Ideas Untrapped
Ideas Untrapped
PRODUCTIVITY, EXPORTING, AND DEVELOPMENT
0:00
-48:52

PRODUCTIVITY, EXPORTING, AND DEVELOPMENT

A Conversation with Eric Verhoogen

We often speak of economic development as a phenomenon of sovereign national countries, but the process by which that happens is through what happens at individual firms in the economy. The decisions by firms to upgrade their products (services), export, and adopt new technology are the most important determinants of economic development. The incentives and conditions that shape these decisions are the subjects of my conversation with my guest on this episode. Eric Verhoogen is a professor of economics at Columbia University school of international and public affairs. He is one of the leading thinkers and researchers on industrial development.

TRANSCRIPT (edited slightly for context and clarity)

Tobi;

Usually, in the development literature, I know things have changed quite a bit in the last few years. But there is a lot of emphasis on cross-country comparisons and looking at aggregate data, and a lot less focus, at least as represented in the popular media on firms. And we know that, really, the drivers of growth and employment and the source of prosperity usually are the firms. The firms in an economy, firms are the ones creating jobs, they are the ones investing in technology, and doing innovation. So firms are really important.

One of the things you often hear a lot is that one of the reasons poor countries are poor is that the firms are not productive enough. So that's sort of my first question to you, how exactly do we define and also measure productivity, you know, for us to be able to distinguish why firms in the developed countries are more productive than the lower income countries?

Eric;

Yeah, this is a big important question. So I agree, in principle, that firm productivity is very key. So countries that are going to be doing well are countries that are populated by firms that are being very innovative, and their productivity is rising, they're learning how to do new stuff, they're producing new products, etc. And so there's a reason why people are very focused on this conversation about firm productivity. The sort of, I would say, dirty secret of economics is that it's very hard to measure productivity well, right? And so the productivity measures we have, I think, are very noisy, and most likely fairly biased. But basically, the way you estimate productivity is you run a regression of like sales on inputs, okay, so on how much you're spending on labour and how much you're spending on materials, and then the part that's left over, we call that productivity. So it's like unexplained sales, you know, sales that can't be explained by the fact that you're just purchasing inputs and purchasing workers.

But that is actually a very noisy measure of productivity. And so I've been working on a review paper, and a separate research paper kind of pointing out some of the issues with productivity estimation. So in principle, it's exactly what we want to know; in practice, it's very hard to measure. So one argument I was making in that paper is we should go to things that we can actually directly observe. Okay, so sometimes like technology adoption, we can often directly observe whether the firm has adopted this particular new technology, or if they're producing new products, we can directly observe that. Sometimes we can observe the quality of products that can be measured. Now, the standard datasets that we have typically don't have those things. It is possible now, in many countries, to follow manufacturing firms or even other sorts of firms, [to] follow them over time, which is great, at a micro level. But those that have the technology, they don't have quality, they do it now increasingly have like what products they're producing, often they don't have the product people are producing and so it's harder, you have to go out and you have to talk to people, you have to access new sorts of data, there's a lot more work, a lot more shoe leather - we'd say you wear out your shoe is going to talk to people trying to get access to other datasets in order to have these measures that you can observe directly.

But I think there's a big advantage to that. Just in terms of measurement. Like, can we measure these things, and record that technology quality and product innovation together? I'm not sure that's answering your question. But, you know, I mean, I totally agree that what firms are doing, that's crucial, right? So the big macro question is, why are some countries rich and some countries poor and how can we make poor ones richer? That's the big question. I think that's kind of too big to be able to say much about. The much more concrete thing, which we need to be focusing on is how can you make firms in countries more innovative and productive. That's the absolutely right question. But that's just hard. There are challenges and research about, you know, how you actually analyze that, and it has to do with these issues of measurement.

Tobi;

I understand the measurement problem, and of course, TFP, the residual, and so many things like that. But practically, I want to ask you, what can you say, maybe if you have a handy checklist or something? what distinguishes firms in rich countries from firms in poorer nations?

Eric;

Yeah. So let me say what I don't think first, and then I'll say what I think. So it's become increasingly common to say that firms in poor countries are just poorly managed. The firms in rich countries have better management, and the firms in poor countries have poor management, right? And partly that's coming from the influential paper by Nicholas - Nick Bloom - and others, and David McKenzie and John Roberts. You know, they had consultants go to some factories in India. In some they camped out for four months, some they were there for only one month, and the ones where they camped out for four months ended up doing better, right? And they say that that's because these consultants improve the management of the firms and management matters. And I do agree that sometimes these management practices matter, but I don't think... sort of, one kind of implication of that line of work is somehow, like, the firms in a developing country are just making mistakes. They haven't gone to business school in the United States, and so, therefore, they don't know what they're doing. And I think that's incorrect. I think that's incorrect. I think the problem is, firms in developing countries face many, many constraints that firms in rich countries don't face. Right. So often, for instance, gaining access to high-quality inputs can be very difficult, right? That you just don't have the supply chains domestically producing high-quality inputs. Often skilled workers are very expensive relative to unskilled workers, and even relative to the price that you might pay in rich countries. Having skilled workers, including skilled managers, is very expensive. In addition, you have all these frictions on trying to get your goods to market or trying to, you know, trying to access export markets, often there are, you know, their costs involved in that.

In addition, being productive requires know-how and often firms lack that know-how, right and so the question is, how do you get that know-how, you know, like, the distinction I'm trying to make is, it's not that they're making mistakes, it's just that they're doing the best they can given know-how they have, and given the constraints that they face. And so in that sense, I would sort of point to those constraints, right, those constraints both in know-how and both in the input and output markets, rather than just failure of management. So now, one of the constraints I should say, actually, so is often, you know, legal and regulatory institutions are much weaker in many countries. It is true in Nigeria, and it's true in many places, right? And so then that does create a complicating factor also when you're trying to do business with somebody, but you don't have the legal recourse of going to court to enforce whatever contract you write down. And so that creates friction. So then you have to do things differently in part because of that. And so you're likely to be much more based on, like, networks of various types. It might be ethnic networks, or it might be people that you know or that you have long-term relationships with. But then that means you can't necessarily just find the best supplier of something, you actually have to find someone that you trust, and that can complicate your life, basically, if you're trying to do business and develop.

Tobi;

So one thing I want us to discuss is the issue of firm upgrading. I mean, one of the things that have helped me in reading your work and taking this firm-level view of development is that, okay, on the one hand, if you look at a country like Korea, we can say the average income, the income per capita for Korea 40 years ago versus now and compare with say Nigeria, but also we can look at Korean firms 40 years ago versus where they are today. Today, Korea have global firms that are at the very frontier of technology. Companies like Samsung are innovating and making chips and making electronics and making smartphones and you compare with firms in Nigeria who have not been able to upgrade their products over that same period. And now what I want to ask you is how important is a firm's ability to upgrade productivity. I take your point on the measurement but controlling for that, how important is a firm's ability to upgrade its output? Its products on its productivity?

Eric;

No, no, I think upgrading is crucial. And upgrading in various ways, you know, more specifically technology, producing higher quality products, producing new products, new innovative products, you know, you might be reducing costs, right, all those things. I do think that's crucial. I think that's crucial to the development process. I mean, much of the conversation in development economics has often been not about firms. It's about, you know, social policy, or it's about education. It's about human capital accumulation. But I'm with you on that, the firm-level upgrading is totally crucial. You know, the question of like, why isn't it happening? Or how could you promote upgrading? That's a very difficult question. There are lots of papers that are sort of speaking to that subject. And this review article I was trying to write was basically all about that. So Alexander Gerschenkron way back in 1962, is a historian writing about late industrialization had this phrase, not very politically correct phrase, but basically, advantages of backwardness. So in principle, if you're a developing country, you should benefit from the fact that technologies have been developed in rich countries, and you should be able to go and adopt them off the shelf. But for some reason, that's difficult, right? It's hard to do. Partly, it's difficult because of, you know, know-how reasons. So I'd say that often, much of the knowledge that you need in order to implement these technologies is not written down anywhere, it's not really in the manual, right? You have to kind of talk to people who know it, rather than just downloading the instruction sheet. That's one reason.

It's also true that many times, machines or processes, actually, may be context specific. So like the picker machine, in a very humid environment, they operate differently than in a non-humid environment. And so, you know, there are things that you need to learn. So I'd say that kind of like gaining the know-how is an important kind of constraint on upgrading. And partly that happens through networks or through... there's a ... Juan Carlos Hallak, who's in Buenos Aires (who would be a good person for you to have on your show, actually, I think that he'd be an interesting person to interview) as a very interesting paper. It's basically on like Argentina, looking at industries that have done well, they've been able to upgrade essentially and looking at what was it about them that made it possible, and especially the leading firms, what were the leading firms doing? And what we're basically finding is that often the key person in the firm, like, had been embedded in markets in rich countries, maybe in the US or in Europe or someplace. So they understood very much how those markets work and what consumers want. So one was like making boats, sailboats, or motorboats right, that was one of the interesting things he focused on. But knowing sort of what the people who are buying those boats really want to see in their boats ended up being important for what they're doing. And so that's an important part of the know-how. It's like, yeah, understanding the customer understanding also how if there are firms that are producing there, understand what the competition is. And so that's know-how that often has to be sort of gained in person rather than, you know, just reading a book or talking to somebody on the phone. And so when I think about... I don't know Nigeria very well, but when I imagine, you know, Nigerian producers, I think, partly what might be holding back is, sort of, maybe not having the understanding of what are the requirements, what are the expectations of consumers in the export markets, right, in the rich countries that they may be selling to?

We've talked about the barrier, we can talk about the driver of upgrading. So then, like, gaining know-how would be a driver. So that's one. I think, and part of a lot of my work has been about quality upgrading, you know, producing higher quality. And I think that's in part driven by who you're selling to, right? So Mexican firms, you know, if they're selling to Mexican consumers, they produce different products than if they're selling to us consumers, which is their main export market, right? And so, you know, and if you're selling to Mexican consumers who have a certain willingness to pay for quality, we would say, right, they have a certain level of, you know, demand for certain characteristics, the optimal thing to do is keep producing that kind of lower quality stuff, right, rather than producing the higher quality. So I had this famous example of a big Volkswagen factory in Puebla, Mexico, which for a long time, it stopped in 2003, but for a long time been producing the old beetle. The old beetle that had first been produced in 1940, or certainly the 1950s. But for a long time, in the Mexican market, that was the main car that people were buying, and they were happy with that because it was cheaper. It was like, you know, it's very reliable. But that same factory started producing the New Beetle, basically, for the US market, right, for the US and European market, which is much more sophisticated, but also much more expensive. So it depends a little bit on which market you're selling into and whether you're going to upgrade or not. And so accessing export market can, in some sense, like pull the upgrading process, you know, once they access these export markets, they'll start producing higher quality stuff for these consumers. And that I think, actually, generates some learning, and I can talk about one paper that shows that a bit. But it seems to be that by gaining access to markets and producing high quality, then firms learn how to do stuff better. And so that can be an important driver of upgrading.

And conversely, not having access to export markets or having a hard time breaking into export [markets] can be a reason why firms failed to upgrade. Let me tell you about one paper that, you know, demand effects can drive learning.

Tobi;

Yeah. Go ahead.

Eric;

Okay. It's a paper by David Atkin, Amit Khandelwal and Adam Osman. It's in Egypt. Okay, it's an RCT experiment, a randomized controlled trial. And it's among rug producers, producing rugs. What they did is they randomly allocated initial export contracts, right? So if they work with an intermediary, like a buyer of rugs, you know, among several hundred rug producers, they say, Okay, some guys are gonna get an initial contract, and some guys not. And so that was a way, this is a way of investigating basically what's the effect of exporting on the decisions and in a very clean way, and they found a couple of things. So one is those guys who had the export contracts and started producing higher quality stuff. So that's sort of consistent with my Volkswagen story, too, right? So increasingly, export markets produce higher quality and they did lots of measures of, you know, how thickly packed the rugs were and how straight the edges were - the very dimensions of quality of rugs. That was one thing. And then the other thing that they found which is very interesting is that you know, these weavers of rugs got to be better at producing rugs, basically. So then, when they took them into a laboratory, and they say, okay, produce this identical rug to a whole bunch of producers, both in their treatment group, and in their control group to produce this identical rug, and they found that the guys who had gotten the export contracts were better at producing that rug, they produce sort of higher quality rugs than the other guys. This suggests that demand can drive upgrading, right, in the sense that it induces firms to produce higher quality, but there's also learning involved in that process. These Egyptian rug producers became more productive as a result of having access to these export countries.

Tobi;

Yeah, I mean, listening to you, I can think of a few things that click in place. When I look at, say, a country like Nigeria, I think about the way the central bank has been running the exchange rate policy, which is messing seriously with the way firms actually source inputs. Some firms actually don't have access to the foreign exchange quota to actually source quality inputs. I mean, from manufacturing firms to agribusinesses who want to buy high-quality seeds overseas, I see how that can be a constraint. But two things I want to get at. Also, if you look at Nigeria whose industrial policy is really about domestic self-sufficiency, you could see that there isn't really an incentive for upgrading, and therein lies my question. If we talk about upgrading and how important it is, even though it's not really discussed as it should, what role do you think industrial or state-directed policies can play in this? Why because industrial policy is back in fashion, you know, it's being discussed everywhere... but usually, at least in my experience and in my opinion, what most scholars and advocates are focused on are [things like] state investments, you know, how the state can put money in one sector or the other. There really isn't so much focus on this sort of micro-level detail and what happens at firms, which your work is about. So for practical purposes, do you see industrial policy as something that can really, really, play a role and incentivize domestic firms to upgrade? For example, something like export quotas, you know, for firms?

Eric;

I mean, in terms of your question, do I think industrial policy can be helpful? I do. I do think that industrial policy can be helpful. Basically, I think that learning generates spillovers that firms themselves can't fully capture. And so I think there is a role for government to promote learning, basically, in a way. To subsidise learning such that - the socially optimal, or - the best sort of amount of investment in learning for society is more than individual firms to do on their own. And so there's a role for industrial policy. But I agree that it's got to be smart industrial policy, it's not just any old industrial policy. And so many countries have this idea...it's a little bit of nostalgia for import substitute industrialization, or it's very much like inwardly focused industrial policy. We're going to try and guarantee a domestic market for our producers, something like that, right? I'm not a fan. I'm not a fan of imports substitute industrialization or these very inward-focused strategies because then you get to the point where there's just not a lot of pressure on domestic firms to be more productive. They become kind of in a comfortable situation where they have kind of protected markets, not very competitive, they have a lot of market power in that market, and so that is a recipe for stagnation over the long term.

So I think the crucial thing is that the targets for industrial policy be export-oriented, you know, outwardly oriented. You want your firms to be successful in world markets, right? I think that should be the key, rather than domestic self-sufficiency. Or rather than just the government investing in well, okay, so I don't have a problem with the government investing in infrastructure, investing in things as long as the aim is always ''what's going to facilitate our firms being successful in world markets'', right, I think that's a good target. Because those world markets are competitive [and] for firms to be able to be successful there, they're going to have to up their game and be more productive and be more innovative, subject to the measurement constraints we talked about, right and to upgrade. And so I think that the smart industrial policies are going to be things that sort of push firms to learn and to be more innovative and to be successful as exporters.

Now, the other thing we have to keep in mind in thinking about industrial policy, is that [for] the governments, it's just very hard to [know] in the future what are the sectors that are going to be successful. What are the activities that are likely to have a future? It's just very hard, it's very hard for people who are, you know, private equity firms embedded in the sector... it's very hard to know, it's gonna be even harder for a government official or someone making government policy to do that. So I think we need to think about policies that have this effect of promoting learning or subsidizing innovative activities, but that, you know, don't require too much knowledge and understanding of the future on the part of the people setting the policy. Right. So things like collaborations between universities and firms for, you know, how to train workers to have the skills that the higher tech firms in your country need. That's something that seems like a good idea that's probably going to promote upgrading without having to pick and say, I think this product or this sector is the future of the Nigerian economy and therefore we're going to subsidize that thing. And you also want policies that are somewhat flexible, right, so that if something happens... so I'm working on a project in Tunisia, where the Tunisian Government was trying to promote exports. But the issue that they've had, and it's a matching grant program where sort of half of the costs of exporting of a certain category of costs of exporting will be paid by the government. The problem with that program, though, has been that it was somewhat inflexible. So basically, if something happened, you know, there's a big shock, and in fact, COVID shock, you know, and that changes what firms want to do. And it's very hard for them to switch gears and say, now I want to spend money on something else, can you please subsidize this other thing, and there were a lot of frictions in the program. And so that's often the case for government programs. The government sets a policy and then the world changes, firms want to do something else, but the policy is still stuck, you know, in the old world. So we need to think about how to build in, you know, flexibility into the programs so that if firms decide, actually, the market is moving in this direction, rather than this direction that we were expecting, that the support that they receive could move in the same direction.

Tobi;

Yeah, I agree. And I don't mean export quotas as hard targets. So I'll give you an example. Nigeria has this policy that we've been running for about six to seven years now, where there are multiple exchange rate windows for different parts of the economy or sectors that the government deems should have priority, you know, to import. And I recall a paper where Korea had a similar arrangement, but it was focused on firms that export. Firms that export to world markets sort of get priorities so that they can source inputs at a very low cost and seamlessly, you know, but it's not just something that we really think about in Nigeria, because we are so focused on the domestic market and how large the population is not minding, you know, how much of that population is poor.

Eric;

Yes, no, absolutely. So, certainly, Korea did this. But the Korean model, a key part of it, and they definitely picked sectors in a way that, you know, it's, there's a little bit of tension with what I just said about, you know, the government officials are not going to be very knowledgeable, there they seem to have done a good job of picking sectors to advance. But the key part was it really was oriented towards success in export markets. And the industries that were not successful on the export markets, they pulled the plug, they removed the, you know, they removed the support, which is politically hard to do, you need a fairly insulated, like, secure government in order to be able to do that. Because, otherwise, you start providing support, and then the industry lobbies a lot to maintain that support, you know, and so then it becomes politically very difficult to remove it. But I think if the government is committed to ''if these industries are not successful, we're gonna pull the plug on the support'', then this can work. Right. But you're absolutely right, in the Korean model, the key thing is the export orientation rather than the import orientation. And what you mentioned about exchange rates, I didn't comment on that. But I think it is an issue, you know, especially for a resource-rich economy, that the exchange rate can be, you know, highly valued, arguably overvalued, which makes it hard to develop the domestic industry. And so I think that's a real issue that, you know, some countries seem to be able to handle that, you know, ''what do we do with the natural resource wealth a little better than others'', if you just let it accumulate and people are going to spend and that leads to devalues your currency to increase that's going to make it harder to achieve export success in export markets for manufacturing goods or other exporting services. And so that is something that needs to be a focus of thinking about how to upgrade.

Tobi;

Yeah, I want to talk about technology for a bit. You had this very, very, an interesting paper on the soccer ball, we call it football, the soccer ball producers in Pakistan. And in a bit, you're going to tell me some of the interesting things you learned about that study. But first, Dani Rodrik and Margaret McMillan had this interesting paper about industrialization in Africa, and how domestic manufacturing firms are now shifting more towards capital-intensive technology. So hence, manufacturing firms are not creating jobs as much as historical patterns should suggest, do you see this as sort of a problem? I know so many other people have this worry about automation and how this technology can be exported everywhere, which is really a concern for maybe a continent like Africa with a large, jobless, and young population. So do you see this as a trend that we should worry about, you know, more capital-intensive technologies, or are there opportunities?

Eric;

Yeah. So I do see it as a trend. I do think it is something to be worried about. You know, Dani Rodrik recently organized a panel with the International Economics Association I participated in, along with Daron Acemoglu and Fabrizio Zilibotti and Francis Stewart from Oxford. And I sort of had two points there. One point was, yes, I think this diagnosis is correct. Basically, economists refer to it as appropriate technology. But the idea is that many technologies are developed in rich countries, you know, given factor proportions, we would say in those rich countries, so basically, skilled workers are more abundant, unskilled workers are less abundant, and so people develop machines that kind of conserve on unskilled workers. That's, in part, the background to the story that Dani Rodrik and Margaret McMillan are saying that in Africa, many firms are using this technology that's been developed in rich countries, that's very skill intensive, but it's not generating a lot of them. Right. So I think the diagnosis there is correct that that happens, right? And so the technology often is inappropriate for poor countries given, you know, their supply of unskilled labour, given how many workers they have that could use employment. On the other hand, the other question, though, was, what do you do about it? And so I was less convinced. So my worry about that. There are two versions of that concern about what you do about it. One is, given the set of existing technologies, you could try to encourage firms to use more labour-intensive technologies. Okay. But the problem is that you may encourage them to be less productive. Maybe they might generate more employment, but they'll be less productive, right? There was an interesting paper that I cited in Brazil by Gustavo D'Souza, which was sort of saying the Brazilian government basically put a tax on international technology licensing. And he shows that sure enough, firms were less likely to use International Technology. They're more like to use domestic technology. They actually generated employment, but they were less productive. Right, and they overall did worse. So there's a worry that you're gonna make firms less productive in an immediate sense. The other worry is that, like, if the Nigerian government starts encouraging Nigerian firms to develop new technologies, which are more labour intensitive, you know, then they'll generate more employment, the worries that you're gonna get sort of fall behind the world technology trajectory, I'll call it that. Like, you can think about the world frontiers moving in whatever, pick an industry, and the world frontier is moving at a particular place, and then, you know, firms are competing with each other and they're, you know, someone gets a patent, someone comes up with a new idea and sort of technology moves in a certain direction. And then Nigeria says, no, no, we want to be on a different trajectory that generates more employment, right? The problem is, you're going to be permanently behind where the technology curve is, right? Where the world frontier is. And I feel like that's worrisome, right, you're likely to have less learning, right, there's gonna be a gap between where the Nigerian firms are and where, you know, the world frontier is that it's gonna be hard for them to catch up afterwards. So in the short term, you might generate more employment, but you're gonna have a less dynamic industry as a result. And so I think, my own view, and this is, it's a feeling rather than something that's very research based at this point. But my own view is, even though it means that firms are not going to generate that much employment, they have to try and stick as close to the technology frontier as possible, or, you know, catch up as quickly as possible to where the world technology frontier is.

Tobi;

And so talk to me a bit about lessons from your walk with the Pakistani soccer ball manufacturers. What did you learn from that particular experiment, especially on the role of appropriate technology and technology use and the incentives that surround it for firms and investors?

Eric;

Yeah, so it was a study of technology adoption, what are the factors that encourage technology adoption? And what made it possible was that the football producers, I'll use that word football instead of the soccer ball, these football producers, there are a lot of producers using the same simple technology, right? And this football design is, you know, 85 or 90% are just these hexagons and pentagons. If you can imagine a, you know, a football, it's got hexagons and pentagons. And so the simple technology involves cutting out hexagons and pentagons and then stitching them together. And there were a lot of those and what made the project possible is we came up with a new improved technology, which is basically a way of cutting pentagons from these sheets. The main costs, you know, 50% of the cost are the sheets, they call it rexine. It's like artificial leather, that's the exterior of the ball. But they were cutting pentagons in a way that was wasting some material. Wasting more than they need to and so the new technology is a way of cutting these pentagons so that you can fit more into a given sheet so that you can get basically 8% more pentagons which ended up being about a 1% reduction in total costs. Which wasn't enormous but on the other hand, it's a pretty competitive industry, profit margins are about 8% so we felt like they shouldn't have been paying the 1%. And actually, when we started out, we thought we were gonna be studying technology diffusion, right, which is, you know, one person adopts, then is that their neighbours who adopt or is it their cousins? Or is it the, you know, people who share suppliers, and what are the channels of diffusion, right, and we're trying to keep everything secret, and we thought, okay, when we let it out, it's obviously the people we give it to who are gonna adopt right away, and then it's gonna spread. And so then we gave out this technology, for free, we gave it to 135 firms.

And then, you know, we had a few firms adopt, and they started using it, and including one big firm that was producing - I can tell you the name later, but basically had like 2000 employees and is producing for Nike, and as a big producer adopted this technology, and, you know, is basically cutting all of its pentagons using our design and our die for cutting rather than the old one. So after, you know, 15 months, there were six total firms that had adopted. And that was puzzling and thought, you know, why is that? So then we started asking firms, we started talking to people and basically, it was revealed that the reason was that the guys doing the cutting... so the cutters are basically paid piece rates, they're paid per pentagon or per hexagon, or essentially per ball like, which is, you know, 20 hexagons and 12 pentagons they're paid. That was what their salaries were based on. And they didn't have the incentive to reduce waste, like, they weren't penalized if they wasted the material, right? And so they just wanted to go fast. And our die was slowing them down, right, made them go more slowly because they had to be more careful how they placed it and also, it was a different design, it was the design that they were used to. Now, it turns out that within about a month, they could get back up to speed, to the speed they were at before but they didn't know that, and in any case, for that month, their salaries would be way down, they'd just be slower and knowing that if the firm didn't change the contracts, their salaries would be lower. And the workers were figuring this out, the cutters are figuring this out, they said, this is not good for me, right, that my salary is gonna go down if I use this thing, I have no incentive to use this new technology. And so then they started telling their firms, you know, this is bad, bad technology, it doesn't work, it's dangerous, it has all these issues.

Okay, so then we realized that this was happening and we said that we were going to do a second experiment. So, you know, half of the people we originally gave the technology to who hadn't yet adopted, we did a second experiment where we said to workers, we're gonna give you a month's bonus, which is not very much it is about $150 US dollar. So these guys are not paid very much we said ''a month's bonus if you can demonstrate to us and the owner of the factory that the technology works.'' And actually, that was enough. The workers were excited about that, you know, they got paid for doing this. Everybody who did it then subsequently passed the tests. So they demonstrated that the technology is working, and then a statistically significant share of the firms that they worked at ended up adopting the technology as a result. So those were the two experiments, those were the facts. What are we learning from that? I think we're learning that, basically, the lack of information flow from workers to their owners, to their managers, was what was getting in the way of technology adoption in this case. Like, the workers knew that the technology was working, but the owners didn't know because they sort of delegated the process of cutting the pentagons to the workers, and given the contracts, the workers didn't have the incentive to share the information. Right. So I think those sorts of, like, information flows or barriers to information flows are actually very important in the learning process. And kind of what our second experiment did when we did this bonus of a month's pay, which induced the workers to share the information and that was sufficient to make the technology be adopted. And so I think the punch line or the one-sentence version of this is, workers need to see that they're going to benefit from the adoption of new technology or from upgrading generally in order for the process to work well. They have to buy into the process. And they have to see that they have the incentive to do so. One recommendation coming out of that would be some sort of profit sharing, or some sort of gain sharing between workers and firms would actually be very useful. And will it help there be more innovation?

Tobi;

It brings me in a way to another very interesting paper of yours which [they] also had a summary essay about, I think, in VOX or something, which is about wages in poor countries. And I mean, thinking about the soccer ball story and the lesson. One issue and this has generated quite a number of debates between I think Rodrik and a bunch of other scholars who are thinking about Africa, is that the reason Africa is not really industrializing, or firms are not creating jobs is because wages are too high relative to the level of income. But what I learned from your paper, and you can correct me if I'm wrong, is that paying higher wages in poorer countries is not really a disincentive to creating employment and even generating productivity and profit. Tell me a little bit about how that works. Because, usually, we've gotten familiar with this logic that for you to be able to industrialize, if you think about China, and so many other countries, you need to have access to low-wage workers, you know, you need to be able to do very cheaply, and labour is where you can really cut a lot of your costs. And then it becomes a problem if your domestic wages are too high for the level of your income or what firms and investors are willing to pay. So tell me this high-wage, low-wage dynamics, especially... I remember the famous Paul Krugman was it article defending sweatshops in Bangladesh, where if you force firms who are outsourcing to pay higher wages or impose certain conditions, poor people in those countries will lose jobs, and they will lose their livelihoods. And so you should not mess with that process. What are your thoughts on these [issues]?

Eric;

Yeah, very interesting. So I think the article you were thinking of, it's related to the specific case of the football producers and seal coats. In Pakistan.

Tobi;

Yeah.

Eric;

There was a very interesting thing that happened. I mentioned that one firm adopted this new technology. And you know, one very large firm and it was producing for Nike, it's called Silver Star. The interesting thing about that firm is that because they're producing for Nike, which had had sweatshop scandals in the past, Nike required them to do a bunch of things, basically, so that Nike wouldn't be vulnerable to a further scandal, right? And among the things that they had to do was make sure they were paying the minimum wage in Pakistan. And the only way this firm could guarantee that they were paying the minimum wage in Pakistan, which many firms were violating basically, the only way they could is to say, we're not going to pay a piece rate, we're going to pay a fixed wage. Right. So this firm was paying a fixed wage rather than a piece rate. And actually, we talked to them about when they first won the Nike contract. They said their labour costs went up 20 to 30%. So they did a bunch of things. They had this fixed wage, there was a medical clinic on the factory grounds. They had sickness pay, they had some retirement benefits. So a bunch of things, they did raise wages. But the advantage of that was that the workers were much less likely to block the adoption of this new technology. Because in a specific way, they did not have a disincentive, you know, their wage was going to be their wage no matter what happened, rather than in other firms [where] what was happening is that the worker can see if they adopt this technology, their wage would go down. And so we believe, and I wrote this in an article that you saw in the Harvard Business Review, I think that's where it was, that those wages, you know, higher wage payments and fixed wage payments, which were imposed by Nike actually contributed to the process of innovation. The title of the article is how labour standards can be good for growth, and also in the process of upgrading. So that's an example of how having higher wages can actually be good for this upgrading process.

Now, there are factors going in both directions, right? On the one hand, you know, the 20 or 30% higher labour costs, I think they did contribute to innovation. On the other hand, 20 or 30% higher labour costs may mean that firms will hire fewer workers or that the industry will be less competitive. So it's not that, you know, this innovation effect is all powerful and it's going to overwhelm anything that's about labour costs. But I think it is something that we need to take into account. And so, you know, labour market institutions that, you know, maybe promote profit sharing with workers, that promote longer-term employment so you have people who are around for longer, that have some job security, the sorts of things that often labour unions want to negotiate, can actually be good for this innovation process. And that's one factor that should be weighed against this issue of, you know, how higher labour costs and how competitive is the sector going to be. You often hear, like, the World Bank or the USAID, the development agencies will often say, you just have to be cheap. Like, you know, the competitive advantage of Nigeria is cheap labour and therefore, you should be focusing on having low wages and producing, you know, garments and textiles and toys and low-end manufacturing. But I think that's kind of a low-road model. You know, and I think that there are viable high road models, which would involve somewhat higher wages, some sort of gain sharing or profit sharing, and being more innovative at the same time. I can't tell you I have it all worked out exactly what that model would look like, I think it's going to vary by country. But I think we need to try to think about and push in that direction of where you can have, it may not be high wage, but it's gonna be higher wage than the market by itself maybe would bring about. So I am optimistic that that can happen. But again, the devil's in the details, you know. So Nigeria needs to think about what are we relatively good at doing right now and let's think about how can we be more innovative and move up to the quality ladder, the technology ladder in those industries. And then how can we get our workers on board to the process of moving up that ladder? And that will probably involve paying those workers more, rather than just trying to cut wages to the extent possible.

Tobi;

Before I let you go, let me... I know you're a relatively quiet person so let me draw you in a little bit... yeah, I know you're not active on Twitter or anything like that. Let me draw you into a little bit of professional controversy. And one of the things that I admire most about your work, I should confess, is that it's methodologically diverse. You know, you do structural econometrics, you do RCT, you do regular modelling and so many things. So there's this huge debate currently that I think, a lot of my colleagues may not think so but I think has important consequences for the policymaking process on development, which is that - is development research right now focused on the right things? You know, RCTs are like the standard tool for the investigation of development questions. Empirics have sort of taken over the field.

But on the other hand, you have folks like Lant Pritchett who are constantly pushing back that this is encouraging researchers to think too small, they are researching cash transfers, and so many other key interventions, whereas we really should be focused on the big questions. And in my experience, these have real-life implications, especially in poor countries where they have budgetary constraints. We might say this is due to corruption, and that will be true, but sometimes they have a real balance of payment crisis, because a lot of these countries are resource-dependent, and it's often cyclical. So a policymaker may really want to know where to spend the most resources to have the maximum benefit for the citizen. So I find these questions very important. What do you think about this debate? As someone who transverses the field very often in your work, how have you been able to navigate this debate? And what do you think is the, maybe right is not the right word here...what do you think is the useful approach going forward?

Eric;

Yeah, good question. Yeah, in my own work, I've been very question driven rather than methods driven. Right. So I've always thought, you know, I'm interested in this question of from upgrading, what are the barriers to upgrading? What drives upgrading? How can we, you know, learn about that, and if we can learn about that using an experiment, that's great. If we're in about that using other methods, that's great, too. So I, sort of, don't have a dog in the hunt, as Bill Clinton would say about, you know, the methodology. And I'm kind of in the middle of the road, I think, in terms of this debate between, you know, J-PAL and Esther Duflo and Abhijit Banerjee and Lant Pritchett or others on the other side. I think, you know, in situations where you can run an experiment, I think that is the most credible source of information. Okay, so I'd rather have a randomized experiment than do a correlation and put some causal interpretation on a correlation. At the same time, I do think that there are many questions, either that can't be answered with an experiment, or that are just very, very costly to answer with an experiment, right? And so it's very hard to run, you know, it's running experiments on firms. I've tried to do it, but it takes a long time. It can be very costly. You have to give much bigger shocks to firms to get them to react, etc. And so, I've heard Abhijit Banerjee articulate that, like, we should never do a policy that hasn't first been evaluated by random experiment, I think that's too strong. Because we're gonna be waiting years and years and years to get the experiments and with a huge investment of resources in order to get the experiments that would then inform the policy. So we're going to have to make policy and, you know, make decisions based on other sorts of information. And so there, I do think we need to be like small ''c'' Catholic, allow for lots of different types of methods, quasi-experimental methods, you know, even structural methods, and then also experiments. There's this famous joke about the drunk guy with a streetlight, you know, he's looking for his keys, and he's looking under the streetlight, because that's where the light is, maybe not where the keys have been lost. And so I take that point, like, maybe we really care about these big questions about, you know, what's going to drive growth, then in that sense, I'm sympathetic to the sort of the Lant Pritchett view. On the other hand, under the lamppost, we actually are learning stuff, right, I feel like we're more confident that we're making progress by looking under the lamppost. And so I think the, you know, the trick, the art here is to sort of stay near the edge of the lights and we're getting closer to the big questions, but in a way that's still credible, and that we're still, you know, we can believe the answers that we're actually given.

To sort of counter the Lant Pritchett view, you can post these big questions, and you can, you know, think big thoughts. But at the end of the day, you have to be able to convince, you know, you have to show us the data, right, you have to show that this is really correct. And that's just very hard to do for many of these big questions. So we need to incrementally build up based on this work. That's why I kind of like this work on firms, we're getting towards these big questions about growth, but in a way that you can actually have some confidence that you understand what's going on.

Tobi;

In your experience doing this work, what are misconceptions that you have encountered in the field that either the professional development industry, so I'm talking about aid and the think-tank and all the other folks, or it may even be your academic colleagues, what are the common misconceptions that you have encountered?

Eric;

Yeah. I mean, so one big thought [is] I think that the of field development agencies, right, it's like, how are we going to spend aid dollars in a way that's going to have a positive effect? And I think there's value to that. All right. I'm all in favour of spending, you know, aid dollars, in the most effective way. But I think that you know, a set of questions does limit to some extent the impact of the field of development on the development process. So I actually think we could spend every aid dollar in an optimal way, and would it have a meaningful effect on the material standard of living of people in poor countries? I'm not sure. I mean, maybe a little bit, maybe marginal, right? I think what's really going to matter is, do these countries start getting industrialization happening? Are they getting upgrading? Are they growing? And so in that sense, I sometimes get a little bit frustrated with the development discussions, it's all about this, you know, how do we spend aid dollars, and let's do RCTs to figure out how to spend the aid dollars, rather than these bigger questions, which are going to have a longer-term effect on people's living standards. You know, that's changing a little bit. I'm encouraged. There are more and more people talking about firms, there are more and more people taking sort of industrial policy ideas seriously. They're talking about bigger-picture questions in a kind of micro-founded way. So there are some encouraging signs. But I think a lot of development is still about that issue of like, what's the right way to do social policy? What's the right way to do, you know, aid spending, rather than trying to understand deeply why is it that Korea was able to make this transition from a poor country to a rich country, essentially, in a generation? And why is it that many countries in Africa are not? What is it that's actually getting in the way? And for that, that's not really like how to spend aid dollars question that's more about how firms behave. What are the factors that constrain them? And those sorts of things.

Tobi;

This is a show about ideas. So I want to ask you, what's the one idea? Just one. One idea that you think everybody should think about and adopt, that you would like to see spread everywhere. What's that idea? It may be from your work, or it may be from other things that inspire you. What's that one idea?

Eric;

I think the one idea I would choose is, uh, workers have a brain. This goes back to the soccer ball study, that there's knowledge and information that, like, workers have or people who are lower down in the hierarchy have, which is not being taken advantage of. Right, the soccer ball thing was an example. The workers were understanding the technology, but because of the way they were paid, and because of the, you know, institutional arrangements, they didn't have the incentive to share that. And I think the world, including the economics profession, tends to undervalue the intelligence that people have. Even the people who are actually, you know, on the frontlines doing the work. And if we can figure out ways to harness that knowledge and give people incentives to share it and give people incentives to develop their own intellectual thinking about whatever it is they're doing, I think that'll have a big payoff. And so I'm interested in sort of investigating what are the sorts of arrangements, what are the sorts of policies that can lead that to happen more?

Tobi;

Yeah. Thank you so much, Eric. I mean, tell me a little bit about what you're working on right now.

Eric;

What am I working on right now? I mean, so one thing related to what we've been talking about that I'm excited about is, again, a paper on technology adoption. This is in Bangladesh, with an energy-efficient motor like sewing machines. They're different sorts of motors that the traditional ones they're kind of spinning all the time. And then people have the foot pedal they like to press the foot pedal and then the needle comes down and stitches right but they're actually wasting a lot of energy because these motors are spinning all the time. And so there's a new type of motor called a servo motor which spins Only when the needle is moving, right, so it's energy efficient, energy efficient motor, but it can just replace the old motor, you don't have to change anything else about the machine, you just put this new servo motor to replace the old clutch motor. And we're studying when new managers or when new owners, when do they make those decisions. And so we're trying to track we're giving them information in different intensities, like including installing the machines in their factory one is just showing a video when it's just providing information, but one is actually installing their machines. And we're seeing how they react to that information. So I think that's a big topic. It's like what's getting in the way of the adoption of energy-efficient technology? These are the people who are making mistakes, or they just don't have good information. Or that basically, maybe if they have the right information, they actually will adapt very quickly. So that's one thing I'm thinking about.

Tobi;

It's been fascinating talking to you, Eric. I enjoyed it so much.

Eric;

Thank you, Tobi. Good questions.

0 Comments
Ideas Untrapped
Ideas Untrapped
a podcast about ideas on growth, progress, and prosperity