Methods: We conducted a cluster-randomized trial of community-level mask promotion in rural Bangladesh from November 2020 to April 2021 (N=600 villages, N=342,126 adults). We cross-randomized mask promotion strategies at the village and household level, including cloth vs. surgical masks. All intervention arms received free masks, information on the importance of masking, role modeling by community leaders, and in-person reminders for 8 weeks. The control group did not receive any interventions. Neither participants nor field staff were blinded to intervention assignment. Outcomes included symptomatic SARS-CoV-2 seroprevalence (primary) and prevalence of proper mask-wearing, physical distancing, and symptoms consistent with COVID-19 (secondary). Mask-wearing and physical distancing were assessed through direct observation at least weekly at mosques, markets, the main entrance roads to villages, and tea stalls. At 5 and 9 weeks follow-up, we surveyed all reachable participants about COVID-related symptoms. Blood samples collected at 10-12 weeks of follow-up for symptomatic individuals were analyzed for SARS-CoV-2 IgG antibodies.
Results: There were 178,288 individuals in the intervention group and 163,838 individuals in the control group. The intervention increased proper mask-wearing from 13.3% in control villages (N=806,547 observations) to 42.3% in treatment villages (N=797,715 observations) (adjusted percentage point difference = 0.29 [0.27, 0.31]). This tripling of mask usage was sustained during the intervention period and two weeks after. Physical distancing increased from 24.1% in control villages to 29.2% in treatment villages (adjusted percentage point difference = 0.05 [0.04, 0.06]). After 5 months, the impact of the intervention faded, but mask-wearing remained 10 percentage points higher in the intervention group.
The proportion of individuals with COVID-like symptoms was 7.62% (N=13,273) in the intervention arm and 8.62% (N=13,893) in the control arm. Blood samples were collected from N=10,952 consenting, symptomatic individuals. Adjusting for baseline covariates, the intervention reduced symptomatic seroprevalence by 9.3% (adjusted prevalence ratio (aPR) = 0.91 [0.82, 1.00]; control prevalence 0.76%; treatment prevalence 0.68%). In villages randomized to surgical masks (n = 200), the relative reduction was 11.2% overall (aPR = 0.89 [0.78, 1.00]) and 34.7% among individuals 60+ (aPR = 0.65 [0.46, 0.85]). No adverse events were reported.
Conclusions: Our intervention demonstrates a scalable and effective method to promote mask adoption and reduce symptomatic SARS-CoV-2 infections.
Democracy is widely believed to contribute to economic growth and public health. However, we find that this conventional wisdom is no longer true and even reversed; democracy has persistent negative impacts on GDP growth since the beginning of this century. This finding emerges from five different instrumental variable strategies. Our analysis suggests that democracies cause slower growth through less investment, less trade, and slower value-added growth in manufacturing and services. For 2020, democracy is also found to cause more deaths from Covid-19.
Background: A growing body of scientific evidence suggests that face masks can slow the spread of COVID-19 and save lives, but mask usage remains low across many parts of the world, and strategies to increase mask usage remain untested and unclear. Methods: We conducted a cluster-randomized trial of community-level mask promotion in rural Bangladesh involving 341,830 adults in 600 villages. We employed a series of strategies to promote mask usage, including free household distribution of surgical or cloth masks, distribution and promotion at markets and mosques, mask advocacy by Imams during Friday prayers, role modeling by local leaders, promoters periodically monitoring passers-by and reminding people to put on masks, village police accompanying those mask promoters, providing monetary rewards or certificates to villages if mask-wearing rate improves, public signaling of mask-wearing via signage, text message reminders, messaging emphasizing either altruistic or self-protection motives for mask-wearing, and extracting verbal commitments from households. The primary objective was to assess which of these interventions would increase proper (covering nose and mouth) wearing of face masks, and secondarily, whether mask promotion unintentionally creates moral hazard and decreases social distancing. This analysis is part of larger study evaluating the effect of mask-wearing on transmission of SARS-CoV-2.
Results: There were 64,937 households in the intervention group and 64,183 households in the control group; study recruitment has ended. In the control group, proper mask-wearing was practiced by 13% of those observed across the study period. Free distribution of masks along with role modeling by community leaders produced only small increases in mask usage during pilot interventions. Adding periodic monitoring by mask promoters to remind people in streets and public places to put on the masks we provided increased proper mask-wearing by 29.0 percentage points (95% CI: 26.7% - 31.3%). This tripling of mask usage was sustained over all 10 weeks of surveillance, which includes a period after intervention activities ended. Physical distancing, measured as the fraction of individuals at least one arm’s length apart, also increased by 5.2 percentage points (95% CI: 4.2%-6.3%). Beyond the core intervention package comprised of free distribution and promotion at households/mosques/markets, leader endorsements plus periodic monitoring and reminders, several elements had no additional effect on mask wearing, including: text reminders, public signage commitments, monetary or non-monetary incentives, altruistic messaging or verbal commitments, or village police accompanying the mask promoters (the last not cross-randomized, but assessed in panel data). No adverse events were reported during the study period.
Conclusions: Our intervention demonstrates a scalable and cost-effective method to promote mask adoption and save lives, and identifies a precise combination of intervention activities that were necessary. Comparisons between pilots shows that free mask distribution alone is not sufficient to increase mask-wearing, but adding periodic monitoring in public places to remind people to wear the distributed masks had large effects on behavior. The absence of any further effect of the village police suggests that the operative mechanism is not any threat of formal legal sanctions, but shame and people’s aversion to a light informal social sanction. The persistence of effects for 10 weeks and after the end of the active intervention period, as well as increases in physical distancing, all point to changes in social norms as a key driver of behavior change. Our cross-randomizations suggest that improved mask-wearing norms can be achieved without incentives that require costly monitoring, that aesthetic design choices and colors can influence mask-wearing, and that surgical masks with a substantially higher filtration efficiency can be a cost-effective alternative to cloth masks (1/3 the cost) and are equally or more likely to be worn. Implementing these interventions – including distribution of free masks, and the information campaign, reminders, encouragement – cost $2.30-$3.75 per villager, or between $8 and $13 per person adopting a mask. Combined with existing estimates of the efficacy of masks in preventing COVID-19 deaths, this implies that the intervention cost $28,000-$66,000 per life saved. Beyond reducing the transmission of COVID-19, mask distribution is likely to be a cost-effective strategy to prevent future respiratory disease outbreaks.
Algorithms produce a growing portion of decisions and recommendations both in policy and business. Such algorithmic decisions are natural experiments (conditionally quasirandomly assigned instruments) since the algorithms make decisions based only on observable input variables. We use this observation to develop a treatment-effect estimator for a class of stochastic and deterministic algorithms. Our estimator is shown to be consistent and asymptotically normal for well-defined causal effects. A key special case of our estimator is a high-dimensional regression discontinuity design. The proofs use tools from differential geometry and geometric measure theory, which may be of independent interest.
The practical performance of our method is first demonstrated in a high-dimensional simulation resembling decision-making by machine learning algorithms. Our estimator has smaller mean squared errors compared to alternative estimators. We finally apply our estimator to evaluate the effect of Coronavirus Aid, Relief, and Economic Security (CARES) Act, where more than $10 billion worth of relief funding is allocated to hospitals via an algorithmic rule. The estimates suggest that the relief funding has little effect on COVID- 19-related hospital activity levels. Naive OLS and IV estimates exhibit substantial selection bias.
This study presents the design and results of a rapid-fire survey that collects labor market data for individuals in the United States. The purpose is to test online panels for their application to social, economic, and demographic information as well as to apply this approach to the U.S. labor market. The Yale Labor Survey (YLS) used an online panel from YouGov to replicate statistics from the Current Population Survey (CPS), the government’s official source of household labor market statistics. The YLS’s advantages included its timeliness, low cost, and ability to develop new questions quickly to study unusual labor market patterns during the COVID-19 pandemic. Results from the YLS track employment data closely from the CPS during the pandemic. Although YLS estimates of unemployment and participation rates mirrored the broad trends in CPS data, YLS estimates of those two rates were less accurate than for employment. The study demonstrates the power of carefully crafted online surveys to replicate expensive traditional methods quickly and inexpensively.
Countries with more democratic political regimes experienced greater GDP loss and more deaths from COVID-19 in 2020. Using five diffferent instrumental variable strategies, we find that democracy is a major cause of the wealth and health losses. This impact is global and is not driven by China and the US alone. A key channel for democracy’s negative impact is weaker and narrower containment policies at the beginning of the outbreak, not the speed of introducing policies.
We examine the potential for exploiting retailer location choice in targeting health interventions. Using geospatial data, we quantify proximity to vaccines created by a U.S. federal program distributing COVID-19 vaccines to commercial retail pharmacies. We assess the distributional impacts of a proposal to provide vaccines at Dollar General, a low-priced general merchandise retailer. Adding Dollar General to the federal program would substantially decrease the distance to vaccine sites for low-income, rural, and minority U.S. households, groups for which COVID-19 vaccine take-up has been disproportionately slow.