Research

Randomized controlled trials (RCTs) enroll hundreds of millions of subjects and involve many human lives. To improve subjects’ welfare, I propose a design of RCTs that I call Experiment-as-Market (EXAM). EXAM produces a welfare-maximizing allocation of treatment-assignment probabilities, is almost incentive-compatible for preference elicitation, and unbiasedly estimates any causal effect estimable with standard RCTs. I quantify these properties by applying EXAM to a water-cleaning experiment in Kenya. In this empirical setting, compared to standard RCTs, EXAM improves subjects’ predicted well-being while reaching similar treatment-effect estimates with similar precision.

Manufacturers perform the majority of US patenting and R&D. The decades-long decline of US manufacturing employment raises concerns that US innovation will falter. We investigate the relationship between between physical production and innovation by constructing a new dataset linking all US firms and their establishments to location geocodes and innovative activities. Pre- liminary results indicate that while firms with both manufacturing and innovation establishments exhibit higher patenting when these facilities are more spatially proximate, manufacturing firms’ overall contribution to US innovation declines steadily and substantially over time. Moreover, cohorts of firms permanently exiting manufacturing in the 1990s and 2000s continue to patent at their prior pace.

We evaluate the quantitative effects of trade policy on the location of firms across space and over time. We develop a multi-country, multi-sector dynamic general-equilibrium trade and spatial model with forward-looking decisions of workers on where to supply labor, forward-looking decisions of firms on where to locate produc-tion, endogenous capital structure accumulation, and trade in intermediate goods with sectoral linkages. We bring the model to data using trade, production, and data on firm demographics across sector and locations. We use the model to study if trade protectionism can revert the declining trend in the U.S. manufacturing em-ployment and firms; and its impact on the location of production across space and over time. We feed into the model the raise in import tariffs between the U.S. and its major trade partners in the year 2018. We find that these changes in trade policy can result in a persistent increase on manufacturing employment and firms. However, these effects do not revert the long run decline in manufacturing employment and firms. Importantly, the relocation of production comes at the cost of higher prices, lower welfare for households, and heterogeneous effects on firm entry across space.

Economists often point to the superiority of cash transfers over in-kind assistance as a means of redistribution because recipients can choose how to use these resources. However, among the trillions of dollars of annual U.S. transfers, redistribution is mostly in-kind. We conducted a survey experiment—using a choice between a cash transfer and a transfer that could be spent only on a bundle of “necessities”—to help explain why. We show that the general population overwhelmingly prefers in-kind redistribution, largely for paternalistic reasons. This preference was common to a majority of virtually all segments of the general population, though not to a sample of educational elites. A persuasion treatment on the value of choice, while impactful, did not change this overall preference for in-kind. For an equal-sized program, below-poverty respondents preferred receiving cash. But they appeared to prefer the larger in-kind transfer to the smaller cash transfer that the general population was willing to support. This suggests that an in-kind transfer may be preferable to both recipients and the general population.

We uncover political dynamics that reward and reinforce increases in US health spending by studying the passage of the 2003 Medicare Modernization (MMA). We focus on a provision added to the MMA, which allowed hospitals to apply for temporary Medicare payment increases. Hospitals represented by members of Congress who voted ‘Yea’ to the MMA were more likely to receive payment increases. The payment increases raised local health spending and led to suggestive increases in health sector employment. Members of Congress representing hospitals that got a payment increase received large increases in campaign contributions before and after the program was extended.

The CARES Act expanded unemployment insurance (UI) benefits by providing a $600 weekly payment in addition to state unemployment benefits. Most workers thus became eligible to receive unemployment benefits that exceed their weekly wages. It has been hypothesized that such high benefits encourage employers to lay off workers and discourage workers from returning to work. In this note, we test whether changes in UI benefit generosity are associated with decreased employment, both at the onset of the benefits expansion and as businesses look to reopen. We use weekly data from Homebase, a private firm that provides scheduling and time clock software to small businesses, which allows us to exploit high-frequency changes in state and federal policies to understand how firms and workers respond to policy changes in real time. Additionally, we benchmark our results from the Homebase data to employment outcomes in the Current Population Survey (CPS). We find that that the workers who experienced larger increases in UI generosity did not experience larger declines in employment when the benefits expansion went into effect. Additionally, we find that workers facing larger expansions in UI benefits have returned to their previous jobs over time at similar rates as others. We find no evidence that more generous benefits disincentivized work either at the onset of the expansion or as firms looked to return to business over time. In future research, it will be important to assess whether the same results hold when states move to reopen, and to analyze the effects of high UI replacement rates on reallocation of labor both within and across firms.

Using city-level crime data for six major U.S. cities from Jan 21 to May 30 2020, we document an approximately 20% average reduction in reported crimes during March, simultaneous with sharp economic downturn and heightened social distancing restrictions. We also decompose trends by crime type and location. Our key findings are:

• Since the steep 20% crime drop in March, overall rates have steadily risen but remain below pre-pandemic levels on average.
• Crimes committed in commercial and street settings (as opposed to residential areas) account for most of the drop in crimes.
• Violent crimes decline in similar proportion to nonviolent crimes.
• Though larcenies fall by one-third, other kinds of theft like burglary and auto theft rise.

Caveats to our findings include the possibility of simultaneous changes in reporting and policing activities.

We propose a method for identifying exposure to changes in trade policy based on asset prices that has several advantages over standard measures: it encompasses all avenues of exposure, it is natively firm-level, it yields estimates for both goods and service producers, and it can be used to study reductions in difficult-to-quantify non-tariff barriers in a way that controls naturally for broader macroeconomic shocks. Applying our method to two well-studied US trade liberalizations provides new insight into service sector responses to trade liberalizations as well as dramatically different responses among small versus large firms, even within narrow industries.