Skip to main content

COVID-19 Finance Working Papers


Click on the working paper titles below to find out more about the research co-authored by the Yale SOM finance faculty on Covid-19 related topics.

Effects of a large-scale social media advertising campaign on holiday travel and COVID-19 infections: a cluster randomized controlled trial

  • Emily Breza
  • Fatima Cody Stanford
  • Marcella Alsan
  • Burak Alsan
  • Abhijit Banerjee
  • Arun G. Chandrasekhar
  • Sarah Eichmeyer
  • Traci Glushko
  • Paul Goldsmith-Pinkham
  • Kelly Holland
  • Emily Hoppe
  • Mohit Karnani
  • Sarah Liegl
  • Tristan Loisel
  • Lucy Ogbu-Nwobodo
  • Benjamin A. Olken
  • Carlos Torres
  • Pierre-Luc Vautrey
  • Erica T. Warner
  • Susan Wootton
  • Esther Duflo


During the Coronavirus Disease 2019 (COVID-19) epidemic, many health professionals used social media to promote preventative health behaviors. We conducted a randomized controlled trial of the effect of a Facebook advertising campaign consisting of short videos recorded by doctors and nurses to encourage users to stay at home for the Thanksgiving and Christmas holidays (NCT04644328 and AEARCTR-0006821). We randomly assigned counties to high intensity (n = 410 (386) at Thanksgiving (Christmas)) or low intensity (n = 410 (381)). The intervention was delivered to a large fraction of Facebook subscribers in 75% and 25% of randomly assigned zip codes in high- and low-intensity counties, respectively. In total, 6,998 (6,716) zip codes were included, and 11,954,109 (23,302,290) users were reached at Thanksgiving (Christmas). The first two primary outcomes were holiday travel and fraction leaving home, both measured using mobile phone location data of Facebook users. Average distance traveled in high-intensity counties decreased by −0.993 percentage points (95% confidence interval (CI): –1.616, −0.371; P = 0.002) for the 3 days before each holiday compared to low-intensity counties. The fraction of people who left home on the holiday was not significantly affected (adjusted difference: 0.030; 95% CI: −0.361, 0.420; P = 0.881). The third primary outcome was COVID-19 infections recorded at the zip code level in the 2-week period starting 5 days after the holiday. Infections declined by 3.5% (adjusted 95% CI: −6.2%, −0.7%; P = 0.013) in intervention compared to control zip codes. Social media messages recorded by health professionals before the winter holidays in the United States led to a significant reduction in holiday travel and subsequent COVID-19 infections.

All or Nothing? Partial Business Shutdowns and COVID-19 Fatality Growth

Matthew I. Spiegel, Yale University - Yale School of Management, International Center for Finance

Heather Tookes, Yale University - Yale School of Management; Yale University - International Center for Finance


Using a hand-collected database of partial business closures for all U.S. counties from March through December 2020, we examine the impact of capacity restrictions on fatality growth due to COVID-19. For the restaurant and bar sector, we find that several combinations of partial capacity restrictions are as effective as full shutdowns. Point estimates indicate that, for the average county, limiting restaurants to 25% of capacity and bars to outdoor service reduces the fatality growth six weeks ahead by approximately 41% while completely closing them reduces fatality growth by about 32%. For gyms, we find that, while full closures reduce the COVID-19 fatality growth rate, partial closures may be counterproductive relative to leaving capacity unrestricted. For salons and other personal services, we find mixed evidence that limiting them to 25% of capacity reduces fatalities. However, other constraints are either ineffective or even counterproductive.

A Mega-Study of Text-Based Nudges Encouraging Patients to Get Vaccinated at an Upcoming Doctor’s Appointment

Katherine L. Milkman – University of Pennsylvania - The Wharton School

Mitesh S. Patel – University of Pennsylvania - Perelman School of Medicine, Department of Medicine

Linnea Gandhi – University of Pennsylvania

Heather Graci – University of Pennsylvania

Dena Gromet - University of Pennsylvania

Hung Ho - University of Pennsylvania - Operations & Information Management Department

Joseph Kay - University of Pennsylvania

Timothy Lee - University of Pennsylvania

Modupe Akinola – Columbia University - Columbia Business School

John Beshears – Harvard University - Business School (HBS); National Bureau of Economic Research (NBER)

Jon Bogard – University of California, Los Angeles (UCLA), Anderson School of Management, Students

Alison Buttenheim – University of Pennsylvania

Christopher Chabris – Geisinger Health System

Gretchen B. Chapman – Rutgers, The State University of New Jersey

James J. Choi – Yale School of Management; National Bureau of Economic Research (NBER)

Hengchen Dai – University of California, Los Angeles (UCLA) - Anderson School of Management

Craig R. Fox – University of California, Los Angeles (UCLA) - Anderson School of Management

Amir Goren – Geisinger Health System

Matthew Hilchey – University of Toronto

Jillian Hmurovic – University of Pennsylvania - The Wharton School

Leslie John – Harvard Business School

Dean Karlan – Northwestern University

Melanie Kim – University of Toronto

David Laibson – Harvard University - Department of Economics; National Bureau of Economic Research (NBER)

Cait Lamberton – The Wharton School

Brigitte C. Madrian – Brigham Young University Marriott School of Business; National Bureau of Economic Research (NBER)

Michelle N. Meyer – Geisinger Health System, Center for Translational Bioethics and Health Care Policy

Maria Modanu – Columbia University

Jimin Nam – Harvard Business School

Todd Rogers – Harvard University - Harvard Kennedy School (HKS)

Renante Rondina – University of Toronto - Rotman School of Management

Silvia Saccardo – Carnegie Mellon University, Department of Social and Decision Sciences

Maheen Shermohammed – Geisinger Health System

Dilip Soman – University of Toronto - Behavioural Economics in Action at Rotman (BEAR)

Jehan Sparks – University of California, Los Angeles (UCLA)

Caleb Warren – University of Arizona

Megan Weber – University of California, Los Angeles (UCLA)

Ron Berman – University of Pennsylvania - The Wharton School

Chalanda Evans – University of Pennsylvania - Perelman School of Medicine

Christopher Snider – University of Pennsylvania - Perelman School of Medicine

Eli Tsukayama – University of Hawaii at West O'ahu

Christophe Van den Bulte – University of Pennsylvania - Marketing Department

Kevin Volpp – University of Pennsylvania - Perelman School of Medicine, Department of Medicine

Angela Duckworth – University of Pennsylvania - Department of Psychology


Many Americans fail to get life-saving vaccines each year, and the availability of a vaccine for COVID-19 makes the challenge of encouraging vaccination more urgent than ever. We present a large field experiment (N=47,308) testing 19 nudges delivered to patients via text message and designed to boost adoption of the influenza vaccine. Our findings suggest text messages sent prior to a primary care visit can boost vaccination rates by up to 11%. Overall, interventions performed better when they were (a) framed as reminders to get flu shots that were already reserved for the patient and (b) congruent with the sort of communications patients expected to receive from their healthcare provider (i.e., not surprising, casual, or interactive). Our most potent intervention reminded patients twice to get their flu shot at their upcoming doctor’s appointment and indicated it was reserved for them. This successful script could be used as a template for campaigns to encourage the adoption of life-saving vaccines, including against COVID-19.

What Explains Differences in Finance Research Productivity During the Pandemic?

Brad M. Barber, University of California, Davis

Wei Jiang, Columbia Business School – Finance and Economics; ECGI; NBER

Adair Morse, University of California, Berkeley – Haas School of Business; NBER

Manju Puri, Duke University – Fuqua School of Business; NBER

Heather Tookes, Yale University – Yale School of Management; Yale University – International Center for Finance

Ingrid M. Werner, The Ohio State University – Fisher College of Business; CEPR


How has COVID-19 impacted faculty productivity? Does it differ by characteristics such as gender and family structure? To answer these questions, we conduct a survey of American Finance Association (AFA) members. Overall, faculty respondents report lower research productivity with less time allocated to research and more time allocated to teaching. There is also heterogeneity: 14.5% of respondents report an increase in productivity. We find the negative effects on research productivity are particularly large for women and faculty with young children regardless of gender. Thus, the pandemic has the effect of widening the gender gap for women and creates a “family gap” in productivity for both men and women with young children. Lower research productivity for faculty with young children is explained, to a large extent, by increased time spent on childcare. Our results suggest the need for deliberate policy to factor in these underlying mechanisms. We caution that a one-size-fits-all tenure-clock extension can have unintended negative consequences of increasing disparity.

Comparison of Knowledge and Information-Seeking Behavior After General COVID-19 Public Health Messages and Messages Tailored for Black and Latinx Communities

Marcella Alsan, MD, MPH, PhD

Fatima Cody Stanford, MD, MPH, MPA

Abhijit Banerjee, PhD

Emily Breza, PhD

Arun G. Chandrasekhar, PhD

Sarah Eichmeyer, MA

Paul Goldsmith-Pinkham, PhD

Lucy Ogbu-Nwobodo, MD, MS, MAS

Benjamin A. Olken, PhD

Carlos Torres, MD

Anirudh Sankar, MMath

Pierre-Luc Vautrey, MSc

Esther Duflo, PhD


The paucity of public health messages that directly address communities of color might contribute to racial and ethnic disparities in knowledge and behavior related to coronavirus disease 2019 (COVID-19).


To determine whether physician-delivered prevention messages affect knowledge and information-seeking behavior of Black and Latinx individuals and whether this differs according to the race/ethnicity of the physician and tailored content.


Randomized controlled trial. (Registration:, NCT04371419; American Economic Association RCT Registry, AEARCTR-0005789)


United States, 13 May 2020 to 26 May 2020.


14 267 self-identified Black or Latinx adults recruited via Lucid survey platform.


Participants viewed 3 video messages regarding COVID-19 that varied by physician race/ethnicity, acknowledgement of racism/inequality, and community perceptions of mask-wearing.


Knowledge gaps (number of errors on 7 facts on COVID-19 symptoms and prevention) and information-seeking behavior (number of Web links demanded out of 10 proposed).


7174 Black (61.3%) and 4520 Latinx (38.7%) participants were included in the analysis. The intervention reduced the knowledge gap incidence from 0.085 to 0.065 (incidence rate ratio, [IRR], 0.737 [95% CI, 0.600 to 0.874]) but did not significantly change information-seeking incidence. For Black participants, messages from race/ethnic-concordant physicians increased information-seeking incidence from 0.329 (for discordant physicians) to 0.357 (IRR, 1.085 [CI, 1.026 to 1.145]).


Participants' behavior was not directly observed, outcomes were measured immediately postintervention in May 2020, and online recruitment may not be representative.


Physician-delivered messages increased knowledge of COVID-19 symptoms and prevention methods for Black and Latinx respondents. The desire for additional information increased with race-concordant messages for Black but not Latinx respondents. Other tailoring of the content did not make a significant difference.

Primary Funding Source:

National Science Foundation; Massachusetts General Hospital; and National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases.

Business Restrictions and COVID Fatalities

Matthew I. Spiegel, Yale University - Yale School of Management, International Center for Finance

Heather Tookes, Yale University - Yale School of Management; Yale University - International Center for Finance


We hand-collect a time-series database of business closures and related restrictions for every county in the United States since March 2020. We then relate these policies to future growth in deaths due to COVID-19. To our knowledge, ours is the most comprehensive database of U.S. COVID-19 business policies that has been assembled to date. Across specifications, stay-at-home orders, mandatory mask requirements, beach and park closures, restaurant closures, and high risk (Level 2) business closures are the policies that most consistently predict lower 4- to 6- week-ahead fatality growth. For example, baseline estimates imply that a county with a mandatory mask policy in place today will experience 4- week and 6- week ahead fatality growth rates that are each 1% lower (respectively) than a county without such an order in place. This relationship is significant, both statistically and in magnitude. It represents 12% of the sample mean of weekly fatality growth. The baseline estimates for stay-at-home, restaurant and high-risk business closures are similar in magnitude to what we find for mandatory mask policies. We fail to find consistent evidence in support of the hypothesis that some of the other business restrictions (such as spa closures, school closures, and the closing of the low- to medium-risk businesses that are typically allowed in Phase I reopenings) predict reduced fatality growth at four-to-six-week horizons. Some policies, such as low- to medium-business risk closures may even be counterproductive. To address potential endogeneity concerns, we conduct two tests. First, we exploit the fact that many county regulations are imposed at the state-level through Governors’ executive orders. Following the intuition that smaller counties often inherit state-level regulations that are intended to reduce transmission and deaths in more populous regions, we remove the 5 most populous counties in each state from the sample. In the second test, we match counties that lie near (but not on) state borders to counties in different states that are also near (but not on) state borders and are within 100 miles of that county. Absent policy differences, these nearby counties should see similar trends in virus transmission; making them good controls. We continue to find that stay-at-home, mandatory masks, beach and park closures, restaurant closures, and high risk business closures all predict declines in future fatality growth.

Interacting Regional Policies in Containing a Disease

Arun G. Chandrasekhar, Stanford University - Department of Economics

Paul S. Goldsmith-Pinkham, Yale School of Management

Matthew O. Jackson, Stanford University - Department of Economics; Santa Fe Institute

Samuel Thau, Harvard University


Regional quarantine policies, in which a portion of a population surrounding infections are locked down, are an important tool to contain disease. However, jurisdictional governments - such as cities, counties, states, and countries - act with minimal coordination across borders. We show that a regional quarantine policy's effectiveness depends upon whether (i) the network of interactions satisfies a balanced-growth condition, (ii) infections have a short delay in detection, and (iii) the government has control over and knowledge of the necessary parts of the network (no leakage of behaviors). As these conditions generally fail to be satisfied, especially when interactions cross borders, we show that substantial improvements are possible if governments are proactive: triggering quarantines in reaction to neighbors' infection rates, in some cases even before infections are detected internally. We also show that even a few lax governments - those that wait for nontrivial internal infection rates before quarantining - impose substantial costs on the whole system. Our results illustrate the importance of understanding contagion across policy borders and offer a starting point in designing proactive policies for decentralized jurisdictions.

Messages on Covid-19 Prevention in India Increased Symptoms Reporting and Adherence to Preventive Behaviors Among 25 Million Recipients with Similar Effects on Non-Recipient Members of Their Communities

Abhijit V. Banerjee, Massachusetts Institute of Technology (MIT) - Department of Economics

Marcella Alsan, Harvard University - Harvard Kennedy School (HKS)

Emily Breza, Harvard University

Arun G. Chandrasekhar, Stanford University - Department of Economics

Abhijit Chowdhury, Institute of Post Graduate Medical Education and Research - School of Digestive and Liver Diseases

Esther Duflo, Massachusetts Institute of Technology (MIT) - Department of Economics; Abdul Latif Jameel Poverty Action Lab (J-PAL); National Bureau of Economic Research (NBER); Centre for Economic Policy Research (CEPR); Bureau for Research and Economic Analysis of Development (BREAD)

Paul S. Goldsmith-Pinkham, Yale School of Management

Benjamin Olken, Massachusetts Institute of Technology (MIT) - Department of Economics


During health crises, like COVID-19, individuals are inundated with messages promoting health-preserving behavior. Does additional light-touch messaging by a credible individual change behavior? Do the features of the message matter? To answer this, we conducted a large-scale messaging campaign in West Bengal, India. Twenty-five million individuals were sent an SMS containing a 2.5-minute clip, delivered by West Bengal native and 2019 Nobel laureate Abhijit Banerjee. All messages encouraged reporting symptoms to the local public health worker. In addition, each message emphasizes one health-preserving behavior (distancing or hygiene) and one motivation for action (effects on everyone or just on self). Further, some messages addressed concerns about ostracism of the infected. Messages were randomized at the PIN code level. As control, three million individuals received a message pointing them to government information. The campaign (i) doubled the reporting of health symptoms to the community health workers (p = 0.001 for fever, p = 0.024 for respiratory symptoms); (ii) decreased travel beyond one’s village in the last two days by 20% (p = 0.026) (on a basis of 37% in control) and increased estimated hand-washing when returning home by 7% (p = 0.044) (67.5% in control); (iii) spilled over to behaviors not mentioned in the message – mask-wearing was never mentioned but increased 2% (p = 0.042), while distancing and hygiene both increased in the sample where they were not mentioned by similar amounts as where they were mentioned; (iv) spilled over onto nonrecipients within the same community, with effects similar to those for individuals who received the messages.

Inside the Mind of a Stock Market Crash

Stefano Giglio, Yale School of Management; National Bureau of Economic Research (NBER); Centre for Economic Policy Research (CEPR)

Matteo Maggiori, Harvard University; National Bureau of Economic Research (NBER); Centre for Economic Policy Research (CEPR)

Johannes Stroebel, New York University (NYU) - Leonard N. Stern School of Business; National Bureau of Economic Research (NBER); Centre for Economic Policy Research (CEPR)

Stephen P. Utkus, University of Pennsylvania; Center for Financial Markets and Policy, Georgetown University


We analyze how investor expectations about economic growth and stock returns changed during the February-March 2020 stock market crash induced by the COVID-19 pandemic, as well as during the subsequent partial stock market recovery. We surveyed retail investors who are clients of Vanguard at three points in time: (i) on February 11-12, around the all-time stock market high, (ii) on March 11-12, after the stock market had collapsed by over 20%, and (iii) on April 16-17, after the market had rallied 25% from its lowest point. Following the crash, the average investor turned more pessimistic about the short-run performance of both the stock market and the real economy. Investors also perceived higher probabilities of both further extreme stock market declines and large declines in short-run real economic activity. In contrast, investor expectations about long-run (10-year) economic and stock market outcomes remained largely unchanged, and, if anything improved. Disagreement among investors about economic and stock market outcomes also increased substantially following the stock market crash, with the disagreement persisting through the partial market recovery. Those respondents who were the most optimistic in February saw the largest decline in expectations and sold the most equity. Those respondents who were the most pessimistic in February largely left their portfolios unchanged during and after the crash.

Designing the Main Street Lending Program: Challenges and Options

William B. English, Professor in the Practice of Finance, Yale School of Management

J. Nellie Liang, Senior Fellow, Hutchins Center for Fiscal and Monetary Policy, Brookings Institution


Unlike the 2008 financial crisis, the current economic crisis brought on by the COVID-19 pandemic reflects fundamental cash flow problems for many businesses as revenues have almost completely stopped. Businesses will need substantial financial resources—from previous saving, direct government grants, or credit—to pay bills, survive the shutdown, and be ready to rehire workers quickly and restart spending. Extending credit can help some businesses manage the near-term shortfall in revenues, restructure operations, and prevent unnecessary failures at a time when bankruptcies will be costly.

The Main Street Lending Program (MSLP) is set up to provide loans to small and mid-size firms and large below-investment-grade firms that were financially sound before the onset of the pandemic. The CARES Act authorizes the Federal Reserve to establish the program under its emergency authorities with capital provided by the U.S. Treasury.1However, the MSLP is a big step for the government, and a difficult one. Lending to risky firms is a significant challenge given U.S. aversion to government equity stakes in private businesses and the Fed’s legal requirements to be secured to its satisfaction and to lend only to solvent firms.

Predicting Initial Unemployment Insurance Claims Using Google Trends

Paul S. Goldsmith-Pinkham, Yale School of Management

Aaron Sojourner, University of Minnesota


Understanding changes in national and state-level initial unemployment insurance (UI) claims has value to markets, policymakers, and economists. Initial claims measure the number of Americans filing new claims for UI benefits is one of the most-sensitive, high-frequency official statistics used to detect changes in the labor market. However, official federal data on UI claims comes out at a weekly interval and at a lag. While last week was a record-setting week, this week’s UI numbers doubled that record, with the largest rise in new unemployment claims in U.S. history, due to widespread quarantines. In advance of each week’s release, we constructed harmonized news-based measures of UI claims in a state over various sets of consecutive days. We also build a daily panel on the intensity of search interest for the term “file for unemployment” for each state on Google Trends. Changes in search intensity predict changes in initial claims. We forecast state and national UI claims using the estimated daily model. These models are new and partially validated.