Today, I'm very pleased to welcome Professor Dean Karlan,

Professor of Economics at Yale University, and

also President of Innovations for Poverty Action.

Which he's going to tell you what that organization does.

He's also the author of a book on this area, More Than Good Intentions.

So Dean, welcome to Princeton and to the course on Effective Altruism.

>> Thanks.

It's great to be here.

>> So you're an economist.

I think a lot of people would be surprised to find that an economist is working

on issues about the effectiveness of interventions to reduce poverty or

to help the global poor.

Would you like to tell us how you came to be in this situation?

>> Sure.

So I came about it first as a Humanitarian.

I was very interested in issues of international human rights

and development.

But actually was, my undergraduate major was in foreign affairs.

I was more interested in things in political philosophy and

things of this nature.

And I, I took an internship two years out of college, just before going back

to business school in El Salvador to work for a Microcredit Organization.

This was 1992, so it was before the days of the internet.

But so I'd read some of the glossy brochures of MicroCred and

it really appealed to the, the market instincts that I had.

And kind of a banker, I'd done two years of investment banking at the time.

And here's this idea that there's women who are not getting access to credit,

they have great opportunities for some sort of small enterprise.

And what they need is a loan to help them, help themselves and

help build up income and with that income, improve the health and

education of their children, et cetera et cetera.

So it really appealed to me.

So I went down for a year as an intern, totally unrelated to research and

what they were doing.

I was developing software systems for them.

But what really struck me, when I got down there was, it just wasn't obvious to me,

why this was working?

Why this was actually doing anything to solve problems of poverty?

And I saw some stories that made sense.

But I also saw a lot of stories that didn't make sense to me.

And I asked them, what's your, where's your evidence?

What, you know, does this work?

And they showed me their impact evaluations which they had given to USAID

and everybody was happy.

And it was no different than others were doing that I, you know,

I had other friends working for other non-profits in El Salvador and, you know,

similar types of studies.

But they were really, really bad studies to be perfectly blunt.

They asked pers, participants in the studies.

Are you doing better than you were a couple years ago?

And I didn't have any formal training, but

I knew enough to know that that doesn't make too much sense.

>> Right. >> As an actual, actual metric of success.

>> So that when considering any other people who hadn't

got the loans as to whether they were also doing better,

maybe the whole economy was doing better or whatever.

>> Maybe the whole economy's doing better >> Maybe, you know,

maybe the Microcredit prorgam which says their choosing

entrepreneurial of people was doing just that.

>> Right, right.

>> And so entrepreneurial people do better over time than people who

are less entrepreneurial.

So that got me excited to try to figure out what's, you know,

does this work or not?

And I, I had a very math-y background that kind of always my, my,

my thing when I was younger.

So when I started looking into what disciplines I could do,

I could study more about, learn more about what works and what doesn't?

It was clear to me that economics was kind of a had the toolkit,

both analytically and theoretically to understand these markets,

to understand human behavior in these contexts.

And also the empirical toolkit to do the statistical work.

To be able to set up to understand whether,

whether things are working or not.

So that's what kind of inspired me to go back to graduate school and

get a PhD in economics.

>> And so

then from then on you were always thinking of this idea of how can we decide what is

really going to be effective in terms of, of helping the global poor?

>> Exactly.

Exactly.

And in, in graduate school at that time, there'd be a fairly long history of,

of research which compared one country to another.

We gave more aid over here and less aid over here, and

is this country doing better?

And it was a lot of frustration with how,

how much you can really learn from that kind of study?

The, because there's just too many things that are going on across countries to be

able to isolate aid and say what the causal effect of aid in general?

The second concern with that is that it's in some sense asking just too

grand a question.

There's no such thing as aid in just a monolithic way.

There's good aid and bad aid and ultimately the question is

about does this particular project work and does this idea work?

Not about the mere concept of aid.

And so when I was in graduate school it was right when there was a new, right when

people were kind of taking that turn away from these kind of macro level studies,

and saying let's go to the ground and set up a test.

>> Right. >> Let's just, just,

rather than sit there and debate this endlessly with data that are questionable,

let's just go the ground and work with a group that's doing something and

let's just set up a simple task and find out.

Does a work better than b or not?

>> Right.

Now since you mentioned this idea of people thinking of aid as one thing maybe

this is a good opportunity to, to ask you about some of the aid

critiques- >> Mm-hm.

>> That people often talk about.

So there's William Easterly book the White Man's Burden.

There's Dambisa Moyo with Dead Aidem my Princeton colleague, Angus Deaton

has a chapter on which is really somewhat critical of aid in The Great Escape.

So what's your response to,

to this view that essentially I guess you could sum it up as aid doesn't work?

>> So I've, I've two critiques to it.

One is that, that it is the wrong question to pose.

That the question is, should be about a particular aid.

That there is no simplistic answer.

That there are certainly situations where you could imagine aid has done harm

and, but

I can point to situations where we've seen long term beneficial effects from aid.

And so the first is just that, the, you know, that,

the, the, the, the, it's kind of like asking is, you know,

is all are, is, is, are, are all health services good, right?

>> Right. >> So certainly there's some

quack doctors out there that are bad and some that are good.

Are all schools good?

Well no. We like education.

Education's good but some are good and some are not.

Some might just waste people's time.

>> Right. >> So in that sense it's, you know,

the question is too grand.

The, the other sense is a lot of the que, a lot of the criticisms

are on a theoretical level about the way people might respond to aid.

So maybe you give something, and we give you a very micro example.

Big debate always took place on bed net policing.

Should you give them away for free or should you sell them?

Right, and this is something that Easterly wrote a lot about in his book.

And others have proposed saying no, we should give them away for free.

And there's actually good theoretical reasons to make the argument and,

you know, [CROSSTALK] if you listen to the theory, you're like wow,

these are good competing arguments, what's right?

And, and that's where, that's where our approach is to say,

well these are good theories one of,

[LAUGH] you know, some of, one of them is going to overpower the other.

>> Yup.

>> Let's go to the ground and set up a test.

And so one of research servers that's in our group Pascaline Dupas and

Jessica Cohen set up a test at Innovations for

Poverty Action in Kenya where some households got free bed nets and

others got bed nets that were sold to them and it was a more complicated test.

And I'm not giving you the full, the full details, but the short

of it is they found really strong evidence that giving bed nets away for

free led to increased long term demand for bed nets and higher usage rates.

>> Uh-huh, that's interesting, because a lot of people would have

thought that if people get something for free, they may not use it.

If they have to pay for it, they will use it.

>> And, and they also might not buy, there's that effect and then the second

effect is that leader down the road, when someone offers to sell it to them, they'll

be like why should I buy it, people give this away for free all the time!

>> Right.

Yeah.

>> We set up a separate test of different health products in Uganda one with a,

a headache medicine.

And we did find long term effects.

Well, I shouldn't say long term, it was about three, two months.

So give, give a headache medicine away for free, and

two months later people were less likely to buy the headache medicine.

>> Uh-huh.

>> Right. Whereas we didn't find this for

some other products where there was positive learning.

So one key theme that we think is true here is, if there's learning that needs to

happen, where people don't know the value of the product.

>> Right. >> Giving away for

free actually makes a lot of sense.

>> So there never had it been there, it's kind of like a free trial.

See, this will keep mosquitoes off you, from biting you.

>> Exactly.

>> And then people want that.

So then they. >> The reason why you,

we can go into fancy grocery store and get free samples.

>> Right. >> Because there's, you know,

needs to be some learning.

>> Right.

>> And then people learn and, and in the learning might be about the benefits,

it might be about it not being as uncomfortable with bed nets is a lot of

times a concern about people think that you know,

the bed nets not comfortable to sleep under.

But if you get it and you get used to it then it's not so bad.

>> Right.

>> Yeah.

>> Now you mentioned your, your group.

But you, you were telling me about how you got into this area.

>> Right. >> But

you didn't actually tell me about how or

why you thought it was necessary to fund Innovations for Poverty Action.

>> So yeah.

So when I finished graduate school, it was 2002,

there were two things that, that immediately struck me were, were missing.

One was that to do the field work that we were doing, that we were,

had, had just started doing really, required a lot of time in the field.

And a lot of expertise that we don't get trained in as economists or

other disciplines either.

Then it's about managing people.

It's about managing surveys,

design of surveys, coordination with organizations, to setup randomized trials.

And this requires a lot of time and effort, and

skills that that you acquire, and you want to keep using.

And so the, the typical model at that time was to use graduate students,

to get them involved in the project.

But then it's, every year it's like a new flow of people.

It would be much better and much higher quality work if we had an organization

that was trained and skilled in helping to support this type of field work.

So that was the first gap that I felt like was, existed and so we needed an entity.

And the, and the other thing we realized is that this gap is,

shouldn't sit within any one university because we want to be able to provide this

type of support to researchers at other universities.

So this is why we created it outside of, outside of the university.

The second reason for it was the, the,

the gap that lies at the heart of our incentives as academics.

So my incentives as an academic is to write an academic paper,

get it nicely published, and move on to some other paper.

I don't have the academic incentive to work on the policy side.

Not nearly as much at least.

I mean not to, sit down with organizations that are in that space and help show them

and, you know, this is what we've learned and this is how you can use it.

And in some cases that might actually require incubating the actual idea

at large scale to see how it works.

So the second motivation for the organization was to work on that policy

sides, to make sure that the knowledge doesn't just get created and

get published and then we just kind of hope that people read it.

>> Right, right.

>> But to actually think about, okay, what are the steps we have to take,

to make this take this to action?

Do we have to do it ourselves?

Do we need to bring it to government?

Do we need to bring it to other organizations?

Who do we do that with?

>> And IPA has now become quite a large organization, right?

>> Yeah, we.

>> Quite a number of people.

>> We have around 1000 employees.

About half of those are full time long term employees and

about half of them are short term hire for management of a particular project,

like a survey or something of this nature.

>> And how many of them living in the United States or other affluent countries?

>> Not too many. Most of them are in the, in the field.

So I think in the United States we have something in the ballpark

of 70 to 80 employees.

And the rest are all international, living overseas.

We have we have about 18 countries that we have focused on.

So one thing that we have found for sure is that our im, ability to impact

policy is much stronger when we have a long-term presence in the country.

>> So you're employing the local people native to that country and they're.

>> Employing, employing locals and

also just building long-term relationships with local policy makers, donors, NGOs.

So that it's not just going in, working on one study and

then moving onto some other country.

So we have actually done projects in 51 countries.

>> Right.

>> But we're focused on 18 in terms of our development and

our efforts to influence policy.

Why 'Does Aid Work?' is the Wrong Question to Ask

Effective Altruism

Meet the Instructors