This post is written by Elizabeth Bryan, Muzna Alvi, Claudia Ringler and Ruth Meinzen-Dick, IFPRI.

The impacts of COVID-19 are being felt widely across the globe as most countries and localities urge residents to remain home to slow transmission of the disease. This global health crisis is particularly threatening to the global poor, who may be more susceptible to contracting the virus, have limited access to healthcare, and are more vulnerable to economic impacts. Yet, poor male and female farmers in developing countries may not experience this crisis in the same way. In order to understand how the pandemic is differently affecting men and women in developing countries, IFPRI is implementing a series of phone surveys with poor rural men and women in selected Feed-the-Future countries as part of the Gender, Climate Change, and Nutrition Integration Initiative (GCAN). The same survey is being carried out in India with the grassroots women’s organization, SEWA, as part of the BMZ-supported project on Reaching Smallholder Women with Information Services and Resilience Strategies to Respond to Climate Change. 

Research shows that men and women are differentially affected by shocks and stressors and have different capacities to respond. The current COVID-19 crisis is likely to have similar gender-differentiated effects. Policymakers, people implementing programs on the ground, and other decision-makers need to understand differential risks and impacts so that they can target programs to effectively support the specific needs of men and women. 

The GCAN Framework illustrates how the ultimate impacts of climate shocks and stressors on men and women depend on their respective exposure, resilience capacities, decision-making authority within the household, community or other spaces, and the responses that are taken to cope with, reduce risks from, or adapt to shocks and stressors. This same framework can be applied to any type of shock that rural farm households may face, including the health and economic shocks related to the COVID-19 pandemic.

Following this framework, possible gendered outcomes of the COVID-19 pandemic include:

• A loss of control over income by women — either due to a loss of women’s sources of income or because men take over greater control over household spending decisions as incomes decline.
• A widening of the gender asset gap if women’s assets are sold as a coping response. It is, therefore, important to track whose assets are being sold or whose savings are depleted as a result of the crisis.
• Shifts in labor allocation — for example, women may go in search of work if their husbands lose their jobs, or they might take on an increased burden of care (either due to having children home from school or caring for sick household members). If schools do not re-open in the near term and if income declines result in girls being taken out of school, past gains in women’s employment might be lost for years to come. 
• Changes in the mobility of men and women under lockdown conditions depend on who leaves the home for essential activities, such as fetching water or fuelwood, or obtaining food for the household. This has important implications for different family members’ risk of exposure to the disease.
• Changes in food insecurity or dietary diversity. Each household member may experience these changes differently.
• Increased stress brought on by the crisis and having all family members confined to the home may increase the incidence of conflict and intimate partner violence.

As we explore these issues through phone surveys with rural men and women farmers in Ghana, Nepal, Nigeria, Senegal, and India, we have gained several insights on the logistical challenges of successfully reaching women with phone surveys. 

One key challenge is that some households do not own phones, in which case we may miss the most vulnerable households. When phone numbers are available, most households have only one phone and that is more often controlled by the male head of household, making it difficult to reach women with phones. 

This image shows gender gaps in mobile phone ownership by region. There are clear regional differences with the largest gap found in South Asia followed by sub-Saharan Africa. Even though women may have access to a phone through their husbands or other family members, getting access to the phone to participate in a survey can be a challenge. The family members who own the phones are likely to take it with them to the field, for example, so finding the right time to call the mobile number is important and may take several attempts.

A second key challenge is that some of the questions — namely those related to household conflict — are sensitive and some women respondents may have difficultly taking the call in private. For example, 65 percent of women in the India survey reported that they were taking calls on speakerphone, and thus, only 220 out of the 627 participating women were asked sensitive questions about intra-household conflict because most were unable to take the call in private. Results from more than 500 respondents in the GCAN Nepal phone survey suggests that close to half of all phones were also on speaker. 

Other challenges we encountered include: women’s phones were switched off because of COVID-19 income losses and the lockdown reduced the need for more than one phone in the household; language challenges with respondents in remote villages where only local dialects are used; cellphone network interruptions or low bandwidth in some rural areas; and challenges engaging elder women in some of the questions. 

Addressing these challenges requires careful planning and thought. First, it may be helpful to build on existing surveys or interventions where contact and rapport are already established. For example, reaching women through trusted women’s groups and using female enumerators who speak the local language and dialects may make women more comfortable about participating in the survey and can increase the likelihood that family members who might take the call first agree to women’s participation.

Second, it is important that researchers omit sensitive questions from phone surveys, such as direct questions about domestic violence. Rather, they should opt for less sensitive options, such as questions that focus on intra-household conflict. These questions should be carefully framed so that answers would not reveal much to those who overhear the conversation. Moreover, any sensitive questions should be dropped if the call cannot be taken in private (e.g. when speakerphone is on).

Lastly, it is important to acknowledge that these challenges introduce bias in the results, given that fewer poor and vulnerable households can be reached by phone compared to face-to-face interviews. The response rate is also lower for phone surveys than in-person interviews and phone surveys miss visual cues that enumerators would act upon in face-to-face surveys to keep the conversation going. 

Ultimately, women are more difficult to reach over the phone and many may have difficulties taking the call in private. This may bias the responses to sensitive questions that can only be asked of women who have more privacy and are presumably more empowered. Despite these biases, important lessons can be learned on the gendered impacts of COVID-19 to strengthen policy responses that support both men and women in rural settings.