GiveDirectly leverages the low costs of mobile money to deliver cash transfers to poor households in select African countries. Initial results are encouraging. The money is not being spent on “sin goods”. On the contrary, it is being – for the most part – directed into productive investment that helps these poor families get ahead.
It is worth noting differences between UCTs and CCTs (conditional cash transfers). CCT programs provide cash payments to poor households, but they impose conditions on recipients before they get the money, mostly related to children’s health care and education (e.g. enroll the kids in school). UCTs put no such conditions. This is why there is such enthusiasm about UCTs. “No conditions” means such programs tend to be cheaper to administer. At least that is the theory. Note that UCTs and CCTs are similar in that neither has any conditions on how the money (once obtained) is spent.
This posting is focused on UCTs because of the current buzz around them. Although they are showing impressive results, let’s be realistic about the potential and limitations of UCTs. There is a lot that we do not know about the conditions under which UCT schemes lead to sustainable poverty reduction. Nor are we clear about how such programs can be scaled effectively. To the credit of organizations like GiveDirectly, they have partnered with Innovations for Poverty Action to carefully evaluate the results of their actions through rigorous randomized control trials.
It is worth noting that GiveDirectly is doing more than just sending cash to the poor; they are also spending resources carefully identifying, evaluating and selecting beneficiaries, and on monitoring and evaluation. This leads me to one of three points I think are worth making about UCTs.
First, the idea behind UCTs may be simple, but the more successful UCT schemes are complex. The “U” in UCTs does not mean that all you are doing is giving poor people money and stepping back. Research done by ODI and funded by the UK’s Department for International Development (DfID) suggests that UCTs work best when accompanied by information, education and communication efforts, careful targeting and selection of participants, and constant feedback and interaction. In other words, you need to consider who will be selected, what complementary efforts/services will enable and facilitate a good response, and you need to constantly invest in citizen feedback channels that allow you to learn and adapt as better information about program impact comes in. This is not much different than what a good INGO needs to do in order to deliver effective programming (UCT or not).
Second, the media coverage ignores how much variation exists among UCTs schemes. As the World Bank’s Berk Ozler has highlighted, there is a world of difference between “waking up one morning and finding $500 in your M-PESA account” (GiveDirectly) and the interventions being carried out in Liberia for unemployed youth, or what the DfID-funded ODI studies describe. Again, it is too early to tell what kinds of effects on poverty reduction we can expect from such schemes and we are miles away from understanding how scheme design details are related to sustainable paths out of poverty.
This leads me to a third set of questions: for whom are UCTs working? How do program results compare in urban vs. rural areas, for different income levels? We have years of data on CCTs, particularly a lot of data from Mexico, Brazil and other middle income countries where these programs have been scaled up nationally. Yes CCTs have problems (what development and social safety net programs do not?). But there is plenty of research demonstrating the conditions under which CCTs work. UCTs are much less well studied.
But the importance of these innovations, as Chris Blattman has already said, is that it forces (or should force) development organizations and donors to think about “top and bottom lines.” In other words, is what we are doing working? And even if it is working, at what cost? More importantly, we should always ask: are there other options for delivering the same (or similar) results more cost effectively?
As the CEO of a child sponsorship organization, I am drawn to the idea of UCTs. In fact, our initial child sponsorship efforts decades ago bear important similarities to today’s UCT programs. But Plan (like most other child sponsorship organizations) stepped away from such direct transfers, as concerns with sustainability and dependency grew. It is perhaps time to take a new look at the evidence around cash transfers, invest in reviewing results of past sponsorship programs and the lessons learned from that experience that may be applicable to a new generation of UCTs.
In the private sector, publicly quoted companies live and die by the share price, and the pressure to innovate and stay ahead is always present. For public charities like Plan, the rewards – and risks – of innovation are much less clear. But ignoring disruptive technologies and innovations, and failing to continuously push to experiment and learn will lead to irrelevancy. The jury may be out on UCTs, but they need to be taken seriously. GiveDirectly and others like it are pushing us all to do better.