Artificial intelligence, once confined to the realm of science fiction, is changing our lives. Cars are driving themselves. Drones are being programmed to deliver packages. Computers are learning to diagnose diseases. In a recent book, the economists Erik Brynjolfsson and Andrew McAfee describe these recent advances as examples of the beginning of what they call “the second machine age.”
The very name – the first machine age was the Industrial Revolution – suggests an epochal shift. And, indeed, if the predictions are to be believed, these technological advances could have profound implications for the way we live.
One common forecast is that as ever-more advanced robots substitute workers, the cost of labor will become less important, and manufacturing will move back to rich countries. Another is that increasingly intelligent machines will reduce the demand for advanced skills, and that the economic advantage of having these skills will decline as a result.
The first of these two hypotheses remains far-fetched. But there is evidence that the second has already started to come to pass, with serious implications for the way that modern economies have attempted to meet the challenges of globalization.
To be sure, there has been plenty of anecdotal evidence of “reshoring” – the relocation of jobs from low-wage countries to high-wage economies. Apple is shifting some manufacturing from China to Silicon Valley; Airtex Design Group is moving part of its textile production from China back to the United States. In a recent survey of 384 firms in the eurozone by the management consultancy PricewaterhouseCoopers, two-thirds of the respondents said that they had reshored some activities during the past year, and 50% plan to do so in the next.
But when one looks at the data, there is no sign of reshoring. Indeed, the trend is continuing in the opposite direction. Offshoring dropped during the Great Recession that followed the 2008 global financial crisis, but quickly rebounded, accelerating past pre-crisis levels. For the moment, the return of manufacturing to rich countries remains a prediction, not an outcome.
Evaluating the second hypothesis is more complicated. At first glance, the evidence supports the possibility that demand for advanced skills is falling. With the exceptions of the US and Germany, the wage gap between skilled and unskilled workers has been declining in all Western countries in the last 17 years.
One possible explanation is that education levels in Europe outstripped the pace of technological change, oversupplying the market for advanced skills. In Austria, the share of people with a university degree or its equivalent increased by 250% between 1996 and 2012. In the United Kingdom and Italy, it almost doubled. In Spain, it jumped by 70%, and in France by 60%. By comparison, in the US and Germany, the share of the population with a tertiary education rose by a more modest 25%.
It is also possible, however, that the drop in the wage gap between skilled and unskilled workers represents competition from increasingly intelligent machines. Here, the US is a case in point. Because educational attainment has been advancing only modestly in the US since the turn of the century, we would expect the wage gap to be rising steeply, as it did in the 1980s and 1990s. Instead, it has remained largely unchanged, and it is the unemployment rate among skilled workers that is on the rise, doubling in the US and the United Kingdom from 2000 to 2012.
Until the 1980s, about 70% of income went to labor income and 30% to capital income. But, since then, the share of income going to labor has declined in all rich countries. It is now at about 58% of GDP. According to research by the economists Loukas Karabarbounis and Brent Neiman, half of this decline is the result of cheaper information technology, which has enabled firms to replace workers with computers.
The implications are serious. If these are indeed the first signs of the second machine age, it is possible that we have been fighting the wrong battle. As the scarcity of human capital declines in importance, the rapid expansion of education may not be the answer to the challenges of globalization that we hoped it would be.
Published in collaboration with Project Syndicate
Author: Dalia Marin is Chair of International Economics at the University of Munich
Image: Machines in a manufacturing plant. REUTERS