Public Policy Forum Blog

With income inequality, Milwaukee is not San Francisco

A recent Brookings Institution report on rising inequality within major cities indicated that Milwaukee is one of the top six metropolitan areas in which inequality is growing. A follow-up article in the Milwaukee Journal Sentinel referred to the study and reports that “Milwaukee was singled out as one of the US cities where income inequality widened most rapidly during the recession and its aftermath.” The other five areas are San Francisco, Atlanta, Miami, Sacramento, and Jacksonville.

Surprised to find ourselves in a group that includes the relatively more prosperous metro areas of San Francisco (where median household income in 2012 was $73,802) and Atlanta ($56,557), we decided to take a closer look at how the study measured inequality and why Milwaukee ($35,823) happened to rise to the top six.  

We found that the Brookings’ report measured inequality in both 2007 and 2012 as a ratio of the median income of the top 95% of households to the median income of the bottom 20% of households. The change in inequality was then measured as the change in the ratio. The problem is that the ratio can go up or down for a number of mathematical reasons that have to do with the relative change for each income level, but do not really measure growth or decline in inequality or relative incomes. For example, if household income in a particular metro rises $2,500 for the top 95%, and for the bottom 20% rises $500, there would be no change in the inequality index. Yet the top gained five times as much income as the bottom.     

We believe a better way to look at the numbers is to consider Milwaukee’s peer group of cities in 2007 and look at our combined experience. We define a peer group as cities where the bottom 20% of households made the same in 2007 as did those in Milwaukee: $16,809. Give or take $1,000 and using the data from the Brookings report, that group includes Kansas City, Missouri; Tucson; Tulsa; Baltimore; and Boston.

What becomes clear through this exercise is that inequality can decline, or grow less rapidly, if “everyone suffers.” In four of the six peer cities (Kansas City, Tucson, Baltimore and Boston), income declined for both the “20%” and for the “95%.” For example, Boston saw incomes at the 95% level drop by $15,000.  When we compare Boston to Milwaukee, where the 95% income held steady, it becomes clear that Milwaukee’s inequality ratio change outpaces Boston’s only because we did not experience the “drop at the top” that Boston did. In fact, the median income for Milwaukee’s “95%” is the lowest of our six peer cities. Is this what we want?

This challenges our perception that changes in income inequality are being driven by a transfer of income from those with less to those with more. Instead, it seems, most families are losing ground. 

Virginia Carlson