Knowledge that Transforms

This paper proposes a robust method for semiparametric identification and estimation in panel multinomial choice models, where we allow for infinite-dimensional fixed effects that enter into consumer utilities in an additively nonseparable way, thus incorporating rich forms of unobserved heterogeneity. Our identification strategy exploits multivariate monotonicity in parametric indices, and uses the logical contraposition of an intertemporal inequality on choice probabilities to obtain identifying restrictions. We provide a consistent estimation procedure, and demonstrate the practical advantages of our method with Monte Carlo simulations and an empirical illustration on popcorn sales with the NielsenIQ data.

The effect of health insurance on consumption risk depends in part on how it interacts with other risks beyond health care cost risk, such as income risk. Using a variety of approaches, I find that for U.S. households, the interaction with other risks transforms the risk protection from health insurance. Standard contracts amplify the impact of other risks, due to both subsidizing normal goods and undoing the protection against other risks from discounts, charity care, and bad debt. Alternative contracts that account for other risks, such as contracts that limit out-of-pocket spending relative to income, can provide better risk protection.

This paper studies the causal effect of foreign currency dependence on international trade by exploiting Brazil and Argentina's 2008 introduction of a bilateral payments system that eliminated the U.S. dollar as vehicle currency. I identify causal effects using a triple-difference design comparing exports across municipalities with varying bank access and across destinations to control for contemporaneous shocks including the financial crisis. Firm-level analysis finds that local currency adoption increased export values significantly, with effects concentrated among non-commodity exporters. These results demonstrate that foreign currency dependence constitutes a meaningful barrier to emerging market trade.

Most of China's fertility decline predates the famous One Child Policy — and instead occurred under its predecessor, the Wan Xi Shao, or Later, Longer, Fewer (LLF) campaign. Studying LLF's contribution to fertility and sex selection behavior, we find that LLF i) reduced China's total fertility rate by 0.95 births per woman (explaining 30.6% of its fertility decline), ii) doubled the use of male-biased fertility stopping rules, and iii) promoted postnatal selection (implying 180,000 previously unrecognized missing girls, or 19% of the total during our study period). Considering Chinese population policy to be extreme in global experience, our paper demonstrates the limits of population policy — and its potential human costs.

This paper proposes a spatial model of imperfect competition in markets with adverse or advantageous selection. The model shows that a reduction in competition exacerbates the inefficiency created by adverse selection but can ameliorate the inefficiency created by advantageous selection. However, reduced competition never corrects the inefficiency perfectly because it introduces an allocative inefficiency. By contrast, the inefficiency can be corrected perfectly through a corrective tax when there is perfect competition. Our results have implications for competition policy in credit and insurance markets, as they caution against viewing imperfect competition as a solution to the inefficiencies created by selection.

After remaining flat for decades, labor productivity at U.S. restaurants surged 15% during the COVID pandemic. The surge has persisted. We explore this using mobile phone data tracking visits and spending at 100,000 limited-service restaurants. It cannot be explained by scale economies, market power, or mechanical consequences of COVID demand fluctuations. However, restaurants’ productivity growth is strongly correlated with the share of their customers visiting for 10 minutes or less, which rose considerably during COVID and persisted. This restaurant-level relationship between labor productivity and customer dwell time is large enough to explain most of the aggregate productivity increase.

We use administrative data to measure sibling spillovers on academic performance before and after the introduction of Free Secondary Education (FSE) in Tanzania. Prior to FSE, students whose older siblings narrowly passed the secondary school entrance exam were less likely to go to secondary school themselves; with FSE, the effect became positive. A triple-differences analysis, using geographic variation in FSE exposure, shows that FSE caused the reversal. Mechanism analyses suggest that changes in parental investments were a more likely channel for this reversal than direct sibling interactions. By alleviating financial constraints, FSE allowed households to invest in more children.

No. Elite institutions offer high salaries because they hire the most valued faculty. Moreover, in contrast to the broader labor market, faculty are equally likely to move up and down the prestige ladder, and they increase their salary either way. We speculate that these facts reflect the visible nature of faculty productivity and the sporadic nature of academic job openings.

Economics papers increasingly report balance, pretrend, placebo, and other “sniff tests,” rejection of which is bad news for authors, undermining the credibility of their main results. We derive nonparametric bounds on the latent proportion of significant sniff tests removed by the publication process (whether by p-hacking or relegation to the file drawer) and the proportion whose significance was due to true misspecification, not bad luck. Using a hand-collected sample of nearly 30,000 sniff tests, we estimate a removal rate of more than 30% for balance tests in randomized controlled trials and a misspecification rate of more than 40% for other tests.

We show that teachers’ value added (VA) depends on the VA of the teachers who preceded them. To do so, we use administrative data from North Carolina and find that a one-standard deviation increase in last grade’s mean teacher VA causes a 0.08σ decrease in teacher VA. Controlling for prior teacher assignment eliminates this bias. Under a benchmark policy that releases teachers in the bottom 5% of the VA distribution, 32% of teachers are incorrectly released using traditional techniques. Our results highlight the importance of incorporating dynamic features of education production into the estimation of teacher quality.