Latest Research Publications

This paper studies the effects of the Green Revolution, one of the largest development projects in the late 20th Century, on individualism. We use a difference-in-differences approach to investigate how rapid technological and organizational changes in traditional agriculture reshaped cultural norms in Indonesia, where Green Revolution rollout was largely driven by centralized government interventions. Regions with greater exposure to Green Revolution policies experienced significant and persistent increases in individualism (as measured by a higher prevalence of infrequent names) decades after the end of the program. Results are not driven by pre-trends and hold across a battery of robustness checks. We further document suggestive, correlational evidence of a persistent fall in community funding for public goods, a key measure of collective action. Our findings suggest that technical change can lead to rapid and persistent changes in culture.

In this paper, we study optimal monetary policy in an economy where firms’ price-setting decisions are distorted by a signal-extraction problem: they cannot perfectly distinguish aggregate from idiosyncratic shocks based on noisy local information. This systematic misattribution of aggregate shocks to firm-specific factors generates an additional price response, implying that strict price stability is no longer optimal. Instead, the optimal policy fully stabilizes the output gap, which necessarily requires accommodating fluctuations in the price level. The cyclicality of the optimal price level depends on the source of firms’ incomplete information: learning from local demand signals implies a procyclical price level, whereas learning from local productivity signals yields a countercyclical one.

We study a signaling game where agents signal their type by choosing when to quit pursuing an uncertain project. High types observe news about project quality and quit when bad news arrives. Low types who do not observe any news may mimic high types by quitting continuously over a phase of time. The reputation dynamics may exhibit nonmonotonicity with respect to quitting time. Our analysis offers a unifying explanation for how and when both early and late quitting can enhance reputation and suggests novel welfare and policy implications.

Using information in returns, we identify the stochastic process of consumption. We find that aggregate consumption reacts over multiple quarters to innovations spanned by financial markets. This persistent component accounts for over a quarter of consumption variation. These shocks command a large and significant risk premium, driving a large share of stocks' and a small yet significant fraction of bonds' time-series variation. Nevertheless, we find no support for stochastic volatility of consumption driving time-varying risk premia. Finally, an otherwise standard recursive utility model based on our estimated process explains equity premium and risk-free rate puzzles with low-risk aversion.

This study investigates how firms engage in information-seeking lobbying to address trade policy uncertainty. I argue that lobbying enables firms to gain early insights into forthcoming tariff actions, allowing them to strategically stockpile products likely to be targeted. Using shipping records of US firms during the 2018 US–China trade war, I find that lobbying firms increased imports of soon-to-be-tariffed products before tariff lists were publicly released, compared to non-lobbying firms. This selective stockpiling pattern disappeared after tariff announcements. Further analysis shows that lobbying firms were less likely to request tariff exemptions for products they had preemptively stockpiled, suggesting that information-seeking lobbying during policy formulation provides an additional benefit by reducing the need for costly government engagement during the implementation phase.

We study whether access to local pollution information causes investors to make greener portfolio allocations, exploiting the rollout of air quality monitoring stations in India. Using a triple-differences framework on the trading records of 19 million investors, we show that retail investors’ holdings in “brown” stocks become more negatively related to local pollution after a nearby station appears. This effect is more pronounced on “alert” dates when air quality is reported to be harmful. The effect is strongest among tech-savvy investors likely “treated” by real-time pollution data, and younger investors, who may be more sensitive to environmental concerns.

Problem definition: With the advance of data analytics, many disease prediction models have been developed with the intent of detecting diseases earlier and improving patient outcomes through earlier treatment. The operationalization of interventions and care based on these predictive models is critical to attaining these goals. We study the real-world effects of a machine learning-guided colorectal cancer screening outreach program deployed at a health system in Pennsylvania. Methodology/results: Using a regression discontinuity design based on the predicted risk score for having cancer, we find that the program increases the likelihood of colonoscopy uptake in three and six months by 6.0 percentage points (214% increase relative to the control sample within the bandwidth) and 6.9 percentage points (117% increase), respectively. Importantly, we also find significant effects on mortality. We estimate that the program decreases two-year mortality by 6.2 percentage points (43% decrease). Managerial implications: Our finding suggests that a proactive cancer screening outreach program where individuals are selected for intervention based on a machine learning algorithm could significantly improve patient outcomes in addition to achieving higher disease detection rates. Our analysis demonstrates an analytical framework for rigorously evaluating machine learning-aided outreach programs for other cancers and diseases. Establishing unbiased estimates of the impact of machine learning-aided screening outreach is critical for capacity planning of screening resources, such as colonoscopies.

Problem definition: To enhance supply chain collaboration, major retailers increasingly grant suppliers access to their data platforms, leveraging data analytics to deliver valuable demand insights. However, such access raises the risk of channel encroachment by suppliers. This paper examines the retailer's strategy for deploying data analytics in the presence of supplier encroachment. Methodology/results: Under the Bayesian persuasion framework, the retailer's data analytics strategy is formulated as a signaling rule that maps market states into predictions with partial informational granularity. We further compare two organizational structures of data governance: a shared data platform, which delivers identical demand predictions to both the retailer and the supplier, and a separate data platform, which provides accurate predictions for the retailer, offering less precise insights to the supplier. On a shared data platform, the retailer adopts perfectly accurate analytics when the supplier's encroachment cost is either very high or very low. At moderate encroachment costs, the shared platform generates informative yet biased forecasts, which surprisingly benefit both firms because of the ease of double marginalization. By contrast, under a separate platform, no informative predictions are generated for the supplier at moderate encroachment costs. Overall, comparing the two organizational structures, the retailer prefers a shared platform when the supplier's encroachment cost is relatively high. Managerial implications: Our findings reveal a nonmonotonic relationship between the accuracy of data analytics and suppliers' encroachment efficiency, offering practical guidance for retailers in implementing effective learning algorithms. Moreover, we highlight the benefits of transparent and integrated data analytics within vertical distribution channels, providing managerial insights into the organizational choices underpinning retailers' data governance.

Designing centralized matching systems on two-sided platforms entails a tradeoff between matching efficiency and agent autonomy. We investigate this tradeoff in the ride-hailing market by comparing two prevalent centralized dispatch systems: a driver-accept system, where drivers decide whether to accept e-hail requests, and an auto-accept system, where drivers are automatically assigned to these requests. We develop a dynamic structural model of a two-sided market where strategic drivers optimize acceptance and relocation decisions over a shift, and riders choose between transportation modes. We apply this model to Singapore’s taxi market, where a leading operator used a driver-accept dispatch system for e-hail trips. We develop an iterative algorithm to solve for the market equilibrium as a fixed point of a nested loop. Our counterfactual analyses reveal that in the short-term equilibrium with fixed demand, switching to an auto-accept system increases driver earnings through higher vehicle utilization, despite the loss of private information caused by removing driver choice. In the long-term equilibrium, drivers’ earnings increase further due to market expansion, although benefits are unevenly distributed between e-hail and street-hail services. Consumer surplus increases in both scenarios. Our findings highlight that centralized assignment can enhance overall welfare by mitigating search frictions, even when agents value autonomy.