We analyze a large-scale survey of owners, managers, and employees of small businesses in the United States to understand the effects of the early stages of the COVID-19 pandemic on those businesses. The survey was fielded in late April 2020 among Facebook business page administrators, frequent sellers on Facebook's e-commerce platform Marketplace, and the general Facebook user population. We observe more than 66,000 responses covering most sectors of the economy, including many businesses that had stopped operating due to the pandemic. The survey asks 136 questions covering topics such as changes in business operations and employment, changes in financing patterns, and the interaction of household and business responsibilities. We characterize the adjustments implemented to survive the pandemic and explore the key challenges to continue operating or to reopen. We show how these patterns differ across industry, firm size, owner gender, and other firm characteristics.

BackgroundConducting surveys in low- and middle-income countries is often challenging because many areas lack a complete sampling frame, have outdated census information, or have limited data available for designing and selecting a representative sample. Geosampling is a probability-based, gridded population sampling method that addresses some of these issues by using geographic information system (GIS) tools to create logistically manageable area units for sampling. GIS grid cells are overlaid to partition a country’s existing administrative boundaries into area units that vary in size from 50 m × 50 m to 150 m × 150 m. To avoid sending interviewers to unoccupied areas, researchers manually classify grid cells as “residential” or “nonresidential” through visual inspection of aerial images. “Nonresidential” units are then excluded from sampling and data collection. This process of manually classifying sampling units has drawbacks since it is labor intensive, prone to human error, and creates the need for simplifying assumptions during calculation of design-based sampling weights. In this paper, we discuss the development of a deep learning classification model to predict whether aerial images are residential or nonresidential, thus reducing manual labor and eliminating the need for simplifying assumptions.ResultsOn our test sets, the model performs comparable to a human-level baseline in both Nigeria (94.5% accuracy) and Guatemala (96.4% accuracy), and outperforms baseline machine learning models trained on crowdsourced or remote-sensed geospatial features. Additionally, our findings suggest that this approach can work well in new areas with relatively modest amounts of training data.ConclusionsGridded population sampling methods like geosampling are becoming increasingly popular in countries with outdated or inaccurate census data because of their timeliness, flexibility, and cost. Using deep learning models directly on satellite images, we provide a novel method for sample frame construction that identifies residential gridded aerial units. In cases where manual classification of satellite images is used to (1) correct for errors in gridded population data sets or (2) classify grids where population estimates are unavailable, this methodology can help reduce annotation burden with comparable quality to human analysts.

No abstract

We analyze a large-scale survey of small business owners, managers, and employees in the United States to understand the effects of the COVID-19 pandemic on those businesses. We explore two waves of the survey that were fielded on Facebook in April 2020 and December 2020. We document five facts about the impact of the pandemic on small businesses. (1) Larger firms, older firms, and male-owned firms were more likely to remain open during the early stages of the pandemic with many of these heterogeneities persisting through the end of 2020. (2) At businesses that remained open, concerns about demand shocks outweighed concerns about supply shocks though the relative importance of supply shocks grew over time. (3) In response to the pandemic, almost a quarter of the firms reduced their prices with price reductions concentrated among businesses facing financial constraints and demand shocks; almost no firms raised prices. (4) Only a quarter of small businesses had access to formal sources of financing at the start of the pandemic, and access to formal financing affected how firms responded to the pandemic. (5) Increased household responsibilities affected the ability of managers and employees to focus on their work, whereas increased business responsibilities impacted their ability to take care of their household members. This effect persisted through December 2020 and was particularly strong for women and parents of school-aged children. We discuss how these facts inform our understanding of the economic effects of the COVID-19 pandemic and how they can help design policy responses to similar shocks. This paper was accepted by Tomasz Piskorski, finance.