Last Friday, I posted a not-so-serious note on how people think any new technology developed after their birth is worse than technology that was already around when they were born.
I’ve spent time both in academia and in industry and I believe I have some ideas on why. Through industrialization of research, pretty much any low hanging fruit in parts of science and engineering has already been picked. What is left is either intractable or too difficult (and sometimes there are physical limitations, such as in semiconductors). And those that can be solved, need significant time. These days, neither in industry nor in academia, there’s practically anyone willing to support that kind of research and in order to get a paycheck/tenure/promotion/funding you need to show steady and continuous progress. That leaves very little incentive for anyone to even try. Even worse, funding agencies like US National Science Foundation expect you to provide a path to the solution in your grant application. So you have to have already made some progress before you can even apply. Incentives and rewards matter
While I agree with you about the state of academic research, there is a problem with this explanation for low productivity growth: we have always had underwhelming total factor productivity growth. It's just been obscured by the garden-variety growth of the post-war era. That growth is plateauing everywhere (yes, even in emerging markets!), so we are now paying more attention to TFP than previously.
I'm really looking forward to Joachim's exploration of this topic!
Hmm, can you point me in the direction of some research or data on this. I would be very interested to explore this more. The data I have from the Maddison database shows that there have been two major boosts to TFP growth in the past 150 years or so. One from the rollout of electricity and electric light and one from the mass adoption of the internal combustion engine. There was a tiny temporary boost in the late 1990s from the internet and computers, but that was really just a few years.
Having been in academia for eight years and now in business, I tend to agree with you that funding for research has been drifting more and more towards projects that show immediate practical relevance and application. Next Wednesday, I will dive a little deeper in the current research practices and what that means for research productivity and the ability of new ideas to become mainstream and I will make the argument that even if we have breakthrough ideas, it becomes harder and harder for them to gain widespread attention, let alone adoption. So stay tuned for that.
Yes, but the challenge is that in the past, we haven't seen a decline in the rate of progress. If I take the Enlightenment as the starting point of modern day "research", you only see a decline in research productivity in the post-WWII era, not before. So maybe, as someone else in t he comments suggested, we have picked all the low hanging fruit by now, but I am not so sure about that.
No, I can't see a connection between low productivity growth and stock picking. Just look at Japan where productivity growth has been very low for decades and stock pickers have not been better or worse than in the US or other regions.
Joachim. my question was only indirectly linked to productivity growth. It was more about the opportunity to create outsized returns being harder to achieve as technology has the ability to pounce on these mismatches much more quickly, thus keeping stocks and prices much more closely aligned to available information...
Two broad thoughts: (1) perhaps academic research has become so specialized that researchers are less likely to have the breakthrough ideas often cross not only segments of their discipline but also other disciplines. Industrial research has always focused on topics with immediate relevance and application, but perhaps all researchers are more short-term and narrowly focused than in the past. (2) Many of the areas of greatest research application do not add to measured productivity. Specifically, we must have a green energy transition for reasons you all know, but the total amounts of resources needed to produce a given quantity of energy will increase, so measured productivity will fall. Similarly, bright minds are drawn to create social media applications and internet advertising, which may be neutral (cynics would say negative) to productivity.
I’ve spent time both in academia and in industry and I believe I have some ideas on why. Through industrialization of research, pretty much any low hanging fruit in parts of science and engineering has already been picked. What is left is either intractable or too difficult (and sometimes there are physical limitations, such as in semiconductors). And those that can be solved, need significant time. These days, neither in industry nor in academia, there’s practically anyone willing to support that kind of research and in order to get a paycheck/tenure/promotion/funding you need to show steady and continuous progress. That leaves very little incentive for anyone to even try. Even worse, funding agencies like US National Science Foundation expect you to provide a path to the solution in your grant application. So you have to have already made some progress before you can even apply. Incentives and rewards matter
While I agree with you about the state of academic research, there is a problem with this explanation for low productivity growth: we have always had underwhelming total factor productivity growth. It's just been obscured by the garden-variety growth of the post-war era. That growth is plateauing everywhere (yes, even in emerging markets!), so we are now paying more attention to TFP than previously.
I'm really looking forward to Joachim's exploration of this topic!
Hmm, can you point me in the direction of some research or data on this. I would be very interested to explore this more. The data I have from the Maddison database shows that there have been two major boosts to TFP growth in the past 150 years or so. One from the rollout of electricity and electric light and one from the mass adoption of the internal combustion engine. There was a tiny temporary boost in the late 1990s from the internet and computers, but that was really just a few years.
Having been in academia for eight years and now in business, I tend to agree with you that funding for research has been drifting more and more towards projects that show immediate practical relevance and application. Next Wednesday, I will dive a little deeper in the current research practices and what that means for research productivity and the ability of new ideas to become mainstream and I will make the argument that even if we have breakthrough ideas, it becomes harder and harder for them to gain widespread attention, let alone adoption. So stay tuned for that.
Isn't it natural that it is getting more and more difficult to improve the better you get?
Regression to the mean!
Yes, but the challenge is that in the past, we haven't seen a decline in the rate of progress. If I take the Enlightenment as the starting point of modern day "research", you only see a decline in research productivity in the post-WWII era, not before. So maybe, as someone else in t he comments suggested, we have picked all the low hanging fruit by now, but I am not so sure about that.
....is this an argument as to why it is becoming more difficult for professional investors to beat the market?
No, I can't see a connection between low productivity growth and stock picking. Just look at Japan where productivity growth has been very low for decades and stock pickers have not been better or worse than in the US or other regions.
Joachim. my question was only indirectly linked to productivity growth. It was more about the opportunity to create outsized returns being harder to achieve as technology has the ability to pounce on these mismatches much more quickly, thus keeping stocks and prices much more closely aligned to available information...
Two broad thoughts: (1) perhaps academic research has become so specialized that researchers are less likely to have the breakthrough ideas often cross not only segments of their discipline but also other disciplines. Industrial research has always focused on topics with immediate relevance and application, but perhaps all researchers are more short-term and narrowly focused than in the past. (2) Many of the areas of greatest research application do not add to measured productivity. Specifically, we must have a green energy transition for reasons you all know, but the total amounts of resources needed to produce a given quantity of energy will increase, so measured productivity will fall. Similarly, bright minds are drawn to create social media applications and internet advertising, which may be neutral (cynics would say negative) to productivity.