New Hampshire courts should cull jurors solely from lists provided by the state Division of Motor Vehicles to ensure minorities and young people have a fair shot at serving on juries, a demographer testified yesterday at a pretrial hearing in the Michael Addison case. The current practice of picking jurors partly from DMV lists and partly from voter registration lists isn't representative, he said.

Addison's lawyers are asking the court to dismiss his capital murder charge and overhaul the state's jury system. They say the system routinely leaves minorities and young people out of the jury pool, and Addison is entitled to a jury chosen from a "fair cross-section of the community."

Addison, 28, is charged with capital murder in connection with the shooting death of a Manchester police officer. Jury composition is especially important in his case for two reasons, his lawyers say.

First, Addison could face the death penalty is he's convicted of killing Manchester Officer Michael Briggs. Second, Addison is black and Briggs was white. Studies show the presence of black male jurors in a case such as Addison's reduces the likelihood of a death sentence, Addison's lawyers say.

Dr. Andrew Beveridge, a sociology professor at Queens College in New York who specializes in demography, testified yesterday about a study he did at the request of Addison's lawyers. It found that minorities and young people were underrepresented among jurors called last year in Hillsborough County Superior Court in Manchester, where Addison is scheduled to be tried in September.

Beveridge said the state's practice of using two separate lists to form the jury pool is partly to blame. A person who has a driver's license and is registered to vote is twice as likely to be called for jury duty, he said, which makes it less likely that someone who appears on only one of the lists will be picked.

And those who appear on only one list are probably young people and minorities, Beveridge said. Studies show they're less likely to register to vote, which means they'd only show up on the DMV list.

Beveridge had several suggestions on how to improve the state's system. One suggestion was to solely use the DMV list because it seems more representative. Another was to blend the DMV and voter lists but purge them of duplicate names and the names of people who recently died or moved.

"I'm not saying it would cure the system," he said, "but it would be a step in the right direction."

Beveridge said he can't pinpoint the source of the under-representation, because he hasn't had access to the master DMV and voter registration lists from which potential jurors are chosen.

State prosecutors questioned Beveridge's methods and the extent of his research. Since the DMV and voter registration lists don't contain information about gender or race, Beveridge used a method called "geocoding" to guess the ethnicity of potential jurors. He compared that data with Census data on the number of minorities in the county to see if they were represented fairly in the jury pool.

Senior Assistant Attorney General Will Delker questioned the accuracy of geocoding, which guesses a person's race based on where they live. He also focused on the numbers in Beveridge's report. Delker asked him to reiterate his finding that even if the jury pool was equally proportional to the makeup of the population in terms of race, only 11 more minority jurors would be included.

The hearing is scheduled to continue today. Dr. Stephan Michelson, a researcher who specializes in jury composition, will testify about a report he prepared at the request of state prosecutors. Prosecutors argue that the state Supreme Court has already ruled that the jury system is fair.