Monday, December 17, 2007

Factchecking the WP's "Factchecker"

Those who make the rounds of the econ-blogs might know that the Washington Post's "Factchecker" Michael Dobbs has some issues in understanding statistics. His post on Mexico's growth rates ("checking" grossly wrong figures from a Post editorial) was the kind of thing that makes economists wince. It's too much to go into here, but here's Dean Baker's overly polite critique.

It was with this background that when I saw Dobbs' writeup of "Obama's Most Revealing Fib" I decided I should check his figures.

Here's the quote which Dobbs claims is a lie:

"I don't want to wake up four years from now and discover that we still have more young black men in prison than in college."

Dobbs seems to agree that this 2002 Justice Policy Institute study does show that there are more black men (of all ages) in prison than in college. But he says that the "data in that study have been challenged," linking to a website that has a dead link to a Tech Central Station column.

Next Dobbs cites a bunch of figures which suggest that the number of young black men in college is around three times the number in prison. So what it comes down to is basically Dobbs' assertion that Sen. Obama was wrong to include the word "young" in his statement. This seems like thin gruel for an accusation lying, but let's take a look at this.

Clearly, the counts will depend on how you define "young men." I know from my past work that there is no universal definition of "young adult" or "young men," and that the range of ages included goes from 15 up to 44. In the crime and public health literature (see for example many studies by the World Health Organization), you will commonly see "young adult" defined as 15-34.

If we take this as our definition of young men, there were 710,000 young black men in college, according to this spreadsheet tabulation from the 2005 Current Population Survey (710K is the sum of cells B20, B30, and B40). Dobbs cites the same figure, so no dispute there. Dobbs implies that this figure is for the age bracket 18-34, but it's actually 15-34. If you add in people over 35, you find that a total of 864,000 black men were in college.

How many are in prison? Dobbs cites a 2005 report with federal and state prisoner populations and then throws in a separate figure for local jails from 2006, which I couldn't track down. In any case, it's much easier to take this 2006 report from the U.S. Department of Justice which gives you the combined federal, state, and local populations by race and age group. This report only includes prisoners 18 and older, however. For the population of those in juvenile facilities, you have to go the Census of Juveniles in Residential Placement. The racial breakdown for juveniles with the most recent data (2003) is here.

Combining these two data sources, you get the following:

Age Black male prisoners
15-18 30,311 (juveniles)
18-19 33,000
20-24 160,000
25-29 156,200
30-34 132,400
35-39 120,500
40-44 103,000
45-54 101,000
55+ 79,000
___ ______
Total 867,000 (rounded to nearest thousand)

This confirms that there are, just barely, more black men of all ages incarcerated (867,000) than black men in college (864,000). Dobbs suggest this is not a valid comparison, because the age profiles of the two groups are different. However, the two age profiles are not so different (the bulk of prisoners are in their 20s, and there are many older black men in college), and it's a shocking statement any way any slice it.

If you add the figures up for the preferred 15-34 age group, you see that there are 512,000 incarcerated young black men in as compared to 710,000 in college. So it is the case that the number in college does nudge out the number in prison. But the figures are of the same order of magnitude, rather than differing by a factor of three, as Dobbs' writeup suggests.

All in all, given that the numbers are pretty close, it seems extreme to label this a "hoary myth" or a "fib." No one would argue that Obama is fundamentally misrepresenting what is happening.

Two other points: I think it's probably the case that in the late 80s and early 90s, when Obama was working as a community organizer and law professor in Chicago, the number of young black men in prison was higher than the number in college. (This was after the huge Reagan-era increase in incarceration rates and before the Democrats had enacted measures to expand college access.) This may be a stylized fact Obama picked up back then and has been repeating ever since. Additionally, the statement is undoubtedly true among particular African American populations, e.g. those living in inner-city neighborhoods.

If this is, as Dobbs says, "Obama's Most Revealing Fib," what does it reveal? I'd say it shows that he is dedicated to bringing about change to avoid the tragic circumstances that put a half million young black men in prison, and that his understanding of the challenges is rooted in his own experience trying to effect change in Chicago.

No comments: