Effective administration of unemployment insurance (UI) is central to its ability to smooth consumption and act as an automatic stabilizer. The federal government’s method of allocating funds to administer UI gives the states no incentive to provide quality service at reasonable cost. We first document the deteriorating performance of the UI system in recent recessions and present estimates of a descriptive model relating state workloads to performance. We then characterize UI administration as a standard principal-agent problem, which leads to a method of allocating funds that would motivate states to adopt new technologies and improve performance. 

We examine the impact of public sector salary disclosure laws on university faculty salaries in Canada. The laws, which enable public access to the salaries of individual faculty if they exceed specified thresholds, were introduced in different provinces at different times. Using detailed administrative data covering the majority of faculty in Canada, and an event-study research design that exploits within-province variation in exposure to the policy across institutions and academic departments, we find robust evidence that that the laws reduced the gender pay gap between men and women by approximately 30 percent. There is suggestive evidence that higher female salaries contributed to the narrowing of the gender gap. The reduction in the gender gap is
primarily in universities where faculty are unionized.

Using newly digitized data from the Federal Trade Commission, I examine the evolution of executive compensation during the Great Depression, before and after mandated pay disclosure in 1934. I find that disclosure did not achieve the intended effect of broadly lowering CEO compensation. If anything, and in spite of popular outrage against compensation practices, average CEO compensation increased following disclosure relative to the upper quantiles of the non-CEO labor income distribution. Pay disclosure coincided with compression of the CEO earnings distribution. Following disclosure there was a pronounced drop in the residual variance of earnings—computed with size and industry controls—that accounts for almost the entire drop in the unconditional variance. The evidence suggests an upward “ratcheting” effect whereby lower paid CEOs given the size and industry of their firm experienced relative gains while well paid CEOs conditional on these characteristics were not penalized. The exception is at the extreme right tail of the CEO distribution, which fell precipitously, suggesting that disclosure may only have restrained only the most salient and visible wages.

In a classic paper, Schelling (1971) showed that extreme segregation can arise from
social interactions in preferences: once the minority share in a neighborhood exceeds a
"tipping point", all the whites leave. We use regression discontinuity methods and
Census tract data from the past four decades to test for the presence of discrete nonlinearities
in the dynamics of neighborhood racial composition. White mobility patterns
in most cities exhibit tipping-like behavior, with a range of tipping points centered
around a 13% minority share. These patterns are very pronounced during the 1970s
and 1980s, and diminish but do not disappear in the 1990s. We find similar dynamic
patterns in neighborhoods and in schools. A variety of specification checks rule out the
possibility that the discontinuity in the initial minority share is driven by income
stratification or other factors, and underscore the importance of white preferences over
neighbors ' race and ethnicity in the dynamic process of segregation. Finally, we relate
the location of the estimated tipping points in different cities to measures of the racial
attitudes of whites, and find that cities with more racially tolerant whites have higher
tipping points.

Alternative work arrangements, defined both by working conditions and by workers’ relationship to their employers, are heterogeneous and common in the U.S. This article reviews the literature on workers’ preferences over these arrangements, inputs to firms’ decision to offer them, and the impact of regulation. It also highlights several descriptive facts. Work arrangements have been relatively stable over the past 20 years, work conditions vary substantially with education, and jobs with schedule or location flexibility are less family-friendly on average. This last fact helps explain why women are not more likely to have schedule or location flexibility and seem to largely reduce hours to get more family-friendly arrangements.

A great deal of urban policy depends on the possibility of creating stable, economically and
racially mixed neighborhoods. Many social interaction models- including the seminal
Schelling (1971) model- have the feature that the only stable equilibria are fully segregated.
These models suggest that if home-buyers have preferences over their neighborhoods' racial
composition, a neighborhood with mixed racial composition is inherently unstable, in the
sense that a small change in the composition sets off a dynamic process that converges to
either 0% or 100% minority share. Card, Mas, and Rothstein (2008) outline an alternative
"one-sided" tipping model in which neighborhoods with a minority share below a critical
threshold are potentially stable, but those that exceed the threshold rapidly shift to 100%
minority composition. In this paper we examine the racial dynamics of Census tracts in
major metropolitan areas over the period from 1970 to 2000, focusing on the question of
whether tipping is "two-sided" or "one-sided". The evidence suggests that tipping behavior
is one-sided, and that neighborhoods with minority shares below the tipping point attract
both white and minority residents.

Although economists acknowledge that various indicators of educational attainment (e.g., highest
grade completed, credentials earned) might serve as signals of a worker’s productivity, the practical
importance of education-based signaling is not clear. In this paper we estimate the signaling value
of a high school diploma, the most commonly held credential in the U.S. To do so, we compare the
earnings of workers that barely passed and barely failed high school exit exams, standardized tests
that, in some states, students must pass to earn a high school diploma. Since these groups should, on
average, look the same to firms (the only difference being that "barely passers" have a diploma while
"barely failers" do not), this earnings comparison should identify the signaling value of the diploma.
Using linked administrative data on earnings and education from two states that use high school exit
exams (Florida and Texas), we estimate that a diploma has little effect on earnings. For both states, we
can reject that individuals with a diploma earn eight percent more than otherwise-identical individuals
without one; combining the state-specific estimates, we can reject signaling values larger than five or six
percent. While these confidence intervals include economically important signaling values, they exclude
both the raw earnings difference between workers with and without a diploma and the regression-adjusted
estimates reported in the previous literature.

This paper studies the effect of labor relations on product quality. We consider whether a
long, contentious strike and the hiring of permanent replacement workers by
Bridgestone/Firestone in the mid-1990s contributed to the production of an excess
number of defective tires. Using several independent data sources we find that labor
strife in the Decatur plant closely coincided with lower product quality. Count data
regression models based on two data sets of tire failures by plant, year and age show
significantly higher failure rates for tires produced in Decatur during the labor dispute
than before or after the dispute, or than at other plants. Also, an analysis of internal
Firestone engineering tests indicates that P235 tires from Decatur performed less well if
they were manufactured during the labor dispute compared with those produced after the
dispute, or compared with those from other, non-striking plants. Monthly data suggest
that the production of defective tires was particularly high around the time wage
concessions were demanded by Firestone in early 1994 and when large numbers of
replacement workers and permanent workers worked side by side in 1996.

We estimate the effect of new unionization on firms’ equity value over the 1961-1999 period using a newly
assembled sample of National Labor Relations Board (NLRB) representation elections matched to stock market
data. Event-study estimates show an average union effect on the equity value of the firm eq uivalent to a cost of at
least $40,500 per unionized worker. At the same time, point estimates from a regression-discontinuity design –
comparing the stock market impact of close union election wins to close losses – are considerably smaller and close
to zero. We find a negative relationship between the cumulative abnormal returns and the vote share in support of
the union, allowing us to reconcile these seemingly contradictory findings. Using the magnitudes from the analysis,
we calibrate a structural “median voter” model of endogenous union determination in order to conduct
counterfactual policy simulations of policies that would marginally increase the ease of unionization.

This paper documents that rotation group bias — the tendency for labor force statistics to vary systematically by month in sample in labor force surveys — in the Current Population Survey (CPS) has worsened considerably over time. The estimated unemployment rate for earlier rotation groups has grown sharply relative to the unemployment rate for later rotation
groups; both should be nationally representative samples. The rise in rotation group bias is driven by a growing tendency for respondents to report job search in earlier rotations relative to later rotations. We investigate explanations for the change in bias. We find that rotation group bias increased discretely after the 1994 CPS redesign and that rising nonresponse is likely a significant contributor. Survey nonresponse increased after the redesign, and subsequently trended upward, mirroring the time pattern of rotation group bias. Consistent with this explanation, there is only a small increase in rotation group bias for households that responded in all eight interviews. An analysis of rotation group bias in
Canada and the U.K. reveal no rotation group bias in Canada and a modest and declining
bias in the U.K. There is not a “Heisenberg Principle” of rotation group bias, whereby the bias is an inherent feature of repeated interviewing. We explore alternative weightings of the unemployment rate by rotation group and find that, despite the rise in rotation group bias, the official unemployment does no worse than these other measures in predicting alternative measures of economic slack or fitting key macroeconomic relationships.