The booming American cities of the 21st century are not just the ones you think

Madison, Wisconsin skyline

By Timothy Bartik, Brad Hershbein, Kathleen Bolter and Kyle Huisman

It’s common wisdom that U.S. economic growth in recent decades has been concentrated in a few large, high-tech coastal or Sun Belt centers, while former industrial areas have suffered. Our new research shows that this view isn’t entirely wrong, but it is incomplete.

For example, while the economic success of big, coastal cities such as San Francisco and Seattle has been widely celebrated, less attention has been paid to the growth in mid-size “heartland” cities such as Canton, Ohio; Madison, Wisconsin; and Omaha, Nebraska. 

The trouble with the common wisdom is that raw differences between places don’t compare apples to apples. It’s relatively easy to calculate how fast earnings have grown in different labor markets, but these superficial comparisons can produce an inaccurate picture of which areas have been succeeding—and why. 

For a better comparison, we created a measure of real earnings growth that adjusts not only for overall inflation but also for changes in local prices and the labor market’s demographic composition. Adjusting for prices and demographics is critical because we are interested not just in the number of dollars people earn, but in their overall economic well-being, which can be heavily influenced by the characteristics of where they live.  

Prices matter because earnings go farther in Manchester, New Hampshire, than in neighboring Boston, for instance. Skyrocketing housing costs can swiftly cancel out increased earnings, leaving residents no better off. Growth in real wages and real earnings—adjusted for local prices and inflation—can look quite different from the more visible trends in unadjusted wages and earnings. 

Demographics matter, too. For instance, if the share of college graduates in a place rises and nothing else changes, average annual earnings in that place will likely rise, too, because people with college degrees tend to earn more than those without. But these averages don’t tell us whether college graduates (or non-graduates) actually earn more than they did before—just that there are more graduates living in that area. 

Our new measures allow us to make comparisons between places on a more equitable basis. We examine changes in real earnings in US commuting zones (CZs), groups of counties with commuting and economic ties. We adjust earnings to effectively equalize age, gender, race, and education across areas, and we further adjust for differences in housing prices, to create measures of a CZ’s economic success between 2000 and 2019.  

As the map below shows, we find some expected patterns—and some surprises. Among the more populous CZs, many of the booming areas are high-tech centers such as Seattle, San Francisco, and Boston-adjacent Manchester. However, the large CZ with the highest earnings growth is Pittsburgh, which grew more than 10 percent. Although Pittsburgh does have a substantial tech presence thanks in part to Carnegie Mellon University, it's often overlooked in stories about economic growth. 
Among mid-size CZs, many prospering areas are heartland centers, including Omaha, Nebraska; Des Moines, Iowa; Madison and Milwaukee, Wisconsin; Canton, Ohio; and Johnstown, Pennsylvania. This eclectic set of CZs has diverse reasons for growth, including government (state capitals), natural resources, tourism, and other sectors.

Map showing economic growth by communting zone

Comparing apples to apples upends the traditional understanding about economic growth during the past two decades. And the picture it reveals is heartening: while big, high-tech centers have indeed thrived, America’s economic growth has been broader and more widespread than that caricature suggests.

Our new interactive data tool allows researchers and economic development organizations to use these new measures to benchmark how their community did in the first two decades of the 21st century compared to peer communities in the US. Users can explore comparisons both for overall populations and for groups defined by age, gender, race and ethnicity, and education. (The data tool uses core-based statistical areas (CBSAs) as the unit of geography, as these might be more recognizable to many users than CZs. Our results are similar for CBSAs and CZs.) The entire database, which contains data at the level of both CBSAs and CZs, can be downloaded for further research and policy analysis.

Details of how the measures of economic growth are constructed are available in the report, “Broadly Shared Local Economic Success Since 2000: New Measures and New Lessons for Communities,” by Timothy J. Bartik, Brad Hershbein, Kathleen Bolter, and Kyle Huisman.