Disbursing emergency relief through utilities
Background, challenges, and context
Emergencies like pandemics, natural disasters, and economic crises test a government’s resilience and preparedness in supporting its population. How can governments best distribute aid to citizens? In countries with established financial infrastructure, such as the US, governments can send money directly to citizens. But in many countries, strong financial connections between the government and individuals may not exist. In these contexts, what can governments do? In response to the COVID-19 crisis, 186 countries implemented direct cash transfers to households, and 181 introduced in-kind programmes that lowered the cost of utilities such as electricity, water, transport, and mobile money.
Evidence shows cash transfers dampened the economic impacts of the pandemic and subsequent lockdowns. In many African countries, mobile money is an important tool for providing cash transfers, particularly for populations excluded from formal financial institutions.
But in-kind transfers may be preferable if they circumvent savings constraints or lower transaction costs. In-kind transfer programmes can be attractive to governments lacking infrastructure to distribute cash broadly, beyond vulnerable populations already receiving government aid.
In March 2020, the Government of Kenya expanded its existing social safety net, Inua Jamii, which provides cash transfers to its most vulnerable populations, and launched urban public works employment schemes.
In May 2020, Ghana provided an additional round of cash transfers to existing recipients of the Livelihood Empowerment Against Poverty (LEAP) Program, which targets around 5% of Ghana’s poorest households.
The government of Ghana also launched an electricity relief programme, citing the importance of reliable electricity during economic downturns. All households with an electricity connection were eligible to receive a monthly transfer, with transfer amounts pinned to consumption in March 2020 before the programme was announced. In practice, not all households received transfers and there was significant variation in the timing and consistency of monthly transfer receipt.
The design and implementation of social protection programmes can meaningfully affect their overall impact. During an economic crisis, should policy makers distribute cash transfers or in-kind transfers such as electricity subsidies? And how does a country’s financial infrastructure affect the relative efficiency of these types of government aid?
Research overview and objectives
This research, part-funded by EEG as an extension to the GridWatch project, asked:
What programmes most effectively address the economic consequences of a nationwide emergency such as a public health crisis?
How does a country’s financial infrastructure affect the distribution of government aid?
The project also assessed the efficiency, distributional, and political implications of Ghana’s electricity relief programme, asking:
How does programme design affect the efficiency and distributional implications of such policies? And what design features determine their political popularity?
The project combined household surveys in urban Ghana and urban and rural Kenya, and a randomised experiment in urban Kenya.
Over a study period of seven months, between May and November 2020, 3,211 respondents in Kenya (983 urban and 2,228 rural) and 1,245 urban respondents in Accra in Ghana were surveyed by phone between one and three times each. The Kenyan urban sample was broadly similar to electricity customers in Accra, based on observable characteristics like housing quality and appliance ownership, and comprised households located in urban areas connected to the grid and using pre-paid meters to pay for electricity in advance by purchasing electricity credit, or tokens. The team elicited respondents’ preferences to compare demand for and impacts of transfers in the form of mobile money and pre-paid electricity credit.
All the respondents in Ghana were eligible for electricity transfers through the government’s COVID-19 relief programme for at least three months in 2020. More than two-thirds had received at least one electricity transfer in the first three months of the programme, and almost half had received a COVID-19 water subsidy transfer (the other main government pandemic relief programme). As Kenya did not have a similar government electricity transfer programme, the team implemented a randomised experiment with a subset of 2,053 respondents. Each respondent was randomly assigned to one of four groups: a control group, a group that received pre-paid electricity tokens, and two groups that were given a choice between cash or electricity transfers at different rates.
Both surveys collected data on household composition, electricity connections and expenditures, electric appliances, expenditures in the previous seven days, food security, credit, government relief, and perceptions of the government.
Research results, key messages, and recommendations
The main findings were:
Cash and in-kind transfers have different advantages that make each suitable for specific contexts. Distributing relief through electricity transfers enables an immediate government response to the crisis because it leverages the existing financial infrastructure between the government utility and households. Moreover, theoretical efficiency concerns about in-kind transfers are mitigated because the transfers are inframarginal for most households and electricity credit can be stored, with many even preferring electricity transfers over cash.
In Ghana, these advantages did not preclude delays in transfer receipt and the exclusion of some eligible households, and the programme was regressive in both design and implementation. The households least likely to receive relief were those who use less electricity, pay a landlord or other intermediary for electricity, or share an electricity meter with other users – all common among low-income electricity consumers in urban settings. Concerns around disbursing relief through utility transfers in this context thus arise not from efficiency loss, but from regressivity, distributional challenges, and politicisation. A uniform rather than a proportional electricity transfer would be more progressive, and possibly easier and less costly to implement. Another advantage is that all households would know the amount they were entitled to receive.
Although urban Ghana and Kenya have similar levels of education, mobile phone ownership, and electricity connectivity, responses vary significantly. The effectiveness of cash transfer programmes hinges on the financial infrastructure used to disburse government aid, and its integration with payment platforms. Where mobile money infrastructure is less advanced, recipients are limited in what they can do with these transfers, reducing their value. This may also be the case when the use of mobile money incurs high transaction fees.
In urban Kenya, in line with economic theory favouring direct cash transfers, 95% of urban recipients prefer mobile money over electricity transfers of a similar monetary value. Many are willing to forego significant value to receive mobile money rather than an electricity transfer. But Kenya is an outlier with high mobile money adoption (97% of Kenyan households have at least one M-PESA mobile money account, and 68% live within 1km of an M-PESA agent): this increases the value of mobile money and reduces the transaction costs of buying electricity credit.
In Ghana – where mobile money is less widespread than in Kenya (only 39% of adults use it) and the transaction costs for buying electricity are higher – 49% prefer pre-paid electricity over a mobile money transfer of the same value, and many are willing to forego significant value to receive electricity rather than mobile money. 10% of respondents in Ghana state that they “worry about mobile money charges/costs.”
Customers of Kenya Power can buy electricity credit easily using mobile money. But 93% of customers in the Ghana sample must travel to an ECG office or official electricity vendor to purchase credit – only 7% have a ‘smart’ pre-paid meter that allows mobile top-up. Respondents who choose electricity credit avoid the time and cost of travelling to an electricity vendor to purchase credit and any fees or costs associated with using mobile money.
If Ghana reaches the high levels of mobile money adoption that currently characterise Kenya’s financial landscape, cash transfers will likely become a cheaper and more effective channel through which to disburse government aid. At the same time, mobile money penetration may remain persistently low in some communities – including those with low financial literacy, and the ultra-poor, for whom mobile banking fees can be prohibitive. There, in-kind transfers may continue to be preferred.
The team found no impacts of electricity transfers on a wide range of socioeconomic outcomes in both Ghana and Kenya.