The City/Suburb Income Gap- Bigger or Smaller?

Michael Lewyn's picture

The Brookings Institution's "State of Metropolitan America" database (at ) contains a wealth of information both on central cities and their metropolitan areas.  One issue I was curious about was the economic gap (or lack thereof) between cities and their suburbs.   In the late 20th century, many central cities (especially in the Northeast and Midwest) fell behind their suburbs economically, becoming the warehouses for the region's poor.  Did this continue to be the case in the 2000s?

To answer this question, I looked for information on city household income as a percentage of regional household income.  (I note that because cities tend to have smaller households than their suburbs, this measurement tends to make cities look poorer than would a focus on per capita income).*  In particular, I was interested in northern cities that had lost population for large chunks of the 20th century, as opposed to constantly-growing cities such as Columbus, Indianapolis, and many Sun Belt cities.

Generally, older central cities fell into two categories.  One category was cities, mostly larger cities, that had rebounded from their 20th century difficulties.  In these cities, the city/suburb income gap narrowed during the 2000s.  For example, in the District of Columbia, urban household income increased from 75% of regional household income in 2000 to 82% in 2009.  This pattern was common in the largest northern regions, such as Boston (74% in 2000, 83% in 2009), New York (from 76% to 78%) and even some population-losing cities such as Chicago (77% to 81%) and Pittsburgh (77% to 80%). 

A second category is a group that former Albuquerque Mayor David Rusk labelled as cities that have passed at "The Point of No Return"- cities that had, as of 2000,** never closed the economic gap with their suburbs.  Rusk labelled ten core cities of major (population of over 1 million) metropolitan areas in this way: Cleveland, Detroit, Birmingham, Buffalo, Baltimore, Hartford, St. Louis, Rochester and Providence.   How did these cities do?

The picture here is more mixed.  Some dying cities continued to die: Cleveland's median household income nosedived from 61% of the regional average in 2000 to 54% in 2009, Detroit's from 63% to 58%, Hartford's from 46% to 42%.  Some other cities sustained smaller losses: Birmingham from 70% to 68%, Baltimore from 60.2% to 59.5%, Philadelphia from 64% to 61%, and Rochester from 61.6% to 60.5%.   But two cities that were supposedly past the point of no return sustained income rebounds: St. Louis from 62% of the regional average to 67%, Providence from 63% to 68%.  (Buffalo also gained ground by an insignificant margin, from 63.6% to 64%).

A middle group of cities that suffered losses during the late 20th century but were not quite at the "Point of No Return" generally continued to become poorer than their suburbs, though rarely at a rapid rate.  Milwaukee's median household income declined from 71% of the regional average in 2000 to 67% in 2009; similar results occurred in Cincinnati (66% to 64%), and Minneapolis (70% to 69%). 

So what do we get out of this data?  I would say that as a general matter, well-off cities tended to become more well-off: the rebounding cities all had median household incomes at or above 3/4 of the regional household income.  These cities were sufficiently appealing to attract more affluent people.  By contrast, cities that had "passed the point of no return" mostly continued to lose wealth. 

The correlation between population growth and income growth was not 100 percent: some growing cities (like Philadelphia) became poorer relative to their suburbs, and some population losers (like Pittsburgh) did not.  Nevertheless, there seems to be at least a modest correlation between population and income: of the eight "Point of No Return" cities that continued to lose income, only two (Philadelphia and Hartford) gained people during the 2000s. 

*William Lucy's new book, Foreclosing the Dream, contains a wealth of data using per capita income.

**See David Rusk, Cities Without Suburbs: A Census 2000 Update, at 78-82 (2003 ed.)

Michael Lewyn is an assistant professor at Touro Law Center in Long Island.



Regional sharing arrangements = one approach?

Interesting look at interesting data. As planners, where do we go from here? To me, it suggests regional tax sharing or regional government mergers to bridge the gaps, especially in those areas where the core city has lost population and share of regional household income. Of course, these can be controversial proposals that oversimplistically can be labeled as "rich vs. poor" when in fact the arrangements can provide much more complex benefits.

Margins of error

I like the way you did this, by comparing the city to the metro area over time. But there are reasons to be careful with income stats from the ACS (and even the Census long form). The confidence intervals around some estimates are pretty large. For example, the 90% confidence interval around Cleveland's 2009 estimate of $24,687 is $1,410, which means that the median probably lies somewhere between $23,277 and $26,097. If you take the high end of that range and compare it to the low end of the metro-area range ($45,395 +/- $868), you get about 57%. The 2000 data are also estimates. Of course all this also implies we believe what people tell the Census about their income; if you don't, or not completely, it helps to look at how other things about the city compare to the metro area as a whole: levels of education, for example, are less likely to be unreported or mistakenly reported than income. (If you've filled out the ACS, as I have, you know it's hard to answer the income questions even if you want to do it honestly.) The confidence intervals around the tract-level estimates (2005-09) are truly mind-boggling, however (not to mention the impact of five-year averaging in such a tumultuous economic period). Proceed with utmost caution using those tract estimates.

Rolf Pendall

Prepare for the AICP Exam

Join the thousands of students who have utilized the Planetizen AICP* Exam Preparation Class to prepare for the American Planning Association's AICP* exam.
Starting at $245
Planetizen Courses image ad

Planetizen Courses

Advance your career with subscription-based online courses tailored to the urban planning professional.
Starting at $16.95 a month
Red necktie with map of Boston

Tie one on to celebrate your city

Choose from over 20 styles of neckties imprinted with detailed city or transit maps.