Cash transfers as COVID-19 relief: Evidence from Ghana
Well-targeted mobile money transfer programmes can be an effective policy tool to smooth consumption during pandemics, especially among the neediest populations.
The COVID-19 pandemic has affected economic activity and imperilled the livelihoods of people across the globe. Its effects on the poor in developing countries has been particularly severe, as these citizens are already vulnerable and largely without access to formal government social protection. Moreover, while many rich countries have expanded social protections, low- and middle-income country governments have responded with limited financial and institutional capacity to the pandemic shock.
Mobile money transfers offer a rapidly scalable and low-cost approach to providing resources and support during such crises and can be used to target individuals with minimal direct contact. But how effective is this method of providing support and encouraging virus mitigation measures (such as social distancing) in the context of a pandemic? To answer this question, we evaluated a mobile cash transfer programme in Ghana that was launched in mid-2020.
We found that the transfers led to an increase in weekly food expenditures of approximately one-third of the weekly value of the transfers and also increased recipients’ compliance with social distancing measures. Importantly, the transfers did not notably impact labour supply, contrary to what critics of cash transfer programmes frequently claim. That said, however, the transfers did not appear to relieve the psychological stress associated with the pandemic.
Importantly, many of these effects are strongest amongst the poorer households, and those belonging to the non-Christian minority population in Ghana.
Survey design
Our study considers a representative sample of 1,500 relatively low-income Ghanaians in the nine months following the spread of the pandemic to Ghana[1]. The sample of households participating in this experiment was drawn from the Ghana Socioeconomic Panel Survey (Ghana Panel) — a nationally representative survey that has been administered to a sample of around 5,000 participants every four years since 2009 by a team based at the University of Ghana, Northwestern University and Yale University.
Prior to the cash transfer distribution, we administered a short baseline phone survey to the sample in May-June 2020. Shortly afterwards, all households in the sample received an initial transfer of GHS 90 (approximately US$15). Half of the sample was randomly selected to receive seven additional cash transfers of the same amount (those selected were informed about this), while the other half served as a comparison group and did not receive additional transfers[2].
Implementing the study during a pandemic
It is important to note that the transfers represent a substantial income boost, and account for 65% of the median weekly food expenditure reported by participants in the baseline survey. This amount is also significantly larger than transfers from the Livelihood Empowerment Against Poverty (LEAP) programme, Ghana’s flagship cash transfer social protection scheme.
Delays in contracting forced a slowdown in the provision of transfers, so the eight disbursements were completed over a period of six months (between June and November 2020) instead of in two months as was previously planned. We note that this delay may have likely introduced uncertainty and loss of confidence among the recipients about the timing and provision of future transfers.
We conducted the first follow-up phone survey in July 2020. At this point, all households in the study had received a single transfer, and the treatment sub-group had been informed about the subsequent transfers.
Finally, we conducted three additional follow-up phone surveys in August, October, and November-December 2020.
Findings: cash transfers increased food expenditures and promoted social distancing
Because the pandemic and the resulting economic fallout is a dynamic and fast-moving situation, we were interested in assessing both the periodic short-term effects and the longer-term impacts of the transfers. By pooling (averaging) disbursements and surveys 2-4[3], we found that transfer recipients are more likely to comply with social distancing (an increase of .08 standard deviations on an index) relative to the comparison group. We also found that household food expenditures among households that received the cash grants increased by an estimated 10.45 GHS per week (about US$1.80). While the reported income of recipient households increased by approximately the value of the grant, we did not find any significant evidence that labour supply changed.
There was no significant evidence of any effect of the grant on an index of self-reported COVID-19-related symptoms, and there are few effects on perceptions of COVID-19, the government response to it, and the psychological impacts of the pandemic on respondents. The primary exception is that the receipt of cash grants significantly reduced respondents’ estimates of the impact of the pandemic on the economy of Ghana.
Critically, there are differential effects of the transfer among recipient households, and we found that poorer households are the primary drivers of many of these findings. Specifically, households with below-median food expenditures (the indicator we used to measure poverty) reported a .15 standard deviation increase in their index of social distancing, while households with above median expenditures did not increase social distancing in response to the transfers. In fact, transfer recipients with above-median expenditures reported an increase in work hours and total income, potentially in conflict with the policy goal of social distancing.
We also found that people belonging to the non-Christian minority population in Ghana (about 30% of the sample; primarily Muslim) responded more strongly to the transfers than members of the Christian population. Importantly, the entire effect of the grant on food expenditure is driven by the non-Christian minority who report a weekly increase of 32.32 GHS (approximately US$5.60) in response to the grant. This pattern suggests that the non-Christian population has limited access to financial markets or other consumption smoothing opportunities.
Policy implications
We believe these results provide strong support for the use of mobile money transfers as a policy response to global emergencies like the COVID-19 pandemic in the developing world. Across the study sample, one third of the transfer amount was spent on food, and it increased social distancing.
These effects were found to be concentrated amongst poor households and members of the non-Christian minority, which highlights the importance of targeting transfers to the needy, especially if the goal is to influence behaviours to reduce transmission. Our research team will continue to engage with key stakeholders in the Ghanaian Ministry of Finance and Ministry of Gender, Children and Social Protection, to discuss these findings and explore ways in which they can inform government programming. We will also conduct another follow-up survey later in 2021 to assess whether the impacts of these cash grants persist over time.
[1] Ghana saw its first confirmed cases of COVID-19 in March of 2020. Roughly a year later, the country had seen nearly 87,000 confirmed cases and 656 deaths in addition to widespread economic and social disruptions.
[2] Participants were informed that they would receive these amounts at roughly one-week intervals, which was the initial schedule designed for the programme.
[3] These are the disbursements for which the comparison households did not receive transfers, and the selected treatment households did.