How much do colleges boost innovation?

Yes, yes, I know patents are not the right measure, that is what we’ve got:

I exploit historical natural experiments to study how establishing a new college affects local invention. Throughout the nineteenth to the mid-twentieth century, many new colleges were established in the U.S. I use data on the site selection decisions for a subset of these colleges to identify “losing finalist” locations that were strongly considered to become the site of a new college but were ultimately not chosen for reasons that are as good as random assignment. The losing finalists are similar to the winning college counties along observable dimensions. Using the losing finalists as counterfactuals, I find that the establishment of a new college caused 32% more patents per year in college counties relative to the losing finalists. To determine the channels by which colleges increase patenting, I use a novel dataset of college yearbooks and individual-level census data to learn who the additional patents in college counties come from. A college’s alumni account for about 10% of the additional patents, while faculty account for less than 1%. Knowledge spillovers to individuals unaffiliated with the college also account for less than 1% of the additional patents. Migration is the most important channel by which colleges affect local invention, as controlling for county population accounts for 20-40% of the increase in patenting in college counties relative to the losing finalists. The presence of geographic spillovers suggests that colleges do cause an overall net increase in patenting, although I find no evidence that colleges are better at promoting invention than other policies that lead to similar increases in population.

That is from new research by Michael J. Andrews, via the excellent Kevin Lewis.


"I find no evidence that colleges are better at promoting invention than other policies that lead to similar increases in population"

So the best way to increase inventions is to increase the formation of cities. Which means: increase immigration.

Of course we have essentially given up on trying to create new cities (at a time when young people want to live in cites).


We have transferred responsibility for creating all the infrastructure needed by new cities to the occupants of the new cities.

The old methods of distant governments creating the infrastructure that enables new cities has been rejected due to the failure of such processes to create cheap housing with low density but everything readily at hand.

The lack of new cities is blamed on liberals. They refuse to create millions of jobs in Iowa, West Virginia, eye to create great cities with low living costs, lots of space, lots of opportunity. Instead liberals only create jobs where big government central planners make everything costly.

Seriously, new cities are created all the time. Announce plans to build a new subway or light rail and build new stations where no one lives, and quickly new cities pop up. Same with limited access highways and interchanges. Gentrification is basically new cities being built on the rubble of old cities, promoted by government policy favoring the new city on top of the old.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Colleges grads are like bricks. A mason (inventor) needs to put them together to work. But you can't build a house without them. Thus, think about systems, not individual components.

Getting immigrant inventors was luck, you can't expect that kind of luck to repeat. It's not the 19th century anymore.

"It’s not the 19th century anymore."

Says you. Meanwhile, America is more polarized than ever since the Civil Wars, the Know-nothings are stronger than ever, the monetary system is in shambles inviting dangerous schemes (bimetalisn, bitcoin). In Russia, a new czardom casts its mighty shadow over mankind. Asian hordes start to be amassed and threaten our species with a new dark age made more sinister, and perhaps more protracted, by the lights of perverted science.

Comments for this post are closed

Comments for this post are closed

Thanks. Interesting analysis. Clever methodology.

I always like examples: what are some losing alternative sites for famous colleges?

For example, UCLA almost went to Pasadena instead of Westwood, but of course that's all Los Angeles County.

In general, the UC campuses are situated on superb land. Too bad the architecture usually isn't worthy of the site.

By the way, Disneyland was originally going to be in Burbank rather than Anaheim, but Walt Disney couldn't assemble a big enough tract of undeveloped land in Burbank. I think that was much the same reason Pasadena didn't get UCLA: Pasadena was the most developed suburb of Los Angeles from 1887 onward. Many people moving to Los Angeles before the development of penicillin preferred the warm, dry inland areas like Pasadena over the cool moist coastal areas like Westwood because of fear of TB. (Even when I moved to Santa Monica in 1981, my parents were fearful for my respiratory health.)

If that's a common pattern, the losing finalist might often have been the more developed site where it was harder to find a big parcel of cheap land, which makes the higher achievement in patents of the winning site in the present even more noteworthy.

What are some other places that should have a college?

For instance, my feeling has always been that Oak Park, IL, with all its Frank Lloyd Wright architecture and Hemingway hometown, seems surprising not to have a college the way Evanston and Hyde Park have colleges.

Comments for this post are closed

Many years ago I attended my state's flagship state university. It was located in a remote small town, what is often referred to as a "college town" (because that's all there was). I say was, not is, because today that same university is located in a mid-sized city. No, the university didn't move, the town grew, and grew exponentially. The university has excellent programs in engineering, medicine, and computer science, which helps explain the growth around the university. Williams College is located in Williamstown, Massachusetts. It is considered one of the best colleges in America, and is one of the most difficult to gain admission. Williamstown was, and is, a remote small town, a "college town" (because that's all there is). Williams College has excellent programs in the liberal arts, which helps explain the absence of growth around the college.

Comments for this post are closed

Interesting stuff! I wonder what would happen if the comparison were expanded to include all patent applications by community, which might provide a better indication of overall innovative activity, rather than successfully granted patents by community, where only the "winners" of patents are counted (data for patent applications is available!). Presumably, if universities really do boost innovation more than simply providing a vehicle for increasing population in a community, the presence of a university would provide additional access to skilled resources that would increase the likelihood of a patent being awarded, where the difference might be seen in comparisons of patent success rates, which would make for a neat extension of the study.

Comments for this post are closed

If we didn't have patents we would have no innovation? I suspect most innovations are never patented. I also suspect probably more important than innovation itself is adoption of innovation. The British had steam engines, so did the Romans. In Rome, though, it seems like it was only used as a novelty to produce some special effects inside temples. The English quickly started figuring out how to incorporate the engine into every aspect of life they could.

So patents are at best the tip of the innovation iceberg, with most innovations never being formalized as a patent. Big assumption, though, do patents track innovation? If there's twice as many patents does that mean there's twice as much innovation going on? Maybe but then maybe it means there's little real innovation going on and business is just doubling down on lawyers to toss in patent applications to make up for a lack of innovation below and above the 'patent line'.

Second big question are innovations being quickly adopted and tweaked into all aspects of the economy or do they just linger in the areas they originated at?

Comments for this post are closed

Are they really saying that migration to non-college towns correlates to patents at the same rate as migration to college towns?

That would surprise me. I would expect a different character to the migration.

(of course, if colleges "work" by way of migration, that still works)

While I am open to "top-down" colleges mattering in some ways to patent creation, I think "bottom-up" is equally possible. Who demands patents? People engaged in economic activity. (opening a college, of course, is also economic activity, but of a different variety than most industries.)

Migration to locations would also be because "there are jobs there" or some other form of economic activity.

Let's say you had to make a bet on whether a college town would have more patents or a town centered around an industrial activity: say steel mills. or mining.

I would bet on the practioner's having more patentable ideas than the eggheads, because they have far more incentive and experience. Heck, a lot of patents are novel ways to make old products!

Now, would colleges produce people more likely to be able to create a patent? Yes. Over time, would technology advance so that eggheads in an office had better ideas than the engineer in the field? Yes.

Just think of software, which doesn't have to be located near the ore or whatever. It can stay right up cozy to its colleges that produce the labor.

Sounds good.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Patents may be worse than not the right measure. I believe that in software the correlation between patents and innovation is negative. Maybe it's different in different fields.

Comments for this post are closed

Patents are all economists have because economists lack the innovation to imagine better metrics, and then raise money to fund building the capital assets of new, better, metrics for innovation.

Economists doing research is like physicists trying to measure gravity waves by climbing the leaning tower to drop lead balls.

Economics would hardly exist if not for the FDR administration needing metrics to guide his big government policies, and the British government needed metrics to guide its historic administrative state, and gain a better understanding of what industrial policy changes are needed, employing people like Keynes.

Marx was funded by the inherited wealth of Engels, the Zionists by wealthy Jews, and the investments made in Russia and Palestine, and the counter investments by Kaiser which led to US employer health benefits to counter socialist health care.

Innovation is a function of labor input. Colleges and universities are places where labor is focused on new ideas.

Even liberal arts, non-STEM, colleges foster innovation, but its not measured in patents.

Eg, "The Program in Creative Writing, at the University of Iowa in Iowa City, Iowa, is more commonly known as the Iowa Writers' Workshop..." to pick a center of innovation in nowhere.

What of metrics like number of Pulitzers, Pritzkers, etc to measure innovation?

Steve Jobs biggest innovation were related to "art", not STEM, focusing on form to drive function.

Asian cars were classically inferior to Detroit cars because Detroit innovated on style, form, while Asia focused on function, and form suffered.

Comments for this post are closed

The presence of geographic spillovers suggests that colleges do cause an overall net increase in patenting, although I find no evidence that colleges are better at promoting invention than other policies that lead to similar increases in population.

Colleges are mixing places where you get people who are new to a topic mix with people who are experts AND where people mix with experts in fields totally outside their primary area. Classic story here is Steve Jobs sitting in on a calligraphy class and realizing fonts matter a lot. Without mixing, computer science types might have seen typefaces as a waste of computer power and memory while a 'better computer' would concentrate on optimizing those other outputs.

Comments for this post are closed

How much innovation does a blog by a professor promote? Empirical support would be much appreciated.

Comments for this post are closed

Not the right measure? Like using chest size to "measure" IQ? Lamppost thinking. Only from Social "Scientists". What is the "rule" for economics publications? Two researchers use it, and it is a school and if more than 5 use it, it is canon?

Comments for this post are closed

Comments for this post are closed