Innovation Shortfall, Low Hanging Fruit, and the Frontiers of Knowledge

Tyler Cowen has a new ebook coming out this week. It’s called the Great Stagflation.  I’m eager to read it, but based on Tyler’s own description, it appears he echoes some claims Peter Thiel has been making about the lack of innovation we’ve seen since the early 70s. For Thiel’s take here’s a TED talk on the topic and a WSJ interview. iPhones and laptops are palpable, so people tend to doubt these arguments. After all the unseen–what hasn’t happened–is difficult to adduce. But back in 2009 Michael Mandel pointed to some telltale signs:

Cowen explains why median wages in the U.S. have been stagnating since the 70s:

I’m also persuaded by the median income numbers because they are supported by related measurements of other magnitudes. For example, another way to study economic growth is to look not at median income but at national income, gdp, or gross domestic product, the total production of goods and services.  Charles I. Jones, an economist at Stanford University, has “disassembled” American economic growth into component parts, such as increases in capital investment, increases in work hours, increases in research and development, and other factors. Looking at 1950–1993, he found that 80 percent of the growth from that period came from the application of previously discovered ideas, combined with heavy additional investment in education and research, in a manner that cannot be easily repeated for the future. In other words, we’ve been riding off the past.

I’d like to add to Cowen’s argument. An economist named Benjamin Jones at Northwestern University looked back over the last 100 years and delved into some interesting data sets on Nobel Prizes, patent filings, and almanac entries. He collected data on the age at which an inventor brings about his first discovery, the average age of a scientist’s greatest accomplishment, and the productivity lifespan of innovators and scientists.

Here are some facts worth considering from the Jones research, that I believe supports the Thiel-Cowen thesis that technological progress has not been proceding apace:

• Innovators today are older than they used to be. Over the last century, the mean age at greatest achievement for both Nobel Prize winners and great tech inventors rose by about 6 years.
• The mean age at which innovators launch their first inventions has increased by 8 years over the course of the century, rising from a mean age of about 23 in 1900 to approx 31 in the year 2000.
• Despite the later and later start, there has been no compensating shift in the productivity of innovators beyond middle age, meaning the late start has truncated careers. We do not see any rise in innovating among 50 and 60 year olds, although people are now more productive in their 40s.
• His data suggests innovation potential is highest approx between the ages of 27 and 42. But the peak has shifted from about age 32 in 1900 to 40 by end of century.
• He estimates a 30 percent decline in life-cycle innovation potential over the 20th century.
• Death of the renaissance inventor: specialization has increased in patent filing. More and more people are working in teams that require a greater division of intellectual labor. And fewer and fewer inventors are making jumps from one patent category to another.

Big takeaway: “The shorter the period that innovators spend innovating, the less their output as individuals over their lifetime. If innovation is central to technological progress, then forces that reduce the length of active innovative careers will reduce the rate of technological progress. This effect will be particularly strong if innovators do their best work when they are young.”

These facts ought to give us pause. Jones postulates that the main reason we see an age shift is because the frontiers of knowledge are farther and farther away. The shoulders of giants are higher. The low fruit has been taken. Because of how complex and specialized our knowledge has become, Jones says the young must spend more and more years training and acquiring knowledge to reach the cutting edge.

Now it may be that the “burden of knowledge” is heavier than it used to be.  But my own take is that we shouldn’t confuse the accumulation of credential crud for the difficulty of reaching the frontiers of knowledge. On the macro scale, we often cite the Mancur Olson thesis that special interests sap the dynamism of economic growth over time. I believe the Jones data and Cowen’s arguments present further evidence supporting the Olson story.  For instance, why should we believe PhD programs are designed to efficiently move students from ignorance to the frontier of knowledge as quickly as possible? Academics are not mainly truth-seekers. They’re incentivized to pursue other aims such as pleasing their advisors, their tenure review panel, or a grant making body. These aims do not coincide with discovering new and useful knowledge. Secondly, we can go back further. A lot of education from K-12 to college core requirements are a waste of time. How much of this knowledge is relevant to future work making discoveries? If anything, Jones should suggest we try to make education more efficient.

But I digress. In all this discussion, Cardwell’s Law looms in the background. No country stays technologically superior forever. As yesterday’s innovators becomes today’s vested interests, stagnation follows. There are reasons for that. All of these arguments–Thiel’s, Cowen’s, Jones’s–suggest we may need a thousand nations sooner than we think. We will soon find not the singularity, but the stagnation is near.

Read more at Let A Thousand Nations Bloom