Randomized Controlled Trials (RCTs) are the gold standard for evaluating the effects of interventions. However, the comparison of outcomes between treatment and control groups relies on the assumption that the assignment process itself—including how participants are assigned—does not affect outcomes. In this paper, I challenge that assumption by introducing and testing the concept of an "awareness state": the subject's knowledge of the study, their treatment status, and the assignment mechanism. I design a field experiment in India around a soil testing program that exogenously varies how participants are informed of their assignment. Villages are randomized into two main arms: one where treatment status is determined by a public lottery, and another by a private, computerized process. My design temporally separates assignment from treatment delivery, allowing me to isolate the causal effect of the assignment process itself. I find that estimated treatment effects differ across assignment methods and that these effects emerge even before the treatment is delivered. The effects are not uniform: the control group responds more strongly to the assignment method than the treated group. These findings suggest that the choice of assignment procedure is not neutral and that failing to account for it can threaten the interpretation and generalizability of standard treatment effects in RCTs.

We study how the saliency of past authoritarian regimes affects privacy concerns, leveraging Germany’s strong culture of Holocaust remembrance. We use detailed street-level data from Berlin to show the effect of Stolpersteine—individual memorials for victims of Nazi persecution—on privacy concerns, measured as blurring requests on Google Street View. To isolate causality, we leverage the quasi-random variation in Stolpersteine location after controlling for victim agglomeration patterns around each address. We show that Stolpersteine cause a localized increase in blurring, with the effect concentrating within 10 meters of a Stolperstein. We also find that Stolpersteine seen while commuting increase blurring. Furthermore, through an experimental survey we show that when Germans are primed to think about the Stolpersteine and Nazi persecution, they respond by spending more time on the experiment’s final consent form. This experimental design together with our blurring measure constitute two novel measures of privacy concerns.

[SSRN] [Extended Abstract at EC'24]

EC'24 Exemplary Empirics and Experiments Track Paper

2022 Sean Buckley Memorial Award for Best Second-Year Paper, Stanford University, Dept. of Economics

We investigate the role and impact of the county agricultural agents organized into the Cooperative Extension Service (CES) in the development of the Black rural population from its creation in 1904 to the Civil Rights Movement era. The project sheds light on the historical significance of Black county agents, their networks, their interaction with discriminatory practices, and their impact on migration and land loss. We digitized and created a new dataset of county agents by year and county they worked at, linking them not only to their socioeconomic characteristics but also to the population they worked with and to the state or federal agents they responded to. With this in hand, together with publicly available data on migration, discrimination, government spending, and agricultural development, we study the effects of the program on Black and white farmers, analyzing how discrimination and local control affected it. Moreover, we study whether Black county agents were able to slow down the Second Great Migration, allowing access to government programs designed during the New Deal era.