Does Human Capital Tend to Cluster in Center Cities or the Suburbs?

It's now conventional wisdom that human capital (what economists call educated people) is a key factor in the growth of cities and metro regions. Cities are engines of economic development, and the skilled people are the high-powered fuel that drives them.

Most studies of the role of human capital in regional economic growth track its effects on and across metro regions. But metros vary widely in size, shape, and spatial and demographic compositions; human capital doesn't always cluster in the same places at the same densities. The revitalization of the urban center in cities like New York, San Francisco, Boston, and others has fueled in-migrations of more skilled and affluent people. Other cities, like Los Angeles and Detroit, still suffer from the proverbial "hole in the donut" effect, with their more educated, higher income populations spread out across their suburbs while their urban centers lag.

"Human Capital in Cities and Suburbs" [PDF] a study I co-authored with my Martin Prosperity Institute colleagues Charlotta Mellander and Kevin Stolarick, looks at the distribution of human capital between the center cities and suburbs of United States metros and the effects this has on their economic performance.

To get at this, we used data from the 2000 Census to identify the percent shares of human capital (measured as adults with bachelor's degrees or above) in both the central city and the suburbs of 331 Metropolitan Statistical Areas.

Map by MPI's Zara Matheson

The map above, by the MPI's Zara Matheson, shows how U.S. metros stack up on the percentage of college grads in their center cities. And the table below shows the top 10 large metros (those with more than one million people) on this metric.

Table data from U.S. Census

Knowledge-based high tech metros top the list. Raleigh-Durham is first, with 45.9 percent of adults in the center city holding bachelor's degrees or higher, followed closely by Seattle (45.2 percent), and San Francisco (45 percent). Greater Washington, D.C. is fourth with 43.5, and Austin is fifth with 40 percent. Oakland in the Bay Area is sixth, Boston seventh and, San Jose eighth. West Palm Beach-Boca Raton, a resort destination which has been trying to bolster its high-tech base, is ninth and the Twin Cities of Minneapolis-St. Paul round out the top 10. More than 35 percent of the adults in their center cities hold bachelor's degrees.

Perhaps not surprisingly, college towns predominate when we add smaller metros (with populations of less than one million) to the list. With nearly 70 percent of adults holding bachelor degrees, Ann Arbor comes in first, followed by State College, Pennsylvania (69.2 percent), Iowa City (55.9 percent), Bloomington, Indiana (54.8 percent), Corvallis, Oregon (53.1 percent), Boulder, Colorado (50.9 percent), Columbia, Missouri (50.4 percent), Madison, Wisconsin (48.1 percent), Lawrence, Kansas (47.6 percent) and Champaign-Urbana, Illinois (47.4 percent).

Map by MPI's Zara Matheson

Which metros have the highest concentrations of human capital in their suburban fringe? The map above charts the percentage of college grads in the suburbs and the table below shows the top 10 large metros on this metric.

Table data courtesy of U.S. Census

Once again, knowledge-based, high-tech metros top the list. San Jose takes first; more than half (50.2 percent) of the adults in its suburbs hold bachelor degrees. San Francisco is next with 42.4 percent. Greater D.C. is third (41.5 percent), followed by Boston with 40.3 percent. These metros also have larger percentages of human capital in their center cities, as noted above. Rounding out the top 10 are the suburbs of New York (39.6 percent), Newark, New Jersey (34.7 percent), Denver (34.2 percent), Raleigh-Durham (34 percent), Oakland (33.9 percent), and Orange County, California (33.5 percent).

Several smaller metros have even higher concentrations of college grads in their suburbs. Suburban Stamford–Norwalk, Connecticut is first with 65.8 percent. The suburbs of Boulder are second with 54 percent. Danbury, Connecticut's suburbs have 45.9 percent and the upscale suburban areas surrounding Trenton, New Jersey (which include Princeton) have more than 41.3 percent. Not surprisingly, the suburban areas of college towns like Corvallis, Oregon, Burlington, Vermont, Iowa City, Iowa, and Gainesville, Florida also have high concentrations of human capital.

The graphic above by MPI's Michelle Hopgood enables us to compare metros across both dimensions, plotting their percentages of college grads in the suburbs on the X-axis and in the center city on the Y-axis. Metros in the upper right hand quadrant have high concentrations in both. These include college towns and knowledge-based metros like Boulder, Corvallis, Ann Arbor, Raleigh-Durham, San Francisco, Washington, D.C., Boston, Austin, Santa Fe, and New York, among others. Metros in the upper left hand quadrant have high concentrations of college grads in the center but lower percentages in the suburbs. Metros in the lower right hand quadrant have high concentrations of college grads in the suburbs and lower percentages in the center city. And metros in the lower left hand quadrant have low concentrations in both.

The main thrust of our study was to develop statistical models of the distribution of highly skilled people between suburbs and city centers so we could see how each affected economic performance. Basically, we found that human capital in suburbs and cities play different but important roles.

First, we found that suburban human capital is important to metros of all sizes, with the percentage of college grads in the suburbs having a positive effect on their income levels and housing values across the board. Suburban human capital is especially important in in smaller and medium-sized metros, those with fewer than one million people, according to our analysis.

But as metros increase in size, human capital in the center city plays a bigger role. For example, metros with more than three million people have roughly twice the density of college grads than those between one and three million — 443 versus 227 college grads per square kilometer on average. The percentage of college grads in the center city plays a greater economic role in large metros (those with more than one million people), especially with regard to housing values, according to our analysis. But, it has no statistically significant effect on either income or housing values in small and medium size metros.

This makes intuitive sense. As metros grow larger and more congested and commutes become more arduous, higher-skilled people seek out more central locations to live in. This post on the blog Old Urbanist highlights a tradeoff between location and commute times, citing a classic study [PDF] which argues that "individuals and firms mutually co-locate in response to congestion costs, and thus reshape those costs." The theory thus implies, as the post makes clear, that "commutes beyond a certain length of time are undesirable despite any other advantages that might be gained from the location (e.g. housing cost, school quality, taxation level, crime)." The post also points out, drawing from data on the New York metro area, that commutes tend to level off at about 30 minutes. When metros get large and congested, a sort of "Manhattan" or "Brooklyn Effect" sets in, as higher income, more affluent households seek more central locations, increasing the concentration of human capital in the center city.

For all the importance economists and urbanists accord to human capital, surprisingly few efforts have been made to examine its workings at a granular level. Not just academics but place-makers of all stripes — planners, politicians, developers — need to have a better understanding of where and why human capital clusters — and of the different roles that those different kinds of clusters play in economic development.

Lede image: gui jun peng/Shutterstock