For the most part, high-income places 40 years ago remain high-income today, and low-income places 40 years ago are still relatively low-income.  This blog post looks at how little change there has been in how places rank by income over the last several decades. There are a few exceptions, and those metros that have climbed the ranks might offer lessons for economic development strategies and place-based investments.

This is the second in a series of blog posts on regional economic trends and place-based efforts for the Department of Commerce’s Regional Economic Research Initiative, which links place-based program data with local economic data to improve the design, implementation, and evaluation of these programs. The first blog post documented the rise in geographic income inequality, showing that the gap in incomes between richer places and poorer places has grown since 1980. This second post builds on the first one: not only has the gap between richer and poorer places widened, but – in general – there is little change in which places are richer and which are poorer. Economic opportunity tends to be concentrated in some places and not others, and most places with less economic opportunity for their residents remain so over time.

Local income rankings have changed little

Places with higher average incomes today tended to have higher average incomes in the past. Looking across all metropolitan and micropolitan areas in the U.S., the correlation between local per-capita income in 1980 and in 2021 is 0.81 – a very strong relationship. A scatterplot showing 1980 and 2021 income for the 110 largest metropolitan areas – all metros with population of 500,000 or more -- shows several places with high income in both years.

Among these 110 largest metros, six were among the top ten for income in both 1980 and 2021: San Jose; Fairfield County, CT; San Francisco; Seattle; Washington, DC; and Denver. Among the ten large metros with the lowest income in 1980, four remained in the bottom ten in 2021, and another three remained in the bottom 20.

Furthermore, the places with higher income in 1980 tended to have faster income growth from 1980 to 2021. This contributed to the growing gap between higher- and lower-income places, as described in our previous blog post on geographic income inequality.

These patterns hold when looking at a broader set of places than just the 110 largest metros. Among smaller metros, Midland, TX, and Naples, FL had the highest per-capita income in 2021 and were among the highest-income smaller metros in 1980 as well.

Going further, we looked at two other measures that closely track local average income with less year-to-year volatility: (1) educational attainment and (2) occupational mix. Unsurprisingly, local incomes tend to be higher in places where a higher share of adults has a college degree and where a higher share of jobs is in managerial, professional, or technical occupations. See methodology section for more on why we chose these two measures.

Educational attainment and occupational mix confirm how persistent local economic characteristics are:

• The correlation between local college degree share in 1980 and 2021 is 0.87 – slightly stronger than the correlation for average income. The correlation between local share of jobs in managerial, professional, or technical occupations in 1980 and 2021 is similarly high at 0.86.
• Local shifts in educational attainment and occupational mix reflected growing inequality.  The college degree share rose more between 1980 and 2021 in places where college attainment was higher in 1980; same with the share of jobs in managerial, professional, or technical occupations.

The places that have made big economic gains

Even though differences across places in income, education, and occupation change little over time, there are places that bucked the trend and experienced transformative positive economic changes between 1980 and today, climbing substantially from a relatively low level of income or education attainment.

As noted above, of the 110 largest metros, six were among the top ten for income in both 1980 and 2021. Two others in the top ten in 2021 were in the top third in 1980: New York (14^th of 110) and Boston (31^st of 110). But two others made much bigger leaps into the top ten: Austin, which was middle-income in 1980, and the Northwest Arkansas metro of Fayetteville-Springdale-Rogers, which was among the lowest-income metros in 1980.

Among large metros that were not already in the top third in 1980, Northwest Arkansas had the biggest percentage increase in per capita income. Its inflation-adjusted income nearly tripled between 1980 and 2021. Metro Provo-Orem UT had the second-largest increase, with inflation-adjusted income 2.4 times higher in 2021 than in 1980. Nashville and Austin had the next largest increases in income among large metros not already in the top third in 1980.

Applying the same logic to educational attainment, two additional large metros stand out for transformative positive change since 1980. Both Charlotte and Pittsburgh saw outsized increases in educational attainment over the past four decades from a relatively low level in 1980. In both metros, the share of adults with a college degree rose from 14% in 1980 to 37% in 2021. These metros also had notable jumps in the share of jobs in managerial, professional, and technical occupations.

Future blog posts in this series will explore case studies of local economic transformation.

Economic opportunity expands where jobs AND income grow

Since 1980 many places across the U.S. have had astounding job or population growth, while others have had little or have even shrunk. Job growth is important for expanding local opportunities, but job creation on its own does not ensure incomes rise too. In fact, there’s little correlation between employment (or population) growth and rising incomes.

Some places with high-income growth from 1980 to 2021 also had strong job growth, including northwest Arkansas, Provo-Orem, UT, Austin, Nashville, and Raleigh. Income and job growth over the past four decades often went hand-in-hand in places like these that started at middle- or lower-income levels. But other places with similarly fast-rising incomes since 1980 had relatively slow job growth, like San Francisco, San Jose, Fairfield County, CT, and Boston: these places had relatively high incomes in 1980.

On the flip side, some Sunbelt metros experienced rapid job and population growth but with relatively slow per-capita income growth, like Las Vegas and Riverside-San Bernardino-Ontario CA. Still, other places had slow job and population growth along with little income growth, such as Cleveland, Detroit, Rochester, NY, and other metros in the Great Lakes and Midwest region.

Overall, the correlation between employment growth and income growth from 1980 to 2021 across all metropolitan and micropolitan areas was 0.08 -- small but statistically significant -- and a not-significant -0.01 correlation between population and income growth.

Why positive economic transformation is hard

Places with higher incomes in 1980 tended to have faster income growth between 1980 and 2021, contributing to widening geographic inequality and little change in the income ranking of places. Only a few metros stand out for dramatic income growth from a low starting point during these decades, and rapid employment or population growth did not necessarily lead to especially strong income growth.

One reason few places have transformed upward is that faster-growing, higher-paying sectors in recent decades tend to cluster in the same places over time. Tech, finance, and professional services have become a growing share of the economy and pay high wages. They exhibit “agglomeration economies”, which mean that firms and workers are more productive when located near others in the same industry, and this effect is especially strong for knowledge-based industries with highly educated workers who have specialized or technical skills. Agglomeration economies keep those industries clustered in specific places. Also, since 1980 rising housing costs in higher-income places contributed to the widening inequality across places. Looking farther back in time, the highest-income metropolitan areas in 1950 included several in the Midwest that slipped in the rankings by 1980, while many of the lowest-income places in 1950 remain low-income today.

The difficulty of transformation suggests that successful place-based policies require careful strategies and possibly significant resources. As future blog posts looking at local case studies will show, leaps in local incomes or education levels often accompany a new or rapidly growing local industry cluster. Local agglomerations can be self-sustaining once established, but public investments can help support new or nascent agglomerations. Biden-Harris Administration place-based programs – totaling $80 billion in authorizations by one estimate – take these challenges and lessons into account. The Coal Communities Commitment and the Build Back Better Regional Challenge focus on shifting or diversifying the local industry mix. The Regional Innovation Engines and Tech Hubs programs aim to encourage agglomerations of innovative, high-paying industries in places outside leading technology centers.

One last word on these data: the local income and education data presented in this post show how uncommon positive transformation is and how persistent local economic gaps are, but these data are not the complete picture. High-income places often include areas of distress, and some of the highest-income metros like Fairfield County, CT, and San Francisco, are among those with the most income inequality across households within the metro. Furthermore, increases in local average income can be driven by the in-migration of higher-income residents or the out-migration of lower-income residents; changes in local average income do not necessarily reflect changes in long-term residents’ well-being. Understanding who benefits from local economic development is a rich area for deeper research.

Download Data Spreadsheet

Methodology notes

Income and population data are from the Bureau of Economic Analysis. Education and occupation are from the 1980 Decennial Census and 2017-2021 American Community Survey. 1980 education and occupation data were downloaded via IPUMS NHGIS.

All correlations, regressions, and residual calculations are population-weighted. All correlations, regressions, and residual calculations for income use the natural log of per capita market income, which includes wages and other earnings as well as interest, dividends, and other capital income, but not transfer income.

Local average income is an imperfect measure of local economic status, for several reasons. First, average incomes can be volatile year-to-year. Second, local cost of living varies, and the best cost of living measures are available only for 2008 onward. Third, local average incomes are affected by the mix of working-age adults, children, and retirees in the population. The two alternative measures of local economic status we analyzed were strongly correlated in 2021 with average income: 0.86 for college-degree share, and 0.81 for managerial/professional/technical occupation share.