We've planned some pilot details.

• Location: East Africa, where we have existing operations, there are robust mobile-payment systems and high concentrations of poverty where we can both generate sufficient sample size for a feasible budget while helping vulnerable communities in the process.
• Targeting: All full-time residents of treatment villages. Payments continue if subjects migrate out and are not delivered to new subjects who migrate in.
• Amount: At least a portion of the group receiving payments sufficient to cover basic needs. We may want to test a smaller variant to inform questions about the proper sizing of a safety net.
• Frequency: We'll want to structure at least some of the payments as monthly payments, so it's hard for people to fall through the cracks between disbursements. We may also want to experiment with different payment structures or with allowing recipients to choose the structure that works best for them.

We've also come up with a tentative budget.

To provide a true test of basic income, we want to (1) provide payments for a long enough time to allow people to make long-term investments or take risks, (2) provide payments to whole communities universally, and (3) provide payments that would be big enough to cover basic needs. We combined those pieces and then determined what sample size would best allow us to detect the impact of basic income, as well as potentially understand the differences between possible variations. None of these estimates are precise, and we'll refine them with our research team over the next few months, but they come out to needing about $30M for our desired study.

Our preliminary estimate of the percentage of this that will be cash transfers works out to a similar fraction to current transfers: approximately 90% of budget delivered to the poor. Overall management costs per recipient will be higher, but the total amount transferred will also be higher given the long duration.

And we're planning to do rigorous research with a top-notch team.

As we have in the past, we'll collaborate with top scholars, including Abhijit Banerjee of MIT.

We're considering several different types of outcomes to gauge the impact of basic income:

• economic status (income, assets, standard of living)
• time use (work, education, leisure, community involvement)
• risk-taking (migrating, starting businesses)
• gender relations (especially female empowerment)
• aspirations and outlook on life

What is a basic income?

A basic income guarantee is a public policy that would provide all people a basic floor—an income that is enough to live on and that is provided irrespective of work simply because the recipient is a member of that community. It is provided to everyone, regardless of need, forever.

Since the time of Thomas More, people from across the political spectrum have expressed interest in the idea, from Dr. Martin Luther King Jr., to the conservative economists Milton Friedman and F. A. Hayek. Some argue this is the moral thing to do; others argue replacing a patchwork of existing government programs with a basic income is more efficient; technologists argue the coming robotization of the workforce makes it necessary. And it's not just words; in countries where basic income is up for debate, trillions of dollars of social services are at stake.

What is the evidence?

The deep evidence base for unconditional cash transfers provides plenty of reasons to be intrigued by basic income. We know people who receive cash transfers don't blow it on drinks or stop working but rather increase their earnings, their assets, and their psychological well-being.

But, we know less about basic income. We haven't yet seen a rigorous evaluation of a program that is:

• universal within well-defined communities (not means tested or targeted towards a specific group),
• long-term, and
• enough to cover a minimum standard of living.

A number of studies provide helpful information, and there are some pilots coming online, but none so far has met all of these criteria.

Additionally, a basic income trial should use rigorous experimental research—randomized controlled trials (RCTs)—to measure impact.

New and upcoming basic income pilots have been mentioned in the news.

• Planned pilots include: Finland, Netherlands, and Canada, the Y Combinator project.
• There are active debates on basic income in: Switzerland, France, New Zealand, Scotland, Namibia, Germany, India, and Vermont.
• Recently we learned more about the Finnish pilot planned for 2017. The next report on their plans is due out November 2016.
• Ontario included provisions for investigating a provincial pilot of basic income sometime this year, and it appears the idea is gaining some traction there.

Basic income claims supporters from very different schools of thought:

• Basic income as a response to technological unemployment: Albert Wenger has written in this vein. This New York Times article explores this perspective as well.
• Feminist perspectives on basic income: Judith Shulevitz fleshed this out well in an op-ed. There's separately good developing world evidence on the impact of (non-basic income) cash transfers on young women.
• Some libertarians support basic income for a mix of reasons including anti-paternalism and its non-bureaucratic and non-distortive nature. To some, it's just the "least bad" version of redistribution. The Cato Institute held an internal debate series on the policy in 2014.
• Many supporters just generally want more redistribution or view freedom from basic deprivation as a right society should provide.

At GiveDirectly, we're convinced enough by the evidence for cash transfers to be intrigued but think the debate is important enough to merit a rigorous test of the program.

There are other variants of basic income, but unconditional cash transfers are the uniting factor.

Not everyone who advocates for a “basic income” prioritizes or even wants all of the components (unconditional, universal, lifetime, sufficient for basic needs) equally. The most common variants either restrict the receiving group (via a negative income tax or otherwise targeted approach) or lower the guaranteed income “floor” below basic needs. Someone's preferred variant depends on the arguments and perspectives that are leading them to basic income in the first place or considerations about political feasibility. Similarly, there are a wide variety of proposals for funding basic income programs in different countries. You can get more of a sense of the types of variants people are thinking about at "What is the Basic Income Guarantee?"