Innovation: It takes a village

Why America needs to keep gathering the world's brightest people in one place

"Math team" by believekevin, CC BY 2.0

If you spend any time watching scientists or engineers actually do their work, you’ll quickly notice how collaborative of an enterprise innovation is. They’re constantly talking to their peers, reading their peers’ work, etc. And if you try to trace the history of a particular technology or breakthrough, you’ll generally find tons of incremental improvements. There are a few notable, famous exceptions — Newton inventing calculus all on his own, and so on. But these are very rare.

Anyway, here’s a new paper by Agarwal, Ganguli, Gaule and Smith that vividly illustrates how powerful that communal production process is. Tracking the International Congress of Mathematicians appointments of very high-skilled people, they observe that conditional on teenage math test performance, these people are six times as productive when they immigrate to the U.S. as when they stay in their home countries or immigrate to countries other than the U.S. Six times!! For scientific publications, immigrants to America are four times as productive as stayers, and twice as productive as immigrants to the U.K. and other rich countries.

That means, first of all, that there is something very special about the U.S. Our network of universities, tech companies, and government labs constitutes the largest concentration of top-tier research talent on planet Earth. Talented people flow between these centers of innovation, exchanging ideas, building on each other’s ideas, cross-fertilizing, etc. The U.S. stopped being the “workshop of the world” long ago, but we are still the world’s research park.

China is trying to challenge us for this position, but it’s not there yet. For example, in A.I. research — a field where China is pushing hard to overtake us — the U.S. still holds an advantage in “top-tier” Chinese-born researchers alone, according to the think tank Macro Polo. As in, more Chinese-born A.I. researchers work in America than work in China! And of course, the U.S. also has tons of A.I. researchers from other countries, as well as all our home-grown folks. And as a result, and because of the productivity advantage reported above, the U.S. still retains the lead in this crucial strategic field.

It’s not immediately clear how much of the U.S.’ massive scientific productivity advantage comes from our special traits as a country — openness, freedom of thought, mobility, capitalism, whatever — and how much comes from scaling effects. But it seems highly unlikely that America’s system is six times better than other countries’ systems. And either way, the policy implication is clear: Get the smart people here in America.

(Now, for the people who bristle at calls for high-skilled immigration, realize that recruiting the world’s top talent doesn’t mean we keep out low-skilled immigrants! Don’t think about immigration policy in those zero-sum terms.)

Now, there are two main objections to bringing in lots of talented people from abroad. The first is the idea that foreign innovators squeeze out American innovators. This idea is exactly what’s being expressed every time someone says “there isn’t a talent shortage” or “we can train enough scientists ourselves”. It’s the idea of a zero-sum competition for spots.

But what the Agarwal et al. paper suggests is that innovation just doesn’t work like that. Instead of squeezing out home-grown science heroes, skilled immigrants complement them. Because innovation takes a village, the more foreign innovators we get, the more innovating our home-grown folks are able to get done.

Also, the more they will be inspired to become innovators in the first place. Anton Howes (who studies the British Industrial Revolution) has a recent blog post in which he explains that the most powerful predictor of becoming an inventor during that famous time period was simply knowing other inventors. As I always say, the most valuable type of human capital is simply ideas for what to do with your life.

This is almost certainly true in the modern say as well. Some 2017 research by economist Raj Chetty has found that even after you control for natural ability (test scores), people’s chances of becoming innovators as adults are hugely dependent on their exposure to other innovators early in life. This is the famous “lost Einsteins” result that you may have read about a few years ago.

In other words, the more foreign innovators we get in this country, and the more we put them in contact with native-born Americans, the more we inspire native-born Americans to also become innovators!

The other big argument you hear against high-skilled immigration is “brain drain”. Why, people ask, should we be selfish and rob the world of their talented people? (Obviously this isn’t the real concern of anti-immigration folks, who generally could not care less about other countries; they just want to keep the immigrants out, and are reaching for any rhetorical device available.)

But the Agarwal et al. paper shows that the world’s most talented people will get a lot more done in America than they would in other countries. By excluding skilled immigrants, we would be refusing to be the world’s research park. That would lower global innovation and global productivity. What a waste!

(Actually there are other good reasons not to worry about brain drain; for a great overview, check out the book The Gift of Global Talent by William Kerr.)

So basically, everything we know about innovation suggests that it takes a village. And the United States has two choices: Either be that village, or don’t be that village.

I want us to be that village.

Update: There’s some good news on the policy front, by the way. Biden has already reversed Trump’s efforts to make it impossible for spouses of H-1b workers to work. He has also proposed legislation reducing green card wait times, raising country caps (the last remnant of the exclusionary system of the early 20th century), and — most importantly — exempting STEM PhD students from the green card cap. That would all be very good stuff. One additional thing Biden should do is to re-implement Obama’s “International Entrepreneur Rule”. Many scientists are also entrepreneurs, so this rule would help bring top research talent into the country (in addition to simply boosting business dynamism and competition). Here’s a long and detailed briefing on the IER, courtesy of Caleb Watney.

---