Blog post

What’s so special about GiveDirectly’s basic income pilot?

5 min read
2 Jul 2016

Give Directly has decided to test a universal basic income intervention in Kenya: they’re planning to provide about 6,000 people with unconditional transfers for at least 10 years, transferring money to all residents of the villages which will be targeted by the intervention. This has received a lot of attention from the media, with the trial being mentioned, among others, on Slate, Vox, The Independent and Der Spiegel.


In this blog post I’ll try to give this news some context, by comparing this intervention with some alternatives. I will start by considering other types of transfers that can be made in the context of development interventions, looking at in-kind transfers and conditional cash transfer. Then I will focus on unconditional cash transfers, and specifically the ones that GiveDirectly has been delivering up until now, and explain what’s different and interesting about their new basic income pilot.

Other types of transfers

Part of international development aid is about re-distributing resources to people living in poverty. But while this seems pretty straight-forward, it is in fact not easy to work out exactly what type of resources we should move, and how they should be moved.

In-kind transfer

Traditionally, much development aid has focused on in-kind transfer, that is, on the provision of specific goods and services. This might include very disparate things: from cattle, books or food, to schools and hospitals. However, some argue that providing cash might be better that providing in-kind aid. There are several arguments for this thesis, but here I’ll just mention two points. First, providing cash can be cheaper than distributing in-kind transfers.1 This is becoming increasingly true thanks to new electronic ways of moving cash: for instance, GiveDirectly makes transfers through mobile phones, which can be cheaper than in person distributions.2 Secondly, providing cash generally leaves recipients more flexibility than in-kind transfers.3 This is good because recipients might have better information about what would be most useful for them.

However, some do not think that in-kind aid should be substituted by cash transfers. Part of the disagreement is rooted in worries about recipients spending money on alcohol or tobacco (what are sometimes referred as ‘temptation goods’). In fact, however, evidence suggests that this worry is misplaced: the evaluation of GiveDirectly’s cash transfer program shows no increase in spending on temptation goods.4

Other worries seem to be more pressing: supporters of in-kind transfers argue that some essential services (such as hospital, roads and schools) bring benefits to people other than the recipients (they lead to what economists call “positive externalities”): for example, hospitals and school are likely to increase levels of health and education, and healthy and educated workers are likely to be more productive, which would in turn benefit not only them, but also the wider population. Goods which have positive externalities are likely to be underprovided by markets, so might not be adequately provided through cash transfers.5

Conditional cash transfers

Interestingly, however, cash transfers are not always offered as an alternative to public services. Sometimes, transfers are used with the specific intent of increasing the uptake of these services. This is the case of conditional cash transfers: transfers which are conditional on the satisfaction of certain behavioural requirements, generally concerning the health and education of children resident in the recipient household. For instance, Mexico’s “Oportunidades” conditional cash transfers programme, provides cash transfers to households conditional on their children regularly attending school and making health clinic visits.6 Alternatively, cash transfers can be unconditional, which means that transfers are made without any condition being set.

We have written extensively about the advantages and disadvantages of conditional and unconditional cash transfers in our new report on economic empowerment. The (really short) summary is that conditional cash transfers lead to better investments in human capital than unconditional cash transfer: more people employ education and health services, which increases workers’ skills, knowledge, and experience. However, unconditional cash transfers have the advantage of not denying a safety net to poor households simply because they may not meet certain behavioural requirements and allow vulnerable households to identify their own needs7. They also have lower administrative costs than conditional cash transfers, since there is no need to verify that the conditions have been met.Because of this, unconditional cash transfer have received increasing attention as a tool for international development.

Other types of unconditional cash transfers

GiveDirectly provides beneficiaries with unconditional cash transfers - they transfer low-income households a certain amount of money, without any condition attached. They have been doing this for a while, and have mainly worked in Kenya and Uganda.

Yet, many people seem to be pretty excited about GiveDirectly’s news that they will be testing a basic universal income and their announcement of the pilot has received a large amount of media attention. So what is so different about their new basic income program?

First of all, the basic income pilot will provide transfers to all full-time residents of treatment villages.8 Up until now, GiveDirectly has been only providing transfers to some people in target villages. However, some evidence suggests that an increase in the mean wealth of the village is correlated with a significant decrease in life satisfaction among individuals in households that did not receive transfers.9 By targeting all residents, universal basic income might avoid this problem.

Secondly, the basic income pilot will provide smaller and regular transfers for 10 years. While the details of the pilot are yet to be defined, the payments are expected to amount to somewhere between $250 and $400 a year per person10. In contrast, GiveDirectly’s has so far made one-off transfers of about $1000. This change is interesting because having a transfers for longer time might give people the security needed to take more risks and invest a larger portion of their resources, which might enable them to increase their income in the long run.11

GiveDirectly universal basic is unlikely to be able to substitute the provision of public goods, or lead to as much investment in human capital as conditional cash transfers would. However, it is likely to be a cheap and flexible way of transferring resources, that can provide an inclusive safety net for people living in poverty. It will also avoid decreases in life-satisfaction for households not receiving transfers (since all households in target villages will receive transfers) and might give recipients more financial security to make investments, and increase their longer term income.