Chris Blattman on Sweatshops
Dec 26 2016

textiles.jpg If you were a poor person in a poor country, would you prefer steady work in a factory or to be your own boss, buying and selling in the local market? Economist Chris Blattman of the University of Chicago talks with EconTalk host Russ Roberts about experimental evidence on how poor people choose in the labor market and the consequences for their income, health, and satisfaction.

Mike Munger on the Division of Labor
Mike Munger of Duke University and EconTalk host Russ Roberts talk about specialization, the role of technology in aiding specialization and how the division of labor creates wealth.
William MacAskill on Effective Altruism and Doing Good Better
How much care do you take when you make a donation to a charity? What careers make the biggest difference when it comes to helping others? William MacAskill of Oxford University and the author of Doing Good Better talks with...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Dec 26 2016 at 4:25pm

Russ, if you haven’t already heard about it I think you might enjoy taking a peek at a new kind of direct charity via mobile phone app called DonorSee. I heard about it on the Tom Woods podcast Ep. 751 Help People Around the World by Going Over the Heads of Governments and Aid Groups, though of course they also have a website. The premise and slogan is “Give directly. See impact.” and they pitch it like so:

DonorSee is the best place to give your money. Whenever you give on DonorSee, you get to see exactly where your money goes via our aid workers who take pictures and videos of your money making an impact.

I’m not affiliated with them at all, though you’ve mentioned your interest in related ideas several times and I know you’re skeptical of many existing approaches, so I thought I’d share. Maybe this one is bunk, too, though so far it seems to be working (at least at a modest scale), and the idea is interesting in any case. (late) Merry Christmas!

Dec 28 2016 at 10:45am

An interesting podcast. Notes:

– The middle level management observation was unexpected and enlightening.

– The Chinese that Russ mentioned have a similar problem with factory workers. Young people from barely above subsistence agricultural areas go to the factories for the promise of more money. But the nature of the work and after expenses return are such that during the annual Golden Week where everyone returns home, a significant percantage never come back to work.

– When studying cohorts, people often forget that they are different people. Over the course of their lives, people generally move a few income quintiles, both up and down (think of an eighteen year old starting out, a fifty year old in prime earning years, and a retiree). This study gets it right by following people.

I first learned this distinction from reading Thomas Sowell, who my father introduced me to 30 years ago by telling me he should be President. I started reading him then and he has taught me more than any other person in my life, aside from my father. I note this as this week is his last column (here). I know that no one does this, but reading his old columns will result in a wealth of knowledge about economics and life. He will be missed.

– Throughout the discussion, I kept wondering what a US McDonalds franchise owner would say. “Of course 20% don’t show up the first day”, “of course we turn over 100% of our employees a year”, “of course sick days are through the roof”. And on top of that, they have to compete with a generous welfare state that doesn’t exist in Ethiopia (never mind minimum wage laws, US regs and corporate franchise regs).

– Based on back of the envelope annualized incomes, the US equivalents would be about $16K for a minimum wage job cohort and $20K for the lump sum investment cohort. Would a US lump sum cohort spend 1/3, save 1/3 and be able to invest 1/3 in an owner operated business and then make $21K/yr off of that investment? An interesting (although expensive) future study.

J. D.
Dec 29 2016 at 7:47am

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]

William Pocklington
Dec 29 2016 at 9:52am

For an American audience, the lead of the Blattman/Dercon paper is buried five pages from the end: The managers… made efforts to add non-wage benefits, such as free transport, free clinics, and in some cases a free meal at a canteen. They appeared less likely to raise wages, however, and often failed to keep pace with inflation and keep real wages constant.

While the authors are right to question the generalizability of the study, it’s hard not to draw comparisons to practices even in developed countries.

Instead of seeking real wage improvements or the opportunities from “informal employment”, US workers are quick to opt for the perceived benefit of stability from institutionalized formal employment offerings.

Dec 29 2016 at 2:23pm

At first, I was surprised and a little disheartened by the revaluation that the factory jobs are not better than other alternatives but it is not a bad thing. Many people say that x country needs to industrialize, get factories but I have for at least 20 years said no they just need to find some way to earn. I was thinking of places like the Cayman islands, they never had much of what most people think of as industry but they prosper. Maybe this story shows that I have been correct.

Robert Swan
Jan 2 2017 at 7:26pm

Interesting, but I think Prof. Blattman might be a little too sure of the results. On the surface, the trial is randomised, but I wonder if some people who would really have performed well were eliminated by the qualifying hurdle.

In the early 1900s, my grandfather diversified his flour milling business to generate electricity for the local town (Buncrana, Ireland). One of his most valued staff members was known around the town as being dull-witted. His job was mostly to watch a small number of gauges, to turn various knobs to keep them all in the happy zone, and to let other people know if any of them were nearing a limit (so they might fire up another generator or whatever). A boring job in which “brighter” people quickly became inattentive, and their inattention occasionally led to expensive machinery being damaged.

These African factory workers were required to have a Grade 8 or 10 education. I gather that was considered a relatively high level for the area. You’re taking some of the brighter people and giving them a job where there isn’t much call for initiative; should you be surprised that a fair number walk away? Perhaps the first randomised trial ought to have been on how good the qualifying test actually was.

Liked the quote on being exploited vs. not being exploited at all. Exploitation is in the same bucket as discrimination: once benign words that are now pretty much always steeped in evil.

Greg Alder
Jan 4 2017 at 12:07am

Excellent comments, Robert Swan. I taught at a rural secondary school in Africa for three years, and as I was listening to this interview I was thinking of the myriad paths that my former students have taken over the last decade. One went to work for one of the Chinese-owned garment factories in the country (Lesotho), and she is now a manager. Another worked in one of those factories for a day and quit. He thought he was overqualified for the dull work, which he also thought paid too little. He turned to selling cowbells as a small-time entrepreneur. He wasn’t making much money, but at least it was of his own initiative and he wasn’t working for pennies under the thumb of a Chinese boss, as he put it.

Jan 8 2017 at 3:10pm

I couldn’t help to wonder during the podcast whether the opening up, or expansion, of a factory had an indirect effect on those not working in the factory.

If an industrial area is expanding then there is more income generated in the area. The income is then spent in the area making entrepreneurial endeavors more likely to succeed.

Kind of makes me think of the ’49ers and the gold rush. It was more profitable to sell shovels and jeans than to mine for gold. In areas of expansion it may be that it is better to offer services/goods to those involved in the expansion than to be a direct part of the expansion itself.

To test this a better control group may be to use an area not undergoing expansion and see how their lives improve, etc.

Comments are closed.


EconTalk Extra, conversation starters for this podcast episode:

This week's guest:

This week's focus:

Additional ideas and people mentioned in this podcast episode:

A few more readings and background resources:

A few more EconTalk podcast episodes:



Podcast Episode Highlights

Intro. [Recording date: November 16, 2016.]

Russ Roberts: Chris Blattman... previously appeared on EconTalk in July of 2014 talking about giving cash to poor people as a way to fight poverty, as opposed to less direct methods. And today he's back talking about a different research project he's been working on, trying to figure out whether working in a factory is better for the poor--what is sometimes called a sweatshop--than informal self-employment. Chris, welcome back to EconTalk.

Chris Blattman: Thank you.

Russ Roberts: Now, you begin your paper, which is co-authored with Stefan Dercon, pointing out that a lot of anti-poverty programs, in the name of development, try to help poor people become more successful entrepreneurs: they are typically self-employed. Let's talk about what some of those programs are and what we've learned about their success or failure.

Chris Blattman: Sure. So, I mean, I think a lot of these programs fall into two categories. A lot of them give--and I think the best ones--give capital to the poorest people. And some of them give skills. They are basically giving some kind of input into production, because these people are producing. And one of the reasons, I think is just because in most of these very poor countries there are very few firms. And so, being employed means being self-employed. And that could be your farm; it could be being a petty trader. It could be some really crude manufacturing. And so, so, basically giving people inputs that they otherwise have difficulty getting.

Russ Roberts: Yeah. In these kind of situations I see somebody sitting on a stool outside a poor dwelling offering something for sale or service inside. Right?

Chris Blattman: Right. Or one of these, you know, you sort of walk around a village and a city and you see just shop after shop selling the same crummy goods, the same maybe 20 things--matches and soap--and you kind of wonder. And that's an employment of a sort. You kind of wonder how any of them are making any money.

Russ Roberts: Or presumably they are not making very much.

Chris Blattman: No.

Russ Roberts: So, when we think about adding capital to that activity--'capital' is a big, fancy word--but sometimes that would mean a factory. Sometimes it means a wheelbarrow. And sometimes it just means some cash to borrow so that they can invest in something, a tool, a very basic tool. Right?

Chris Blattman: So it could mean--so, a lot of the time it means livestock. And it means inventory. And is basically what people do. So, if you give them those things directly, or if you give them cash, which is what we talked about last time, people tend to actually take these things that don't look super-profitable on the face of it. They expand the stock of the store from, you know, 10 crummy items to 30 crummy items. Or they got a goat, or two. Or they get some chickens or maybe a cow. And that's how they sort of expand what they are doing a little bit. And so these--the funny thing is these programs tend to be pretty successful. So, when we've done randomized evaluations of giving people cash, of giving people goods that they can use in a business by helping them expand their inventory, it turns out these shops, you know, they are not super-profitable, but they do a little bit better. And so, they go from earning $2 a day to $4 a day. Or they go from earning $5 a day to earning $10 a day. And that's not a whole lot. But when you consider how cheap some of these interventions can be --not always, but how cheap they can be--that's a pretty good return. And if you only earned $2 bucks a day, an extra $2 bucks makes a really big difference.

Russ Roberts: For sure. And also, it's a cushion. If things don't go well for a while. There's illness, there's all kinds of uncertain and unexpected events.


Russ Roberts: One question I have about this whole research agenda is: As an outsider to the development world--meaning this is not my expertise. You are in it, and I'm an outsider who is very, very interested in it; but I am not in it--

Chris Blattman: Yup.

Russ Roberts: And the actual kinds of aid that occur--well, some of them are happening as part of an experiment. Which is what we are going to talk about today. You get, or an economist gets a grant or has some funds to find out how people respond to various things. And that's interesting. It's interesting about the human condition; it's interesting about how work is in poor countries. But when I think about "global aid"--when I think about, say, money that the U.S. government spends or an aid agency: Is that the kind of thing they are going to actually do, is go around to poor merchants on the streets of a terribly poor city and give them a little bit of money? In other words, what kind of infrastructure is there to implement these kind of lessons?

Chris Blattman: Right. Well, there are just so many forms of aid. Even if we don't give very much, as say a percentage of income here, that's still a small percentage turns into big numbers. So, aid gets spent on a lot of things. There is a significant amount that turns into these kinds of employment programs or cash transfer programs. Here's a couple of the examples. One, of all the millions of Syrian refugees throughout that region, whether they are in Lebanon or Jordan or Turkey, have ATM (Automatic Teller Machine) cards. In fact, they probably have a wallet with 6 ATM cards: one is from the U.N. High Commission for Refugees. And one might be from the International Rescue Committee. One might be from another organization. And when those organizations, when they--instead of giving out sacks of grain or tarps like they might have in the past or might still in a really remote region--these are refugees who are not in a remote regions. You don't need tarps or grain they need to survive. And so they transfer cash onto these ATM cards. So, infrastructure there is extremely simple in some ways. And it needs to be simplified. But that's very effective. And then in places where that doesn't exist, or even in the recent past we didn't have so many ATM cards--some of the cash transfer programs [?] in Uganda--the Ugandan government gave out, had a $300 million dollar program, which was basically cash transfers to groups of youths and to communities. Some of that was for starting businesses; some of that was for livestock purchases; some of that was for community goods building like little bridges or a teacher's house or something in that nature. And that was the second-biggest program in the history of the country. So, countries are doing this on a large scale, either with their own money or with aid money.

Russ Roberts: I can't help but note that we aren't as likely to give a tarp to a poor person in a Third World country, we are still giving 'tarps' to rich people in America. That's a little finance joke. I couldn't help myself. TARP--Troubled Asset Relief Program--I think is what it stood for. The $300 million that the Ugandan government gave--that's out of their tax revenue? Who is funding that?

Chris Blattman: Well, that was a credit from the World Bank. And a lot of aid comes in the form of subsidized credit. And so, in this case the World Bank provides a very low-interest loan to Uganda; and the Ugandan government has conceived of this program. They've done so with some advising from the World Bank, because the Bank has said, I think, they've done this in Tanzania; and the Tanzanian government was very happy with it. And so they sort of help transfer some of these programs. And so the idea is they'll pay this back eventually. Now, if it fails, probably those loans would get forgiven. But if they succeed--and this program seems to have succeeded--then the idea is that this helps you grow your economy now; and then you have the tax revenues in the future to make taking that loan worthwhile.

Russ Roberts: Boy, that sounds optimistic. But I guess we'll find out. It's a nice idea.

Chris Blattman: Yeah. I mean, you know, a lot of the average African country has been growing--I can't remember the exact percentage but somewhere between probably 6% and 8%--not per capita but overall over the last decade, maybe even for 15 years. So, that's not bad. I personally think aid is a part of that. I don't think it's a major part of that. Which is why I ended up taking an interest in factories and industrialization. But I do think it's been an important part.


Russ Roberts: So, before we turn to industrialization, I wanted to ask you about private efforts. Are there private charities that are trying to fund these kind of microfinance or small infusions of capital for an entrepreneur in a very poor setting like we've been talking about?

Chris Blattman: So, I think most of the innovation has been with private organizations. And they've often then been picked up and scaled by governments. And so, whether it's organizations--there's a whole host of organizations that are giving out cattle and livestock along with various skills where there have been randomized control trials in countries. So, I think BRAC is an example of an organization that does this all over the world. The International Rescue Committee is a private organization that's doing this in the Syria region as well as other conflict-affected places. Give Directly is a famous example of an organization trying to give cash in the simplest, cheapest way possible. Which is arguably the best way to make sure that this organization passes some kind of cost-benefit test.

Russ Roberts: At least that people don't starve to death. I think there's a big difference--they're both important--but there's a big difference between keeping someone from death--making sure they are "above subsistence," that they don't fall into tragedy, on the one hand--and growth, which is a much more ambitious and I think more challenging thing. So, if you think about an undeveloped--there's not an easy word to describe; let's call it a barter economy, where people are making small amounts of something and then trading it with their neighbors. Perhaps they are reselling things they purchased, so they are acting as a middleman. That's better than being self-sufficient. We know that as economists. But it doesn't translate into growth. There's no engine there to push that higher. Right? So, if I have access to a supply of matches and you have access to a supply of soap, we might swap; and we might use cash to make that easier so we can trade more effectively. But it's not likely to lead to growth.

Chris Blattman: So, under some circumstances that's true; under others, it's not. And I don't know if we know exactly what kind of a world we're in, in a lot of these places. But one of the things is that if you think that there are actually a lot of opportunities for these really small entrepreneurs to go from selling or manufacturing a few matchbooks to manufacturing more matchbooks, or from producing a small amount of milk to a larger amount of milk--but if you think that there are really fundamental breakdowns in the credit and the insurance markets, which is basically a perfect description of a lot of rural villages, then providing a one-time amount of capital, be it like a cash grant or providing regular cash transfers, which acts like as a form of insurance as well as providing capital, can lead people to grow businesses that they otherwise wouldn't have been able to grow. And maybe take risks that they otherwise would not have sensibly taken. And so you can get growth. But again, even then--when you look at--the thing that stimulated my interest in industrialization in general was, while working on this as a graduate student, I actually spent a little bit of time helping write a report for the World Bank on the Kenyan manufacturing sector. So, I was in Kenya; I was doing other things; and I was roped into this. And the thing that still sticks with me to this day was some statistic--and I should probably remember exactly what this was, but this was something like 14 years ago--was something like 5% of the country's workforce was in manufacturing. But they produced something like half the national income, or the Gross Domestic Product (GDP). And the reason was because these manufacturing firms were just so productive--they added so much value. They combined all of this capital and technology and people's labor to produce so much more value than these farmers. So, you could double the income of every farmer in the country, but because they make very, very little and they are producing very little real income and they are not very productive, that would make them a lot better off. These are the kinds of programs that I work on all the time. But you are not going to have a huge impact on national income. You are not going to transform the economy. So, you can get growth, if you resolve these credit constraints and insurance and you can make people a lot better off. And that's a really important service of aid. But the growth you are going to get is--I don't know that there's many people who think that's going to be very transformative or dramatic.

Russ Roberts: Yeah. Just a couple of comments. The 5% making 50% of GDP is partly a result, probably, of the inability to count some of the economic activity that's taking place in informal structures.

Chris Blattman: True.

Russ Roberts: But still, the point I'm sure is real: We're really talking about Adam Smith's pin factory--that specialization drives productivity; division of labor is limited by the extent of the market. If you don't have a lot of people to sell to, it's not worth it to invest in capital to make more soap or more matches. If you have a bigger market--if you are in a bigger village or in a city--you can obviously change your production process; and eventually you get to industrialization: you get to a factory.


Russ Roberts: The other thing I just want to comment on--I don't have a way to say this very well, and so I'm going to ramble here for a minute or so and see if you can clean it up for me. But, it seems to me we don't, as economists, have very good understanding of the nature of the dynamics of an economy. So, if you think about a village--I'm trying to invoke--I called it a barter economy. It's not really the right word. But people are--they are self-producing. They are not in factories. They are not cooperating in large numbers to produce lots of stuff. Each person makes something. Or resells something. Or adds a couple of things together and makes something. And they trade it. And they buy it. And they buy other stuff from other people and they sell the stuff they can make, whether it's livestock, output, or food from farming or whether it's small, small-scale production, like clothing made by hand. And there's just an inherent limit--and this is really, again I'm coming back to Adam Smith--there's an inherent limit to how much prosperity that kind of situation can produce. You can shower them with money: it's not going to have a big impact--if it's a small group of people interacting. But if it's a larger group of people interacting, there's the potential for what we call economies of scale, or industrialization, or factory productivity--then, yes, there can be barriers of capital that keeps entrepreneurs from creating those institutions, those type of companies and firms. But there just seems to me to be a big difference between a village, which is going to inherently be somewhat poor, and a city, that is going to have the possibility of the kinds of economies of scale that Smith and we are interested in. Does that make sense?

Chris Blattman: So, I think it's true. But at the same time--if you traveled across different countries in Africa or even different countries in Latin America, for argument's sake, which are the places that I know better, and you went from rural area [?] village to village, it's true that you might--you'll never find, like a Singapore there. You'll never find a--you'll very seldom find an enormous amount of, you know, First-World-level production. But here's the thing: You will find that some villages in some places, or even in the same country, are 10 or 20 times as wealthy as other villages. And a lot of that is because they've become better at producing and better at organizing and just more productive in general. And so there is a lot of room for variation. So, you can get a lot of growth. And you can make a lot of strides, in terms of quality of life, and infant mortality, and all of the things that we care about. Things like--just the marginal utility, like the benefits we get a little bit on the margin from those advances are so huge--

Russ Roberts: Yeah, I agree--

Chris Blattman: that you can--so there's a lot of room. I don't want to leave people with the impression that it's all about urbanization, industrialization. That there's a lot of benefits to being able to become more productive in agriculture and small-skill production.

Russ Roberts: Yeah, that's cool--well said. I just--I think there's a discontinuity of sorts there. There's a--I like to think of it as: You put the hundred most talented people, and you can pick them. I'll let you pick who they are. And you put them on an island with no outside exchange possibilities. You can give them all the technology you want. You can give them all the resources. It can be an island full of oil underneath it. And all kinds of metals underneath it. And beautiful, rich soil. But 100 people can't be very wealthy. Now, 100 people trading with, you know, a billion people can be wealthy. So I'm not suggesting that we have to get everybody into the cities. Obviously a well-developed urban sector can help a well-developed rural sector thrive through their interactions. But I just think--when we think about "growth," as opposed to, again, a decent standard of living or a minimal standards above subsistence, there's something a little different going on there. But your point about the variation is huge. Obviously, small villages can be much better off than other small villages if they do things well.

Chris Blattman: And I've been very surprised. Some of the projects I've worked on, and things I've studied--I've often assumed that a lot of these isolated rural villages are really little economies unto themselves and are somewhat cut off, and the markets aren't very integrated across these villages with the city. And I've generally been surprised. I think that there's a lot of exchange going on. And so, that's in some sense a secret. When I say that there's actually a lot of potential in these villages, I think it's in part because there's a lot of potential for these villages to, say, turn to cash crops or animal production or things and then sell into the cities. And so engage in this kind of trade. So they are not 100 people on an island. There's definitely some costs to getting things to market. That is another source of poverty.

Russ Roberts: Transportation, infrastructure, ports.

Chris Blattman: All of these things. Or even just having access to a bicycle or frankly a big challenge, I think in Africa is that there never has been really much in the way of draft animals, because of the disease environment. And so that holds things back, as well. So, as a consequence, there [?], and this is where some of the opportunities lie. There is an ability to trade with the rest of the world. To trade with that billion people, and get some pretty serious gains.


Russ Roberts: So, let's now turn to the other possibility, which is going to be the focus of the paper: the comparison between the informal, self-employment, entrepreneurial environment of small-scale production, versus being an employee. So, we're going to call this a 'sweatshop.' It's--for better or for worse it's become a term to describe relatively low-paying employment in a poor place, in a factory or a larger workspace. What is your--before you did this research, what was your preconception about that kind of opportunity for the poor in very poor countries.

Chris Blattman: Sure. Well, the first thing I should say, sort of by way of disclaimer, is I didn't then, and even to some extent now, not--I'm not much of a factory expert. This has been sort of a new journey for me. And as you know, not only do I spend a lot of my time mostly thinking about poverty alleviation in the poorest places, but I actually--my main thing is studying violence and gangs and rebellion and things of this nature. So, factories have been--something that for better or for worse, I, you know, I'm not sure I had the--I guess I had the idea that a lot of economists and a lot of my colleagues sometimes voiced, which was different than maybe the pessimistic view that some people have about sweatshops. I would occasionally see a factory. And, like I said, I had a little bit of exposure to the Kenyan factories. There were a few firms, a few industrial firms in Uganda. And when those firms had a set of jobs or when they opened up a new production line and, say, they needed 50 people, you'd see a line-up of 300 people for that job. If not more. And so there's this very basic idea of revealed preference: that if people are standing in line waiting for jobs, these jobs must be pretty good. And people are queuing for this employment. And so that was my first impression. And you would even see some sweatshop activist organizations saying things like, 'Well, as bad as these jobs are,' and they have this kind of agenda to improve what Nike is doing, to improve what some other organization is doing in a poor country. They'd say, 'As bad as these jobs are, they are still better than most people's alternatives in the informal sector, this sector where they are self-employed, producing agricultural goods or running these small, crummy little shops.' And so that was my first impression: that as bad as these are, it's better than people's alternatives. And here I am, spending all this time trying to get a program that can help some--you know, poor somebody--go from $2 a day to $4 a day. And maybe I'm missing out on the action. Maybe I need to be spending more time in thinking about how to help that 5% of the population produce 50% of the wealth. Maybe I have to sort of think about how to illustrate that sweatshops are the answer, not cows or cash transfers; and get the international community focusing a little bit more on that.

Russ Roberts: Yeah; I'm reminded of--I think it's from the Cocoanuts, Marx Brothers' movie, where Groucho, who is the manager of a hotel, says, gathers the staff together and says, 'You don't want to be wage slaves, do you?' And they say, 'No.' And he says, 'You know what causes wage slaves? Wages. So, I'm not going to pay you. And that way you won't be a wage slave.' And you echo this in your paper. Your paper is worth reading just for this one line, for me: It said, you write, as the economist Joan Robinson remarked,' and Joan Robinson was a great British economist in the early and middle 20th century--her quote is: 'The misery of being exploited by capitalists is nothing compared to the misery of not being exploited at all.' And this is an issue--you know, it's very--it's really a deep, it seems obvious to most economists that that's true. But I think it's a hard issue for a lot of people to think about and deal with, that your shoes or whatever it is you are wearing or buying is made by people who are very poor and who are paid very little. And it's true that it's better than their alternatives. But it's still not very much. And I think that--that's an issue we are not going to grapple with but I think it has to be mentioned.


Russ Roberts: But the point is that people do line up for these. As you say. And of course they line up for Walmart jobs in the United States. And people complain about that too. And I take the attitude that--I don't know how to solve all the problems of the world at one time, but when I see people desperately trying to do something, I assume they are trying to better themselves, and probably know more about their situation than I do.

Chris Blattman: Right. And I think on top of that I had the sense that, Okay, even if these jobs aren't much better than people's other alternatives, that, you know, a couple of things are going to happen. So, first of all, I thought these jobs were better than the alternatives. And then I thought, over time, this was going to be doubly the case: on the one hand, you would see the manufacturing sector advance over time--one hopes--and the quality of the jobs would improve and the skills they need and the worker commitment and productivity they need. And so you would see wages rising over time, potentially in the sector. So you might send people--the people who are lucky to get these jobs, in that moment the job might not be that much better but it might send them off on this virtuous cycle. And then the second thing: Which is where people like Paul Krugman and Jeff Sachs, two very prominent economists, had come out in the last 10 or 20 years: They'd said things like 'What we need is more sweatshops for Africa.' And their reasoning was not this reasoning that I necessarily thought was the case--that people are lining up because these are better jobs. But their point was that, you know, there's a lot of people who are underemployed in these relatively unproductive things like agriculture and all the other kinds of crummy businesses we've been talking about. And the more that factories grow and suck them up and export to other countries, then the more opportunities there are for farmers to--and the more incentives there are for farmers and these small business people to actually invest in capital and improve their opportunities. Because they are not competing with nearly so many farmers. And the returns to investing in that capital--like getting a farm combine or getting a better store are going to become more attractive. They get to specialize in producing those things, while these other people go off to work in factories and then buy their food and buy their haircuts and buy all of their other things. So there's that double effect. And, you know, we weren't in a position to measure that big macroeconomic effect where you are sucking up all this labor. We weren't going to be able to do that in the study. But we could test this idea of: How do people choose between occupations? And: What are the benefits and risks that come with this different occupations?


Russ Roberts: The last point I want to make before we get to the actual details of the research is that: It's clear that hundreds of millions of people--hundreds of millions of people--in China, in India, have improved their standard of living by leaving agriculture, coming to the cities, and working in factories, producing goods for people around the world. And that's just a reality. That's just phenomenal. Whether we can make that happen in Ethiopia, or Uganda or Kenya is a different question: whether someone can make that happen is a different question. But it's certainly true that industrialization is a, at least on a material basis, which is not the only thing that counts. You can call a business 'crummy'--it may be crummy by our standards. But there may be some dignity involved there for the person and their family that we're not aware of. And I guess what we're talking about. That's a--you can react to that if you want. And why don't you start talking about the study?

Chris Blattman: Well, I guess I'll only say, I'll give some context: it is happening. So, for example, I, when I was working in Ethiopia, when I first started working there in about 2009, what was interesting: First, there were a lot of, like, Ethiopian-owned factories that were expanding operations; and they were starting to export things to Europe. So, there were a lot of companies that were starting to send chives or spinach or, you know, t-shirts, to European markets; and shoes. And these were just growing and growing and growing. In part, in response, I think, to basically like a stable--well, basically, they are close to Europe; there's a lot of, you know, Ethiopia, a lot of people think of Ethiopia as--they remember these, you know, bomb killed off and Ethiopians singing about starving children. And there is a little bit of that: in the periphery of Ethiopia, toward Somalia, toward Sudan. These are lowland areas. They are very hot. They are very dry. And people--there's a lot of drought. And so there's a lot of problems. People--there's a lot of famines, from time to time. But the middle of Ethiopia, the highland area, is incredibly prosperous by comparison. It's green, it's the source of the Nile. There hasn't really ever been any kind of problem with getting people fed. And the levels of state organization and the levels of education and the levels of sort of capabilities of firms and civil society are very, very high. And so there are all these companies. And then not only were these domestic companies screwing up, but you had--what was so interesting to me at the time was you had all of these Indian, Bangladeshi, and Chinese companies--and some American and European companies--starting a toe-hold. So, you'd see this huge, vast area, this industrial park that was empty. And there was one factory. And this factory basically was an Indian entrepreneur, a Chinese entrepreneur. And they had maybe 500 or 1000 employees. And you'd say, 'Why do you have all of this land?' And they'd say, 'Well, we're getting ready. We're trying this out. We think this is a pretty good place to be. There's lots of people. There's a domestic market. We can export to Europe; it's very close by.' Especially at the time oil prices were high, so being closer to Europe was helpful. 'And the wages are going up in China. And the wages are going up in India. Just like the wages went up in Mexico, as these areas got more productive. And so, this is where we are going to plant our foot next, we think. And so we are going to test it out; and if things go well, in a few years you'll see 10,000 or 20,000 people working here.' And there's a lot of that going on. And it's not just going on--you know, the capital is going on in sort of the rural hinterlands, and in this northern city and in that small town, in agriculture and textiles. So it's--you know, you kind of have this feeling of, 'Wow, we are on the cusp of something maybe really big.'

Russ Roberts: And we are talking about Ethiopia. I don't know anything about Ethiopia, or very little. What kind of government do they have? And did it change, to make this kind of investment imaginable? Or is this more--why is it Chinese or Indian? Why isn't it internal? Something's going on there.

Chris Blattman: Right. Well, so, this is happening in a number of countries. So, the short answer--it's difficult to give a short answer on Ethiopia, there. But I'll try. I will say that it's going on in a lot places where there's different kinds of government. So, Ethiopia is relatively authoritarian. It's sort of taking as its model, trying or to some extent Vietnam, where it's going to have some--you know, it's sort of a Communist and Socialist influence, government that came to power and as part of a revolution, as part of a civil war, and is repressive in many ways. But is also very serious about sort of this marriage of private sector and public sector development. And some state influenced development, but one where there is a humongous amount of private investment. But it's also happening in Kenya, where you have a much more democratic regime--with other problems, corruption and things of this nature--but it's happening across a wide number of regimes. It tends to be happening in places where there is really good infrastructure, where there is proximity to Europe. You talk to some of these factory owners and the thing they complain about is not the quality of [?] or the quality of roads--they do complain about the quality of roads and the quality of infrastructure and things. They complain about everything: but this is the thing. It's always a struggle and they are always taking risks. But one of the things they really complain about is the availability of good managers. They need middle managers; they need accountants; they need somebody who can run their factory floor; they need someone who can run the HR (Human Resources) department; they need someone who can handle mergers and acquisitions; they need someone who can help them do due diligence on buying and acquiring new companies; they need all of these skills that we don't usually think about as poverty or development economists. And that's where they feel really tight. They say, 'I can only do so much. I have all this capital; I made all this money,' in real estate or trading or mine money [?] or ill-gotten gains or wherever they got this money. And they only have so much time. So they really need people who can organize. And they need people who can help them execute transactions, whether it's buying a company or helping them build sales contacts and do things overseas. So that's why places like Kenya and Ethiopia I think are really attractive, because these are places that are pretty politically stable--whether they are democratic or not--but they also have a, like, a middle class, however small, and an upper class who are pretty educated, pretty capable people who can do this kind of a thing for them.

Russ Roberts: That's just fascinating. I don't know why it came to my mind, the analogy with the football team: you can have a great offense, a great defense, but if your special teams get all their punts blocked and everything gets thrown back for a touchdown, or everything gets run back for a touchdown, you don't win any games. You've got to have that--there's a certain supportive cast that you don't think about when you think about a factory, because we're only economists and academics: We tend to think about, 'Welp, there's a manager and there's a bunch of workers.' But of course there's that huge level of other folks who help make the thing go. And that's very interesting.

Chris Blattman: I think that's one of the biggest constraints on these companies--it's that, at the end of the day, I think some of the reasons that we saw some of these factories doing well and some doing very poorly is they really struggled. Especially the middle management, the people who would manage human resources and organize the factory floors. I think they just really struggle to get good people. It's also an early stage: because these firms are new there's just a lot of learning going on. So, we're really careful to sort of scope what we're doing. We're talking about an early industrializing society, a place that has--in some ways it's had a long history of manufacturing. Ethiopian firms have been making shoes for 70, 80 years and exporting to Italy. An awful lot of shoes have also been made in Italy; but there's also a decent chance that if you bought a pair of Italian shoes, it was made in Ethiopia, and then somebody in Italy stitched on the tag that said 'Made in Italy.' And there you have it. So there's been a long history of private sector production in Ethiopia, but not on this scale. That's what has been new. And indeed, even in the last 8 or 10 years there have been ups and downs. Not just because of the ups and downs in the global economy. That's affected Ethiopia, Africa less, in part because they haven't been so integrated with the world economy. It's been going up and down with domestic politics, which actually right now, for example, which look really stable are actually not right now. We've actually--some stuff I'm doing in Ethiopia right now we've had to stop, because there's a lot of unrest right now, some popular unrest and opposition with the government. And so this is actually not just crippling research projects, it's crippling a lot of industrial production.


Russ Roberts: So, let's move to the study. What were you trying to figure out here? What was the goal? What were you trying to understand?

Chris Blattman: Well, so, in some sense--it was pretty simple. If we have this belief--if I'm right that this--if I was right that these jobs are better than the alternatives, what we wanted to do is to say, 'Well, let's test that out.' Since you have 300 people lining up for these jobs, why not look at--take the ones who are qualified and instead of taking the first 50 in line who are qualified for the job and hiring them and everybody goes home--which was really what was typically done--why not see if we can find a factory owner who will say, 'Okay, let's find the 150 who are qualified and instead of taking the first 50, we'll flip a coin. And we'll take the 50 people out of those 150 qualified applicants as random.' And we'll follow them over time. And we'll look at what happens to their incomes and their health and their career trajectories over time. I mean, if we have to, we'll follow them for 10 or 20 years. And we'll see: Is it hard to get a job if you don't get this job? Are they really queuing and if they aren't lucky enough to get the job from that queue, do they find another job or are they forced to go back to agriculture? How long do they stay in this job? Are there short term risks? Are there long term risks? And let's sort of get something that both the activists and the boosters can say, 'Yeah, we're curious what the result is.' And so that's what we did. Yeah. Well, I guess you can say--what I did is I had this idea as a graduate student 10 or 12 years ago. And I always thought, 'Well, every time I meet a factory owner, I'm going to feel him out.' And I did. Once in a while I got on a plane to Uganda to work on one of my projects, usually related to poverty or conflict, and I'd maybe sit beside a factory owner. And I'd say, 'Oh, here's this idea that I have.' And they'd usually sort of look at me a little funny, or--they wouldn't leap at the possibility. But also, I was just this person they met on a plane; and I was a graduate student. I probably didn't approach it well. And so it never really materialized.

Russ Roberts: And then?

Chris Blattman: Oh, and well, so, I wanted to pause because I feel like you might want to jump in with something else.

Russ Roberts: You want to jump in with something? What should I have asked?

Chris Blattman: No, no, no, no. I--no, I think that's probably--so what happened is, I was at a conference in London, and there was an Ethiopian businessman who--I used to say he was the Donald Trump of Ethiopia because he was sort of a real estate mogul--

Russ Roberts: Can't say that any more--

Chris Blattman: and that just seemed inappropriate. Yeah. That's no longer the first thing that people associate with Donald Trump any more. But back in 2009, he'd made a lot of his money in real estate. He'd also--actually originally he'd made a lot of his money in investment banking in the United States. He sort of grew up as an expatriate Ethiopian. And he was giving a talk to a group of Development Economists, something called the International Growth Centre. And he was saying--the thing I just told you about where I said what firm owners want, is they want managers; and they want M&A (Mergers and Acquisitions) specialists: he was lecturing this group of economists, saying, 'You guys, you don't quite get it. This is what we--this is the big deal and this is what we need and I just want to help you understand what I think the real problem is.' And so, so I just channel him and many other firm owners I met after that. And I approached him afterwards and said that was terrific and I really enjoyed talking to him; we kept chatting, and I said, 'I've had this idea: you know, I think that [?] firms not only help achieve growth but I they might actually be tools of poverty mitigation. Here's an easy way to answer the question.' And he said, 'That sounds great. Let's do it.' And so, literally, 5 or 6 weeks later we were on the ground in Ethiopia, launching, doing the first randomization. He had a water-bottling plant. And he was adding a new line. I think he had been producing like the small bottles and the 1-liter bottles, and I think he was adding like a 5-liter-bottle manufacturing line and he needed 30 new people. And so, there we had it. We just scrambled and it all came together.

Russ Roberts: So, the idea--the 3 categories, what's important for people to keep in mind--so, you have a group of people you hope are similar, in that they all are qualified to work at the plant. A third of them are going to work at the plant. A third of them--talk about how you divide up the sample. How many people [?] end up being across so many firms.

Chris Blattman: Right. So, well, the first thing I should say because I love this story is, even--this is an example of where things are. There weren't a lot of manufacturing firms. Beverages are like one of the first things that people produce locally--and not for export, for domestic consumption. You know, it's really heavy to send water around; and water is usually around. So bottling water is like Step 1. And this wasn't done--when he went to Ethiopia sort of around 2005, 2006, back, and thought, 'I'm going to become this mogul,' and when he drifted from real estate into this stuff, he would go and he would buy a bottle of water at the store: and it was $3. And it would come from Yemen or Israel or something like that. And he said, 'We are the source of the Nile. And we are importing water from the Middle East. This is patently ridiculous.' And so he started the first water bottling plant. And he did phenomenally well, and then like 6 other water bottling plants grew up. And so at the time he was expanding, he was expanding into new product lines because he was trying to diversify away from this competition, but he's also struggling to maintain his profits, because he was facing a lot more competition. So, he was also trying to open a whole bunch of other firms. Which is where he thought he would get more and more participants. He had visions of opening 6 or 8 or 20 firms. Now, as a matter of fact, he didn't. He ended up deciding that, 'Actually, my profit margins,' gosh, you know, he says, 'I'm actually much better at being a real estate mogul. And of the 40 cranes you see around the city, 20 of them are mine; and maybe I'll focus on 30 of them being mine rather than just--people are just going to copy me if I go into manufacturing. But people aren't copying me on this other much more capital-intensive activity that they can't enter. So, now that I've helped ignite this thing, I'm going to step back.' So we had our first little cohort of 30 people--initially I think it was 30 people. We almost call it our pilot--30 people in this water factory and 30 people who by the flip of the coin didn't get into the job. And we started tracking them. But then we started to have to say: Well, we're not going to have this study of 60 people. That will be a case study: it won't be a statistically precise or generalizable kind of exercise. So, we basically started beating the bush--went around; and there may be, I would say there's probably between 500 and 1000 manufacturers in the country, and we probably talked to at least half of them over the next year or so. And the first surprising thing to me was that, when we brought up this crazy idea of randomizing who they hire, most of them thought, 'Oh, yeah, that sounds interesting; we'd be willing to do that. ' Provided--and this is important--provided they were basically the kind of company that has huge unskilled workforces. So, you're not going to do that with your engineers, if you are like a pretty sophisticated engineering company. But if you are the kind of company--and this is the kind of company that is in a lot of low-income countries and has the most manufacturing employment--if you are doing garments, which is just a lot of cutting and sewing and packing, or if you are bottling water, or if you are spraying and packing and cutting almost an industrial style of production in greenhouses of chives or spinach, or all sorts of other really low-skilled things where people are basically interchangeable--then yeah, whatever: we're already ad hoc. You're going to answer an innocent question. You're also going to collect all this data that's going to tell us where people go. Do they get jobs at your competitors'? Why do they leave so often? And so forth. And so a lot of them were interested in it. But the problem was that not all of them were adding a production line--like, say, 50 people.

Russ Roberts: Right. You're going to hire 3 people.

Chris Blattman: Yeah. You're going to hire 5, or 6 people. It just wasn't worth our time. Even 10 people. It's a pretty big expense. If you have a firm of 150 people, adding 10 people every 6 months is a serious growth. But that wasn't going to work for us. So we needed these firms that were moving ahead in leaps and bounds. And we found maybe 10 or 12, and 6 of them actually happened. This was the other thing I [?]--it was so interesting, first to see certain plans go awry: so, a lot of firms have big plans. But they went awry for interesting reasons. You'd get market changes, and all sorts of other events. But also, you'd see--they'd say, 'Well, as soon as the machine arrives from India--but the government, it's been sitting in the port for 3 months because the government won't let it pass through, for some reason.' One version of the different kinds of sometimes control and repression and ineptitude that you would get, from this government, or really maybe any government. And so there were a lot of frictions that kept them from realizing these plans. And so I think we had in the end, 5 firms; some of them hired big chunks of people multiple times over that first year. So, I think we ended up having, what we call 8 cohorts of people. And so, in total maybe about 300-350 people who got hired by a factory.


Russ Roberts: And the other folks you followed were doing what?

Chris Blattman: So, the other folks--so, initially we thought, 'We'll just follow people--we'll have an equivalent number of people who just then go about their lives. And maybe they try to get a job at a factory elsewhere, but they probably just go back to whatever they were doing previously.' Which is, mostly they were unemployed. Or mostly they were just doing a little bit of odd jobs, and they were searching for odd jobs. Then we thought, 'You know what? Why just compare this factory job to people's miserable alternative? Or potentially miserable alternative? What if we could make their outside option a little bit better? How would they prefer that to a factory?' And so, what we did is we said, 'Well, let's have this other intervention where some of these people who don't get the job will get $300 cash and about 5 days of business planning and consulting with some people who understand how to, who understand this sort of self-employment and this informal sector stuff. And we'll help people think through a really basic plan for spending that $300 however they please.' And so, in a given factory, if there were 300 people lined up, maybe 150 of them met the minimum qualifications--meaning they were between--typically that just meant they were between 18 and maybe 28; they had a minimum education of maybe grade 8 or grade 10. And they were not obviously physically unhealthy. That was the extent of job qualifications. And if you met that--say, the 150 who met that qualification--50 would be offered the job; 50 would be offered this, what we call this 'entrepreneurship intervention,' which was the grant in this basic training. And then 50 would not receive either and would then go about their business, whatever that would be.

Russ Roberts: That's the control group.

Chris Blattman: Yeah, that's what we call the pure control group.

Russ Roberts: So, what did you find? And over what time period? I think it's the first year, right?

Chris Blattman: Right. So, we sort of followed people a little bit over, qualitatively and a little bit quantitatively over the first couple months, with some of the factory data and some phone calls. But then we did a big survey after a year, where we collected really detailed information about--from them and from their family. A lot of these were young women. So, the first thing I should say, I think about 80% of them are women who are in their early 20s. They are generally unmarried. And this is often their first job in the formal sector. And they've been, maybe like--a lot of them are living at home and they are, they've been mostly unemployed for the last month; and to the extent they are unemployed, they are doing informal work. So, they've been working in a little shop or they've been helping on the family farm or helping in the family shop or doing some petty trading of their own. And so they don't have a--they don't look like they are really that productive or have that many good options. And so you can imagine, this is--so they are sort of looking forward to this factory job. They know it's not particularly good. They know it's not particularly well-paid. The wages aren't great. But it's going to give them 40 or 50 hours of work a week, at a low wage; that's going to give them some stability. So, the big thing is not just the rise in income--this is what we were thinking at the beginning. The big thing is not just the rise in income. Even if that rise in income is not high, it's just an end to the volatility and the uncertainty: that they are no longer not going to know what they are doing the next week. And if they do find something good in the informal sector, they are often doing it for, say, 30 hours a week. But that might only last for a week or two. And then they never know when it's going to end. So, it can mean [?] unemployment spells. So that's at the outset we're thinking is the big impact. And then, basically, everything I think I know about the world turns out to be wrong.

Russ Roberts: Hey, hey. Welcome to my world.

Chris Blattman: Yeah.

Russ Roberts: The world of humility.

Chris Blattman: Yeah. And I, you know, we can come back to why I was wrong. I think we--you know--stuff that had been working in Ethiopia for 20 years, it's like one of the country's most foremost professions, foremost like agricultural and development economists. So, we were fooled.

Russ Roberts: Before you say, 'fooled,' before you go into that, let's make it clear, again, what you were expecting. You were expecting that, for the factory--let's call them the 'factory group,' the 'control group,' and the 'entrepreneurial group--so the factory group gets the job. The entrepreneurial group gets $300 bucks and a little bit of training. The other people we just say, 'Well, we don't know what will happen to them; we'll watch them.' So, in those three groups, your ex ante thought, your before-the-fact thought, was that the people who get the factory job are now going to have a more stable--it's not a great life, but it's more stable. At the end of the year, they are going to look better, financially--certainly than the control group--because they've gotten the job. And they're going to be happier. They are going to have more stability, more regular employment. Is that a good summary?

Chris Blattman: I guess the--and so the intuition that I had is, even if they are not making that much more money, who wants to be an entrepreneur? I don't want to be an entrepreneur. I don't want to not know where my income is coming from a few weeks down the road. And so just the stability of this work would be appealing. Um, and so, so there were lots of grounds to think that--so I didn't know precisely what the advantage of this job would be. I didn't know if it would send them off on oversized[?], on like rising, rising wages, and so their friends competed for you, because you had this scarce experience. I didn't know if the firms were going to pay you so much better than you could get in the informal sector. Or I didn't know if it was going to be the same income but just more stable. And so, anybody who is risk-averse--this is like Econ 101--I'd rather have $10 with certainty rather than a lottery where I could either get $5 or $15 dollars by flipping a coin. And so, if we think most people are risk averse--especially when you are poor, because getting a bad shock when you are poor means--

Russ Roberts: death--

Chris Blattman: means terrible things. Well, death--for these guys, not death. These are, if you have a grade-8[?] education in Ethiopia and you have a family who can support you and so on, there are some options in the end. It's sort of like living at home and not having anything to do and not being able to anything to contribute to the family, not having any spending money; and maybe having a harder time finding a husband or a wife. And maybe also bad things happening in the household. Like, maybe you are contributing to your younger brother going to private school or something. So, bad things. But these people are not on the margins of death. This isn't who sweatshops are hiring. At least in this [?] case.


Russ Roberts: Okay. So, what did you find?

Chris Blattman: So, people really hated these jobs. And they tended to quit really quickly. At least, some, a lot of them. And they quit partly because it was really hard work for kind of a crummy wage. And it turned out that their outside options in the informal sector, even if we didn't give them a grant, their outside options in the informal sector were actually much better than they looked like at baseline. For someone starting out in the labor market like these young women, their outside options actually look pretty good over time, compared to the factory. And the ones for whom that wasn't true tended to stay in the factory. So, some people tend to stay, but very few. So, by the end of the year, maybe. First of all, 10% didn't show up on the first day. Which was our first clue, really: What's going on? Why aren't they showing up? And we tried to find out. And over time we realized that something was going on. And then after a month, like, people were quitting all the time, we are like, 'Why are they quitting? What's going--' and we talked to them, and we tried to figure it out. And then, by the end of the year maybe 2/3rds had not just quit that factory, but they'd left the industrial sector altogether. And they were doing not bad in the informal sector. They, they, and then--but the amazing thing, I think, for us--maybe the most important thing--is that: If you--and the thing that we want to go back to understand better--if you got, if you were offered this factory job, even though you quit, even though most of you quit within a month or two, if you got this factory job, a year later the rate at which you were reporting some kind of serious health problem had doubled.

Russ Roberts: Hmmm.

Chris Blattman: Just from getting the offer. Just from working there for like 20 weeks. You had, maybe, something like, so the people who in the control--at the beginning, everybody was healthy. So there were no reported health problems. And at the end of the year, the people in the control group, [?] 4%--on the particular day we interviewed them, 4% were reporting various kind of health problems. And that could have been a cold; or it could have been they'd lost a limb; or it could have been chronic back pain. Often these were moderate--I wouldn't say they were serious conditions. I think they were like moderate to discomfort to things that would inhibit your enjoyment of daily life, but not cripple you. So, 4%. And in the group that was offered the factory job, it was 8%. And so that's a big difference.

Russ Roberts: Yeah. I have to say--a piece of folk wisdom I've always carried around--an economist told it to me so it must be true: that, if you change jobs, get married, get divorced, move--any of those three things, you are much more likely to get sick. If you do all 3--so, if you get divorced and move to a new city to restart your life, and change your occupation, say--you are not just transferring--you are really much more likely to get sick. So, you know, some of that--

Chris Blattman: Why is that?

Russ Roberts: I don't know.

Chris Blattman: I haven't heard that before.

Russ Roberts: I don't know. But let me tell you, every time I see a sick person I confirm my bias, because I say, 'Oh, Oh, they just moved here. Or they'--I always find some--but I think the casual intuition--it may not be true, but the casual intuition is that all those things are stressful. Right?

Chris Blattman: Right.

Russ Roberts: Change is stressful for most people. And so their immune system is compromised to some extent. And so, it's not--you don't lose a limb if you get married or divorced or move. But you are much more likely to have the flu, to be run down; you don't sleep as well. I can imagine--it may not be again--I hope some listener out there can confirm this is a true statement. Someone can give me the study that my economist-friend quoted to me that I somehow took as true, even though I probably wouldn't if I had looked at it more carefully. But--

Chris Blattman: So, that could be true here. I don't think that's what's going on, here in Ethiopia for a couple of reasons. One is: We basically, we ask a lot of people about their levels of both financial anxiety and also symptoms of depression anxiety. So it's also supported. But we don't actually see self-reported higher stress levels or anxiety levels in either group. Partly I think, being in this--I actually thought that the group, whatever stress change brings, I actually thought that having the stable income was going to reduce your anxiety. And that being an entrepreneur would raise your anxiety. That was my--we actually saw the opposite.

Russ Roberts: Well, you'd think that. But that's because you don't know anything about the entrepreneur or factory life in Ethiopia, as it turns out. Right?

Chris Blattman: Well, yeah. What did we learn? Like, so, people complain--so first of all, people complain about kidney problems. So, I was like--well, what is a kidney problem? And so we looked into this. And I think kidney problems is one way, one manifestation of sort of back pain. So, basically any kind of sort of joint, back pain in that area can be colloquially referred to as kind of kidney problems. So, you can think of as kind of back pain repetitive stress problems. That was--so, that's not surprising, from physical labor, repetitive physical labor. But the other thing that is--people kept talking about--so, there's a lot of chemicals. So, this is--first thing I want to say is you go into these plants: There's safety plans on the wall. There's usually a clinic on site. They are well-lit places. They are clean. It's not the best work environment in the world--

Russ Roberts: But it's not frightening.

Chris Blattman: Yeah. It's not frightening. It's no worse--I've worked in a kitchen, and I've worked in a warehouse; and things of this nature in the past. And on the surface it looked no worse than any of those kinds of environments. And, but it turned out--I think there was just a lot of problems either with chemicals or with particulate matter. So, with particulate matter: If you are in a giant room cutting cloth all day, there's lots of little particulate matter. Like, lots of microscopic and not so microscopic bits of stuff floating around that gets in your lungs. And then you have respiratory problems. Or, the other thing--chemically, there's a lot of cleaning agents in these commercial pharmacies, factories for certain industrial-style production, for spinach or chives--there's a lot of spraying going on of chemicals. And there's a lot of chemicals involved in the plants. And people are wearing some basic productive gear. But they even have a clinic on site that's monitoring people's toxicity in their blood levels. And if your toxicity in your blood levels go above a certain level, on your, like, bi-monthly test, then they rotate you off into like the packing area or something. And then, 3 months later you can go back to the spraying. [?]


Russ Roberts: This reminds me of when I moved to Los Angeles in 1987. And after about a month there, I realized I had a sort throat. And it didn't go away. And I eventually decided--I don't know if it was accurate or not--it was because of the air. Which, in those days--you couldn't see the mountains in Los Angeles from the city. This is very rare--now I've gone back and things have changed. They've put more restrictions on car pollution--automobile emissions--and you can see the mountains on most days. And it's really a beautiful thing. But I realized--you know, I think it was just my body--doesn't like whatever is in the air, here. And after a while it went away. And I thought: Is that a good thing or a bad thing? Is that good? It's probably not good, because I'm now not even aware that I'm taking in some toxins. So, some of this, of course, is--the change in environment, again. But part of it is also, some of that environment is probably not so good for you.

Chris Blattman: And some people would say, 'The biggest problem I face--one of the reasons I left was I was fainting a lot from the chemical smells.' So, that--this is partly what we're looking into right now as we go back after--we're trying to go back after 4 years. We'll see if that happens because of this political instability. But, it's one thing to sort of get a scratchy throat or inhale particulate matter--which is pretty serious. You know, you can see reports from--New Delhi today is kind of the hot spot or the point where we hear about people developing really terrible health problems. Just as a result of the smog. But now you're fainting from fumes a few times a week?

Russ Roberts: Yeah. Doesn't seem like a good thing.

Chris Blattman: What's--no. It's not a good thing. And, you know, this isn't limited to Ethiopian factories. There's a lot of chemicals in a lot of factories, and in the developed world as well. And most of the time there's protections. But I don't know if they are adequate or not.

Russ Roberts: But the bottom line is that--and we should mention--this is about a thousand people, it's not a hundred, right, over all? Right? It's a pretty big sample.

Chris Blattman: Right. The thing involved is that--but it's also just 5 firms.

Russ Roberts: Correct.

Chris Blattman: So we've had to be really careful. If we could have done this with 100 firms, in 20 different countries, that would be a better study.

Russ Roberts: But, just to summarize: At the end of the year--I want to make sure I get the numbers right--what proportion were still working in the factory that they had started with the year before?

Chris Blattman: Let's see. So, I would say, out of--so let's say there were 300 people who got offered the factory job. It was a little bit more. But 300 is a nice round number. Roughly 100 were in any factory job by the end of the year, of the ones who had been offered the initial factory jobs. So, that meant 200 of them had started, worked for 10 or 20 weeks, and quit. And so, about a third were still around. Now, how many of those were in the actual exact factory job? Most, but not all. People were moving around. And this was the other interesting thing. So, that, in the control group--the group who didn't get a factory job offer, it turns out that 20% of them--so out of that 300, like 60 found a job in a factory, somewhere else. Usually not the study factory. Usually it was some other factory in the area. And so, in the end, what we did, we basically raised the probability you were still in a factory job by about 12 percentage points. Which is substantial, but it is not huge. It turns out that being offered a factory job wasn't so transformational because these other options came along. What we're doing by--after a year, maybe the difference--the way you can think about the treatment and control group is, often they are just slightly to be in a factory job. But they sort of had more exposure to factory jobs. So, if I were remember the numbers right, the average person who was offered the job had spent an average of 20 weeks in a factory that year--because of all these quit rates. And the average person in the control group had only spent 10 weeks in a factory. And so when I talked to--that's what makes these health impacts so amazing to me. Which is that, this health impact is coming just from an added 10 weeks of exposure to a factory. So, basically, like, for every month you are in a factory, you are facing more than a percentage point increase in these sort of self-reported health problems.

Russ Roberts: Well, I would think if you told the factory owners that, which I would think you did--or will[?]--they're going to be very interested in that. I assume.

Chris Blattman: Yeah. And these are--one of the reasons the factory owners were interested in the study was they were genuinely concerned about these issues. These were not sort of the movie villain capitalist types. These are people who genuinely wanted to know if they people's health was--they had these protections because they, and they thought they might be adequate.

Russ Roberts: Well, it's hard to be a good movie villain, capitalist, raking in the dough, if your workers are quitting on you because they are fainting.

Chris Blattman: Yeah--

Russ Roberts: So, that's pretty practical.

Chris Blattman: Well, they were--here's the thing. Like, it turned out, they were aware that there were turnovers. They would always complain about turnover. We didn't know the scale and the scope of it until we started following this panel of people. And they didn't know why people--people would just quit, eventually. And they would fill the ranks. And the information wasn't always moving up the chain to the very top about what was going on and why and what the issues were. And this is this management problem that I mentioned. This is a real issue. But the firms are constantly experimenting with ways to try to get workers to stay. So, they are--they started the clinics before we ever came along, for example. They were offering--they were often offering lunches. They were providing--like, a bus was going around the neighborhood. Most people lived locally. There weren't migrants coming from the countryside. These were people mostly who live locally. But they are helping them with the commute. And though the firm is doing all this helping, the one thing they are not doing is raising wages. Which is interesting. And, there's lots of reasons for that. I mean the obvious one is that it's very expensive to raise wages--

Russ Roberts: And they are still getting workers, even though they don't stay very long.

Chris Blattman: And people are still showing up.

Russ Roberts: And turnover is not--

Chris Blattman: and their production product--

Russ Roberts: And turnover is not so costly that they have high training costs.

Chris Blattman: Yeah. Exactly. They were certainly turning a profit. I mean, the mystery is whether or not, if they could get their act together and have people say maybe they could be more profitable. We could maybe come back to that later. The--I would say the other reason--a lot of these firms had cash-flow problems. Just like these individual entrepreneurs have cash-flow problems. Just paying the wage bill was really hard every month. They are not getting paid all the time; their cash is coming in lumpy. And the wage bill is one of their big cash outlays every month. So it's not just that it wasn't profitable, or not. They were like fundamentally constrained. So, raising the wages would actually exacerbate this cash flow problem that they often had.


Russ Roberts: So, we're going low on time. We haven't talked at all about the third, or 300 or so people or so who got the $300. So, how did they do relative to the people who got the factory job, or who got nothing--the control group?

Chris Blattman: Right. So, basically, they ended up not--they ended up starting small businesses. So, they take about a third of the money and save it, and a third of the money and spend it on this or that. And a third of the money would get invested in some kind of small business. They might buy some charcoal or manufacture charcoal; they might get some livestock or they might start a little shop. Those are very common things that they might do with about $100 bucks. And then they'd go from--they'd basically increase their earnings by a third. So, I think they were earning--like, if I remember right--they were maybe $1 a day; and so they went to earning like $1.33. Those aren't exact numbers. But those are--yeah, that's a big increase. You know, it's not life changing.

Russ Roberts: Pretty close.

Chris Blattman: Yeah, it's pretty close. And they were happier. And they had less financial anxiety.

Russ Roberts: Happier and less anxiety relative to before? Or relative to other groups?

Chris Blattman: Relative to the two other groups. And they weren't necessarily working more hours. They were working a little bit more. Basically, if you didn't--if you worked in the factory, it turns out--I guess the simple way to think about it is: If you weren't in the factory, three out of four weeks you could find some other kind of work. And when you had work, you could probably get 30 or 40 hours of work. But at any moment, you might not have a week of work. And that may be, you work 3 out of every 4 months, or 3 out of every 4 days, or 3 out of every 4 weeks. But you'd have some volatility there, in how much work was available. That didn't change a whole lot with the people who got the grant. It was still pretty volatile. But first of all, it was a lot less volatile than I thought. So, that's a lot more work than I thought people could have gotten, outside of the factory. Certainly it was a lot compared to what they had in the months before the intervention. So, their counterfactual life was actually not so bad. But then basically when they were doing work, because they had this capital, they were more productive. Their profits were higher. And so if they were working more hours, they were getting more cash. And that's where that extra $0.33 on the dollar was coming from.


Russ Roberts: So, we're out of time. I just--I found this extremely interesting, every aspect of it, our general conversation, the specifics of this experience. One lesson you could say I as the outside observer has learned from your results is that Chris Blattman learned more about himself than he learned about the workers of Ethiopia. That--the workers, poor people find ways to make do. Some things are a little better than others. Certainly, you get $300 bucks is really great when you are poor. And it helps a little bit, and makes a difference. And if working in a factory for some people was worth it, others maybe most of them, it wasn't: and they move around. They try different things. And it sounds like it's a little more dynamic and alive than I might have thought. Maybe than you might have thought. So, why don't you finish by telling us what you think you've learned and where you think you might go next with this kind of work? It's obviously not--you said, there's only 5 factories, a thousand people. It's not the decisive study of sweatshops. It's just provocative and interesting.

Chris Blattman: Yeah. And I do want to stress that. I guess I'd say--I do want to stress that, you know, I greatly hesitate to go beyond 5 firms in Ethiopia. So, who knows? Who knows what's going on in the wider world? What you said, 'And Chris, have you learned something about himself?' It's true. So, field experiments--all field experiments have turned out to be like this, for me. On some level they are like this intellectually dead thing, that sometimes I get very frustrated, that I work on a lot. Because it's not that--you run a very simple regression and you randomize the beginning. And then there's a huge amount of logistically and--there's a lot to think about. And it can be an intellectual enterprise. But on some level it doesn't feel very sophisticated. But then actually on the other level it forces you to interact in the world. And being in the world, and to follow people over time, and to get into--and also, I have to manage--the users learn so much about how the world works, and gets us, it gets me out of my office. And it gets me spending days and weeks understanding how factories work and talking to floor workers and factory owners. And then it turns out that some assumptions that I have turn out to be, maybe not just false but sometimes the opposite turns out to be true. And I actually think--this is one of the great contributions of field experiments to the field of economics right now--is that, it's not the fact that--who cares about the point estimate? I mean, I care about the point estimate; I care about that we've gotten some finding that is important or counterintuitive. But I think it's just getting a huge chunk of people out there interacting in the world, and figuring out how things work or don't work. And it's a not-uncommon sensory important, because so many things, whether it's getting to the point--now I have this sort of stylized fact in my head about how important management is to an organization. Or I have these ideas about how actually, I think that one of the things these firms actually lack is good human resource management. And to, say, we're able to improve their human resource management, I think some of these problems and some of these turnover and some of the profitability of these factories would all be improved. That's actually not where I'm going to go with my research. This has been a nice little side-route for me, and I hope many people pick up this agenda and run with it. I'm still going to keep working on, you know, youth gangs and violence and things; and poverty alleviation in places with sort of gangs and violence. But, I'm--you know, my view of what it means to develop and what the barriers are, where growth is going to come from, is forever changed.

Russ Roberts: Do you need to start an MBA (Masters of Business Administration) program? No, I'm just kidding.