Last Friday I had the great pleasure of participating in a panel at the Federalist Society’s National Convention with Chris DeMuth of the Hudson Institute, David Weisbach of the University of Chicago Law School, and Judge Frank Easterbrook on intergenerational equity and old age entitlements, like social security. My talk had two parts. I first rejected a common claim about old age entitlements: that they transfer resources from a poor generation to a richer one, because young people today will be less well off than their elders. I, then, nevertheless showed that old age entitlements as presently structured raised substantial problems for the young and old. In this post I summarize the first part of the speech.
The history of economic growth in the United States suggests that, as they age, the young today will be much better off than the old are today. Since 1950 – less than a lifetime – real GDP per capita in the U.S. has tripled. And economic growth continues, even if the statistics suggest that it is slowing down. But this slowdown is to a substantial extent an illusion, because it fails to fully account for the two greatest ongoing revolutions of our time—the improvements in health care and the exponential increases in machine intelligence that are rapidly expanding throughout the economy.
First, take improvements in health care and longevity: they do not even show up in the GDP. And yet they are massive: as Larry Summers once remarked, it is not at all clear that one would choose to have the health care of 1950 and the income of today rather than the income of 1950 and the health care of today. Life expectancy continues to increase by almost 2.5 years with every passing decade. And genomics and personalized medicine promise more improvement, perhaps at an even faster rate. If the young live longer and healthier lives than their parents, they are by that consideration alone likely to be better off, because long, healthier life trumps higher income for most of us.
Second, the growth of machine intelligence is the most important phenomenon of our time and its exponential rate of change will likely increase the wealth of younger generations more than any other in history.
Moore’s Law captures a central regularity of our age: computer power doubles every eighteen months or two years. The computational capacity in a cell phone today is 1,000 times greater and 1 million times less expensive than all the computing might that was housed at MIT in 1965. Projecting forward, computing power 30 years from now could exceed today’s by 1 million times.
Moore’s Law will run out when the size of transistors cannot shrink further, something that’s predicted to happen sometime in the late 2020s. But other methods under research today, from carbon nanotechnology to optical computing, are likely to become new platforms for continued growth.
The dramatic increase in hardware capability is only part of the change in computational capacity. Software improvements, while less steady, provide another forceful multiplier for the power of computation. Computers now also interconnect among themselves and with human intelligence, sharing data more seamlessly and increasing effective computational capacity even more.
The kind of fluidity and connectivity provided by this combination enables computing applications in an increasing number of areas. IBM is now deploying Watson, the machine that won at Jeopardy, in medical diagnostics. Everyone has heard of Google Self Driving Cars, but traditional car companies also now think this is a safe bet: Volvo is introducing self-driving cars in Gothenburg in 2017. Drones will be delivering packages.
The story of machine intelligence and its ability to transform activity is always the same. Once it gains a foothold, it improves until it dominates. I could beat the best chess computer when I was on the high school chess team. Now I am humiliated by my smart phone.
The result may not be good in the short run for people whose jobs can be automated, but it will be great for productivity, delivering and higher quality services to consumers at lower prices and providing services and products undreamed of even ten years ago. As a result, we will continue to become a wealthier nation.
GDP statistics in fact have trouble capturing such dramatic changes in products and goods and services. Briefly, inflation calculations require that we compare baskets of goods from one year to the next, but the faster technology changes the harder the comparison. This effect tends to overstate inflation and understate economic growth.
In short, everything we know about technology today suggests that not only will the millennials lead longer, healthier lives than any previous generation, but they will inhabit a much wealthier nation. While there are other good arguments against the current structure of old age entitlements, the claim that younger generation is likely to be worse off than the elderly is not one of them.