“Excitement and horror” – Paul Niehaus on the path to ending extreme poverty
News Editor Gussie Coulter joins development economist and GiveDirectly founder, Paul Niehaus, to explore the tractability of extreme poverty, direct cash transfers and his recent paper, ‘What would it cost to end extreme poverty?‘
By Gussie Coulter
“You guys must be smoking crack” was one criticism Paul Niehaus and fellow graduate students at Harvard and MIT faced when they conceived GiveDirectly in 2008. Others, he recalls, “literally said ‘I hate it’”.
Eighteen years on, GiveDirectly has delivered more than $1 billion in cash into the hands of over 2 million people living in extreme poverty; and a recent research paper, entitled What Would it Cost to End Extreme Poverty? co-authored by Paul, offers even greater promise yet.
It estimates that reducing the poverty rate to 1% through use of direct, unconditional cash transfers could cost a mere 0.3% of global GDP or, as Paul frames it “about a seventh of what the world currently spends on alcoholic beverages.”
Paul is a prominent development economist, entrepreneur and Chancellor’s Associates Endowed Chair in Economics at UC San Diego. As I join him via Zoom from across the pond, we explore his philosophy, philanthropy and findings in this landmark paper. He offers hope in the potential for state-driven change – but also in the power that we wield as individuals.
Influenced by his mathematician mother, Paul has previously indicated a desire to take mathematics as far as possible. But it was at university that he began to synthesise this quantitative background with philanthropy and international development.
“When I came to college… I was thinking about either studying engineering or economics.” Opting for the latter, Paul found that “it seemed like a toolkit that you could use to think about a very broad range of social problems in a rigorous and helpful way.”
[Economics] seemed like a toolkit that you could use to think about a very broad range of social problems in a rigorous and helpful way
Outside of class, Paul engaged with “the most beautiful and inspirational pieces of development economics – writing by people like Amartya Sen, and so forth,” which he was “very excited about.”
At the same time as his studies at Harvard, an “experimental revolution” was beginning to take hold of development economics: “There was lots of energy and excitement building over this idea that you could sort of learn things that you needed to know about the world by setting up controlled experiments.”
I ask Paul how the concept of GiveDirectly originated.
“GiveDirectly actually grew out of the conversations about how to try to reduce corruption in other existing programs,” he explains. “We were talking specifically to some people that were working in last mile payments” to establish whether there were “more secure ways to transfer benefits to some of these very, very poor households.”
This work illuminated a remarkably simple, but as-of-yet untapped source of potential. “If you’ve created bank accounts essentially for some of the poorest people on the planet, it also opens up… this very exciting opportunity for individual donors: We could put money into them.”
“That was sort of the initial impetus.”
And during this “experimental revolution,” Paul started to see some of the results of “good, high-quality” studies on the effects of cash transfers. “Those also to us looked very positive in the sense that you were seeing people using money in a variety of different ways, but all things that were sensible, including things that could have long-term benefits for their lives.”
“Not the sort of systematic abuse or misuse of transfers that I think… we had been led to expect by the rhetoric we had heard growing up.”
These studies employed an almost clinical approach to international development and used methodologies like randomised controlled trials. Something as straightforward as unconditional cash transfers were shown to have tangible benefits: the orthodoxy, to some extent, had been falsified.
GiveDirectly has continued to build upon this approach by assessing the impact of different cash transfer policies, such as by exploring the relative impacts of a Universal Basic Income (UBI) or a program of lump sum cash transfers.
“I think that’s been a pattern that I look for actually in my career when you propose an idea. If everybody’s sort of ‘ehh’, then it may not be that big of a thing.”
“If some people love it, and some people hate it, then that’s probably a good sign that you’re onto something.”
I think that’s been a pattern that I look for actually in my career when you propose an idea. If everybody’s sort of ‘ehh’, then it may not be that big of a thing… if some people love it, and some people hate it, then that’s probably a good sign that you’re onto something
Many may ask why we should give cash to the poor, and why we shouldn’t prioritise other, more paternalistic approaches to development and charity? Paul advocates for a total inversion of this question. “The question should be: ‘Why not cash?’” he explains.
“There’s this funny line, ‘teach a man to fish’, which has become very popular” Paul tells me. He is referring to the proverb give a man a fish, and you feed him for a day; teach a man to fish and you feed him for a lifetime. It is “this idea that the problem is that people have been trained inappropriately, and we know how to train them appropriately – if we can do that, it will fix the problem.”
Throughout the pursuit of poverty elimination in the last century, almost all the money invested into this mission has been spent by people other than those living in extreme poverty. This “teach a man to fish” mindset is “radically counterfactual if you look at what the data says,” Paul explains.
And furthermore, it neglects “intrinsic value in people’s agency and ability to make their own choices… we value that over and above the impacts.”
[The teach a man to fish mindset] neglects intrinsic value in people’s agency and ability to make their own choices… we value that over and above the impacts
And to what degree has the development sector taken on treating cash as a benchmark? “To an extent,” Paul tells me, “and I think it would be good if we did it more.”
“The humanitarian subset of development work is the part that has been most receptive to this… we have seen at USAID, before it was shut down, some really exciting benchmarking studies” within which outcomes, including employment, job creation, enterprise and nutrition were assessed.
“Maybe we can actually get just as much out of them per dollar if we just give the money to people and let them do what they want with it,” they thought, “and that turned out to be the case.”
It was observed that by many of these metrics cash outperformed policies optimised for specific outcomes – i.e. you may be able to yield superior employment outcomes from cash transfers than from alternative, more paternalistic approaches optimised to address the issue of employment.
Having noted his engagement with the Effective Altruism for Christians movement, I ask Paul to what degree his faith has motivated his work. “That’s why I do it,” he tells me modestly. “We are not here, I think, purely for our own enjoyment, but to become better human beings and to contribute to other people’s lives. And so that’s my reason for trying to spend my life the way I do.”
We are not here, I think, purely for our own enjoyment, but to become better human beings and to contribute to other people’s lives. And so that’s my reason for trying to spend my life the way I do
Paul’s recent paper centres on asking the question of how much an actionable cash-transfer policy would cost to end extreme poverty. “We have good estimates of how far below the poverty line most people in the world are, and we call that figure the poverty gap – and that estimate is about 0.1% of global GDP.”
“And that, by the way is, I think, very exciting and a positive thing that people should know if they don’t – because a lot of people have the misperception that extreme poverty has been getting worse, that the numbers of people in poverty have been going up.”
“They’ve actually fallen dramatically, and that’s a very exciting and positive story.”
A problem remains, however: “0.1% isn’t actually the cost of a feasible policy that would end poverty.” The reason being that “when we come up with those estimates we get them from these very, very small, representative but small samples that countries survey to estimate the standard of living.”
Yet, it remains that case that we do not know how far below the line individual households are. There is a “cost of imperfect information.”
This research took information held by governments used to target existing programs, such as where people live, what their houses are built of, or how many cows they own, and considered how predictive those things may be of standard of living.
In collaboration with his co-authors at Berkeley and Stanford, the research uses “modern data science techniques to learn the optimal policy, to learn a transfer amount to give to each household, based on what we know about them, and what that would end up costing.” The work landed on 0.3% of GDP, three times the poverty gap.
It “had to be at least weakly above 0.1%,” Paul explains, but “it could have been much, much higher.” “In the extreme case, if you wanted to completely eliminate extreme poverty for everybody, and you knew nothing about them, you’d essentially have to give everybody $2.15 a day… the universal basic income – an extremely expensive policy.”
“It’s like saying if everybody were willing to give up drinking for one day a week, that would generate enough revenue for us to feasibly approximately end extreme poverty.”
If everybody were willing to give up drinking for one day a week, that would generate enough revenue for us to feasibly approximately end extreme poverty
How did Paul react to this finding?
“Great excitement, because I think this is an incredible opportunity to all give up half a percent… I would even advocate for more than that,” Paul tells me. “But some small amount of our income, if we were all willing to give that up, we could achieve this tremendous moral landmark that people have thought so long is elusive.”
“One person did tell me that they feel both excitement and horror. The excitement because we can do this, and the horror because we haven’t already done this.”
“I think that’s a fair reaction,” he reflects, “this is something we can do… but we should also have done it already.”
Paul’s paper notes that whilst the total is small, it may still be too large for individual poverty-afflicted nations to finance it themselves. I ask how he thinks policy makers need to progress beyond the 0.3% number and pursue an actionable plan.
Paul notes that “there are a few different audiences who can make use of the results in the paper,” including national governments optimising their existing transfer programs. But there is also great opportunity for “large public international donors… who pay a big role in financing many of the transfer programs we have and could do more.”
“Obviously, there have been huge cutbacks in the last year, so many people are sort of looking at this and saying, ‘gosh, we really need to economise and figure out how to best spend the money we have’.”
“But I think we also need to get people excited and rebuild support for some of this.”
Paul suggests that we need to shift from thinking just about what we give up.
“Most of the discourse around foreign aid is about trying to hit a target for how much we give up – 0.7% or whatever it is – which isn’t bad per se, but it’s honestly not all that exciting because it doesn’t say anything about what we get.”
“I think we should be communicating to people what we get from foreign aid, which is some pretty great stuff – we failed to do that.” Paul thinks we should prioritise communicating to people that ending extreme poverty is “actually doable” and “here’s how we’re going to do it”.
I think we should be communicating to people what we get from foreign aid, which is some pretty great stuff – we failed to do that
“I think this is something that anybody who reads the paper could say like, heck yeah, I want to be a part of that. I want to do my bit. It’s like negligible – I’m not even going to notice the difference.”
I ask Paul what his reflections on the path to ending poverty, given the dramatic cuts and closure of USAID.
“I think it raises a question about how likely it is that any of this will get coordinated at the sort of public national level,” he explains. “But I don’t think anybody should let that deter or discourage them.”
Paul insists that the levers to end poverty still exist, and that he doesn’t think “there’s anything that comes close in terms of how exciting [ending extreme poverty] would be.”
“I don’t… feel particularly deterred by the moment we’re in… we should press on.”
Given the interdisciplinary nature of Paul’s paper, involving modern statistical methods, economics and machine learning, I ask Paul what his message to Durham students interested in working on problems like this would be.
“I just feel very, very grateful that I’ve been able to work on problems that feel truly meaningful,” Paul reflects. “We all have to figure out what we’re good at and where we fit into the big picture… for me I think being at the boundaries between different fields or different sectors.”
“I definitely do… encourage people to keep an open mind about what that role might be,” and that one particular specialisation does not “limit you to doing that one thing.”
I definitely do… encourage people to keep an open mind about what that role might be… one particular specialisation does not limit you to doing that one thing
“I was at grad school studying to be an economist, and then this idea came along for a nonprofit that seemed compelling, and so we started that. And then that led to the idea for a couple of for-profit companies to increase capital flows to poor countries. We did that as well.
Paul insists that “things can lead in exciting and unpredictable directions, and I would be open to that.”
Leaving Paul, it became clear to me that the end of extreme poverty is manifestly in sight – even attainable in the near future. Whilst its lack of actualisation did leave me feeling sick, its proximity should indeed be a source of excitement and motivation. And, as Paul insists, we should be invigorated by our part to play in this mission.
Centre for Effective Global Action: What Would it Cost to End Extreme Poverty?
To End Extreme Poverty, Give Cash — Not Advice | Rory Stewart | TED
Image credit: Paul Niehaus / Centre for Effective Global Action
