We all have heard the stories about how the robots are going to take our jobs. The line is that innovations in computer technology will make robots ever more sophisticated, allowing them displace a rapidly growing number of workers. This could leave large numbers of workers with nothing to do, implying a massive amount of long-term unemployment.

There are two basic problems with this story. The first is a logical problem. The story of worker displacement by technology is not new, it goes back hundreds of years and it is ordinarily considered to be a good thing. This is what we call productivity growth. It means that workers can produce more goods and services in the same amount of time. This is the basis of rising wages and living standards.

If we see rapid productivity growth, as robots allow for the same output with fewer workers, this should allow the remaining workers to be paid more for each hour of work. This will allow them to spend more money, creating more demand in other sectors, which will allow displaced workers to be re-employed elsewhere.

Of course we have not seen workers getting the benefits of productivity growth in higher pay in recent years. This is due to policies and institutional changes that undermine workers' bargaining power. For example, trade policy has deliberately put manufacturing workers in competition with low paid workers in the developing world. The Federal Reserve Board routinely raises interest rates to slow job growth when it fears that workers are getting too much bargaining power and could possibly get inflationary wage increases. And, lower unionization rates mean that workers are less effective in demanding higher pay from employers.

For these reasons, most workers have not gotten their share of the gains from productivity growth, but there is another problem with the robots displacing workers story. Rather than robots leading to a massive surge in productivity, in recent years productivity growth has been unusually slow. According to the Bureau of Labor Statistics, annual productivity growth has averaged less than 0.6 percent since 2010. This compares to an average rate of 3.0 percent in the Internet boom years from 1995–2005 or 2.8 percent in the long post-World War II boom from 1947–1973. Even in the years of the productivity slowdown, from 1973–1995, had a 1.4 percent annual rate of growth, more than twice the recent pace. In short, there have not been many gains to share.


Sweden is experimenting with a possible solution to both problems: slow productivity growth and ensuring that workers get their share of the benefits. Many businesses across the country are experimenting with six-hour work days. There is some evidence that workers are more productive per hour working six-hour days rather than the traditional eight-hour days. A growing number of Swedish companies are testing this view by cutting back their work hours.

If this reduction in work hours proves successful, it will address both sides of the problem. It would first mean that we would see a surge in productivity growth as workers produced more hour in their six-hour days than in their eight-hour days. This would justify higher hourly pay. The second part of the story is that by reducing the average amount of work-time per worker, we would be opening up more jobs.

This point is straightforward. If there is demand for the same number of labor hours, and the average worker puts in fewer hours, there will be more demand for workers. This is a way to sustain higher levels of employment, thereby ensuring that workers have more bargaining power. With short enough workweeks/work years we can keep the economy near full employment and make sure that workers have the necessary bargaining power to get their share of the gains from growth.

For these reasons, the six-hour day sounds like a very intriguing idea, in addition to the fact that it will give people more time to do things they enjoy.