Launch a basic income

Treatment Groups: As part of a randomized-controlled trial, each of 295 rural villages in Western and Rift Valley region of Kenya were assigned to 4 groups:

1. Long-term basic income: 44 villages (4,966 people) with recipients receiving roughly $0.75 per adult per day, delivered monthly for 12 years. All payments for all groups are made in Kenyan Shillings.
2. Short-term basic income: 80 villages (7,333 people) with recipients receiving the same monthly amount, but only for 2 years.
3. Lump Sum: 71 villages (8,548 people) with recipients receiving the same amount as the short-term basic income group (in net present value), but all up front as a “lump sum.”
4. Control group: 100 villages not receiving cash transfers.

Pilot villages: GiveDirectly selected an additional 2 villages to receive monthly payments for 12 years. People living in these villages are not part of the formal research study, enabling GiveDirectly and others to have more in depth, qualitative conversations with them about what it’s like to receive a basic income.

Survey Data: Baseline responses have already been gathered in each village, collecting answers from four different groups of people: household heads and spouses to understand household level-conditions; village elders to measure village-level changes; and owners of local enterprises to understand the effects on the local economy.

Mitigating bias: IPA staff conduct all research surveys independently of GiveDirectly. The study’s authors publicly registered the study to ensure they can be held to account for publishing its results, whatever they are. They will also publish a pre-analysis plan that will guide how analysis is conducted before any outcome data are analyzed to prevent cherry-picking.

Program Efficiency: We expect 85% to 90% of funds raised for the program to be delivered as cash to participants in the program. Research costs borne by IPA are funded separately by private funders.

Effects studied: UBI could plausibly affect a wide range of outcomes and behaviors, from what families eat and how they interact with each other, to broader macroeconomic or cultural dynamics. With this in mind, the research team designed the surveys to cover a wide range of outcomes:

1. Economic well-being: consumption, assets, remittances, debt and credit, food security, employment
2. Health: psychological well-being, nutrition, cognition
3. Social well-being: mental health, intra-household gender relations, aspirations, social integration
4. Macroeconomic well-being: prices changes, availability of goods & services, community demographics, village infrastructure, political participation, and crime rates
5. Financial preferences: willingness to take financial risk, invest, start businesses, and migrate

Comparing treatment groups:

1. Comparing people receiving 12 years of basic income with those only receiving 2 years will illustrate how the promise of future basic income payments affects recipients’ propensity to invest, migrate, start businesses, and take financial risks. This may also help researchers interpret results from other, shorter-term studies.
2. Comparing people receiving basic income payments for 2 years with those receiving the same amount but all upfront will help will test how breaking cash transfers into smaller, recurring pieces changes their effects.

What we’ve learned so far: First results on the impacts of a basic income on recipients during COVID-19 showed transfers improved well-being on measures such as hunger, sickness and depression in spite of the pandemic, with modest effect sizes. Researchers also noted that recipients reduced hospital visits and decreased social (but not commercial) interactions, implying potential public health benefits. During the pandemic, recipients lost the income gains they previously accrued from starting new non-agricultural enterprises, but also suffered smaller increases in hunger. Results from the full first endline are expected later this year.

What won’t we learn? No single study can confidently predict how a UBI would work all over the world. Effects may differ in countries that are wealthier than Kenya, or in more urban or more remote areas. The effects of UBI are likely to differ in the developing world and the developed world.

Additionally, the basic income in this study is universal within villages, but not country-wide. The study won’t fully capture the effects of implementing a UBI at a national level.

How is this different from prior studies of UBI?
This study is the largest, long-term, randomized evaluation of a universal basic income:

How are villages selected for the trial?
GiveDirectly worked with the research team to select rural sub-counties in Western and Central Kenya where we had government authorization to operate. Researchers selected a pool of 295 comparably sized, rural villages, randomly assigning each to one of three treatment groups or the control group. Each of those groups was sized to maximize expected statistical power within the constraints of the program budget.

When will we see results?
While payments for the long-term group will continue for 12 years, initial results from the first endline survey will be available later this year. A working paper on the effects of a universal basic income during COVID-19 was released in September 2020. Follow-up research will take place every 2 to 5 years following.