The evidence behind putting money directly in the pockets of the poor
By Dr Kate Orkin
COVID-19 and the related government-led lockdowns have resulted in widespread economic shock and job loss, in the UK and around the world. Governments, including in low and middle-income countries, have responded with economic interventions to cushion the shock. The most widely-used government tool has been cash transfer programmes: the World Bank finds 234 measures involving expanding cash transfer programmes worldwide, as well as 100 food or voucher schemes.
In the past, some governments have been concerned that transfers would increase dependency on the state and, in particular, that cash transfers would not be used well because it is hard to monitor how people spend them. Many governments used instead to provide food aid or subsidise basic food items.
So why have governments shifted to using cash in low and middle-income countries?
Why have governments shifted to using cash in low and middle-income countries?
In most situations, there is strong evidence that money, not food, is the most efficient and effective way to distribute emergency aid and social programmes. This particular question, on how poor people use cash transfers, is one of the most studied in development economics.
The finance ministry in Mexico did one of the first conditional cash schemes in 1997. To evaluate the programme, they ran a randomised trial. They could not roll out the programme all at once, so they randomly selected some communities to receive it first and compared them to those who did not. This is a similar process to how drugs are tested, but with a social programme. Since then, many countries' governments in low and middle income countries have implemented similar studies.
In most situations, there is strong evidence that money, not food, is the most efficient and effective way to distribute emergency aid and social programmes.
Poor people spend cash grants well. The bulk of transfers are spent on food anyway. For example, a review of 165 studies by the Overseas Development Institute found that recipients of cash grants have better dietary diversity and are less likely to face food insecurity. A World Bank review found grants improve growth and cognitive development in small children. Cash also has the added benefit of giving people autonomy to spend on what they need most. It also stops distortions arising in local markets, where bringing in free food can lead to price decreases that hurt local producers.
Poor people spend cash grants well
Emergency, fast cash is a smart investment in long-term poverty alleviation
Numerous studies, from China and India to Ethiopia and Malawi show that economic shocks have severe long-term consequences. Poor households often take short-term decisions that leave them in deeper long-term poverty. The decision most feared is that households reduce children’s nutritional intake.
Setting aside moral arguments, malnourished children have lower schooling attainment and lower earnings throughout their lives. A World Bank review found grants improve growth and cognitive development and later outcomes in small children. It is harder to find a smarter investment.
When households face economic shocks, women may enter into transactional sexual relationships: during the outbreak in West Africa, a study by anti-poverty organisation BRAC found young women had older partners, higher rates of pregnancy and did not return to school. Both responses to poverty could be mitigated by cash transfers.
But studies also show that, when facing a short deep shock, desperate households often sell productive assets such as cows, vehicles or phones or dip into meagre savings which they usually use to search for work. Losing the means of earning can lead to many additional years of poverty. Temporary cash grants can help. Studies in Bangladesh and Malawi found recipients of grants are less likely to sell assets when they face shocks.
There are anecdotes of welfare queens: people spending their welfare money poorly. But the anecdotes just do not bear out the reality
In low and middle income country settings, cash transfers also mostly do not affect whether, or how much, people work. In some studies they increase job search because they give people money for transport costs and airtime. For example, one South African study found youngsters in households with a pension recipient are more likely to find jobs.
There are anecdotes of welfare queens: people spending their welfare money poorly. But the anecdotes just do not bear out the reality in large samples of people. There is really no good evidence of waste. A review of 19 studies by the World Bank found cash grant recipients did not increase spending on alcohol or cigarettes. In some countries, which only give the grants to parents, there have been arguments that the grants are incentives for women to have children. But there is little rigorous evidence of this. For example, trials in Nicaragua and Malawi found that women in households are less likely to fall pregnant; a trial in Mexico found no effects.
An increase in cash grants may also help stimulate the economy. In a trial in Kenya, the NGO GiveDirectly evaluated an intervention which gave the poorest people, in some randomly selected villages, large lump-sum cash transfers. They found that people living near to villages, where the poorest received cash transfers, also had higher food consumption, partly because recipients spent money in their businesses. There were no meaningful inflationary effects.
Most striking of all, the study estimates a ‘fiscal multiplier’ of 2.6 for this area of Kenya, implying that every $1 invested in fiscal stimulus will grow the local economy by $2.60. That is even larger than in such multipliers the US during a recession. These were much, much bigger transfers than most governments in poor countries give out, equivalent to about $1,800, if spent in the US. So we do not know if the same effects occur from smaller transfers, but in principle the mechanism might work in the same way.
Cash is also usually cheaper to distribute than food. It can often be transferred to bank accounts or mobile money accounts. It does not go off
Cash is also usually cheaper to distribute than food. It can often be transferred to bank accounts or mobile money accounts. It does not go off, and governments don’t have to worry about having the wrong type of cash in the wrong place. There are worries that cash might be more fungible and possible to divert, but there is also evidence on how to prevent ‘leakage’, in particular by paying directly to beneficiaries, requiring biometric identity verification and being very clear about who is getting what benefits on what schedule. It is obviously difficult to set these systems up from scratch, but many countries have them in place already. And cash is much better for social distancing than food parcel queues, if it can be sent to bank accounts.
There can be cases where food aid makes more sense, for example if there is disruption to the food supply chain or if prices go up rapidly. Sometimes beneficiaries say they prefer food. But in general, it is fair to say that a cash system can work better than a food system.
South Africa is an interesting recent example. They had an emergency food parcel system before the crisis and it got completely overwhelmed. They just could not scale up fast enough to reach everyone who needed it. The authorities eventually admitted that the food parcel system was not working, so instead they increased the amount of their pension and child grant, where people were signed up already. They have also added a new cash grant for the unemployed.
Dr Orkin is a labour economist at the Blavatnik School of Government, University of Oxford