The Race between Education and Technology

"This book is important because it captures another conventional wisdom with respect to automation: that the best response to the challenge is ‘more’ education. And it does this in a wonderfully clear and engaging way. The book is a reaction to an empirical puzzle that emerged in labour markets during the second half of the 20th century. During that time, particularly in the US, the number of people graduating from colleges and universities increased. And in economics, if the supply of something increases, you would expect the price of it to fall. But the puzzle is that the wages these people received relative to others—the so-called ‘skill premium’—was rising instead. How could that be? The answer is that new technologies were ‘skill-biased’, that people with more years of formal schooling behind them were more able than others to put them to productive use, and so demand rose for their efforts. As a result, even though the supply of skilled people was increasing, the demand for their work soared so much that their wages were still pushed up, relative to those without all that formal schooling. Get the weekly Five Books newsletter This is why the book argues that the best response to technological change is more education: if technology benefits those with skills, then you want to try and make sure that your workforce is as skilled as can be. There is a metaphorical ‘race’ between workers and machines and, as the latter become more capable, you have to give the former more education to keep up. At the start of the 20th century, ‘more’ meant ‘more people’: mass education was the goal, that everyone, whatever their background, should have access to formal schooling. But as the century went on, more started to mean ‘more advanced’; encouraging people to go college and university, to pursue advanced degrees. This book, like my first choice, also supports an optimistic story about the future of work. It suggests that, as technological progress accelerates, there is an effective way to respond: more education. I expect that many policymakers, if they were to read this book, would find themselves nodding along in agreement. My worry, though—and it is a worry increasingly shared by others as well—is that limits to education are starting to emerge. Yes, that is a growing issue. Though the two books actually complement each other quite neatly. A natural question to ask of Katz and Goldin is, ‘Well, why is it that technology was skill-biased over the last few decades?’ And the Levy and Murnane book provides one helpful answer: skilled jobs tend to involve tasks that require faculties like creativity, judgment, empathy, and these are hard to capture in a set of explicit instructions for a machine to follow. In short, to use the old distinction, they tend to involve ‘non-routine’ tasks, and that is why they have proven so hard to automate. That is why there has been so much demand for human beings to do those activities. As an aside, it is worth noting that when I use the word ‘skilled’, I am using it in the economist’s sense of the word: namely, to reflect the amount of formal schooling someone has under their belt. This is a very particular use of the term, and there is clearly a difference between it and a more commonsense use of the word ‘skilled’. That, quite rightly, often has very little to do with education at all. This does not mean either term is wrong, but it is important to be clear exactly what it is that economists have in mind to avoid causing any upset or offence. Finally, remember that is not only activities that skilled workers do that tend to be relatively ‘non-routine’: it is also the case that many of the tasks done by people in lower-paid service roles are ‘non-routine’ as well. For instance, those who work in nursing or social work, or in restaurants and cafés, tend to draw on faculties like manual dexterity or interpersonal skill, and these have been hard to automate until recently as well. This is a very important point. It is a particularly troubling feature of the pandemic that many of these service roles, which have proven so hard to automate in the past, have been hardest hit by the virus. A corner of the labour market that one would have hoped to be an important source of job creation in the future has been completely decimated over the last few months. I want to say a little more, though, about an earlier claim—that there are emerging limits to education as a response to automation. In A World Without Work, I describe two different ways that people might find themselves without work because of technological change. This distinction helps to pin down exactly why I worry about education. One way is what I call ‘frictional technological unemployment’; here, there is still lots of work to be done, the problem is that people are not able to do that work for various reasons. The other is what I call ‘structural technological unemployment’; here, there is simply not enough work to go around, full-stop. The problem with education is that it is an imperfect solution to the former, and a relatively impotent response to the latter. Let me explain. The most familiar reason for frictional technological unemployment is that people don’t have the right skills for the available jobs. Here, more education seems like the appropriate response: we have to give people the skills to keep up in Goldin and Katz’s metaphorical ‘race’ with new technologies. My worry, though, is that policymakers do not always appreciate quite how tough an undertaking this race is for those who want to take part. To begin with, many people in this race are already running as fast as they can: as others have noted, it is very difficult to get more than 90 per cent of people in a country to finish secondary school, or to get more than 50 per cent of people to graduate from university. In turn, the pace of this race is accelerating: basic literacy and numeracy, for instance, might have been enough keep up in this race at the start of the 20th century, as people moved from agriculture into factories, or from factories into offices for the first time, but they are no longer enough. Ever higher qualifications are required to keep up. So that is the first limit to education: that it cannot always solve the skills mismatch that stops some people being able to take up the available work. “If we find ourselves in a world where there are simply not enough jobs to be done, then we will be forced to consider more radical interventions” Crucially, though, there are also other mismatches that might cause frictional technological unemployment: and education, at least traditionally understood, does not adequately address them. One is the ‘place mismatch’, that people simply don’t live in the same place that jobs that are being created. Then there is also an ‘identity mismatch’, where people have a particular conception of themselves, rooted in a particular type of work, and are willing to stay unemployed in order to protect that identity. To see this, think of adult men in the US displaced from manufacturing roles by new technologies. There are some that say they would rather stay unemployed than take up—and it’s a really unfortunate term—so-called ‘pink collar work’, a term designed to capture the fact that many of the jobs that are hard to automate are disproportionately held by women. So, 97.7% of kindergarten and primary school teachers, 92.2% of nurses, and 82.5% of social workers in the US are women. Again, as I said before, is not obvious how education addresses either place-mismatch or the identity-mismatch. This is the second limit to education. The third and final limit to education is that it does not seem to be an effective response to structural technological unemployment. If we find ourselves in a world where there are simply not enough jobs to be done, then we will be forced to consider more radical interventions: a basic income, perhaps, or a job guarantee scheme. So, why is this Katz and Goldin book so interesting? Because it captures, in a clear and elegant way, the prevailing view of how best to respond to automation: through more education. If new technologies are making high-skilled work more valuable and more important, if there is a ‘race’ on between people and machines, then we need to give people the skills required to keep up. But, as I have said, I think this only right up to a limit. A world with less work presents us with problems that more education alone cannot solve."

"What I think is really ingenious about this book is that it squares the circle. In the 1990s Bill Clinton told us that computers had revolutionised what was required from the workforce and that all of a sudden you needed to get a college education to perform in this knowledge-based economy. Goldin and Katz actually point out that the technological changes at the beginning of the 20th century were probably a lot more dramatic. At the beginning of the 20th century you had the spread of electricity, then the advent of radio, the spread of air travel and the growth of television. And as technological change imposed greater and greater skill demands on workers, the skill level of American workers rose and rose. Why? Simultaneously the United States was experiencing the high-school movement. At the start of the 20th century very few people went to high school. Within a few decades most people were going to high school as part of a conscious policy change. Europeans snickered at Americans for the sentimental notion that there should be universal secondary education. They thought that secondary education should be for preparing the small segment of the population that was going to college. Consequently, American workers were able to meet rising technological demands in a way that European workers were not, just when high-school education became more necessary for all sorts of blue-collar work. Factory workers needed to have a fundamental understanding of how electricity worked and farmers needed to be familiar with new agricultural and marketing processes. In the United States the high-school graduation rate rose and rose, as did technological demands, until the graduation rate levelled off in the 1970s. But with the advent of the computer technological demands continued to rise – that bid up the price of skilled labour. That’s Goldin and Katz’s central insight: It wasn’t computers that caused inequality, it was the fact that for this technological revolution, as opposed to the technological revolutions that preceded it, the American K-12 [primary and secondary] education system was no longer keeping up with marketplace demands. Europeans had realised their mistake and started spreading secondary education. Secondary school attendance rates in many Western European countries now exceed those in the United States. College completion rates in many Western European countries exceed those in the United States. Yes. There are no advanced industrial economies in Western Europe that are experiencing economic inequality to the same extent the United States is and there are no countries where inequality is increasing as fast as it is in the United States. The debate is miles behind what needs to be done, but it’s picking up. There’s certainly more discussion of this issue today than there has been in the previous third of a century. For example it’s good that President Obama is calling for raising taxes on incomes above $250,000, but he also needs to talk about laying in some brackets at $1 million and $10 million and $20 million. I think they should go up to 70% top marginal rate. But that’s not really within the political mainstream right now. Here’s something that is in the political mainstream – expanding Pre-K [nursery school] education. Right now just a little more than a quarter of all four-year-olds are receiving Pre-K education. There’s no controversy about the fact that early childhood education vastly improves a child’s chances of succeeding academically. Another issue that I thought was outside the mainstream when I suggested it in my book, but then to my surprise President Obama suggested something very much like it in his State of the Union address, was the idea of price controls on college tuition increases. Obama didn’t put it so bluntly, but he did say he was putting college and universities on notice that if they could not get their tuition increases under control the federal government would use its leverage to force them to do so. Am I allowed to put in a plea for a sixth book here? Larry Bartels’s Unequal Democracy looked at economic performance under presidencies going back to 1948. What he found was under Democrats the greatest income gains went to those at the bottom and tapered off as he went up the income scale. Under Republicans, the greatest income gains went to those at the top and tapered off as you went down the income scale. It correlates perfectly with our cartoon notions of how Democrats and Republicans differ – the idea that Democrats really are the party that cares about people at the bottom and Republicans really are the party that cares about people at the top. Yes I do. I think that for the first time in 33 years there’s a national movement to address the problem. For the first time we have a president who’s not only interested in the problem but willing to discuss it out loud. We have a chairman of the council of economic advisers, Alan Kruger, who not only has looked at this problem extensively in his academic life but is continuing to research and discuss it as chairman. He issued a report a few months ago that looked at the relationship between rising income inequality and falling mobility. Certainly at the discussion level more is going on. But the most difficult task is reviving labour. As I talk about the book to various groups I find that this is the one idea that even liberals resist. They just don’t want to hear that if you really want to address income inequality you have to revive the labour movement – there is no alternative. You can’t achieve this just by sending everybody to college. You want to train as many people as you can to perform more technologically demanding work but to revive the economic condition of the middle class it is going to be necessary for people to have labour unions that defend their interests against bosses. Today it is just about impossible to organise a workplace in which the bosses are determined to keep unions out. In my book I talk about an attempt to unionise a Wal-Mart in Colorado. What that illustrates is how thoroughly the deck is stacked against labour unions today. The laws favour managers and even when the laws don’t favour managers the penalties for violating laws are so pitiful that it’s actually economically irrational for managers to obey the law. This is the most difficult challenge but it has to be done. I guess that gives you a sense of the relative difficulty of various things that I propose. The government has a huge influence over trends towards greater income equality or greater income inequality. This is not just something that the economy does all by itself. There’s a lot of research showing that these trends are heavily influenced by government action. For a long time people didn’t realise that because taxation and government benefits did not have an enormous impact on income inequality – but researchers were looking in the wrong places. Get the weekly Five Books newsletter There are all sorts of other things that the government does that do make a difference. It can’t be a coincidence that Democratic presidents have tended to foster economic equality while Republican presidents have tended to foster economic inequality. Democrats and Republicans govern in a million different ways. Fed policy has enormous impact on income inequality. The fact that Democrats are more tolerant of inflation than Republicans has meaningful distributive consequences. Monetary policy is very influential in this area. Raising the minimum wage is certainly a factor. One aspect of Obamacare that President Obama doesn’t want to talk about is that it’s fairly redistributive. It extends healthcare coverage to lower income people through the expansion of Medicaid. I think the Obama administration is a little nervous about talking about that because they don’t want to sound socialist, but Obamacare will certainly achieve a kind of invisible redistribution – it will change the distribution of benefits and the availability of medical care. The thing to be most hopeful about is that people are concerned about inequality now. They want to talk about it. They want to understand what’s causing this phenomenon. Glib explanations coming from conservatives who would like to deny the Great Divergence aren’t really persuading anybody because they are not only untrue but also unpersuasive. This problem was not created overnight and it’s not going to be solved overnight, but I think that it can be solved and we can start to see some improvement."

"This is a really wonderful book. It gives a masterful outline of the standard economic model, where earnings are proportional to contribution, or to productivity. It highlights in a very clear manner what determines the productivities of different individuals and different groups. It takes its cue from a phrase that the famous Dutch economist, Jan Tinbergen coined. The key idea is that technological changes often increase the demand for more skilled workers, so in order to keep inequality in check you need to have a steady increase in the supply of skilled workers in the economy. He called this “the race between education and technology”. If the race is won by technology, inequality tends to increase, if the race is won by education, inequality tends to decrease. The authors, Claudia Goldin and Larry Katz, show that this is actually a pretty good model in terms of explaining the last 100 years or so of US history. They give an excellent historical account of how the US education system was formed and why it was very progressive, leading to a very large increase in the supply of educated workers, in the first half of the century. This created greater equality in the US than in many other parts of the world. They also point to three things that have changed that picture over the last 30 to 40 years. One is that technology has become even more biased towards more skilled, higher earning workers than before. So, all else being equal, that will tend to increase inequality. Secondly, we’ve been going through a phase of globalisation. Things such as trading with China – where low-skill labour is much cheaper – are putting pressure on low wages. Third, and possibly most important, is that the US education system has been failing terribly at some level. We haven’t been able to increase the share of our youth that completes college or high school. It’s really remarkable, and most people wouldn’t actually guess this, but in the US, the cohorts that had the highest high-school graduation rates were the ones that were graduating in the middle of the 1960s. Our high-school graduation rate has actually been declining since then. If you look at college, it’s the same thing. This is hugely important, and it’s really quite shocking. It has a major effect on inequality, because it is making skills much more scarce then they should be. They do discuss it, but nobody knows. It’s not a monocausal, simple story. It’s not that we’re spending less. In fact, we are spending more. It’s certainly not that college is not valued, it’s valued a lot. The college premium – what college graduates earn relative to high-school graduates – has been increasing rapidly. It’s not that the US is not investing enough in low-income schools. There has been a lot of investment in low-income schools. Not just free lunches, but lots of grants and other forms of spending from both states and the federal government. No, there is no easy solution. But it is important that this becomes a top priority. There are many things that can be done. Reforming schools and spending more on schools are not silver bullets, but they will certainly help. Yes, but it cuts both ways. The US has been a land of opportunity, so it has attracted skilled engineers, scientists, and PhDs. On the other hand, one of the contributors to low high-school and college completion rates is that the US is open to immigrants of all sorts, and the integration process is often slow. If you look at high-school and college completion rates of immigrants, especially Hispanic immigrants, they are much lower that those of non-immigrants."