GiveDirectly Continues to Expand the Cash Transfer Map
The non-profit GiveDirectly has arguably done more to advance the cause of basic income and cash transfers than any other in the modern world. Initially considered highly controversial (“you can’t just give people money!”) GiveDirectly has, with its ever growing list of evidence, shifted the conversation to one in which cash transfers are acknowledged as one of the most effective forms of aid.
As they have expanded the charitable aid conversation, GiveDirectly has also expanded their own geographical scope as well. Starting in Kenya, they have started or will soon begin cash transfer programs in Uganda, Malawi, Liberia, Rwanda, and the DRC. They even worked in Texas and Puerto Rico after Hurricane Maria — a rare entry into more developed countries, where dollars don’t go as far, but attention to the idea of cash transfers is amplified with this work.
We invited GiveDirectly CFO Joe Huston back onto the podcast (a rare second time guest for us) to discuss the latest results from GiveDirectly’s expansive work in Kenyan villages, and the organization’s new frontiers.
Listen to the episode or read the transcript below:
Owen: Hello, and welcome to the Basic Income Podcast. I’m Owen Poindexter.
Jim: And I’m Jim Pugh. Almost two years ago, in one of our early episodes, we talked to Joe Huston. He’s the CFO over at GiveDirectly. He was telling us about GiveDirectly’s basic income experiment that they were getting ready to launch in Kenya at that point. Given it’s been so long, we thought it would be good to catch up with Joe and see how things were progressing so far with the pilot.
Owen: I spoke to Joe Huston about how things are going and some of the new work that GiveDirectly is doing. Here’s my conversation with Joe Huston at GiveDirectly.
Alright, Joe Huston, thank you for joining us on the podcast.
Joe: Thank you, guys, for having me back.
Owen: This will be a review for some of our listeners, but could you just start by giving a quick overview of GiveDirectly’s Village Pilot Program and what stage it’s in right now?
Joe: Yes. GiveDirectly in general only ever does one thing. It delivers unconditional cash transfers. Then along the way, as we’re delivering those cash transfers, something we’re often doing is testing how they work. Do they work at all? Do particular structures for particular types of people work better or worse than other structures? We’re constantly experimenting with different types of cash transfers.
As we saw the debate and conversation about basic income grow a lot over the last two years, we saw it as an opportunity for us to test a universal basic income. When we talked two years ago, we were at the very beginning stages of fundraising for that project and sketching out how it could work from a study perspective and what sizes the cash transfers would be and things like that. Since then, about two years ago, so a few months after we talked, we kicked off the first one-village pilot. This was one village with about 100 adults, each receiving monthly cash payments of about US$ 20, with the promise that those will continue going out for 12 years.
We’ve now been able to follow these people over the course of about two years of those monthly payments. In tandem, we continue to fundraise and work on the research designed for the full study. Enrollment for that full study, which will be over 20,000 adult recipients in over 200 villages, kicked off at the end of last year and wrapped up at the beginning of this year. We should have the first round of follow-up results something like early to mid next year, following up with people after a year of cash payments.
Owen: Great, generally, are things going as planned? Have you had to modify things as they go along or you’re pretty much sticking to the original design?
Joe: From a design perspective, things have largely gone as planned. One design question we got to test with the pilot was whether or not individually targeted payments would be seen as threatening to households. That was probably my biggest, almost operational implementation worry, that it would be seen as trying to pay each individual adult versus paying households or families as a group would be seen as like GiveDirectly is trying to shake up family structures or empower one person over another in a couple.
That was interesting to see it play out in the pilot: when we asked people if that was okay, their answer was that it was better, that it was nice, that people got to receive and spend their own money on the priorities they thought were most important. That it was helpful for family relations and things like that. That was a positive thing out of the pilot.
The main operational hiccup we saw was for the full study, because it’s fairly large, 20,000 people, more than 200 villages, 100 extra included as control group villages, we didn’t want to overlap with the Kenya national elections, which were scheduled for August last year. Basically, we didn’t want to get caught up in the campaigning period and confuse people about what we were doing and what the money was for and some things like that. We stopped a few months short of that August round of elections. Then they ended up having to redo the elections a few months later, so we had to pause our work for about five or six months or so.
That was the biggest operational hiccup which delayed us a bit. Otherwise, things are mostly proceeding as planned.
Owen: Gotcha. You mentioned the finding around couples that they don’t find this threatening to their family. Are there other findings you can share, either anecdotally or from the data you’re collecting?
Joe: Yes. I should caveat that most of what we’re learning so far is anecdotal. The full study, because of how it’s structured in terms of its sample size and that it has a control group, will be able to provide the most rigorous answers on all the questions we care about about a UBI. But from the pilot, it has been interesting to see a handful of things. Another question I was curious about which relates to comparing a UBI versus a negative income tax or other more targeted approaches to welfare cash systems, was how people would perceive a universal basic income.
This is a village where in absolute terms, on average people are very, very poor. But there’s still a decent amount of income inequality. The richest person has a greenhouse, and the poorest person has a family in one room with a house that’s falling over, to give you a sense that there’s still pretty meaningful ranges of wealth and income. We asked people whether it seemed fair that everyone was receiving the same amount regardless of need, that GiveDirectly wasn’t doing any attempt to try to suss out need.
The reaction from people was pretty interesting, that they thought it was better that GiveDirectly not meddle and try to pick winners and losers, that that was fair or more likely to be correct in their eyes, and also better for community relations. That was funny, and interesting to see and relevant for the broader conversations with the UBI.
Related to that, another thing I’ve seen is conversations about the payments and how different people are using them and how people can pull them together are just immediately a little bit less awkward than in programs where GiveDirectly, for whatever reason, has targeted specific individuals within a community. There’s this dynamic that everyone knows that everyone is receiving, and so it’s a lot easier to talk about the issue.
We’ve seen people do things like form savings groups, as an example, where you can imagine writing down a ledger where every month people contribute some portion of their transfer that they’re getting from GiveDirectly. Then every month, one person gets a big payout equal to the sum of all those portions of transfers. It’s a way to turn stream payments into lump sums, so it’s a way of savings. We saw those crop up basically instantly after announcing the program. I wonder how fast that would have developed if we had gone with a more targeted or means tested approach.
Owen: That gets into something I wanted to ask you about, which is if you’re seeing these village-wide effects or synergistic effects of everyone getting it and whether that’s on the economy of the entire village or group projects like the one you’ve described. Are there any more you can share?
Joe: From what I’ve seen so far, it has been mostly those savings groups. Basically, every demographic within this village has created one. The more elderly women, the young people, and so there has been a lot of trust-based savings groups pop up. I haven’t seen as much pulling together in infrastructure or something like that, but that’s definitely something we’ll want to keep watching for both in the village and in the full study.
Owen: I know it’s a little early to say, but do you think this pilot program has affected the basic income conversation more broadly and/or the charitable giving world in a new way that your work hadn’t before?
Joe: I think for the basic income conversation, something we’ve been trying to do, and maybe you guys can tell us how we’re doing, is there aren’t that many people receiving a basic income. Both with the pilot recipients and with GiveDirectly’s other UBI recipients, we’ve been trying to amplify their voices, ask them, “Okay, there’s this academic debate about whether a basic income should be universal, what do you think?” Or, “Should we target individual adults or families? What do you think?” Or, “How are you spending that?”
So much of the debate is very philosophical or theoretical, and we’re in a unique position to give some of the only basic income recipients out there a microphone to ask them, what do you think about these different debates? When we’ve written about the pilot, that has been a big goal of ours, and you can also, on our website, on GDLive, filter for UBI recipients and just hear how they’re describing their priorities, spending, and things like that which I think is a pretty cool perspective to bring to a debate that otherwise feels like a broader political philosophy debate or something like that.
I think that has been good. The other push we’ve been doing which compliments that push is trying to frame the debate around the evidence, which has two prongs. The first one is so much of the questions and assertions about basic income apply to cash transfers broadly. People on the pessimists side are worried that people will stop working or start drinking or spend poorly. We actually have a remarkable amount of evidence on recipients of cash transfers. doing not those things and tons of randomized control trials from all over the world. That’s a largely tested question. It’s largely played out that people don’t end up doing those things people are worried about.
People also have a lot of hopes for a basic income, that the money could get spent on businesses or schooling or health expenditure or whatever it is. We have a lot of evidence on those things as well, testing different structures of cash with different types of outcomes. The first prong of what we’ve been trying to do is contextualize the debate about basic income in terms of what priors we should have, given everything we know about cash transfers broadly, which is there’s a pretty good bet here in terms of at least the household level affects of giving people cash. That leads to the second prong, which is what are we actually testing here? What is unknown about a basic income? Which is often a little bit different from the questions that dominate the conversation; spending on alcohol or whatever it is.
Owen: Yes, your website is a really fantastic resource, both for the evidence and also those anecdotes about how it’s affecting real lives. GiveDirectly is expanding their work in some interesting ways. I want to touch on a few of those. Can you tell us about the work you’re doing with Ugandan refugees?
Joe: Yes. What’s neat about cash transfers, which applies to basic income, but also broadly, is that it can help force a policy question which is, in general, we have a policy goal of helping a certain group of people and because we care about it, we’ve attached a budget to it. Maybe it’s $1 billion a year or something like that. Are our efforts better than what the people we’re trying to help could do if we just gave them the money instead?
We’ve done a few different versions of that in more development contexts. We worked in post-hurricane Texas and Puerto Rico, which tested in a different type of context. The context for how we help refugees is very, very similar, where a lot of our systems for helping refugees globally were set up post World War Two and were set up for refugee crises that were fairly different from the crises we deal with today.
Uganda, you may or may not know, has taken in a ton of refugees, more in total than all of Europe did last year, from places like South Sudan and the DRC. With these types of crises, people end up staying for a decade or more. These are very prolonged crises, and as a result, people are often starting their new lives in a refugee resettlement versus being somewhere temporarily before moving on.
Our model for helping those people is matched to a older, more acute crisis model. We’re very good at keeping people alive with shelter or food or things like that, but not as good at launching people, giving them the resources they need to make big investments in their new lives. Something we’ve been testing in Uganda is giving people large grants. $750 to $1,000, as usual letting them spending on whatever they want, and seeing how that plays out. We kicked off a pilot at the beginning of this year, with about a few thousand refugees in one settlement in Uganda. We’re gearing up for a larger experimental evaluation with more than 10,000 refugees testing the same basic model there.
Owen: Also, you mentioned Texas and Puerto Rico in their hurricane recovery efforts. This is a different context than you’re usually working in. Usually it’s with these very poor villages where $20 a month goes very far, whereas Texas, that’s not necessarily the case. Can you tell us a little bit about why you chose to work there and any results you can share?
Joe: Yes. We chose to work in Texas and then in Puerto Rico once Maria hit, because one, you could see everywhere on TV and the news the devastation that the Hurricanes caused. What was encouraging was that a lot of people wanted to help. There was this groundswell of support and a pretty strong eagerness to try to help out, especially for Americans to help out fellow Americans.
Then what was interesting is that there was also a lot of frustration with the ways that those people had to help. This was a time when people were expressing a lot of frustration about the Red Cross. It felt like an opportunity to at least provide a proof of concept, that one model for helping these people is you could just send them cash and they could buy what they wanted. I think especially for the disaster-relief industry or context that’s helpful. When I went to Texas there were warehouses filling up with goods that were being sent to help people. Whether it was canned goods, or I saw industrial-sized bottles of lotion. Somebody had shipped a couch from Wyoming.
Our traditional model of helping these people is very much sending stuff and that might make a lot of sense in the initial 72 hours after a disaster, but after that, often stores are opening up, ATM machines are opening up and people have pretty varied needs. It’s a good opportunity just to provide a proof of concept there.
We ended up delivering debit cards to people in Texas and Puerto Rico of about $1,500. I think the results were interesting. You’re right that a dollar goes a lot less far in Texas or Puerto Rico. The role of that $1,500 I think was different than the role either of basic income or a lump sum grant plays in East Africa. What I saw it being was much more of a gap filler. That people had access to different types of support, whether it’s immediate support from the Red Cross or FEMA down the road or support from family members.
After all that support, there is often little things falling through the gaps. Whether it was somebody needed new clothes or they wanted to buy a washer-dryer or needed to get a head start on their home and so needed construction materials. That unconditional cash played a pretty good role because people could put it wherever they needed it. The main takeaway we saw from asking people about what they were spending on and things like that was, even with everyone experiencing the same crisis, how varied the spending patterns were across people, which I think demonstrates a lot of the value of cash, that it’s enabling that type of flexibility.
Owen: What was the public reaction to working in a more developed country?
Joe: One thing that was interesting was we have a lot of practice, and we talked about this last time, introducing ourselves to the communities where we’re working in East Africa. I think we weren’t exactly sure how we’d be received in East Texas or in Puerto Rico. It was funny to see how many of the same issues cropped up, that handing out cash is weird, and people think it is going to be a scam. You have to take the same type of tactics, whether it’s introducing yourself through the Mayor’s office or through the local church or whatever it is. With my operations hat on that was the most interesting thing to see was how much of the same communication introduction strategies and respect for the communities that we have practiced a lot in East Africa were required to get people to accept us in Texas and Puerto Rico as well.
Owen: Yes, so basically building community support and then working from there?
Joe: Right, and how universal the weirdness of cash is.
Owen: Lastly, I want to give you a chance to touch on the work you just started doing in Liberia, getting back to Africa. You just started working there. Anything you can share about the unique challenges there or what made you decide, what made GiveDirectly decide to start working in Liberia?
Joe: Yes, I’ll be intentionally a little bit vague here but in Liberia and in a couple other countries we’re entering this year, DRC and Malawi, we’ve been working with one of the larger government aid funders. For them what we’ve been implementing is a very literal application of that question I said cash can help pose earlier: that we have this budget, are we outperforming just letting the people who are trying to help spend it? As a result with this funder we designed, I think something like six or seven randomized control trials across Rwanda, Liberia, Malawi, and DRC.
Testing different structures of cash with different populations and basically seeing in what areas are the programs we’ve chosen to invest in outperforming what the people can do themselves and in what areas are they not? Which is a pretty, I think, exciting use case for cash as a benchmark or tool for large aid funders. That’s what brought us to Liberia in the first place. Liberia I think is one of the more challenging payments environments we worked in. Weirdly maybe after the US because we benefit a lot from mobile money in the other places where we’ve worked. I think the initial challenges in Liberia have been us working through how it will make sense to pay people.
These are environments where many of the roads are impassable for large periods of the year because of the rainy season. While we’ve dealt with different types of remoteness in the past, this has been a more extreme challenge of that. I’m curious and excited to see how we experiment and test with different payment modes as part of delivering cash there.
Owen: Those are the questions I had for you. Is there anything else you would like to add?
Joe: I would give a plug for searching UBI on our website at GDLive and just seeing what basic income recipients are saying about their experience. That’s a pretty cool opportunity and it’s, frankly, also just fun.
Jim: That was Owen Poindexter talking to Joe Houston from GiveDirectly.
Owen: I always find it interesting, the logistics of cash transfers, and also the basic question of what are they good for? If anything, what are they not good for as a default in terms of aid and charitable giving? Does it beat other forms of helping people? I think by and large it does.
Jim: I feel like there has been solid evidence for a while now that cash in at least certain situations adds more value, no pun intended, than a lot of the traditional in-kind approaches to supporting a people.
One thing that I continue to be surprised by is that on one hand, it seems like, we’ve actually done a lot of experiments, now we’ve done these pilots, we have a good sense of what will happen. Yet every time someone does a new one, it seems like there’s still new insights that we gain. There’s still more about the space that we’re being able to understand as we continue with these pilots and with these experiments.
Owen: Yes. Along those lines, I always find that cash works even better than I think it’s going to. One example of that was the one that Joe gave about, will this affect social structures in the village? That was an issue in the US when they were doing trials in the ’70s. You think it could really threaten the social fabric if you give people a new level of independence. From what he said, everyone prefers it this way, that individual adults each get their cash transfer.
Jim: Yes. I thought the takeaways so far or at least the observations so far around universality were particularly interesting, because that is the thing here that is new. That if you look at the existing pilots to date, there has been almost no saturation studies. The fact that they are doing this universally at the village level, from the get-go, the idea was that’s going to help us to better understand and learn more about those effects.
Yes, like you said, I thought it was really, really interesting to see what the reactions on that were so far and that they’re– at least in the context of these Kenyan villages, people really seem to appreciate the universality. What can we extrapolate from that, as far as understanding how folks elsewhere might or might not perceive it the same way?
Owen: I’m hoping that all these questions that GiveDirectly is asking make their way more and more into the charitable aid world generally. The question he kept bringing up of, “Is cash an improvement on what else we might be doing for these people?” I think it takes a lot of keeping your ego in check. If you’re an organization that gives cows or mosquito nets or clothing or whatever it is, to ask yourself, would we be just better off giving people money?
Yes, in some cases, I’m sure there are cases where a mosquito net is more valuable than the cash it costs, but I think in a lot of cases people will have to ask themselves the question that’s threatening to the organization.
Jim: Right. It seems like that is a process that has been happening for a decade now. You’re starting to see some of these more traditional organizations gradually get more on board with that. The fact that the Red Cross is now using cash as part of its way of supporting people and families in struggling situations. We’re on a gradual curve of adoption, and there will probably still be resistance from some fronts for quite a while, but, hopefully, less and less over time.
Owen: That’ll do it for this episode of the Basic Income Podcast. Thank you to our producer, Erick Davidson. Please subscribe on Apple Podcasts and leave us a rating and review if you can there or on the service of your choice. And please tell your friends so we can keep expanding this conversation. See you next week.