Mapping Project Explores Racial Disparity in Juries

Article excerpt

He usually doesn't count heads each time he enters a room, but when Reggie Flowers showed up for jury duty and glanced around at the other 87 people, he couldn't help but notice something.

He was the only black person present.

"I don't understand," he told me later. "It's supposed to be a cross-section of your community."

I, too, noticed it after a few weeks of visiting jury rooms. Though the Census shows 11 percent of Allegheny County's adult population is black, I found the typical criminal court jury room was only 4 percent black.

On the other hand, African-Armerican defendants far outnumber white ones, according to court records. Those defendants frequently have to trust a dozen white people to weigh their actions and decide their fates.

Juries clearly have an important function: They physically demonstrate the founding concept of American democracy that our government is of, by and for the people.

That's what they're supposed to do, anyway. But courtrooms in Pittsburgh, one prominent defense attorney said, look "like South Africa during apartheid."

What's going on?

In Pennsylvania, as in many other states, computers make random lists of potential jurors from voter registration and driver's license databases. Challenges to jury room imbalances in such systems, where there is no overt bias in picking possible jurors, have been struck down repeatedly in the state and federal courts. Rulings follow the U.S. Supreme Court, which has said that while the system may not discriminate, no defendant is guaranteed a racially representative jury or jury pool.

So just counting heads in the jury room wasn't good enough. I had to look at the system. Is it really random? Do blacks and whites have a reasonably equal chance of being called to jury duty?

The answer is no, and that conclusion was the centerpiece of a four-month (Pittsburgh) Tribune-Review investigative project, "A jury of peers?"

Flawed surveys

Pennsylvania's open records law is among the nation's worst, but this time I lucked out and got what I needed. The key public records I relied on were the jury arrays, the computerized daily lists of everyone summoned for jury duty, along with their addresses.

Using mapping software, I plotted the home addresses of the 45,000 people summoned to serve on a county criminal jury in the previous 18 months.

From the 2000 Census, I knew the adult population of each municipality or Pittsburgh neighborhood. With these numbers, I came up with a "jury service rate" for about 150 municipalities and neighborhoods.

After that, the color patterns were clear. Countywide, 44 of every 1,000 adults were called for criminal jury duty over the last year and a half. But in neighborhoods that were at least 98 percent white, an average of 53 jurors were called. In a few white neighborhoods it was 65 or more.

But in majority black neighborhoods, the average was 26. And not one of those neighborhoods had more than 31 jurors called per 1,000 adults.

In other words, people living in overwhelmingly white neighborhoods were twice as likely to be called for jury service, on average, as people living in black neighborhoods.

Call it what you want, but that's not random.

So what's wrong with the system? For one thing, flawed survey methods.

Think of it this way: If you wanted to do a residential telephone poll about the next election, what would happen if your pollsters only made weekday calls? Your respondents would be mostly homemakers and pensioners, and you'd miss most people who are employed.

If you wanted a broader picture, the pollsters would have to call back in the evening to catch the people who work during the day. If they didn't, if they only relied on the people who answered the first time, the survey sample wouldn't represent the public at large, and all the survey conclusions would be suspect. …