New technologies are sold as a net benefit to society as a whole, but the truth is that technological progress is only loosely correlated to the improved welfare of the majority of citizens. This is not to say that technology and innovation are bad—we’re big supporters of both—but when tech CEOs hold all the power to make decisions that affect all of us, that becomes a problem. For a long time, technology has been used by the rich and powerful to further enrich themselves and consolidate their own power. Is there a way to ensure that everyone benefits from innovation—not just the wealthy few? Returning guest Daron Acemoglu shares insight from his new book on the subject, Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity.
Daron Acemoglu is the Institute Professor of Economics at MIT, the university’s highest faculty honor. For the last twenty-five years, he has been researching the historical origins of prosperity, poverty, and the effects of new technologies on economic growth, employment, and inequality. He is an author (with James Robinson) of The Narrow Corridor and the New York Times bestseller Why Nations Fail.
Twitter: @NarrowCorridor
Power and Progress https://www.hachettebookgroup.com/titles/daron-acemoglu/power-and-progress/9781541702530
Website: https://pitchforkeconomics.com
Twitter: @PitchforkEcon
Instagram: @pitchforkeconomics
Nick’s twitter: @NickHanauer
Nick Hanauer:
There is only a loose correlation between technology and the welfare of the majority of citizens, which is not to say that technology is bad or that we shouldn’t have it and we should not have innovation.
Daron Acemoglu:
There are automatic powerful forces that will ensure that as long as we remain innovative, we will all benefit from technological advances. And in fact, we might even have something approaching shared prosperity.
Speaker 3:
From the home offices of Civic Ventures in downtown Seattle, this is Pitchfork Economics with Nick Hanauer, the best place to get the truth about who gets what and why.
Nick Hanauer:
I’m Nick Hanauer, founder of Civic Ventures.
David Goldstein:
I’m David Goldstein, senior fellow at Civic Ventures.
Nick Hanauer:
Goldie, today, I’m so fired up because we get to talk to a returning guest, but definitely one of our favorite economists in the whole wide world, and that’s a short list.
David Goldstein:
One hand. Count them on one hand.
Nick Hanauer:
Yeah. Daron Acemoglu, who is the institute professor of economics in MIT, which is the university’s highest faculty honor, and he has written a ton of incredibly consequential books, including one of our absolute favorites, Why Nation’s Fail. But in addition, The Narrow Corridor. And he has a new book out that is super interesting and consequential called Power and Progress: Our Thousand Year Struggle Over Technology and Prosperity with his colleague Simon Johnson.
And in that book, they argue, as we have argued before, that there is only a loose correlation between technology and the welfare of the majority of citizens, which is not to say that technology is bad or that we shouldn’t have it and we should not have innovation, but they make this really interesting argument about the necessity of power in discussing power in this framework. And I know you love the book.
David Goldstein:
Yeah. And most of our listeners are not economists because let’s be honest, economists, I can’t imagine why most of them would like this podcast.
But a lot of you have taken Econ 101 and you’ve had some sort of introduction to economics. And one thing you might notice was missing, is this notion of power. Orthodox economics assumes it away. There’s nothing about power in neoclassical economics. There’s market power and so forth, but there’s supply and demand and balance-
Nick Hanauer:
Perfectly efficient markets and all that stuff.
David Goldstein:
You just let the market do its magic and the invisible hand will balance all this stuff out. And so if a new technology enriches just a few people and impoverishes a bunch of others, well that’s Pareto optimal or something because it’s inevitable, it’s deterministic, and we couldn’t have that technology if we didn’t allow it to happen that way. And of course, what this book steps you through, with different technologies and different examples over history, are the different types of power and how different technologies have been used or abused, and how none of this is deterministic. It’s the consequence of choices we make at the institutional level, societal level, individual level, and the different types of power that come into play, not the least of which being the power of persuasion, which is something that we hope we have a little love on this podcast.
Nick Hanauer:
Exactly. So with that, let’s talk to Daron.
Daron Acemoglu:
My name is Daron Acemoglu. I am an institute professor at MIT in the economics department, and I research topics of inequality, technological change, and long run economic development. And I guess part of the reason why I’m here is to talk about a new book that I have with Simon Johnson, my colleague and longtime collaborator, Power and Progress: Our Thousand Years Struggle over Technology and Prosperity, which brings my interest in institutions and long run development together with my concerns about inequality and the future of technology together.
Nick Hanauer:
We’re huge fans of your work and we’re thrilled to have you on the podcast. But for our listeners, why don’t you just sort of lay out the general thesis of your new book Power and Progress?
Daron Acemoglu:
Well, actually, let me tell you why we wrote the book, which will also serve as an introduction to the themes covered in the book. We wrote it as a partial counterweight to a type of techno optimism we see in academia, in policy circles, and in of course the tech industry, that there are automatic powerful forces that will ensure that as long as we remain innovative, we will all benefit from technological advances. And in fact, we might even have something approaching shared prosperity. And when this premise is questioned, people always say, “Are you claiming that this time is different? Because in history it’s always been fine. When we have been technologically innovative. The result has been fairly good ultimately, or eventually.”
Of course, ultimately, and eventually there’s a lot hidden in there. But even leaving those words aside, when you look at history, the picture that emerges is much more complex. In fact, I would say this time is no different. This time is just like many other technological transitions, in which who controls technologies of utmost importance. And if technology remains in the hands of powerful leaders, elites, big companies, without any countervailing powers, there is no guarantee that anything approaching shared prosperity will be created. In fact, we could be on a path towards very damaging choices about the future of technology, future of work, future of inequality, and it is doubly so with the advances in generative AI that we are witnessing right now.
Nick Hanauer:
So I just have to ask, not to take you off track, was there some conversation or thing you saw or person you interacted with that pissed you off so much that you thought, [inaudible] I’ve just got to write this book because I am so-
Daron Acemoglu:
Can I plead the fifth on that? No.
Nick Hanauer:
I’m so sympathetic to this motivation, because I’m, as you must know, surrounded by tech bros who are absolutely convinced that every thing they do, every word they utter, every move they make is somehow benefiting all of humanity. It is maddening.
Daron Acemoglu:
Well, since this is pitchfork economics, I have to tell you that going back to the mid 2010’s, there were several meetings with leading tech entrepreneurs or public intellectuals aligned with their views, which essentially gave me variance of the following theme. Of course, technology is going to create inequality. And the main problem is that pitchforks will come out because of the inequality, not seeing that we’re all going to benefit so much that inequality doesn’t matter.
So all we need to do is somehow find a way of convincing people to take the technology on, but no discussion of whether technology really has to bring inequality, whether they really have a right to impose technology on people or a question that there are alternative paths of technologies and institutions. And that’s when issues that ultimately led to this book started coalescing in my mind.
Nick Hanauer:
Yeah, that makes good sense. I didn’t mean to try to out you or anything like that, but it just-
Daron Acemoglu:
Well, I said several meetings. I didn’t give you any names, you see.
Nick Hanauer:
Anyway, okay. So obviously, we can infer from your description that in fact there is not a necessary direct correlation between the amount of technology we have and how great everyone’s life becomes. But tell us more about what you discovered when you examined the historical record.
Daron Acemoglu:
Let me give you some examples. I think that’s going to be the best way of introducing the topic. Here is one from this country’s economic history, the cotton gin, which arguably transformed much of the US south from an economic backwater into the largest exporter of cotton at the beginning of the 19th century. Cotton, of course at the time, was the most important commodity fueling the industrial revolution in Britain.
And cotton gin really underpinned huge fortunes for landowners, complete reorganization of society. But if you look at what it did to the workers who were producing cotton, the answer is not so great. Those were the enslaved people who got shipped to the deep south. Working conditions got much worse on cotton plantations, much greater discipline, much longer working hours. And there is no evidence that cotton plantations which were fabulously profitable led to higher wages.
Why not? Well, you don’t need to be a PhD in economics to see the answer. This was embedded in a very unequal power relation. Slaves versus bosses and bosses had all the power, and if they wanted the slaves to work harder, they could coerce them to do so. They didn’t need to share the gains. That’s just one dimension of the relationship between technological progress and higher wages or the pillars of shared prosperity that turns out to be much more complex than saying, well, average productivity increases, we’re all going to benefit. No workers need to have a voice, some power to make sure that they get a slice of the cake.
But even more intriguing for many people, especially economists, because we tend to think, well average productivity, that’s got to increase the demand for labor. Actually, how average productivity increases, how we become more productive is very important. So the often quoted story or parable of the future of the factory or the future of the modern factory is that it will have two employees, a man and a dog. The man is there to feed the dog, and the dog is there to make sure that the man doesn’t touch the equipment. Well, I don’t think that’s the utopia. It’s more like dystopian to me.
But if we are heading towards a factory like that and in some sectors we’re not completely far off, those factories can become very productive, but they’re not going to rush to hire more labor, more men and their dogs. They’re not going to pay them more because what makes those factories very productive is automation, the ability of machinery, and increasingly algorithms, to perform the tasks that humans use to perform. So that means that they’re not going to need more labor, and if they don’t need more labor, they’re not going to pay more to labor. So there isn’t an even so determinate relationship between productivity and wages, even leaving aside issues of coercion and institutions and labor markets.
David Goldstein:
So who buys the stuff that this factory produces?
Daron Acemoglu:
Well, we’re having a little bit of that problem in the world today, but we are, right now, in an economy in which about 30% of Americans are doing very well. That includes a small fraction of entrepreneurs who are doing fabulously well. It includes probably 10, 15% of the population that have either postgraduate degrees or other specialized skills such as excellent programmers or excellent surgeons or excellent performers. They are doing extremely well.
And if you look at where a lot of spending is, it’s in a few metropolitan areas such as San Francisco, New York, and it’s the type of spending that caters to these ultra wealthy people. So we have more and more sommelier and personal massage therapists and all sorts of individualized services and private wealth managers. We don’t have as much of the spending in the United States going to things that are consumed by low income people because low income people are not seeing their incomes increase.
Nick Hanauer:
I think you make a really interesting point about the limited connection productivity increases and general welfare increases have. And of course I’m not an economist, so I don’t tend to think in these terms, but again, it’s one of those things, where when you look at the averages, it seems like it’s fine, but then when you look at the facts, it’s not. That’s a really great example of how average productivity gains and the average welfare of the typical citizen, or however you want to put it, can diverge so much. It’s really interesting. So-
Daron Acemoglu:
Absolutely. Absolutely right. There’s nobody who consumes the average basket in the United States and has the average GDP per capita as their income.
Nick Hanauer:
No.
Daron Acemoglu:
But actually let me make another point. We argue also in the book that focusing too much on automation and pressing down on middle skill and low skill workers has actually not been for average growth either. GDP growth and productivity have suffered in this country because I think we have not invested enough in the workers.
Nick Hanauer:
Right. And of course, it’s this weird vicious cycle is that the less you pay people, the less fewer things they buy. So there’s less demand. So less need to invest in the future. So then you get these horrible economic artifacts like a trillion dollars in stock buybacks a year, which increase inequality and so on and so forth. All of this stuff is related.
Daron Acemoglu:
Absolutely. Absolutely. But there are so many other vicious circles here. The one that I would also emphasize is that you don’t invest in the worker in your employee’s training. You don’t give them the right tools. You don’t create new tasks and responsibilities for them, and then you say, oh, look, they’re not very productive, so let me automate more tasks instead. So there’s another vicious circle there which becomes a self-fulfilling prophecy.
David Goldstein:
Let’s take this to actually what you really, the main focus in the book, the reason why we are not making these investments and why we’re seeing rising inequality. It’s not something inherent in the technological progress. It has to do with power. That’s the first word of the title. What’s changed? Why did we grow this incredibly large middle class in the three decades following World War II and then we decided to reverse course?
Daron Acemoglu:
I think that is the critical answer. And thank you for emphasizing those three decades because they are not unique, but they are really illustrative of a broader trend that had started sometime in the middle of the 19th century. Perhaps they are the apex of that trend, but they show contrary to what some people would claim that capitalism or market economies will always lead to inequality. They would always involve the rich trampling on the poor. They actually show you have a considerable period during which growth was rapid and people at the bottom benefited even more than people at the top. Inequality actually declined. And overall it looks very much like shared prosperity.
But if that’s right, then it becomes also vital to understand how it came to an end. And I would say it cannot be understood without thinking about both what happened to technology and what happened to institutions and power. And those two are of course interlinked. It isn’t simply a story of a few companies got bigger and more powerful or unions got bashed. Although these did happen. It is also that the way that we started using technologies changed fundamentally.
Work that I have done with Pascual Restrepo from Boston University finds that a critical factor in understanding why some demographic groups such as men without a college degree have experienced declines in their real income since 1980. We’re not just talking about stagnation, actual declines in their real income. Is that much of what they used to do has been automated since.
Nick Hanauer:
Or outsourced.
Daron Acemoglu:
Or outsourced.
Nick Hanauer:
Which is effectively the same thing.
Daron Acemoglu:
Right. They work very similarly. They work very similarly. So actually that’s a very good point, and let me build on that, Nick. What is offshoring? I don’t want to get into too much jargon, but people use the term outsourcing when it happens domestically as well, and it’s got some important implications when it happens domestically. But let me focus on the foreign outsourcing. So people use the term offshoring for that.
What happens if you take some manual tasks such as assembly and offshore them to China? Well, you are taking these tasks away from the workers who used to perform them, and those workers have to go and do something else, either be unemployed or do something else in which they’re not going to receive as much wage, or they might even go and push other people’s wages down. What happens if you automate them? Well, you take those tasks and instead of workers in China, now it’s machinery like robots who are doing it. Same implications. So the two are very, very similar.
So offshoring and automation have been at the forefront of this decline in the real incomes of low education, non-college workers. But automation has been with us forever. If you look at the 1950s, sixties, seventies, there’s a lot of automation in the US then as well. But what’s different is that automation was combined with lots of other users of technologies that were creating new tasks, new responsibilities, new activities for workers. So workers were displaced from tasks, some tasks, and were finding other new ones. That is both technological and institutional. So for example, if you look at Germany, what you see is that when German companies such as BMW or Volkswagen introduce robots, they then take the blue collar workers who used to perform these tasks that are now assigned to robots, and they train them even more so that they become technical employees.
Why do they do that? Well, first of all, because they have an apprenticeship system and these blue collar workers have a lot of firm specific, industry specific, very sophisticated skills, and they don’t want to lose them. There’s labor shortage in Germany. So again, that increases the value of these workers. But also the whole thing is embedded in institutional setup with work councils and unions that encourages firms to find productive users for this labor. In the US, the same robotization processes associated with these workers losing their jobs. That’s the institutional aspect. But you also see the technological aspect. The German companies that are doing this are also finding new software and new machinery to make these workers more productive.
Nick Hanauer:
Whereas we’re doing stock buybacks.
David Goldstein:
Yeah, exactly.
Daron Acemoglu:
Well, so then the question is, why did we get into this different path? So if this alternative path, not everything is rosy in Germany. Germans are struggling with productivity problems, especially in the service sector, but alternative ways of approaching technology and manufacturing is possible. I would go further and I would say alternative ways of choosing the direction of digital technologies was feasible. So we chose one.
Alternative ways of prioritizing different things was feasible for businesses, and they chose a particular path as well. So you have to analyze both together. And so what are the critical events? Well, I think two things happened. Two interrelated things happened in the power dimension. First, what you might call the Friedman doctrine or the shareholder value revolution, which is a symptom as well as a cause. But the idea that the only thing socially responsible right thing for managers to do is to look after their shareholders. Whatever happens to their customers, whatever happened to their workers, whatever happens to the environment. Those are completely secondary.
I think that started taking hold, and it could take hold precisely because countervailing power started getting weaker around the late 1970s and 1980s. Unions were already declining because of some amount of the industrialization, but then there was a turning point with the Petco strike being beaten by President Reagan. Regulations, government as another pillar of countervailing powers. That started getting weaker and weaker with the deregulation movement.
So you have this double whammy in the business world where businesses were becoming ideologically more inclined to say, well, let’s not share the gains with the workers. Let’s even take active steps to cut labor costs, reduce labor force and so on, at the same time as the constraints that would’ve prevented them to do so, were being dismantled. But it’s not all. That by itself would’ve been bad for labor, but it would not have been as disastrous as what we are experiencing had it not been at the same time that these companies also got the tools for doing this very effectively. And that’s where the digital revolution comes in.
And that’s actually a very interesting story if I may, because if you look at some luminaries in the 1960s, seventies, they thought that computers, including personal computers were going to be the mother of all liberating tools. They were going to empower citizens against governments. They were going to empower small businesses against big business. They were going to empower workers because they were tools in the hands of regular people. And I think they weren’t completely delusional. Digital technologies, just like AI today have that sort of capacity, that potential, but at the end, nobody came up with a monetization model for making money from that approach. Instead, all these more decentralization people, they used to rile against IBM because they thought IBM was trying to control… As big companies always do. They were trying to control information. They were trying to control computing power.
Well, actually, IBM became a small big player. Other bigger players such as Microsoft, Oracle and other software developers came up with ways of using the software for office automation. And then we came up with ways of putting this digital technology into other machines such as robots that could then the next stage of, that could do the next stage of automation. And while we did not invest enough in using digital technologies for creating new tasks for workers. So it’s the double whammy here is that the tech industry turned more anti-worker precisely when the business sector had the ability to turn anti-worker.
David Goldstein:
And it didn’t have to turn out this way. These were choices we made.
Daron Acemoglu:
It did not have to turn out this way.
Nick Hanauer:
It’s not deterministic.
Daron Acemoglu:
Absolutely. That’s the choice is the most important concept in the book. If you look at Scandinavian countries where unions remain strong, they have used many of these same technologies, but you don’t see the same increase in inequality. And if you look at places where they’ve made different choices about technology, you see different paths as well. So there are a lot of choices and those choices are feasible and very consequential.
Nick Hanauer:
Yeah, it’s so interesting. And I guess from our point of view is that the lubrication for these transformations really was that sort of generalized neoliberal view about economic cause and effect and that shareholder value max because if you believe… So, the evil part of shareholder value maximization is not claiming that the only purpose of the corporation is to enrich shareholders. The evil part is the second part of that assertion, which is, and by doing so, they benefit everybody. That’s the evil part, right?
Daron Acemoglu:
Right. I think the second part is what fools people, but the first part is evil too.
Nick Hanauer:
Yeah. Okay, of course. But if all you said was screw the poor, business, people just make as much money as you can. Well, that’s one particular social construct. If you’re making effectively… That’s a normative claim. If you’re making a positive claim that when business people screw the poor, that will be good for the poor, that’s a much more powerful claim. And if people believe it, that is what lubricates all of those institutional rearrangements that benefit the few and harm the many. And that’s-
Daron Acemoglu:
Absolutely 100%. But let me make three points to amplifying your arguments, but one sort of putting a somewhat more complex historical perspective on it as well. So the two that amplify your points. First of all, you’re a hundred percent right, but it’s not new. So if you look at other periods in which there were major technological choices and transitions with important distributional consequences, similar arguments were made. So for example, when the enclosures were ongoing in British agriculture where large landowners, with the support of parliament, were compelling others to give up their land or their customary rights to open fields and common lands. Their argument was, this is going to improve agricultural efficiency and everybody is going to benefit. Even slavery was justified on similar terms in the US South. So the sort of the conceit that, okay, we’re going to enrich ourselves, but ultimately we’re going to benefit.
That’s been discovered by many, many people independently. So it’s not a genius to discover that. But the shareholder value revolution was a particularly powerful version of it. And what makes it pernicious in my mind is that I think in functioning organizations, there is a lot of what some economists like George Akerlof called gift exchange. You do well and you share those gains with your workers.
And I think that’s a normal human social cooperative instinct. You don’t say-
Nick Hanauer:
Social reciprocity.
Daron Acemoglu:
Social reciprocity. That’s exactly right. So most of us, unless we are taught to do that or we have a bit of a sociopathy or there’s some special circumstances so that we see the people not as our ingroup, we don’t say, I’m doing so well and screw you, I’m not even going to give you crumbs. Most people would say, well, you are part of this organization and we’re doing so well, so let’s share some of that.
So that was a powerful engine of shared prosperity throughout the 20th century. It was partly selfish because when you share these things, then you motivate your workers. That’s why Henry Ford was at the forefront of introducing such high wages to motivate his workers, to reduce absenteeism, to create a less conflictual environment, which was good for stopping unionization as well.
So shareholder value revolution really gave a justification or an impulse to many managers to say, no, no, we’re not going to do that. The right thing is to hold workers down to as low a wage as we can. That’s the most pernicious part to me.
Nick Hanauer:
Yeah, no, a hundred percent.
Daron Acemoglu:
Here’s the complicating element. It was more attractive and more natural for managers in the United States to do that because management, labor relations in the US have been always conflictual. So the secret sauce in the Nordic countries is strong unions, but it’s also that it’s a cooperative arrangement that has emerged. The corporatist model may have some unpleasant features, but it cultivated that corporation. And our corporation has never emerged to the same extent in the UK. And that is something for which both managers who always fought unions and unions who took sometimes very anti-business attitudes, they’re both to blame for that. And that creates a bigger problem for the US because we need, in the future, in the age of AI, we need labor voice, but how are we going to construct it? How are we going to make management listen?
Nick Hanauer:
Right. Yeah. So one of the other questions I have is… I think we both substantially agree with your thesis that the advance of technology and the advance of human welfare are very loosely correlated.
Daron Acemoglu:
Yeah. It’s a conditional correlation.
Nick Hanauer:
Yeah. And that most of the time when we have a new technology, a few people win and everybody else gets screwed. That’s sort of been the story of history until people rearrange institutions and grab power back and claw a little bit back for themselves. But at the same time, we do want to encourage lots of innovation. We don’t want to be Philistines. That innovation is, of course, at the end, the source of improving human welfare in many ways. And we want as much of that as we can get within the constraints of finding ways to make sure that everybody benefits from it. I guess this is our benevolent dictator question, Daron. If you were in charge and were free of political constraints, what would you do if you ran the zoo? What would you do?
Daron Acemoglu:
Well, so let me actually step back for a second and give one more layer of our theory, which I think holds part of the answer to your excellent question, Nick. So one, we maintain that technology is highly malleable, which means that you can use technology in many different ways. For example, for AI, we can use AI and for more automation, more data collection, more surveillance, or we can use it to make workers more productive, give them better information, more autonomy, more agency in how they choose to use their unique skills. So that malleability means two things.
A, choice really matters, and we have real choices. And B, the key is not to slow down innovation, and Simon and I never call for any slowdown in innovation. It’s to change its direction. So we want more innovation, and I’m actually very open to having more automation so long as at the same time, we also use these technologies to create new tasks and new capabilities for workers. So that automation is accompanied by things that are good for workers and good for the economy, also at the same time. So that’s why our perspective is you have to be very pro innovation, but we have to make sure that that direction of innovation, the direction of that innovation, is aligned with creating more demand for labor and better social outcomes.
David Goldstein:
And in the short term, in the here and now, not in the future. I know you point out in the book that however much… I certainly have benefited from the industrial Revolution, but it was a hundred years of immiseration for English workers before they started to see the benefits.
Daron Acemoglu:
Absolutely. And I think the industrial revolution is so, so important, so critical, and so often misunderstood. First of all, it is true, exactly, like you say, we are all immensely fortunate to be living 300 years, 250 years after the start of the industrial revolution because our lives are so much more comfortable, so much more prosperous, so much healthier because of the application of industrial technologies and scientific knowledge to every aspect of our lives. No question about that.
But those who say, oh, look at that, that was such a smooth process are also wrong. For 100 years, the gains were captured by a few entrepreneurs and capital owners. Real incomes stagnated or improved only very little. Real wages per hour may have declined because working conditions got deteriorated and people started working about 20, 25% more hours a week. Life expectancy fell sharply because people got pushed into incredibly horrendously dirty, polluted, infested with infectious diseases, cities, and just the whole existence was horrible for people.
The modern factory, the origin of the modern factory is a draconian place. It’s like amazingly harsh, very unhealthy, incredibly long hours with huge discipline, very little autonomy. It’s all of these things that people had to put up with. And why? Because the direction of technology was cost saving, cost saving, automation, automation. There was no worker voice. Unions were heavily prosecuted in Britain. There was no democracy. When Chartists wanted to have some rights, they were all put in jail. So in that institutional environment, it’s no surprise that it took 100 years for improvements in the technology to translate into anything approaching higher wages for most people in Britain.
David Goldstein:
And a hundred years after really the worst of it, it actually, the worst of the industrial industrial revolution is around the time that Marx and Engels are writing. But less than a hundred years later, 70 years later or 80 years later, whatever John Maynard Keynes writes the Economic Possibilities for our Grandchildren. Is that what it was called?
Daron Acemoglu:
Yes. Economic-
David Goldstein:
In which he imagines a not too distant future, really now, where he imagines we’d have 15 hour work weeks and everybody would be living in comfort that he could imagine that then is amazing. That it’s actually gotten worse, is incredibly disillusioning.
Daron Acemoglu:
It is. Keynes was an amazing economist, amazing social thinker, and that essay is really brilliant. Everybody should read it. And he got so much right and so much wrong in that essay. And I think it is telling.
Nick Hanauer:
Yeah, that’s so true.
Daron Acemoglu:
He got it bang on the head that industrial productivity would grow about 2% a year, and that there will be a lot of automations that we would need much less labor to perform those industrial tasks. And then from there, he jumped onto two mistaken conclusions, one of them provably wrong, which is that then the only thing that could happen is that there will be less work technological unemployment. And hence he said, either we are going to have many fewer people working, or it would be better to have everybody work 15 or 10 hours. And then in the rest of the time, we would take leisure and we would go to museums and enjoy the finer things in life.
So he was wrong on the first because he did not think about new tasks. He did not think about what Simon and I emphasized so much. The choice element, the fact that we can use technology for creating lots of things in which humans are going to be centrally involved in. He didn’t get that. And that’s really the key story of the next 50 years, the period of shared prosperity. Those were the engines of shared prosperity.
But second, I think he was also wrong, on multiple fronts, in how he imagined a happy society could be created. So there is a book out by a former advisor to James Corbin called Fully Automated Luxury Communism. I think the title says it all. And I think that pushes the Keynsian view to the extreme and says, well, we should invest more in robots and we can have everything automated. And then out of that we can create a version of communism in which everybody has abundance. Well, there are two problems with that. First of all, I think that’s not what most people want.
Nick Hanauer:
Probably more than two.
Daron Acemoglu:
Since time is short. Let me just focus on one.
Well, one is that it’s not feasible-
Nick Hanauer:
Yeah okay.
Daron Acemoglu:
But this isn’t utopia. But one is that’s not what most people want. Most people actually want to be busy. They want to contribute to society. And if you just give them video games and income, that’s worse than bread and circuses in the Roman Empire. But second is if we have the power structure that we have today, do you think Elon Musk and Mark Zuckerberg and Sam Altman are going to be, “oh, well, we are the geniuses. We’re creating everything. We’re going to give everything away.”
Nick Hanauer:
Not likely.
Daron Acemoglu:
Nothing like the political economy of that is right.
Nick Hanauer:
Right.
David Goldstein:
So great. So Keynes, like most economists, he did not acknowledge the role of power in economics. Getting back to the title of your book, so much of this has to do with power, doesn’t it?
Daron Acemoglu:
Absolutely. Absolutely.
Nick Hanauer:
So just a couple of last questions Daron. We hate to put you on the spot, but if you were a benevolent dictator-
Daron Acemoglu:
I would resign. I don’t want to be a dictator.
Nick Hanauer:
Yeah, okay. But if you were, and I realize you’re not a policy person, but what are the two or three things that you would suggest we do about this problem?
Daron Acemoglu:
Actually, we do play that game with Simon. Simon and I play that game. And we do play that game for, first of all, I think people always ask James Robinson and I, when we wrote Why Nations Fail, we made a very explicit decision not to have a policy chapter because-
David Goldstein:
You hear that Nick, no policy chapter.
Nick Hanauer:
I know, I know, I know.
Daron Acemoglu:
Well, because that was about the longer term issues that we thought we could get away with it, and we thought not having a policy chapter would have a chance of making the book more durable and not completely outdated within a month. But on this topic, I think we couldn’t get away with it. And so we obliged. But also there’s a reason. I think we are at a critical point with the advances in generative AI we are seeing. Questions about who is going to control AI, who’s going to control data, what direction of technological change we’re going to have. I think those are really critical. Those are the ones we have to make in real time. They’re not going to be made in 10 years. They’re going to be made in the next few years. So that’s why we do have that chapter.
And let me tell you, here is my summary. I would say 50% of the task ahead is to change the narrative. If Simon and I have contributed in a little way to the view that it is both desirable and feasible to have more pro-human, human complimentary technologies, technologies that elevate humans, that keep humans in the loop and increase their productivity, if that is feasible and desirable and that’s accepted, that already changes everything.
And then once we have that accepted, then we have to talk about how we achieve that. And first of all, you cannot achieve that without tackling the power issue. So you need to have some countervailing powers. The institutional and civil society dynamics have to be such that there are more voices. So no, it doesn’t look good, and it’s not good when finally the US Senate wakes up to the issue of AI, they have a meeting with Guess who the heads of the top five tech companies, no worker voice, but these AI tools are going to shape the future of hundreds of millions of workers in the United States and billions of workers around the world. What about their voice? What about their perspective? So that doesn’t look good and it’s not good. And now that has to change. So you have to have a broader conversation with more countervailing powers so that it’s not just one technological leader who imposes his wish or his will or his ideology on the rest of us.
And then we also come up with some specific policy proposals with the full acknowledgement that these may or may not be the best policy proposals. But once we change the narrative, once we sort of agree on what is feasible and desirable, we can then have a much more productive debate about what are the best ways of going there. So among the ideas we float are a few that, let me mention just very briefly.
First of all, our tax system is screwed. And it’s not just unfair, but it also encourages excessive automation. We tax labor. If you hire more workers as a company, we tax you at 25% plus plus because you also have to pay healthcare costs and everything. If you instead you buy a machine to do the same task, you write off a lot of that, you pay something like 5% tax. So that creates a powerful incentive to use machines instead of humans, and it’s a regressive tax. So I think changing that is a no brainer.
Who’s going to control data? I think that’s going to be critical. So we need to have data regulation and we need to compensate people. If their creative data or other type of data are being used, then we also need to create regulation that protects them. If they don’t want that data being used right now, a lot of profits in the tech sector come from their ability to expropriate other people’s data.
We also recognize that there are natural forces, especially under the influence of the shareholder value revolution that will push towards automation. So there may be a need for more specific policies like the set of policies that that were implemented in encouraging green innovation in the energy sector that specifically targets more human complimentary users of data, more ways in which you use AI technologies, for example, to provide information to people rather than misinformation. And then finally, we also talk about some even more radical ideas such as breaking up of the big tech. We have never seen as big corporations as the tech companies of today in human history. I don’t think breaking them up is a solution by itself, but it may be a very important part of the solution. Again, going back to the issue of power, because if you have such powerful companies, they’re going to shape everything. And one way you have to start thinking about is, can we reduce their power?
Nick Hanauer:
Yeah. No. I could not agree with you more.
David Goldstein:
Really. You don’t trust like Elon Musk to make the right best decisions for the rest of us.
Daron Acemoglu:
Elon, I do. He’s special.
David Goldstein:
Yeah, the same people. It’s amazing. It’s amazing that in one sentence, they’re warning you about the dangers of self-aware AI, and then in the next sentence they tell you, but I’m investing billions in developing it. I’m doing it anyway. So I don’t know. Having these decisions made by a bunch of sociopaths may not be the best choice.
Daron Acemoglu:
Well, and then look, going back to the issue of no countervailing powers. So then what happens is US senators say, “Well, Mr. Sam Altman, can you regulate yourself?” And he says, “Oh, that’s not a bad idea.”
Nick Hanauer:
What could go wrong? What could go wrong? Daron, one final question. Why do you do this work?
Daron Acemoglu:
Oh, it’s so much fun. Why not?
But no, seriously, look, I am absolutely passionate about this. I love doing it. I enjoy understanding the things that I think are formative for our future. This is what drew me to social science. As a teenager, I thought I could study economics and then that would enable me to understand issues of democracy, economic growth, economic development. Little did I know that that’s not what economists did in most of their time.
But these are critical questions. They’re exciting questions. They are consequential questions, and I love studying them. But also, I think these are really important issues for people of all backgrounds to be informed about so that they can be part of the debate. And that’s part of the reason why we write for a general public. I have written a lot for academic journals, for my colleagues. But the driver of this book project was to reach a broader audience. And thanks for helping us do that via this podcast.
Nick Hanauer:
Yeah, that’s fantastic. Well, thank you so much for being with us today. It’s absolutely fascinating. We love the book. We love-
Daron Acemoglu:
Thanks for having me. This was a great conversation. I am really grateful for having been given the opportunity to be here.
David Goldstein:
I think Nick, one of the great things about this book is it raises that issue of power, which as we’ve talked about, is absent from orthodox economics and overlooked and really actually they roll their eyes at. But also, and it’s something we’ve emphasized again and again and again, how much economics is a choice. A lot of the policies that have led to where we are with the radical inequality and the falling wages, and all the other, and really the Trumpism that’s going on now, the threats to our democracy and to our climate. All of those are the results of choices we have made. And these are choices that have been made… And I think this is an important part of the book we didn’t talk enough about… These are choices we have made as a result of the power of ideas and the power of persuasion.
And I think it’s really important to remember that those three decades after World War II. When we built the American middle class, and we had, as he mentioned, wages at the bottom, rising faster than wages at the top. And you had this broad prosperity, not for everybody. Obviously, we had an incredibly racist country, so it wasn’t evenly distributed, but you saw it in the United States and in Europe and throughout the industrialized world. It was not unprecedented, but certainly very different from what we’re seeing today.
That wasn’t inevitable. That was the result of decisions that were made based on a dominant economic and political narrative that was built with intention and by choice, and was extremely deliberate. And where we got to since the mid 1970s was also a choice. And it is based on a narrative that was built deliberately to do the things that it did.
Nick Hanauer:
Right. It was a narrative designed to enrich the few and impoverished the many. And it was very successful.
David Goldstein:
Not just enrich, but empower the few.
Nick Hanauer:
That’s right.
David Goldstein:
And this is an important thing about neoliberalism, is that for all this stuff we hear from the right about all liberal elites and all that, it is an incredibly elitist ideology. It is an ideology that believes that if you are financially successful, it’s because you are more meritorious. If you’re poor, it’s because you’re less meritorious. You get exactly what you deserve. But more importantly, that it is the wealthy who should be running the country. It is the wealthy who should be making these decisions. And if Elon Musk decides that a self-aware AI that will destroy humanity serves his interest, or he’s just interested to see what it looks like if he flips the switch and turns it live, well, he should make that decision, not us.
Nick Hanauer:
Yeah, yeah. No, it’s so true. What I think is exciting is the way in which… We are in this great moment of economic transition from the trickle down neoliberal world to the middle out world. Daron is a middle outer, if I ever saw one, although he may not know it. Because, for sure, when you’re talking about power, you’re talking about the right stuff. When you’re thinking about these problems in the way that he and his coauthor have, this is the central point of middle out economics, is that a thriving middle class is the cause of economic growth, not its consequence. That these technologies and institutions need to be organized in ways that benefit the broad middle, not a tiny minority of people at the tippy top.
David Goldstein:
And in a very real way, Nick, that this book is a continuation of Why Nations Fail. One of the main thesis in Why Nations Fail, which is you want a politically, our economy does better when we have a politically inclusive democracy that that is key to building a broadly prosperous economy. And when you have a politically inclusive democracy, that democracy is a form of countervailing power.
Nick Hanauer:
It is. And that democracy will build economic institutions, which include everybody and generate more broad-based growth.
David Goldstein:
And unfortunately, you can choose between that virtuous cycle, or you can have the vicious circle we’ve seen in the past 40 years, where we have people who are giving up on democracy and turning against it because their lives are going to shit. Or at the very least, aren’t going the way they were raised to think they were going to go. They didn’t get what they were raised to expect. And it can be really dangerous because why nations fail-
Nick Hanauer:
That’s right.
David Goldstein:
Nations fail.
Nick Hanauer:
That’s right. True words were never said.
David Goldstein:
Again, we highly recommend it. Power and Progress: Our Thousand Tear Struggle Over Technology and Prosperity from their Daron Acemoglu and Simon Johnson. You can purchase it at your local independent bookstore or that giant online powerful monopoly who shall not be named. And as always, there is a link in our show notes.
Speaker 3:
Pitchfork Economics is produced by Civic Ventures. If you like the show, make sure to subscribe, rate, and review us wherever you get your podcasts. Find us on Twitter and Facebook at Civic Action and Nick Hanauer. Follow our writing on Medium at Civic Skunk Works and peek behind the podcast scenes on Instagram at Pitchfork Economics. As always, from our team at Civic Ventures, thanks for listening. See you next week.