Race Matters in Milwaukee: How Segregated is Milwaukee? Part 2

It turns out that Milwaukee is not the most segregated metro area after all.

 

Racial Map of Milwaukee

 

***

Racial Map of Detroit

 

(Both graphics by Eric Fisher)

***

The landmark report on segregation by the U.S Census Bureau published five measure of segregation. As previously discussed, this report ranked metro areas with a sufficiently large black population on how racially segregated they were. Then, the U.S Census Bureau averaged these rankings, and used that average to conclude that Milwaukee was the most segregated metro area in the country.

After all of the sophisticated statistical analysis that went into the production of the five segregation measures, it is surprising that the U.S Census Bureau would produce an overall segregation rank by averaging the segregation measure ranks, and not the measures themselves. As the following example shows, this distorts things.

Imagine three people whose wealth is measured in three different ways. You want to rank them in overall wealth by averaging their wealth from each measurement. In parenthesis below is the rank of how wealthy each person is compared to the other two people.



If you’re just averaging the money in each measurement, Aaron is the wealthiest person and would rank number one. But if you average the rankings, Brett’s average ranking (the average of 1, 1, and 2) is better than Aaron’s average ranking (the average of 1, 2, and 2).

Detroit is like Aaron. It has the worst segregation measures, but not the worst average ranking. Milwaukee is like Brett. We do not have the worst segregation measures, but we do have the worst average ranking.

When the segregation measures are standardized and averaged, Detroit comes out as the most segregated metro area in the country. Milwaukee comes out at number two. Here are the top five segregated metro areas using this way to measure:

The U.S Census Bureau may have had a good reason for going with their method. And, none of this changes the fact that Milwaukee is highly segregated, and that this remains a central challenge to our future. There’s little excitement in knowing that Milwaukee is “second only to Detroit” in yet another measure of socioeconomic health. At the same time, the stigma of being the most segregated place in the country is a damaging one. As it turns out, it’s not necessarily legitimate.

Sometime next year, the 2010 Census should be completed and we will be able to see how Milwaukee stacks up in segregation and many other areas. In the meantime, it is still important to look at the impact that segregation has on our health and our future.

Later this week, I will be publishing a look at the connection between segregation and education. Stay tuned.

***

Race Matters in Milwaukee

Part I: How Segregated is Milwaukee?

Part Ia: How Segregated is Milwaukee? Part 2

Part II: The Causes of Milwaukee’s Segregation

Part III: The Effects of Milwaukee’s Segregation

Part IV: Segregation and Education

About these ads

3 responses to “Race Matters in Milwaukee: How Segregated is Milwaukee? Part 2

  1. Pingback: If Chicago is #1 in the Midwest... - Page 9 - City-Data Forum

  2. I just followed the link from the Urbanophile. There’s a valid reason to use the ranking average rather than the metric average: it is less sensitive to outliers. In your example of Aaron vs. Brett, it could well be that wealth measure #1 is just atypical. Ranking average is more like a median, whereas metric average is more like a mean.

    Also, on a side note, I wouldn’t believe anything coming out of the 2010 census, not after how it dropped the ball on counting inner cities. The same methodological problems that led it to undercount New York, Chicago, and Atlanta are going to lead it to undercount urban blacks and Hispanics, and this is going to severely skew segregation measures. (And, of course, this undercount is completely intentional. The census bureau wanted to use sampling, but the GOP shitted bricks and demanded a methodology that would undercount people who vote for the wrong party.)

    • Thanks Alon. I wasn’t able to find an explanation in the Census report for why ranked was averaged rather than the metric. Given the headlines that result from these sorts of rankings, an explanation probably should have been included.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s