There has been a bizarre debate in national policy circles in recent years, with many raising the prospect that the development of robots and other technologies will lead to mass unemployment and lower living standards. The debate is bizarre for two reasons. First, productivity growth has actually been quite slow in recent years; this is the opposite of the “robots replacing workers’’ story. Second, faster productivity growth should mean higher wages and better living standards in a well-working economy.
Dean Baker is the co-director of the Center for Economic and Policy Research.
Taking these points in turn, productivity growth has slowed sharply in the last decade. From 2005 to 2014 productivity growth has averaged just 1.4 percent annually. This is down from 2.9 percent from 1995 to 2005. In the last two years productivity growth has been even slower, averaging less than one percent. By comparison, in the post-World War II Golden Age from 1947 to 1973, productivity growth averaged almost three percent.
Productivity growth is a direct measure of the rate at which technology is increasing output per worker hour. If robots and other technologies are making workers obsolete, we should be seeing sharp increases in the rate of productivity growth, since they are allowing us to produce goods and services with few or no workers. Instead, slower productivity growth means technology is displacing workers less quickly than in 1995–2005.
Productivity could speed up, but there is no obvious reason that it will. It is also worth noting that economists have been terrible at predicting changes in productivity trends. No economists foresaw the slowdown that followed the end of the Golden Age and very few predicted the upturn that began in 1995 or the new slowdown in 2005. This history should be grounds for viewing any stories of an imminent productivity boom skeptically.
The second point is that more rapid productivity growth has historically been associated with rising living standards and low unemployment. From 1947 to 1973 we saw low unemployment and rapidly rising living standards as the gains from productivity growth were broadly shared. The rate of unemployment averaged less than five percent over this quarter-century. Median family income more than doubled.
There is no reason in principle why we should not expect to see comparable gains in living standards if we were to suddenly see an acceleration in the rate of productivity growth to the Golden Age pace or faster. Higher productivity growth could mean that workers both have more income and more leisure. An important fact often left out of discussions on productivity and jobs is that the length of the workweek and work year is not fixed. In the 19th century in the United States, 60- to 70-hour workweeks were standard. Other countries have long surpassed the U.S. in shortening workweeks and offering more vacations and paid days off. According to the Organisation for Economic Co-operation and Development, workers in Germany on average work 25 percent fewer hours than workers in the U.S. This means that if we adopted German work patterns we would have one-third more jobs for the same level of demand in the economy.
The key to ensuring that workers share in the benefits of productivity growth is to have a full employment policy. This requires monetary, fiscal, and trade policies consistent with full employment. For example, a lower-valued dollar that would bring the trade deficit closer to balance could create millions of jobs. Also, the Federal Reserve Board should place a priority on high employment and be prepared to delay interest-rate hikes until there is clear evidence of inflation. Full employment gives workers the bargaining power they need to have wages keep pace with productivity growth, as happened in the low unemployment years of the late 1990s.
Unfortunately, there is little commitment from political leaders to pursue full employment policy. So the gains from improvements in productivity from robots and other technologies might well go overwhelmingly to corporations and a small group of workers well positioned around the new technologies. This could mean that improvements in technology are actually associated with a reduction in the standard of living of much of the population, even if they are not the cause.
This risk is amplified by the fact that the new technologies themselves may facilitate the evasion of rules that have been put in place to protect workers. This would be the case with companies like Uber or Task Rabbit, which claim that labor laws do not apply because their workers are independent contractors. This allows them to evade rules such as minimum wage laws, hours and overtime restrictions, and liability for injuries on the job that would ordinarily be covered by workers’ compensation.
In this case there is no problem in principle created by the new technology; it is simply a political failure to alter labor laws to ensure they apply to these new firms as well. The model for such a massive failure to adjust policy is Amazon, now one of the largest retailers in the world. In most states, for most of its existence, Amazon avoided the requirement that it collect the same sales tax as brick-and-mortar retailers. Its savings almost certainly exceeds its cumulative profit over its existence.
The corruption of U.S. politics may be so great that corporations will be able to use new technologies to undermine labor laws on an ever larger scale as the government pursues macroeconomic policies that are intended to leave much of the labor force unemployed and most of the employed with little bargaining power. This is indeed a very bleak scenario for the future, but it is silly to blame the robots.
For the Future of Work, a special project from the Center for Advanced Study in the Behavioral Sciences at Stanford University, business and labor leaders, social scientists, technology visionaries, activists, and journalists weigh in on the most consequential changes in the workplace, and what anxieties and possibilities they might produce.