Employment is increasingly concentrating, and a few cities lead the way

The Brookings Institution’s Metropolitan Policy Program has a new report out, Where jobs are concentrating and why it matters to cities and regions” looking at the density of employment in the nation’s metro areas, and how its changed in the past decade.  Here are the takeaways:

• The density of employment in cities and metro areas is increasing
• A handful of leading metropolitan areas account for most of the increase in job density
• Job density varies systematically by industry, with knowledge-intensive industries concentrating in dense locations, and goods-producing and distribution industries becoming less dense
• These data confirm an inflection point after 2007; job density which had been flat to declining, has increased steadily since then

A New Metric:  Job Density

Most urban planners will be familiar with metrics of population and housing density, measuring the concentration of population in particular places by counting the number of people or dwelling units per acre or square mile, and using this metric to compare and contrast the relative density of different areas. The Brookings report develops a parallel metric for the economy.

The heart of the Brookings report is a measure of job density constructed from very detailed, block level data on employment.  The report takes a more sophisticated spatially disaggregated approach to computing job density, developing a metric it calls “perceived job density.”  A simple minded measure of density is just dividing the total number of jobs in a metro area by the region’s urbanized land area. What this misses is the highly concentrated nature of job centers. What the Brookings metric does is to compute job density based on an average that is essentially job-weighted, rather than area-weighted. What is the level of density of experienced by the median worker, rather than the level of density on the median acre of urbanized land?  (This approach is analogous to the way the Census Bureau computes population-weighted population density). Brookings calls its measure “perceived” job density, but might want to think of it as “experienced job density”: the level of density of employment experienced by the typical worker.

The report shows that there are wide variations in job density across US metropolitan areas.  Unsurprisingly, New York has the highest job density of any metro area, with nearly 140,000 workers per square mile.  San Francisco and Chicago are a distant second with fewer than 50,000 workers per square mile, and only a handful of other metros have more than 20,000 workers per square mile. Most other metro areas appear to have job density of less than 5,000 workers per square mile.

Big and successful metro economies have high levels of job density

In basic sense, density is at the heart of urban economies.  Cities thrive and are more productive, because they bring lots of people with different knowledge and skills into close proximity. This is especially true for knowledge intensive industries like finance, software, and professional services.  Brookings data confirms that these industries have high levels of job density, while in contrast, goods producing and goods moving industries like manufacturing, transportation and distribution, tend to be much less dense.  Finance, information industries, professional services and company headquarter employ between 40,000 and 70,000 workers per square mile, while manufacturing and logistics, on average, employ about 7,000 per square mile.

Given this relationship, it’s hardly surprising that the greatest gains in density have been recorded in those metropolitan areas with the strongest concentrations of information and professional services.  The Brookings report concludes that just four metropolitan areas accounted for the bulk of the increase in job density in the US.  The following chart shows that job density in these four metro areas increased 40 percent between 2004 and 2015, while in the aggregate, job density increased by only about 10 percent in all the remaining metropolitan areas.

In a sense, this is more evidence for the emergence of so-called “superstar” cities.  These places flourish because they provide for the dense concentration of talent that supports innovation and productivity. Workers in these locations are more productive than they would be if they were scattered more widely among in less dense job locations. Locating in these dense concentrations of workers gives firms a competitive advantage they would be hard pressed to replicate elsewhere.

There’s been a shift to increasing density

This new Brookings report sheds additional light on a subject we’ve been researching at City Observatory for several years.  In 2015, we published our report Surging City Center Job Growth, measuring job centralization–the number of jobs in the center of metropolitan areas–using the same underlying data Brookings used to compute job density. For decades, metropolitan employment has been decentralizing.  Our analysis showed that this pattern of decentralization held during the last economic expansion (between 2002 and 2007); suburban employment grew faster than in the urban core. But after 2007, a different pattern prevailed: urban cores grew faster than their surrounding suburbs. This marked a historic shift in job location.

The Brookings report doesn’t chart trends in job density prior to 2004, but it seems likely that job density was decreasing steadily in prior decades, as jobs and economic activity suburbanized.  While the Brookings data don’t look explicitly at centralization, but the growth that the report charts in concentration, particularly in knowledge intensive industries, and in thriving metropolitan areas, is consistent with the pattern we documented in Surging City Center Jobs.

Quantifying the urban spatial economy

This new metric is an important addition to the set of fundamental statistical measures we can use to characterize and compare the spatial aspects of metropolitan areas.  Brookings has previously developed other measures of urban economies. In 2001, Ed Glaeser published his analysis of job centralization in different metropolitan areas, using a series of radii and zip code level estimates of employment to measure what fraction of metro employment was in the core compared to the periphery in each metro area.  In 2009, Elizabeth Kneebone updated Glaeser’s original work in her report on Job Sprawl Revisited.  And in 2016, Kneebone used data from the LEHD program to compute job accessibility and median commute distance for different metropolitan areas.  Coupled with the new job density data, these metrics let us triangulate the spatial aspects of metro economies.  Job sprawl measures capture centralization and decentralization, job density shows whether employment is concentrating or becoming less dense, and the job accessibility and commuting measures show how this affects demand for transportation and access to economic opportunity.

Further exploration

The new job density metric gives analysts another way to statistically describe the dimensions of urban economies. The next step will be to take a more detailed look at the correlations between this metric and other indicators of urban economic success.

One of the key questions has to do with the effect of the level of job density in a metro area and its economic health and development.  A few areas (New York, Chicago, San Francisco) have levels of job density that are vastly greater than almost all other metro areas. What advantages (and disadvantages) does a high level of job density confer for these cities? Is there some threshold level of job density that is necessary or desirable for promoting a region’s economic health?  There’ll be plenty of grist for further statistical analysis from this new data series.

The report was researched and written by Chad Shearer, Jennifer Vey and Joanne Kim of Brookings Institution’s Anne T. and Robert M. Bass Center for Creative Placemaking. City Observatory’s Joe Cortright is a member of the Bass Center’s Advisory Committee, and reviewed a draft version of the report prior to its publication. The report, including interactive data displaying job density data for 94 of the nation’s largest metro areas, is available here.