When thinking about how to revive economically lagging regions, especially in the Rust Belt, I often talk about the importance of universities. Big, high-quality research universities have been essential for creating technology clusters in Austin, Raleigh and San Diego. But even small colleges in rural areas can have big benefits for the surrounding area.
Why are colleges so great? To use universities as a tool for economic development, it’s important to think about the many ways they contribute to local growth - and some ways they don’t.
Most people think of colleges mainly as educational institutions. Economists argue about whether education improves people’s skills or is just a fancy form of time-wasting credentialism.
But most people seem to believe that higher education is important and useful. This naturally leads them to think of colleges as places that educate the local populace.
At first glance, evidence would seem to bear this out. Economist Enrico Moretti has found that an increase in the supply of college graduates in an area results in higher wages for workers of all education levels. Other studies generally conclude that a more educated populace leads to higher productivity. Moretti finds that places where the U.S. government gave land grants for colleges over a century ago have more educated workforces to this day.
But a closer look at the data reveals some problems with the basic story. The people who get their educations at the local university often don’t stay in the region. While some cities tend to keep 70 percent or more of the alumni of local colleges, others have retention rates below 50 percent. This is also a problem at the state level - career information website Zippia found that some states lose more than half of their college graduates.
Meanwhile, graduates tend to stay in states like Texas, California, and Illinois, which have big cities with lots of potential employers (and lots of potential friends and mates).
Careful research by economists Jaison Abel and Richard Deitz confirms that there’s only a weak relationship between the number of degrees a university produces and the number of educated people who live in the area. Educating locals does lead to having a more educated populace, but there’s not very much bang for the buck.
If colleges are paying to educate the future workforces of far-off cities and states, some might think there’s no reason for a city or state to spend money on a local university. But there is a good reason. Even as universities lose graduates to other regions, they attract other smart, educated people.
Abel and Deitz find that university research expenditures have a strong effect on the number of educated people in a region - over four times as strong as the effect of degree production. This is partly because skilled workers come to do research at the university itself. But most of the effect comes from private-sector activity in the surrounding economy.
When a university spends a lot on research, ideas and technology leak out to surrounding businesses in myriad ways. Universities cross-license technologies to the private sector. Academics consult for local businesses. Grad students, researchers, and professors start local businesses of their own. Companies establish research centers and hire smart people away from their Ph.D. programs or campus jobs. Some universities provide forums for local entrepreneurs, inventors and academics to meet each other, exchange ideas and offer employment.
High-productivity technology businesses therefore tend to cluster around universities, in order to take advantage of the rich flow of ideas and skilled workers.
That, in turn, draws smart educated people from other regions, boosting productivity and raising wages even for less-educated locals.
The policy implication is clear. In order to boost local economies, universities should stop seeing themselves only as educators, and start seeing themselves as platforms for local economic activity. Cleveland State University researchers Richey Piiparinen, Jim Russell, and Charlie Post call the former a “consumer university” model, and the latter a “producer university” model. They apply the distinction to explain the diverging performances of Cleveland and Pittsburgh. Cleveland’s colleges, they say, are still too focused on educating locals, while Pittsburgh’s - especially Carnegie Mellon - have taken an active role in boosting the city’s technology industry.
That doesn’t mean universities benefit cities only if they create technology clusters - even a modest college in a small town can create a pleasant environment that draws in locals from surrounding rural areas, channels government spending to an area and increases the education of the local workforce.
But for top research universities in languishing areas, the primary goal should be to form the seed of a local knowledge-based business community. That means focusing on high-quality research, not just on teaching.
Noah Smith is a Bloomberg View columnist.