A Note on Data: Early 90s Population Estimates and Chicago’s Homicide Rate

This came to my attention this week, and I feel like in the interest of transparency I ought to make a note about it.

Because people don’t read all the way to the end of blog posts, I also feel like I should put the conclusion at the beginning. So:

  • Over the course of the 1990s, the Chicago Police Department used two different population estimates to calculate murder rates. In this post, I used the one that the CPD was using in the years I covered. But there is also a reasonable argument to be made for using the one it chose to use later in the decade.
  • If you use the second estimate, mostly things look the same. In particular, the “skyrocketing inequality of violence” skyrockets just as much, still roughly tripling from the early 90s to the late 2000s.
  • One thing that does change is that some, but not all, of the areas that show a rise in homicide rates with the first population estimate show a modest decline with the second population estimate. This is because the second population estimate suggests that fewer people were living in heavily non-white areas of the South and West sides in the early 90s, and so it gives them a higher homicide rate to begin with. Specifically, of the seven police districts that have rising homicide rates according to the first population estimate, four (the 3rd, 7th, 11th and 15th) show modest declines according to the second, one (the 6th) is roughly flat, and two (the 8th and 22nd) still show clear (in fact, larger) increases.
  • I don’t think it’s clear which estimate is better.
  • Bottom line: it’s somewhat ambiguous exactly how much of Chicago’s South and West sides saw homicide rates rise over the last 20 years – although substantial portions certainly did. It remains crystal clear that the inequality of violence has increased by a factor of about three over those same two decades, and that that gap has extremely serious consequences.

In somewhat more detail:

In the early 1990s, the CPD used population estimates from local planning bodies, either the city itself or the Northeast Illinois Planning Commission, to determine how many people lived in each police district. Those are important numbers for calculating crime rates, since a neighborhood’s safety level is usually described in terms of crimes per number of people – number of homicides per 100,000 residents, in this case. Beginning in 1995, however, with a report that covered 1993-94, the CPD switched to using population estimates provided by the U.S. Census. The Census numbers differed from the city/NIPC numbers mostly in that the Census counted fewer people – in some cases as many as 30% fewer – in the mostly black and Latino sections of Chicago’s South and West sides.

As a result, going back and plugging the Census-derived population numbers into the CPD’s earlier reports makes the South and West sides look more dangerous, because there are the same number of crimes but fewer people. This doesn’t affect the trends in geographic inequality of violence, but it does, as I described above, create a higher baseline against which to compare homicide rates from the late 2000s. In four districts, the baselines are so much higher that what previously looked like increases in homicide rates over 20 years look like modest declines. In another, the increase is wiped out but there is no decline, and in two there are still increases – in fact, the increases get bigger.

Which set of numbers – the city/NIPC’s or the Census’ – are better? I don’t know. On the one hand, the Census is clearly the governmental head-counting of record. On the other, the 1990 Census was the target of a lawsuit from nearly every major city in the country, including Chicago, that alleged severe undercounting (in the hundreds of thousands in Chicago)particularly in high-poverty neighborhoods with few white people – ie, exactly the neighborhoods at issue here. Either way, the scale of the disagreement – again, as much as 30% in some districts – means that at least one of these estimates has to be quite flawed.

In any case, to repeat myself: the bottom line is that there is some ambiguity about exactly how much of Chicago’s South and West sides actually saw increases in homicide rates, although substantial portions undoubtedly did.

And what is still crystal clear is that the inequality of violence has increased by a factor of about three over the last twenty years, and that that gap has extremely serious consequences for the city.


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