Daron Acemoglu on Innovation and Shared Prosperity – Beragampengetahuan

Daron Acemoglu: Thank you, Russ. It’s a true pleasure to be here with you.

Daron Acemoglu: Thank you, Russ, and you summarized it brilliantly, especially with the word ‘necessarily.’ And that’s really the main reason why we wrote the book.

And, Simon and I are convinced that–as economists and as concerned public policy makers, analysts–we have to hold two potentially conflicting views in our mind. One, that today we are extremely fortunate, from a historical sweep, because of the technological breakthroughs that started sometime around the middle of the 18th century. The application of knowledge–later scientific knowledge to industrial processes–that have brought us much better health, much greater comfort, much greater level of prosperity. So, we are huge beneficiaries of technological progress.

But, at the same time, we also have to hold the conflicting thought with that: that technology can be used, and has been used for bad; and there is nothing automatic about technological progress bringing widespread prosperity.

And that automatic–or ‘necessarily,’ as you put it–is the key thing. Because in thinking that somehow technological progress is going to be such a powerful force for lifting all boats automatically, we abrogate our responsibility to shepherd it in the right direction and to make institutional adjustments so that society at-large benefits.

And, the reason why we do all this historical work in the book is to argue that we have benefited from technology precisely when there are certain preconditions that have been satisfied.

Daron Acemoglu: Well, I don’t know whether I can change your mind, but what I’m going to do is I’m going to first claim three critical ingredients for technology to have the type of effect that you explicated. And then, I’ll give you some historical examples. And of course, feel free to interject. I don’t want to go on for too long.

But One is: competition. Absolutely, you are 100% right. But, I think you are not right in thinking that competition is automatic, either in our age or throughout history.

Second: coercion in labor markets. So, that’s the tip of the iceberg in some sense. That means that when employers have too much power over their employees, there is no guarantee that an increase in marginal productivity of labor–okay; sorry, I’m using a jargon, but I think it’s okay for an EconTalk–an increase in the contribution of labor to productivity will translate into higher wages.

And, Third: automation. That, by itself automation may increase capital’s productivity–may increase profits–but it may not increase the marginal productivity of labor; and hence break one of the chains that goes from higher output, higher productivity to higher wages.

We see all three of them in history. So, we spent quite a bit of time on medieval technologies to argue that the so-called Dark Ages weren’t dark in terms of technology. But they were dark in terms of the conditions of the peasants. And, part of the reason was this lack of competition and coercion both of the first two factors I’ve mentioned.

Windmills, for example, increased productivity more than 10-folds in some tasks, but because they were monopolized by lords and abbots who actually forcefully did not even allow any competition and coerced people to use their mills, few, if any of the benefits spread beyond that small, narrow elite.

Let me give you another example of coercion. A transformative technology for the United States’ history is the cotton gin. It turned the United States from–it’s essentially a backwater for agricultural productivity, especially in industrially relevant agriculture–into the largest exporter of cotton. It revolutionized the South’s economy.

But, the workers–black, enslaved people–actually fared much worse under the cotton gin because they were moved into much, much harsher plantations in the Deep South, as cotton plantations spread.

And then, finally, automation. When we see automation being the only or the main focus of technological changes, we see that many workers experienced lower wages or lose their jobs. That’s what happened in the first phase of the Industrial Revolution. It went longer than half a lifetime. It may have gone on for about 90 years or so–perhaps two lifetimes for people at the time, especially since life expectancy at birth declined to around 30 years as people moved into the cities.

So, all three of these are important. And when I meant ‘shepherding,’ I meant ensuring competition, ensuring labor has voice and power, and also encouraging broader applications of technology than just automation.

Daron Acemoglu: Automation versus other things that increase marginal productivity.

Daron Acemoglu: Absolutely.

Well, actually, Russ, you’ve made two points. One is about coercion more generally. And, I want to come back to that because I think it’s very, very important.

But, if we can put a pin on that–and let’s not forget it–let me answer your question about the Industrial Revolution. No, I am certainly not confident about real wage data between 1750 and, say, around 1840. And, there are voices on both sides. You are absolutely right.

But the general statement that the working people did not much benefit from the industrial progress, I think I’m fairly confident. Because, there you have to bring in what was happening to their living conditions? What was happening to their work autonomy? What was happening to their health? What was happening to their children?

I think when you look at the data there, to me there is very little doubt. In 1840 [1834–beragampengetahuan Ed.], there was a Parliamentary Report about the condition of England that reported children as young as five or six working 18 hours in deep coal mines and pushing carts with their heads. 1840. 1840. This is 90 years after the 1750, which is the rough beginning date of the Industrial Revolution.

1720 is the beginning of first big textile mills. So, maybe under 120 years after that.

And, the reason why we have these data is because the British had parliamentary commissions; and there was starting–why so late? That’s a question–was starting some concerns about the conditions of England, because of petitions, because of people talking and newspapers.

So, this is the condition that we’re looking at. 1850s: Still, the living conditions in British cities–where more and more people were living–were horrible. Life expectancy may have fallen below 30 years at birth. Infectious diseases were rampant. There was no sewage system. And, even though we don’t know about real wages, working hours expanded quite significantly under some excellent economic history works that really document this.

So, yes, we can debate about real wages, and it’s a good debate. But, the conditions that we’re talking about were really very, very different from what started to transpiring from around the 1860s, 1870s.

And one important factor there is, you know, worker voice. Trade union activity was very harshly suppressed in Britain. Any kind of democratic movement was super-strongly discouraged.

1840s, when I was just mentioning a second ago, it was also the beginning of the Chartist Movement. Chartists collected 3 million signatures. I mean, let that number sink. Three million signatures for asking for minimal changes in the political system so that people could vote and could participate in the political process. They were very careful not to make even demands about trade union organization so that they could not be labeled Socialists.

And you know what? The British upper-middle classes or middle classes–or whatever you want to call the system–the people who were politically powerful, they completely ignored them and threw all of the Chartist leaders in jail.

So, that’s what I mean that there was nothing automatic about that process.

And when Trade Unions start, you know, organizing after the Master and Servant Acts–which completely disempowered workers against their bosses, and Trade Unions were legalized–that’s when you see conditions in factories improved quite a bit.

Daron Acemoglu: Yeah, you nailed it. You nailed it. I think–exactly. It could have come sooner, and it did come. But again, it’s not the automatic process.

And, I’ll come back to the United States–because I want to come back to the United States in the context of automation issue because I think it’s a great example.

But just as a preliminary answer: Yes, they migrated to the United States because living conditions were better in the United States. Wages were higher in the United States. Wages are key.

And, why wages were higher? Well, that’s about labor shortages and technology, both of them; and we’ll come back to that.

But, the reason why they went to London is mostly push factors. So, labor was pushed out of the countryside in the United Kingdom for two reasons. One of them is semi-good: Agricultural productivity was increasing in a way that reduced labor requirements, so people did not have as much work.

But second, there was also this whole Enclosure Movement, which rationalized parts of agriculture, but in a very harsh way, cutting out many of the means that people had for surviving. Because, in the open fields, they could collect wood, they could hunt, they could do other things in order to support their existence there. And again, that sort of illustrates some of the tensions that we’re talking about.

One of the, sort of, most eloquent observers of the Enclosure Movement was Arthur Young, who started as, you know, almost like a Malthusian before Malthus in terms of his views, and was very–very much–an economist by instinct–thinking about economic efficiency and how important it was to introduce rational property rights, etc.

But, after a while, he changed his mind and he said, ‘Look, through this Enclosure Movement, we’re really treating these people so badly and we don’t need to do it in this harsh way–in this rapid way–that throws them out. And we could at the same time allow them some means of existence.’

You know, again, this is sort of the tension. We are all recognizing that the economy had to be transformed.

And technology was providing the means.

But, how are we going to use these means? And through what institutional means?

And, that’s exactly why you nailed it. With the right kind of regulation, it didn’t need to be three generations. It could have been perhaps one generation in the United Kingdom.

Daron Acemoglu: I’m not claiming that–in the United States, it was regulation. I think the evidence for, quote/unquote–“regulation”–is not even the right word. I just picked that one because you used it.

The institutional framework, is: Is it Trade Unions? Once Trade Unions were in the mix, then you see conditions in factories improve, and worker voice, more investment in education, more investment in public services. So, that’s the sort of the institutional balance.

And, more generally democracy.

You know, the United Kingdom at one level–and this is what for example, in my work with Jim Robinson, we’ve celebrated this–is a huge success of democracy, a peaceful process of democratization–but, on the other hand, it was a slow one. It took, you know, again, 1860s, 1870s, until really truly there was some sort of democracy which provided voice to the working people.

Daron Acemoglu: They even came from England. That’s exactly your point: they came from England.

Daron Acemoglu: So, let me say a couple of things, Russ. And, again, I still want to come to technology in the United States because that’s the critical part for the United States and it’s also a critical part of our argument.

But, first of all, long-run effects of trade unions are very difficult to know. So, if you have a very strong opinion about long-run effects–like, in 30 years time, how will they change wages, etc.?–that’s very difficult to nail.

But, short- to medium-run, I think there are both good theoretical reasons and empirical evidence that when they are effective–and that effectiveness requires both trade unions to behave the right way and the employers to react to them the right way–they can have quite significant effects.

In terms of the theory, just very quickly: You know, the question is whether you start from a labor market where everything is fully competitive except the trade union, or whether you start with a frictional labor market. Like, for example, if you have a frictional labor market where there are, for instance, search frictions–which mean that every employment relation has a quasi-rent in it. Meaning that: I would love to be here and you want me to be here, Russ, and it’s not like I have an outside option that will make me exactly indifferent. And, you have an outside option at this point that will make you exactly indifferent.

I think in the medium run, that’s right. If you planned for next week, you could find as good a guess as me. But, right now we are locked into this relationship. That means that in our relationship there’s a quasi-rent and we have to distribute it. How are we going to distribute it? That’s the bargaining problem. And, if that bargaining for whatever reason–for instance, if I’m an enslaved worker or the rest of the economy is treating me really badly–you have all the bargaining power in there, that’s actually not going to lead to a fair division. It’s going to lead to inequality.

Actually, it won’t necessarily lead to much efficiency, either.

So, that’s the question, which is: How are these employment-generated quasi-rents distributed?

And that becomes a particularly big problem when firms get bigger like the modern factories because they have a lot of power over their workers.

And the way to think about why trade unions did have a meaningful effect, and without creating huge inefficiency–again, that may be very different than from the police unions and the teacher unions–but, in the modern factory system, it’s because they level the playing field to some degree.

And where do you see the evidence for that?

In the United States it’s more complicated, but if you look at Scandinavian countries, there you are actually seeing how unions, when they are acting in a way that’s in concord with employers, but actually, you know, asking for better working conditions or better wages, also compressing the wage structure, they are having massive effects that you can see in the data. [More to come, 27:08]