A Trapped Factors Model of Innovation

That is the new paper (pdf) by Nicholas Bloom, Paul Romer, Stephen Terry and John Van Reenen.  Here is the abstract:

When will reducing trade barriers against a low wage country cause innovation to increase in high wage regions like the US or EU? We develop a model where factors of production have costs of adjustment and so are partially “trapped” in producing old goods. Trade liberalization with a low wage country reduces the profitability of old goods and so the opportunity cost of innovating falls. Interestingly, the “China shock” is more likely to induce innovation than liberalization with high wage countries. These implications are consistent with a range of recent empirical evidence on the impact of China and offers a new mechanism for positive welfare effects of trade liberalization over and above the standard benefits of specialization and market expansion. Calibrations of our model to the recent experience of the US with China suggests that there will be faster long-run growth through innovation in the US and that, in the short run, this is magnified by the trapped factor effect.


I came to the same conclusion some time ago.

As a bike fanatic I noticed that the average bike in the Netherlands did not change at all from a technical point of view for decades. The huge bike market in the Low Countries was controlled by a few big players that kept prices high and stifled innovation.

This all changed very fast when the first cheap bikes from China appeared on the market. In a very short span of time bikes evolved very fast and many of the old players vanished.

Now it seems that we are reaching another saturation point again with the E-bike. Once again there are only a few players who only offer us limited choice.

I can't wait to see what happens when the first African made bikes hit the market!

Comments for this post are closed

Of course, Tyler believes that the Great Stagnation arrived simultaneously with the China Shock.

In contrast, high tariff America went to the Moon.

It is possible for two trends to exist that push in the opposite direction. The America of 1969 went to the moon because there was a lot of low-hanging fruit in rocketry, avionics and integrated electronics, fields that had only recently been invented. In contrast in the emerging fields today, like genomics and artificial intelligence, there's very little low-hanging fruit to be reaped. The nature of these areas proved to be much more frustrating and complex than initially believed.

For example the Soviet Union would have also assuredly gone to the moon at the latest by 1980 had the Americans not beat them to the punch. Certainly no one would claim that the Soviet Union represented the height of economic and technology dynamism and innovation. Nor would anyone but a madman claim that the USSR in 1970 had a superior economic, technological or science system to the globalized economy of 2013.

A far simpler explanation is that rocketry, avionics and integrated electronics offered a lot of very easy to reap innovations at the century mid-point. Nearly any system that wasn't totally dysfunctional (and less dysfunctional than the USSR is a low hurdle) would have gone to the moon within at most a few decades given the political will and state of technology circa 1945.

Trying to claim the superiority of any political or economic system based on different rates of innovations is pointless across historical points. Different points in the technological development space have far different gradients. There was nothing special about American science and technology in 1960. It was playing the Olympia Fields course on a sunny no-wind day. In 2013 we're playing on Carnoustie with heavy wind and rain.

Hey, I've played Olympia Fields on a sunny, no-wind day: it was plenty hard enough.

Haha, fair enough. Golf is tough enough for me on a public easy par 3 course.

Btw, Steve, you need to be commenting on this post!

Comments for this post are closed

Comments for this post are closed

One might also note that the political photo-op value of the moon program was far greater than working on the human genome, even though it's a lot easier to see the possibility of long run effects on human life from the human genome work.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Note, implicitly, that patents are the key to this working. Imagine that the high-wage country, the USA/Netherlands, does not have, nor could obtain, any patents. There would be no reason to innovate to differentiate against cheap Chinese imports. This thinking was also the key behind the theory of "dark imports" that was fashionable a while ago to explain that there is no trade deficit (a neat theory--whatever happened to it? Like cold fusion, it might make a comeback someday?). IP is the 'hidden wealth' that is hard to measure in a society. And for some industries, it is said shipbuilding is one of them (though it's hard to prove), if you outsource your 'hands on' manufacturing, you lose your IP as well, since a lot of IP is not design work on paper but rather 'know-how' from actually doing the manufacturing.

'Note, implicitly, that patents are the key to this working'

What successful part of the Internet itself (BIND, http, TCP/IP, etc.) is patented?

You may want to look at the IETF IPR disclosure site:

For example, a Motorola IPR claim on DHCP:

Ethernet and SONET/SDH also have some patents. SSH and other cryptographically based systems run up against crypto patents.

Every Internet video site (YouTube, Netflix, etc.) is using a video compression scheme encumbered by a large number of patents.

Let's not forget the GIF patent (now expired)

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Your reply seems to be a completely irrelevant. The internet was developed by various government agencies over 20 years ago.

It has absolutely nothing to do with the concept of patents, which are designed to incentivize private companies to publicly innovate and show their work. This also allows the technology to become a public good vs a trade secret since a requirement of patenting is to publicly document your patent.

Comments for this post are closed

"The internet was developed by various government agencies over 20 years ago."

To be precise, ARPANET was built and early Internetworking protocols were developed by DARPA-funded workers at universities and research organizations (such as Stanford Research Institute and UCLA), some private, some public institutions.

Most of the Internet infrastructure (fiber, routing infrastructure, data centers, etc.) that we use today was built by private organizations starting in the early 1990's.

Comments for this post are closed

Comments for this post are closed