COVID-19 pilot study finds signs of economic distress outside India’s metropolitan areas
By Nitya Agrawal and Siobhan McDonough
Since early March, IDinsight has provided analytical and advisory support to governments and civil society organizations to help them respond to the COVID-19 crisis. We have provided economic relief recommendations to the Department of Health in the Philippines, and cash transfer and physical distancing recommendations to government partners in sub-Saharan Africa. In India, we’ve been using our Data on Demand infrastructure to collect data via phone surveys across the country’s poorest districts to help inform government response to the pandemic.
India’s COVID-19 lockdown began on 25 March. That week, we mobilized our systems to conduct a pilot dipstick survey — a survey that provides a quick snapshot of the economic situation, even though it’s not necessarily representative of a population of interest. We conducted the survey end-to-end, in five days with about 300 respondents, nearly half of them female, in four northern Indian states. In each state, we selected one location to get a mix of urban and rural respondents. This quick turnaround was possible because of our previous Data on Demand Innovations work on phone-based data collection and generating sample frames. For this pilot we gathered data on COVID-related knowledge and attitudes, physical distancing practices, and economic impacts of the lockdown. We are currently collecting data on these and other indicators through representative sample surveys with more than 6000 respondents across 27 districts.
We primarily conducted this pilot to assess the feasibility of fast survey turnaround and to pressure-test our questionnaire. Our survey was conducted early in the lockdown and with a small sample size, and our results are not intended to provide a full picture of what is happening in urban and rural India. However, our findings can provide an initial snapshot on the challenges faced by some of the most vulnerable Indians, and will be important to inform future COVID work with our partners.
Preliminary findings on economic impact and health knowledge from four districts in north India
Our 300 respondents in Northern India were among the most vulnerable to health and economic effects of COVID-19, more than 60 percent of our respondents work in informal positions such as day labor, self employment, or agricultural sales. About 22 percent were not employed at the time of the survey. It was vital to take the following steps to reduce the burden on respondents already experiencing distress and upheaval during this time. Most importantly, we wanted to ensure we used respondents’ valuable time well. After having found during pilots 10–15 minute surveys were the optimal length, we kept our surveys to only 15 minutes. Furthermore, it was important to us to limit survey questions to only those we felt could best inform policymakers’s response to improve respondents’ situations. Because the impact of COVID could differ by gender, we sought equal gender representation in our phone survey sample by asking to speak to the primary female decision-maker half the time, which helped us reach close to half female respondents.
About half of all surveyed households reported losing income from the beginning of March to the end of March, about 70 percent in urban areas compared to 29 percent in rural areas. This is in line with other urban surveys currently being conducted in India, suggesting a widespread economic problem.
On March 26, the day after the lockdown was implemented, the Indian government announced a 1.7 lakh crore ($22.5 billion USD) relief package to provide cash and food to India’s most vulnerable residents. However, during the first week of the lockdown, only 48 percent of the households we surveyed were aware of government relief to support households during the pandemic. Only 4 percent of households reported receiving any government relief such as rations or cash transfers in the past week. The level of awareness and actual utilization of government relief were low in both urban and rural areas.
COVID-19 knowledge and messaging
Fighting COVID-19 necessitates massive changes to daily routines and knowledge is most often the first step to behavior change. We asked respondents to list the symptoms of COVID-19 and precautions to take against it. The majority of respondents listed cough or fever as symptoms of COVID-19. One in four respondents listed all three of the most common symptoms: cough, fever, and difficulty breathing. However, 17 percent of respondents (8 percent urban, 27 percent rural) reported that they did not know even one symptom of Coronavirus.
The most common precautions respondents reported taking were to wash hands, maintain physical distance, and wear a mask. About 74 percent of respondents mentioned some kind of physical distancing and 38 percent of respondents specifically reported knowing that they should maintain a distance of 1 metre from other individuals. About 17 percent of respondents (4 percent urban, 22 percent rural) could not list any precautions they should take.
Most respondents knew about Coronavirus and learned of it the week before the survey, likely thanks to the Prime Minister’s announcement of the day-long “Janata curfew” three days before the lockdown. An overwhelming majority of respondents recalled hearing about Coronavirus from television, followed by friends, family, or a community member. Only 29 percent of respondents recalled hearing about Coronavirus from text message or Whatsapp. Not surprisingly, in these early days of lockdown, recall levels were less than 2% from posters, health facilities, and community health workers (ASHAs), reflecting the importance of mass media or digital mediums during physical distancing.
We are using these preliminary findings to inform IDinsight’s future COVID-19 work that will draw from larger, more representative samples in India, to support evidence-based decisions by key officials across government. Although the results are limited to four locations and indicative of only the first week of the lockdown, they point us in important directions. There is an urgent need to ensure India’s most vulnerable populations are aware of and receiving relief measures. In rural areas, where there is less access to mass media platforms, innovative ways to increase the spread of messages are being experimented with and will become necessary. A possible avenue is to leverage community health workers (such as ASHAs) to communicate messages over the phone, while maintaining physical distance.
We are currently engaged in further large scale surveys and potential further work with partners. Please reach out to Dr. Divya Nair at divya.nair[@]idinsight.org if you are interested in learning more about or collaborating with our India COVID-19 response team.
 How did we select these respondents?
In November and December 2019, we conducted in-person surveys for an unrelated project in the 4 states (UP, Bihar, MP, Rajasthan). In each state, we sampled 2–3 Assembly Constituencies (AC), and from each AC, we sampled 2 polling stations. In each polling station, we censused all households that agreed to respond to the survey. During the survey, we requested respondents to share their phone numbers, and asked them whether it was ok to call about other research.
For the pilot phone survey, from the overall sample, we randomly sampled 200 respondents each in 4 locations. These locations were selected to maximize the diversity of our respondent pool: one location per state, and overall, two urban and two rural locations. From these randomly sampled households, we managed to complete surveys with 300 respondents (we didn’t manage to call all the sampled respondents, some didn’t answer the phone, and some refused to participate).
 One location each in Uttar Pradesh, Bihar, Madhya Pradesh, and Rajasthan