in the state of Indiana to estimate the effect of daylight saving time (DST) on residential electricity consumption. Our main finding is that, contrary to the policy's intent, DST increases electricity demand. The findings are consistent with simulation results that identify a trade-off between reducing demand for lighting and increasing demand for heating and cooling. We estimate a cost to Indiana households of $9 million per year in increased electricity bills. We also estimate social costs of increased pollution emissions between $1.7 to $5.5 million per year.

We derive the distribution of a proxy for the risk tolerance in a representative sample of U.S. households. Our measure is deduced from the willingness to bear risk as indicated by the variance of returns of each household's observed portfolio. The estimates, obtained assuming constraints on portfolio composition, show substantial heterogeneity across households. We find that risk tolerance falls with age and increases with wealth. Other variables, such as education, gender, race, and household size, do not have a significant relation to risk attitude. Our findings are robust to changes in portfolio definition, asset returns, and sample composition.

This paper uses a unique new data set on manufacturing firms in Brazil and India to estimate production functions, augmented by information and communications technology (ICT). We find a strong positive association between ICT capital and productivity in both countries that is robust to several different specification tests. The paper also breaks new ground when using the Indian data to investigate the effect of the institutional and policy environment on ICT capital investment and productivity. We find that poorer infrastructure quality and labor market policy are associated with lower levels of ICT adoption, while poorer infrastructure is also associated with lower returns to investment.

We compare the finite-sample performance of impulse response confidence intervals based on local projections (LPs) and vector autoregressive (VAR) models in linear stationary settings. We find that in small samples, the asymptotic LP interval often is less accurate than the bias-adjusted bootstrap VAR interval, notwithstanding its excessive average length. Although the asymptotic LP interval has adequate coverage in sufficiently large samples, its average length still far exceeds that of bias-adjusted bootstrap VAR intervals with comparable accuracy. Bootstrap LP intervals (with or without bias correction) and asymptotic VAR intervals are shorter on average, but they often lack coverage accuracy in finite samples.

Where does adaptation to innovation take place? I present evidence on the role of agglomeration economies in the application of new knowledge to production. All else equal, workers are more likely to be observed in new work in locations initially dense in college graduates and industry variety. This pattern is consistent with economies from the geographic concentration of factors and markets related to technological adaptation. A main contribution is a new measure, based on revisions to occupation classifications, that characterizes cross-sectional differences across cities in technological adaptation. Worker-level results also provide new evidence on the skill bias of recent innovations.

We use two large samples of firms to assess the effects of business environment constraints, competition, export orientation, and ownership on firm performance. We deal with omitted variables, errors in variables, and endogeneity, and find that few business constraints affect performance. Replicating the analysis with Doing Business and Heritage Foundation indicators of the business environment yields similar results. In fact, country fixed effects, reflecting time-invariant differences in the business environment as well as other factors such as health care and education, matter more for firm performance than differences in the business environment across firms within countries.