Why some economists worry when unemployment gets too low
By R.A. | WASHINGTON, DC
IN 1977 America's government gave the Federal Reserve what seems like a straightforward goal: maximum employment. Janet Yellen, the current chairman of the Fed, thinks America is pretty close; at 4.7%, the unemployment rate is quite low by historical standards. But firms continue to hire, and American adults, of whom only about 69% have a job, seem less than maximally employed. Most governments set themselves or their central banks a guideline of full or maximum employment. But what exactly counts as full?
Ms Yellen has a particular definition of maximum employment in mind, built on the economic experience of the past half-century. Macroeconomists reckon that governments can only push unemployment so low. Push the unemployment rate below some (unobserved) “natural rate”, and it will soon rise back up, along with prices. At the natural rate, all the workers who can easily or usefully be hired are working, and new hires can only occur by luring people from other jobs by offering them higher wages. Central bankers take a guess at the natural rate and at how quickly unemployment that is “too low” will spark inflation, then seek the sweet spot: when the labour market is as tight as it can be without runaway price rises. The natural rate of umemployment depends mostly on what economists label “frictional unemployment”: which occurs when people switch from one job to another, or when they enter the workforce (after finishing university, for example, or returning from leave). Barriers to job switching, like occupational licensing, increase friction and push up the natural rate. Other factors grease the gears. The lower natural rate of the 1990s might have been down to more efficient hiring thanks to information technology. The rise of “gig economy” work might also push the natural rate down.
But the boundary between long-term structural unemployment and the temporary, cyclical kind is not clear-cut. In the 1980s and 1990s economists argued that short-term unemployment could become long-term unemployment: a phenomenon known as "hysteresis". As a worker’s time without a job grows, his professional connections might weaken and his skills might become obsolete. Workers not laid off during a downturn might bargain for greater job security, which might in turn make firms reluctant to rehire those who were laid off (since they could not easily be sacked during the next recession). Hysteresis also works in reverse, at least to some degree. As America’s unemployment rate has fallen below 5%, wages have finally started rising at a faster pace. Better pay will entice some people who had given up on the labour force to look for work again. Better economic conditions might lead firms to offer temporary workers permanent positions, or to give more hours to part-time employees. As firms find it harder to hire new workers, they might offer existing workers more hours, or convert part-time or temporary positions to full-time or permanent ones. Bosses might even try to raise output per worker, by investing in training or in new equipment. The rub is that policymakers cannot know how much slack remains in the system until they see inflation spiralling up—and they would rather it didn't. So central banks often end up slowing growth before full employment is reached because of an excessive aversion to inflation.
To treat full employment as a purely economic phenomenon is not quite right, however. If the goal of full employment is a happy society, then the quality as well as the quantity of jobs matters. Employment rates in subsistence societies are extremely high. In richer countries, more people would be in work were governments to withdraw unemployment benefits and repeal the minimum wage. Yet society would be worse off for it. Technological change complicates matters. A scarcity of workers could drive investment in machines, allowing each worker to produce more. Yet it might also lead to more automation. If robots can easily replace lots of workers (and it increasingly seems that they can) then full employment is not simply a matter of making sure the economy is growing fast enough so that all willing workers have a job. It will also depend on the decisions societies make, regarding when workers can be given the means to refuse poor, low-pay jobs which could instead be done by machines.
