THE NEW MAP OF ECONOMIC GROWTH AND RECOVERY
The U.S. economy is in a constant state of evolution, but recessions tend to bookend distinct eras of growth and development. The Great Recession was no different. The ongoing expansion continues to take shape, but it so far has been characterized by deep economic anxiety persisting alongside steady headline growth. Getting to the bottom of that paradox motivated this report. This analysis surveys the economic landscape emerging from the Great Recession and compares it to previous recovery periods. It identifies differences in the strength and geography of countylevel growth in employment and business establishments — two key markers of economic dynamism — and uncovers three significant transformations in the economy. The first and most unambiguously troubling is a collapse in the number of new firms in the economy. The second is the increasing geographic concentration of recovery-era businesses and jobs into a smaller number of more populous counties. The third is the shift in the counties driving the nation’s economic recoveries from smaller to larger ones. Together, the findings capture an economy veering towards a less broadly dynamic, less entrepreneurial, and more geographically concentrated equilibrium — more reliant than ever on a few high-performing geographies abundant in talent and capital to carry national rates of growth. Even in the relatively short period of time analyzed here, patterns have reversed. Large urban counties dominate where they once lagged, while small counties have nearly disappeared from the map of recovery altogether. At the national level, a scarcity of new businesses implies a future of reduced economic dynamism. New businesses play a disproportionate role in commercializing innovations, stoking competition, and driving productivity growth. They also create the bulk of the nation’s net new jobs and provide the extra demand that is critical to achieving wage-boosting full employment. What is more, the geographically uneven nature of the collapse in startups implies that wide swathes of the country will soon be contending with the consequences of a missing generation of enterprise.
I. An unprecedented number of counties continued to lose jobs during the 2010s recovery.
The U.S. economy added 9.1 million new jobs from 2010 to 2014, compared to 7.5 million over the first five years of the 2000s recovery and 9.4 million over the first five years of the 1990s recovery. Those figures represented job growth rates of 10.1 percent in the 1990s, 6.7 percent in the 2000s, and 8.1 percent in the 2010s. As with business establishments, the geography of job growth has narrowed from one recovery to the next. Following the 1991 recession, only 14 percent of counties continued to post job losses over the course of the next five years. That proportion rose to 28 percent in the 2000s and to 31 percent in the 2010s (three-quarters of which lost population at the same time). In other words, the share of U.S. counties participating in national jobs recoveries has fallen from 86 percent in the 1990s to 72 percent in the 2000s and only 69 percent in the 2010s. The share of the country’s population living in bypassed counties varied but was relatively low in each period: During both the 1990s and 2010s recoveries, 11 percent of the population lived in counties that continued to lose jobs. Over the 2000s, when manufacturing employment losses pummeled more populous corners of the Great Lakes, more than one in five Americans lived in counties that were bypassed by the national jobs recovery. Job growth has remained consistently more pervasive than business establishment growth across counties over time. This is likely due to the fact that firms respond to changing economic conditions by shrinking or expanding their workforces before closing or opening entire establishments, which represent costly sunk investments.
Changes in population alone do not explain the geographic shift in employment and establishment growth. In practice, the share of U.S. counties experiencing population decline increased from 22 percent in the 1990s to 35 percent in the 2000s and 54 percent in the 2010s. However, roughly one-third of the counties that lost business establishments and one-quarter of the counties that lost jobs in the 2010s actually saw an increase in population. Furthermore, more than one in five counties that lost both jobs and establishments nevertheless saw their populations rise over the same period.
III. Fifty percent of 2010s job growth accrued to only 2 percent of U.S. counties.
Half of the 2010s recovery’s 9.1 million new jobs accrued to only 73 counties, which together contained only 34 percent of the population and 39 percent of the nation’s employment base. Job creation was much more distributed during the 2000s, when 120 counties together produced half of all net new jobs, and during the 1990s, when 107 counties contributed to this benchmark.
Nevertheless, the number of people living in such counties has held relatively steady between 43 and 46 percent — implying again that the counties experiencing heady growth are increasingly the more populous ones.
In every period, however, a majority of the country’s population lived in counties with lagging job growth. In only 12 states did a majority of residents live in a county that saw employment increase at least as fast as it did nationally from 2010 to 2014. Three of those states — Arizona, Florida, and Nevada — were epicenters of the housing crash. Others — North Dakota, Oklahoma, and Texas — benefitted from the energy boom, which has already started to dissipate. California, Colorado, Minnesota, Oregon, and Utah round out the list — all places benefiting from technology and high-end service sector-led growth.
III. Fifty percent of 2010s job growth accrued to only 2 percent of U.S. counties.
Half of the 2010s recovery’s 9.1 million new jobs accrued to only 73 counties, which together contained only 34 percent of the population and 39 percent of the nation’s employment base. Job creation was much more distributed during the 2000s, when 120 counties together produced half of all net new jobs, and during the 1990s, when 107 counties contributed to this benchmark.
In each period, the counties responsible for half of the country’s job growth together contained roughly one-third of the country’s population. The U.S. economy relied on fewer but generally more populous counties to generate the bulk of the 2010s jobs recovery than in the past. This could represent a deeper shift in the mix of industries driving the country’s economic growth towards ones that thrive in larger urban agglomerations. It could also reflect that larger (and presumably more diversified) counties were better equipped to bounce back from this exceptionally deep recession.
Indeed, many of the counties with the highest volume job growth were also the country’s largest. Others specialized in fast-growing technology or energy industries, and all together, of the 20 counties that generated the largest absolute increase in employment, 19 posted faster job growth than the country as a whole (8.1 percent). As Figure 15 on the previous page shows, job growth far outpaced population growth in all 20 of the counties with the largest absolute increase in employment.
Each map of the counties leading the past three recoveries (Figure 17) captures a distinct era in a constantly evolving U.S. economy. The 2010s map is clearly urban in orientation and the sparsest of the three. The 2000s map, by contrast, is much fuller and heavily oriented towards Florida, the Southwest, and outlying counties on the Eastern Seaboard, while manufacturing losses served as a drag on job growth in the Midwest. The 1990s map is much more balanced geographically. In that era, the Bay Area and Boston reaped the rewards of the first information technology boom; the Mountain West was being settled; and from Little Rock, AR, to Greenville, SC, the Southeast was one of the country’s strongest engines of job creation. To an even greater extent than with business establishment growth, the Rust Belt fueled the nation’s 1990s employment expansion: The counties containing Akron, Cleveland, Cincinnati, Columbus, and Dayton, Ohio, all joined a large national network of peers to deliver half of the economy’s new jobs.
IV. The geography of high-quality job growth was uneven.
Middle-wage workers suffered the starkest change in circumstances due to the 2009 recession — a reality that plays out in local labor markets across the United States. Fully half of the 8.9 million jobs lost from 2008 to 2010 were in middle-wage sectors, 36 percent in low-wage sectors, and 14 percent in high-wage ones. Half of the recovery’s 9.1 million jobs, however, were in low-wage sectors, and only 31 percent were in middle-wage sectors and 19 percent in high-wage ones.4 As a result, by the end of 2014, the economy had recovered only three out of every five middle-wage jobs lost to the recession. That left a gap of 1.8 million missing middle-wage jobs. An analysis of four leading job-growth counties reveals that the quality of jobs generated by the recovery varies geographically, too.
In Orange County, FL (Orlando), for example, almost all new jobs appeared in the low-wage accommodation and food services, retail, arts and entertainment, and administrative services sectors. In Sun Belt cousin Maricopa County, AZ (Phoenix), job growth was more diversified, with high-wage professional services and finance and middle-wage construction, manufacturing, and transportation all growing too. Advanced industry-heavy King County, WA (Seattle), for its part, saw some of its strongest job growth in middle- and high-wage construction, manufacturing, information, and professional services sectors. Meanwhile, white-collar, headquarters-heavy Hennepin County, MN (Minneapolis), saw almost all growth take place in the very high-wage management sector with only modest expansions elsewhere. These four profiles suggest that baseline levels of low-wage job growth were a nationwide hallmark of the 2010s recovery, while the distribution of new high-paying jobs remained much more uneven.
The country’s most populous counties powered the 2010s recovery.
I. The largest counties produced 58 percent of the country’s new business establishments.
II. The largest counties produced more than twice as many jobs during the 2010s recovery as they did in past ones.
III. The most populous counties have become the fastest growing.
IV. Only 15 large counties enjoyed their strongest recovery in the 2010s.
I. The largest counties produced 58 percent of the country’s new business establishments.
New business establishments consolidated into the country’s largest markets over the 2010s to an extent not seen during prior recoveries. In the 1990s, counties with under 500,000 people generated 71 percent of all new business establishments — a total of nearly 300,000. By the 2010s, counties with over 500,000 people had become dominant, generating 81 percent of all new business establishments — despite housing only 47 percent of the country’s population. Were it not for these relatively few pockets of resiliency, the U.S. economy would have seen near-total stagnation in its business landscape.
In total, counties with over one million people added 99,000 of the country’s 165,000 net new business establishments from 2010 to 2014. Meanwhile, counties with under 100,000 people saw a sharp reversal from the 1990s, swinging from a net increase of 135,000 business establishments from 1992 to 1996 to a net decrease of 17,500 establishments from 2010 to 2014. The steep drop-off in new establishment openings in small and mid-sized counties is responsible for much of the national-level collapse.
II. The largest counties produced more than twice as many jobs during the 2010s as they did in past recoveries.
Over prior recoveries, large counties did not play a central role in delivering the country’s employment growth. This is no longer the case. The most populous U.S. counties — those with over one million people — created 3.3 million jobs over the 2010s recovery, more than any other size class of counties and more than twice as many jobs as they created over each of the two prior recoveries. By contrast, counties with under 100,000 people collectively created fewer than one million new jobs in the 2010s — substantially fewer than 2.5 million they created from 1992 to 1996 and the 1.2 million new jobs they created from 2002 to 2006. As a result, the 2010s recovery has served to accelerate a shift in the country’s economic gravity towards populous counties after decades of decentralizing growth that spread economic activity to more locales.
III. The most populous counties have become the fastest growing.
In a recovery that brought economic development disproportionately to those places lacking it, small counties would post higher growth rates than larger ones. On the other hand, in a recovery that reinforced or even accelerated the concentration of economic activity into existing economic centers, large counties would post higher growth rates than smaller ones. And indeed, the 2010s recovery was distinguished by the largest counties registering the fastest growth rates on average. This development directly challenges the conventional wisdom that growth rates in larger and highly developed cities are by nature slower than growth rates in smaller and less developed ones (just as at the country level one does not expect the United States to post as high of growth rates as China). It also challenges the basic tenants of economic geography that suggest that economic activity will spread from high cost, congested locales to low cost ones as a matter of course. Ultimately, it implies that the knowledge- and technology-driven industries that power modern growth have very different — perhaps even more than already recognized — locational requirements from the industries that led past eras of growth.
In the 1990s, the smallest counties added business establishments at the fastest rate, while the largest counties trailed the pack. Those same counties saw growth fall precipitously over the 2000s recovery, though. By the 2010s, all counties with fewer than one million people averaged far lower establishment growth rates than in the past. In contrast, the 39 counties with over one million people posted average establishment growth rates more than twice as high as the national 2.3 percent.
Meanwhile, the smallest counties saw more business establishments close than open, resulting in a -1.0 percent average growth rate — despite the energy boom that lifted many corners of rural America over the period.
The same pattern holds for job growth. The economic expansion of the 1990s was fueled by economic development spreading to more locales — a “rise of the rest” story — that disproportionately benefitted smaller counties. Counties under 100,000 people averaged 16 percent job growth during the 1990s recovery, while counties over one million averaged only 7.7 percent job growth over the same period. Nationally, employment increased by 10.1 percent.
The 2000s recovery was characterized by an overall convergence in growth rates across size classes even as mid-sized counties pulled ahead as the era’s job growth leaders. By the 2010s, the largest counties were growing faster than any other group, averaging 9.9 percent job growth — their fastest in recent history — and reinforcing the concentration of the nation’s economic activity within them. Average job growth rates in smaller counties, meanwhile, had fallen to their lowest levels in recent decades.
IV. Only 15 large counties enjoyed their strongest recovery in the 2010s.
Despite the pervasiveness of large county growth, only 15 counties of the roughly 130 with over 500,000 people posted both their strongest job growth and their strongest establishment growth of the past three recoveries in the 2010s. Four Bay Area and five New York City area counties were in the group, counterintuitively placing the country’s highest cost and most congested locales as some of its fastest growers. Also in the mix are signs of hope in some Rust Belt communities: Erie County, NY (Buffalo), Monroe County, NY (Rochester), and Allegheny County, PA (Pittsburgh) also rose to the list of places thriving most relative to their own pasts, alongside Denver County, CO, and Harris County, TX (Houston). Nevertheless, the relative shortness of the list makes clear that, although the 2010s recovery was led by cities, many still underperformed relative to their own recent histories. The headwinds affecting the national economy — notably the collapse in new business formation — continued to batter most of the nation’s cities, hemming in national growth, as well.
CONCLUSION
The United States has undoubtedly enjoyed more robust GDP and job growth following the global financial crisis than most developed economies. Its relatively swift recovery is a sign of national resilience. However, in light of the findings presented here, that resilience seems due more to the dynamism of its major metropolitan centers than to grassroots economic vibrancy nation-wide.
The geographically uneven nature of the decline in new business starts implies that large swathes of the country will soon contend with a missing generation of firms — ones that should be providing employment opportunities and new foundations for economic growth in the years ahead. The uneven geography of new business formation tracks very closely with that of access to capital — particularly venture and other forms of risk capital. Addressing the former challenge will surely involve tackling the latter. Without mitigating these disparities, the trend towards increasing concentration documented here may even accelerate, given that today’s largest economic centers are the few remaining places producing tomorrow’s new businesses.
The dynamics captured in this report result to some extent from a global trend towards the clustering of knowledge-based economic activity (the modern economy’s growth sector) in large, connected cities. People are following economic opportunity to cities as well — the share of the country’s population residing in counties left behind by economic recoveries has not significantly increased over time. Such macro trends are neither inherently negative nor in need of mitigation. They do suggest, however, that people and places on the wrong side of the trajectory of economic change may need assistance to adjust and cultivate new competitive advantages.
The new map of growth and recovery points to very different futures for American communities. These findings suggest that the gains from growth have and will continue to consolidate in the largest and most dynamic counties and leave other areas searching for their place in the emerging economic landscape. While many will benefit, the new map also calls for a new toolkit for ensuring broad access to opportunity and helping both people and places realize their economic potential.