Last week Bill Gates called for taxing robots. He argued that we should impose a tax on companies replacing workers with robots and that the money should be used to retrain the displaced workers. As much as I appreciate the world’s richest person proposing a measure that would redistribute money from people like him to the rest of us, this idea doesn’t make any sense.
Let’s skip over the fact of who would define what a robot is and how, and think about the logic of what Gates is proposing. In effect, Gates wants to put a tax on productivity growth. This is what robots are all about. They allow us to produce more goods and services with the same amount of human labor. Gates is worried that productivity growth is moving along too rapidly and that it will lead to large scale unemployment.
There are two problems with this story: First productivity growth has actually been very slow in recent years. The second problem is that if it were faster, there is no reason it should lead to mass unemployment. Rather, it should lead to rapid growth and increases in living standards.
Starting with the recent history, productivity growth has averaged less than 0.6 percent annually over the last six years. This compares to a rate of 3.0 percent from 1995 to 2005 and also in the quarter century from 1947 to 1973. Gates’ tax would slow productivity growth even further.
It is difficult to see why we would want to do this. Most of the economic problems we face are implicitly a problem of productivity growth being too slow. The argument that budget deficits are a problem is an argument that we can’t produce enough goods and services to accommodate the demand generated by large budget deficits.
Gates’ notion that rapid productivity growth will lead to large-scale unemployment is contradicted by both history and theory.
The often told tale of a demographic nightmare with too few workers to support a growing population of retirees is also a story of inadequate productivity growth. If we had rapid productivity growth then we would have all the workers we need.
In these and other areas, the conventional view of economists is that productivity growth is too slow. From this perspective, if Bill Gates gets his way then he will be making our main economic problems worse, not better.
Gates’ notion that rapid productivity growth will lead to large-scale unemployment is contradicted by both history and theory. The quarter century from 1947 to 1973 was a period of mostly low unemployment and rapid wage growth. The same was true in the period of rapid productivity growth in the late 1990s.
The theoretical story that would support a high employment economy even with rapid productivity growth is that the Federal Reserve Board should be pushing down interest rates to try to boost demand, as growing productivity increases the ability of the economy to produce more goods and services. In this respect, it is worth noting that the Fed has recently moved to raise interest rates to slow the rate of job growth.
We can also look to boost demand by running large budget deficits. We can spend money on long neglected needs, like providing quality child care, education, or modernizing our infrastructure. Remember, if we have more output potential because of productivity growth, the deficits are not problem.
We can also look to take advantage of increases in productivity growth by allowing workers more leisure time. Workers in the United States put in 20 percent more hours each year on average than workers in other wealthy countries like Germany and the Netherlands. In these countries, it is standard for workers to have five or six weeks a year of paid vacation, as well as paid family leave and paid vacation. We should look to follow this example in the United States as well.
If we pursue these policies to maintain high levels of employment then workers will be well-positioned to secure the benefits of higher productivity in higher wages. This was certainly the story in the quarter century after World War II when real wages rose at a rate of close to two percent annually.
Of course these policies will not ensure that no workers ever suffer from automation. While we can never guarantee that no worker is harmed by improvements in technology in a dynamic economy, we can look to soften the impact.
One obvious policy would be to require severance pay, for example two weeks of pay for each year worked. This would both give displaced workers somewhat of a cushion and changes the incentives for employers. If a company knows that it faces large payout if it lays off a number of long-term employees, then it has more incentive to think about modernizing its facilities and retraining workers. This would be a win-win where the company has an interest in ensuring that its workers are as productive as possible while the workers get to keep their jobs.
In short, there is no reason that productivity growth should ever be viewed as the enemy of workers. We just need the right set of policies to ensure that they share in the gains.