Research Database

We measure college graduate quality — the average human capital of a college’s graduates—using the average earnings of the college’s graduates adjusted to a common labor market. Our implementation uses the database of the website Glassdoor, which has the necessary information on earnings and education for non-migrants and migrants who graduate from roughly 3,300 colleges in 66 countries. Graduates of colleges in the richest countries have 50 percent more human capital than graduates of colleges in the poorest countries. Migration reinforces these differences. Poorer countries do not just lose a higher share of their skilled workers; their emigrants are highly positively selected on human capital. Finally, we show that these stocks and flows matter for growth and development by showing that college graduate quality predicts the share of a college’s students who become inventors, engage in entrepreneurship, and become top executives, both within and across countries.

We document four facts about the COVID-19 pandemic worldwide relevant for those studying the impact of non-pharmaceutical interventions (NPIs) on COVID-19 transmission. First: across all countries and U.S. states that we study, the growth rates of daily deaths from COVID-19 fell from a wide range of initially high levels to levels close to zero within 20-30 days after each region experienced 25 cumulative deaths. Second: after this initial period, growth rates of daily deaths have hovered around zero or below everywhere in the world. Third: the cross section standard deviation of growth rates of daily deaths across locations fell very rapidly in the first 10 days of the epidemic and has remained at a relatively low level since then. Fourth: when interpreted through a range of epidemiological models, these first three facts about the growth rate of COVID deaths imply that both the effective reproduction numbers and transmission rates of COVID-19 fell from widely dispersed initial levels and the effective reproduction number has hovered around one after the first 30 days of the epidemic virtually everywhere in the world. We argue that failing to account for these four stylized facts may result in overstating the importance of policy mandated NPIs for shaping the progression of this deadly pandemic.

Using linked employer-employee data from Brazil, we document a large gender pay gap due to women working at lower-paying employers with better amenities. To interpret these facts, we develop an equilibrium search model with endogenous firm pay, amenities, and employment. We provide a constructive proof of identification of all model parameters. The estimated model suggests that amenities are important for men and women, that compensating differentials explain half of the gender pay gap, and that there are significant output and welfare gains from eliminating gender differences. However, equal-treatment policies fail to achieve those gains.

This paper measures the output costs of sovereign risk by combining a sovereign debt model with firm- and bank-level data. An increase in sovereign risk lowers the price of government debt and has an adverse impact on banks’ balance sheets, disrupting their ability to finance firms. The resulting fall in credit supply impacts firms directly, as they need to borrow at higher interest rates, and indirectly through general equilibrium effects on the price of inputs and other goods. Importantly, firms are not equally affected by these developments: those that have greater financing needs and that borrow from banks that hold more government debt are mostly affected by the change in borrowing rates, while firms that do not borrow are only impacted indirectly. We show that these direct and indirect effects can be recovered using a firm-level regression, which we estimate using Italian data. We calibrate our model to match the measured firm-level elasticities and find that heightened sovereign risk was responsible for one-third of the observed output decline during the Italian debt crisis.

This paper characterizes the allocations that emerge in general equilibrium economies populated by households with preferences of the additive random utility type that make discrete consumption, employment or spatial decisions. We start with a complete markets economy where households can trade claims contingent upon the realizations of their preference shocks. We (i) establish a first and second welfare theorem, (ii) illustrate that in the absence of ex-ante trade, discrete choice economies are generically inefficient, (iii) show that complete markets are not necessary and a much smaller set of securities decentralizes the efficient allocation. We illustrate the relevance of these results in several canonical settings and for measuring how welfare changes in response to changes in prices.

In the U.S, after age 65, households face income and health risks and a large fraction of these risks are transitory. While consumption significantly responds to transitory income shocks, out-of-pocket medical expenses do not. In contrast, both consumption and out-of-pocket medical expenses respond to transitory health shocks. Thus, most U.S. elderly keep their out-of-pocket medical expenses close to a satiation point that varies with health. Consumption responds to health shocks mostly because adverse health shocks reduce the marginal utility of consumption. The effect of health on marginal utility changes the optimal transfers due to health shocks.

A weakening of labor protection policies is often invoked as one cause of observed monopsony power and the decline in labor’s share of income, but little evidence exists on the causal impact of labor policies on wage markdowns. Using confidential Mexican economic census data from 1994 to 2019, we document a rising trend over this period in on-site outsourcing. Then, leveraging data from a manufacturing panel survey from 2013 to 2023 and a natural experiment featuring a ban on domestic outsourcing in 2021, we show that the ban drastically reduced outsourcing, increased wages, and reduced measured markdowns without lowering output or employment. Consistent with the presence of monopsony power, we observe large markdowns for the largest firms, with the decline in markdowns in response to the ban concentrated among high-markdown firms. However, we also find that the reform reduced capital investment and increased the probability of market exit.

Financial integration generates macroeconomic spillovers that may require international monetary policy coordination. We show that individual central banks may set nominal interest rates too low or too high relative to the cooperative outcome. We identify three sufficient statistics that determine whether the Nash equilibrium exhibits under-tightening or over-tightening: the output gap, sectoral differences in labor intensity, and the trade balance response to changes in nominal rates. Independently of the shocks hitting the economy, we find that under-tightening is possible during economic expansions or contractions. For large shocks, the gains from coordination can be substantial.

This paper explores the role of restrictions on the use of international reserves as economic sanctions. We develop a simple model of the strategic game between a sanctioning (creditor) country and a sanctioned (debtor) country. We show how the sanctioning country should impose restrictions optimally, internalizing the geopolitical benefits and the financial costs of a potential default from the sanctioned country.

We characterize the little wage information contained in online job posts. Wage information is rare: only 14% of posts contain any information. Of these, wage ranges are more common than point wages, and are wide on average, spanning 28% of the midpoint (e.g. $32,000-$42,000/yr). Posted wages are highly selected in low income occupations: 40% higher than wages of employed workers. High wage firms are more opaque, with more and wider ranges. We find zero correlation between wage information and local labor market tightness. We provide an example of bias in econometric inference that worsens as wage information falls.