May 23, 2023
When I saw that two of the country’s most prominent economists wrote a book on “our 1000-year struggle over technology and prosperity,” I expected a lot. I was disappointed. To be clear, there is much here to like and I’m sure that most readers will get much from it, as I did. But, the book fails to follow through adequately on the key point in its analysis, which is that the gains from technology are a matter of struggle, not an outcome given by the technology itself.
I’ll start with the positives. The book gives a cursory, but useful, account of the major developments in technology going back more than a thousand years. Some of their discussion deals with the origins of agriculture, an innovation that goes back many thousands of years. However, most of the book does describe events in the promised thousand-year horizon.
It points out that there were important technological innovations in the Middle Ages, which did allow for very modest gains in productivity and living standards across much of Europe until the 13th or 14th century. At that point, the rate of increase accelerated, although it was still much slower than in later centuries.
One of the points it makes, which was underappreciated (at least by me), was the extent to which the gains in this period were siphoned off by the church. The huge cathedrals and monasteries constructed in this period absorbed a huge chunk of the surplus produced in agriculture. This both depressed living standards and prevented resources from going to investments that would increase productivity.
As technology continued to advance in the subsequent centuries, there were modest gains in living standards for large sections of the population as the demands of the church were reined in. However, the book makes clear that there was no automatic transmission from improvements in technology to gains in living standards for the bulk of the population.
In particular, the book points out that the early years of the industrial revolution in 19th century England were associated with a deterioration in living standards for large segments of the working class. Factory workers, and especially children, were forced to work longer hours under worse conditions than ever would have been the case in agriculture. In addition, the living conditions in cities were far more unhealthy than what they faced in the countryside.
When conditions for the English working class did subsequently improve in the second half of the 19th century, it was due to the growing political power of the working class, as well as the growing importance of labor unions.
They also point out that the direction of technology itself is very much determined by power relations. In particular, they note that the enclosure movement in England was not, in fact, about allowing for the transition to more modern crop rotations. There were many places where improved crop rotation systems were adopted in open fields. Rather, enclosure was about giving landlords more control over land and displacing the peasantry.
None of this would be new to people familiar with the history of this period, but it is still refreshing to see prominent economists make these points. The distribution of the benefits of technology is far from being determined by strictly economic factors. It depends very much on the institutional structure and power relations in society.
It is Not Just the Masses Who Rely on Political Power to Secure the Gains from Technology
However, having said this, my major criticism of Acemoglu and Johnson is that they fail to carry this critique far enough. While they discuss extensively the factors that have allowed for gains in the past for ordinary workers, and which could go further in the future, they don’t consider the ways in which the winners have structured the economy to ensure that they get the bulk of the benefits of growth. In effect, the book treats the upward redistribution as a natural outcome of the development of the market and technology, whereas it requires interventions to ensure that gains are widely shared.
Just to take the most obvious case (and my favorite target), the money that goes to the top end of the income distribution due to patent and copyright monopolies is entirely due to the winners’ ability to write the laws on intellectual products in ways that benefit them. If the government didn’t threaten to imprison people who copied Microsoft software without his permission, Bill Gates would still be working for a living. It wasn’t technology that created five Moderna billionaires, it was the fact that the government paid the company to develop a Covid vaccine, and then allowed Moderna to have monopoly control over its distribution.
If it is difficult to understand the idea that government policy on technology, rather than the technology itself, created a huge amount on inequality, imagine a world where we had no patent and copyright monopolies and non-disclosure agreements were unenforceable. Arguably, we would have seen less technological progress, depending on what, if any, alternative incentive mechanism we used; but surely we would see nothing like the inequality that has resulted from preserving and expanding these monopolies over the last four decades.
Similarly, it was not just technology that allowed us to subject manufacturing workers to competition from their much lower paid counterparts, but protected highly paid professionals. There is no technical reason that we can’t all have Zoom consultations with doctors in India who get paid one-tenth as much as doctors in the United States. Globalization has not had as much impact on the pay of highly paid professions because of their power, not because of technology.
And of course, the big bucks made by the top people in finance is not due to the wonders of technology, but the fact that a politically powerful industry was able to rig the rules. The too-big-to-fail insurance on vivid display in the Great Financial Crisis, and again following the collapse of the Silicon Valley Bank, is only part of the story.
We have the technology to make the financial system far more efficient, starting with Federal Reserve banking accounts and running into just about every area of finance, but it doesn’t happen because the big players in the financial industry don’t want to see a hit to their income. No one knows this better than Simon Johnson, who has written extensively on the abuses and corruption in the financial industry, but for some reason this is mentioned only in passing in Power and Progress.
These points are not nit-picking. It is not just the masses who need to change the rules to get their fair share of the gains from productivity growth, it is also the rich who change the rules to their benefit. This point is essential both as a matter of logic and politics.
It is also important to recognize that we improve the living standards of ordinary workers not only by getting them more money, but by giving the rich less. If the rich have less money to buy multiple houses, then housing is cheaper for everyone else. And, if they have fewer personal servants fulfilling their whims, there are more workers available to meet the needs of the rest of the population.
There are no preset given rules of a market economy. We make them up as we go along. If we want to see more equality and a system where the vast majority, and not just a lucky few, share in the benefits of progress, we can go two routes. One is the route Acemoglu and Johnson advocate throughout the book: strengthen workers’ bargaining power and expand the welfare state. But, we can also work the other side; weaken the institutional structures the rich have been putting in place to make themselves richer. We should, of course, be opportunistic and make advances on both sides wherever we can, but it makes no sense to ignore the other side.
Will AI and Related Technologies Change the World?
I have one other small bone to pick with Acemoglu and Johnson. They look at the developments in AI and are largely dismissive of the idea that it will lead to another productivity boom. As they note, we have been hearing the promises of a computer-driven productivity boom for decades and have yet to see it.
I share their skepticism, having listened to the same promises. However, I will cautiously suggest that this time might be different. According to the analyses of Nick Bloom and his colleagues, between 25-30 percent of work days are now remote. This is up from around 5 percent before the pandemic.
This is a huge change in the lives of tens of millions of people. They are saving hundreds of hours a year in commuting time and thousands of dollars in commuting related expenses. These benefits are not picked up in our productivity or GDP data. Commuting time is not counted as work hours, even though most commuters would probably rather be spending that time at work than commuting.
Also, expenses associated with commuting, like gas, train, and bus fares, as well as dry-cleaning services and the costs of business clothes, are just treated like any other consumption. Most immediately, less spending on these items would actually be treated as a drop in GDP.
While the option to work at home is disproportionately available to more educated workers, it clearly goes beyond a small elite. In this respect, it is worth noting that after the surge in remote work and an unprecedented rate of job switching, a Conference Board survey shows workers report their job satisfaction is the highest in the four decades the survey has been taken.
To my view, we have not yet reaped the full benefits of this revolution in the opportunity to work remotely. The option to do things remotely is likely to be extended in many different directions, most obviously with telemedicine.
Again, many of these benefits will not be picked up in GDP and productivity data, but they will make a real difference in people’s lives. The fact that a person in bad health doesn’t have to travel across town or across country to consult with a medical professional makes a huge difference to that person and their family, but does not show up anywhere in the GDP accounts. So, while I respect Acemoglu and Johnson’s skepticism on this topic, I think there is a real basis for guarded optimism.
In short, Acemoglu and Johnson have given us much to think about in this book. It didn’t go everywhere I would have liked to see it go, but that’s what happens when you talk about a thousand years.
 I was very happy to see Johnson and Simon take up one of my recurring themes in this area, Section 230 protection for Internet platforms. This both allows for platforms to have inordinate power in shaping our politics and allows for their shareholders and top execs to get rich. They go in a somewhat different direction for a remedy. I would take away the protection for sites that relied on advertising or selling personal information, requiring platforms to respond to takedown notices, as is currently the case with allegations of copyright infringement. They would take it away for promoting incendiary material, but the key point is that we need a serious discussion on the matter, which has not taken place to date.