Opinion – The End of Progress

Is progress slowly running out of breath? At first glance, the question seems crazy – iPhones now have more processing power than Nasa had at the time of the moon landing, voice recognition and translation work pretty well after decades of unsuccessful attempts, millions of songs can be heard almost everywhere, and new mRNA vaccines against Covid have been developed at record speed.

But if you measure the economically relevant progress, you suddenly don’t see so much of all the improvements – productivity growth has been sluggish for many years. In the long term since World War II, hourly economic output has increased by about 2.5% per year. Since 2005, the average has only been 1.5%, and since 2019 just 0.5%. These trends are even more dramatic in the former Asian Tigers, China and Western Europe. Even if it looks a bit different in every country and because there are fluctuations due to short-term shocks – the trend is unmistakable. As early as 1987, Nobel Prize winner Bob Solow joked that you see computers everywhere but in productivity statistics. Because many new technological devices now benefit end-users (rather than being in factories), the paradoxical situation is that technological progress falters while many people feel that enormous things are happening.

“Research is by its very nature unpredictable. Anyone who seriously wants to promote it has to live with the fact that a lot, maybe even most of the money and time is wasted.”

So are we measuring growth wrong – maybe we’re getting more productive, but the stat misses all the extra benefits of Facebook and Twitter, Spotify and Netflix? In the past twenty years, statistical institutes everywhere have made great efforts to measure improvements in quality, precisely through technical progress, with so-called hedonic adjustments. But not much has changed in the overall picture; the actual engine of long-term gains in prosperity, productivity growth, is clearly weakening.

More effort, less profit

Even where there is progress and it is easily measurable, it is bought at ever higher costs. Nick Bloom, Jon van Reenen and Chad Jones have examined how much research and development is needed to keep Moore’s Law – doubling the number of transistors on computer chips every 24 months – working. You’ve found that steady progress requires exponentially increasing spending. Today, eighteen times as many researchers are employed in the semiconductor industry as in 1971, per capita researcher productivity has declined 6.8% each year. The same research paper shows how common similar relationships are in other sectors: every year of additional life expectancy in the medical sector costs more and more research. While the invention of penicillin saved millions of lives almost overnight and at low cost, today even small improvements require billions.

But how can it be that progress is paralyzed on a planet where more and more people are becoming more educated and governments and private companies are spending a lot of money on research? One possible explanation is that we have already found all the easy solutions, and for that to happen, progress must necessarily slow down—something like penicillin can only be found once. In fact, relative to the tremendous advances in chemistry and physics in, say, 1900-1950, the improvements of the past fifty years do not look so impressive.

Add to this what is known as the “burden of knowledge” – in order to even absorb what has already been achieved and understood in a certain field, young scientists today have to spend much more time than in 1900; by the time they catch up, the prime years for fundamental, creative breakthroughs are over (and the more mathematical a discipline is, the earlier in the researcher’s life such breakthroughs occur).

scientific process in transition

A third possible explanation has to do with organizational and cultural changes. We owe many great breakthroughs in the computer industry, for example, to Bell Labs, the research facility of the American telephone monopoly Bell Telephones/AT&T, which issued the engineers and scientists there a blank check for research and experimentation without expecting much profit. Thanks to deregulation, such monopolies disappeared. But the story is more complicated. When the German IG Farben was dissolved after 1945, the successor companies increased their patenting activities. It is therefore rather unclear whether monopolistic tendencies are generally good or bad for the innovation rate.

What is clear, however, is that the state has withdrawn more and more from funding large, daring projects, not least since the end of the Cold War. But particularly important breakthroughs are often credited to the military-industrial complex, from the Internet to jet aircraft and nuclear energy to liquid-propellant rockets, GPS and insecticides, military uses have come first. Not only has public investment in developed countries been declining for decades, but basic research with its “long shots” is still running on the back burner at best.

In addition, the scientific process itself has changed fundamentally. Where previously research was mostly carried out individually or in small groups, today huge teams are required to use machines that often cost millions of dollars; scientific publications with thousands of authors are not uncommon. Those who spend so much money have to convince committees of themselves and their research; there are colleagues who can only approve a fraction of the applications and do not want to waste the taxpayer’s money. So it quickly becomes obvious that somewhat crazy, rather improbable project ideas are shelved and, above all, that which definitely produces (a little) progress is promoted. This is accompanied by the transformation of many scientists into managers who lead their teams, lead meetings, write proposals, but hardly lend a hand in research, collect and clean data, run regressions, read observations.

No odd heads

But good managers are not necessarily cut from the same cloth as people who come up with new ideas; odd characters such as the nuclear physicist Leo Szilard, to whom we owe the idea of ​​the chain reaction, would have little chance in today’s scientific community. Institutions such as universities, which used to look almost exclusively at brilliant intellect and research achievements and were the last refuge for people with rough edges, are increasingly applying the HR criteria of industry. On the one hand, this increases the probability of keeping a well-oiled machine running in which little “goes wrong”, but the interaction of the logic in the allocation of funds and the personnel policy of the research institutions make real breakthroughs less and less likely.

Research is by its very nature unpredictable. Anyone who seriously wants to promote them has to live with the fact that a lot, maybe even most of the money and time is wasted. In a world increasingly geared towards accountability and documentable results, one of the most important sources of our wealth is drying up, despite often high expenses. For comparatively little money, however, a handful of new “Bell Labs” could be set up to work on new breakthroughs, freed from the constant cycle of funding applications and documentation of results. Where such working conditions beckon, interested scientists would not be far away, and anemic growth could pick up momentum again.

Found a mistake?Report now.

0 comments