While reporting on a different story in Louisiana last month, I spent a spare Tuesday afternoon sitting in on criminal court in Lake Charles, in Calcasieu Parish. Calcasieu, in the southwest corner of the state, was once the largest parish in the state, known as Imperial Calcasieu, until three parishes were broken off in 1912, all named for Confederate leaders.

I made my way through security (no laptops, cellphones, recording devices or smartwatches — even “sketching” is technically banned in Louisiana courtrooms) to the courtrooms on the second floor. The scene was oddly chaotic. In the commons area just outside the courtrooms, about 100 people milled about, many looking somewhat confused. They were wearing name tags. Most had gray, white or no hair. It looked like a tour bus had just unloaded a group of vacationing retirees. Few of them talked or acted as if they knew one another.

I’d later learn that they were part of the venire — the larger population of jurors from which the pools for individual trials are picked. I’d also learn that at about the time I was there, a local judge was considering a motion filed by Natasha George, an attorney with the Calcasieu Parish Public Defender’s Office. George was representing a man who was charged with attempted second-degree murder and was about to go to trial. But as jury selection began, George looked at the possible jurors sitting in the jury box and noticed something peculiar.

“They seemed disproportionately white,” she says. “But more than that, they were overwhelmingly elderly.” Then she thought back to her last trial. “I thought — you know what? — it was the same thing there. Nearly everyone in that jury pool was also over 50.”

George made an oral motion to suspend the trial so she could review the jury questionnaires. The judge — himself the former head of the parish public defender’s office — granted her motion. What she found more than confirmed her suspicions. Of the 50 people picked for the jury pool in her case, 39, or 78 percent, were over the age of 50. There wasn’t a single potential juror under 33.

As in most states, Louisiana parishes draw their population of potential jurors from public databases such as voter rolls, driver’s license records or — less often — public utilities records. In a parish the size of Calcasieu, the initial list of possible jurors will generally have 300 to 400 names, depending on the number of trials on the docket. Depending on how many show up and how many are dismissed for hardship, illness or other reasons, the list is generally pared down to a pool of 200 to 300. From this list, commonly called a venire, about 50 are randomly selected for each trial (death penalty trials usually require a larger pool). This is commonly referred to as the trial’s jury pool. (Attorneys tend to use the terms pool and venire interchangeably, which can add to the confusion.)

George and her colleague Carla Edmondson then obtained questionnaires for the larger venire from which the pool for her trial was selected. The demographics of that larger group was just as skewed. According to the 2010 Census, 17.3 percent of Calcasieu Parish residents are between 18 and 29. Yet of the 318 potential jurors, just five, or 1.6 percent, were under 30. Moreover, 76 percent of the names were over 40, and 60 percent were over 50. Just 45 percent of Calcasieu Parish is over 40, and about 34 percent is over 50.

The venire was also disproportionately white and male. While Calcasieu is about 70 percent white, whites made up 84 percent of the venire. White males in particular made up half the venire by themselves, even though males of all races are less than half of Calcasieu’s whole population.

George and Edmondson then obtained the next two upcoming jury venires, one in July and one in August. Again, the numbers were heavily skewed toward older jurors. In the July venire, more than 60 percent of potential jurors are over 50 (again, vs. 33 percent of the parish), and nearly 80 percent are over 40 (again, compared to just under half for the parish). For August, 70 percent are over 50, and 86 percent are over 40.

If we look at median ages, the discrepancy gets starker. The median age of the June jury venire was 57. For July it was 55 and for August 56.5. Since the minimum age to serve on a jury is 18, we need to look at the median age in the parish among people who are at least 18. The most recent data we have here is the 2010 Census, which puts the median age among those over 18 at roughly 46 — far younger than the median age of the three venires.

I asked statistician Aron Galonsky how likely it would be for a randomly selected jury pool of 200 people from a parish the size of Calcasieu to have a median age approximately 10 years above that of the population from which it was selected. He calculated such a result would be approximately 160 standard deviations above the mean. That means the odds of getting such a result randomly are basically zero, yet Calcasieu Parish did it three times in a row.

As it turns out, Calcasieu draws its population of prospective jurors solely from voter rolls. This isn’t uncommon, but most jurisdictions supplement voter lists with driver’s license and state ID databases, utility records or, a generation ago, with phone books. Active voters tend to be older, richer and whiter, so it isn’t surprising that the parish’s jury lists would overrepresent those groups. Calcasieu Parish also only updates its population of possible jurors every three years, so voters aged 18, 19 and 20 are periodically left out entirely, as are voters of any age who have only recently registered. (Attorneys I spoke to said Calcasieu isn’t the only parish to take this approach.) And when a jurisdiction relies solely on voter rolls, this also means that every time a state passes new laws aimed at making it more difficult for young people or minority groups to vote, those groups will be less represented on juries.

Young people are more likely to move around, which means a summons for jury duty is less likely to reach them. Older people also tend to be more financially secure, or retired, and thus more willing and able to miss work for modest jury duty pay.

But even after taking all of those factors into consideration, the numbers in Calcasieu look lopsided. The three venires that George obtained contain 763 people from which prosecutors and defense attorneys could pick jurors over a three-month period. Just one — or possibly two — were under the age of 29. (One 23-year-old juror appears in the data set for one venire, but George hasn’t been able to find a questionnaire for that juror.)

Assuming there were two, I asked economist Daniel Bier to estimate the odds of getting just two people under 29 in a randomly selected group of 700 people from the population of Calcasieu Parish. To arrive at that number requires considering a fair number of variables, but all agreed that the odds are infinitesimally small. He calculated that, expressed as a percentage, the odds would be something like 2.2e-88 — which is a decimal point, followed by 87 zeros, followed by 2.2. It would be like flipping a coin hundreds of times and getting heads every time.

This isn’t the first time these issues have come up in Louisiana. Defense attorneys I spoke with last month said some parishes have been caught failing to delete duplicates when they dump the latest voter rolls or DMV records into their juror databases, meaning that people who have resided in the parish and at the same residence for the longest time have a far greater likelihood of being called for jury duty — a bias that again would favor older, whiter and wealthier residents.

Last year, the Advocate reviewed six years of jury venires for nine of Louisiana’s 10 busiest courthouses. The newspaper found that in nearly all of them, blacks were far underrepresented in jury pools. (Interestingly, Calcascieu was the one parish where black representation in jury venires was about the same as their percentage in the general population.)

But a problem in East Baton Rouge Parish exposed in April seems most similar to what’s happening in Lake Charles. As jury selection began for accused cop killer Grover Cannon, his attorneys discovered that because of a “computer glitch,” the jury rolls had been not been updated in eight years. No one in the parish between the ages of 18 and 25 — and most of the people 26 years old — had ever been chosen for jury duty, nor had anyone who moved to the parish after 2011. The Louisiana Supreme Court ruled that the glitch amounted to a violation of due process and halted Cannon’s trial. The parish has since suspended all of its criminal trials while it tries to sort out the jury issue.

So does age really bias a jury? It certainly seems to make a jury less skeptical of the criminal justice system. Older people are more likely to support the death penalty, more likely to support longer prison sentences, and more likely to express confidence in the police. In a 2012 study for the National Bureau of Economic Research, three economists looked at the age of jurors in more than 700 felony trials over 10 years in two Florida counties. They found conviction rates in those counties increased by one percent for every one-year increase in the average age of the jury. Randi Hjalmarsson, one of the authors of the 2012 study, notes that juries that include, say, blacks or Latinos tend to treat black and Latino defendants differently from all-white juries, and says she would worry that something similar may be occurring with respect to age. “I would especially be concerned whether young defendants (who are not represented on these juries) are treated differently than older defendants,” she wrote in response to an email query.

In George’s case, the trial judge ruled the jury pool invalid. The Calcasieu Parish district attorney’s office appealed that ruling to the Louisiana Court of Appeal, Third Circuit. In their brief for that court, George and Edmondson noted the severe underrepresentation of people under 30, but they couldn’t argue that everyone under 29 had been systematically excluded from jury duty because even though the odds of so few younger people being selected at random are close to zero, one or two people under 29 were still somehow chosen.

Instead, they argued that because the parish only updates its population of possible jurors every three years, people between 18 and 20 have been systematically excluded. They also asked for more time to obtain more data, to perform a statistical analysis and to try to figure out why so few people under 30 were being chosen more generally.

That appeals court ruled against them, 2 to 1. The majority first decided that residents between 18 to 21 do not comprise a “cognizable group” for the purpose of jury representation, so excluding them from the population of possible jurors isn’t a violation of due process. The court also ruled that even if that age group were a cognizable group, the jury selection process also excluded people of any age who had registered to vote in the past three years. Finally, the court ruled that George’s case is distinguishable from the Louisiana Supreme Court’s ruling on the East Baton Rouge problem because the latter involved a software glitch that had permanently excluded everyone born after a given date. Here, the exclusion of new voters is only temporary.

George says she thinks a glitch similar to the one in East Baton Rouge is behind the numbers in Calcasieu. “We just need more time to look into it,” she says. The math certainly suggests that something is skewing the jury venires. The numbers are too stark to be attributable to randomness and voter registration patterns alone. That would seem to suggest either a bug in the software, an oversight in how the jury lists are populated, or even something more sinister — someone is intentionally rigging the system. George says she doubts it’s the latter.

But it’s notable that neither the third circuit appeals court nor Calcasieu Parish District Attorney John DeRosier seem particularly interested in getting to the bottom of what’s going on. The disinterest from DeRosier’s office probably shouldn’t be surprising. DeRosier has hired prosecutors who have been fired from other parishes for misconduct. He was one of just a few officials in the state who opposed efforts to repeal the state’s Jim Crow-era law allowing convictions from non-unanimous juries. (The repeal measure passed overwhelmingly in the 2018 election.) He was also one of several district attorneys accused of participating in a scheme to redirect funding from the state’s public defenders to the coffers of prosecutors. Louisiana’s prosecutors already have enormous advantages in the courtroom, which is a big reason the state trails only Oklahoma in its rate of incarceration.

Whatever is happening in Calcasieu Parish, it seems clear that anyone who goes to trial is being denied the right to a jury representing the local community, instead of representing just the portion of the community most inclined to hand down a conviction. Redoing all of those trials would be a costly endeavor. But if the system wasn’t producing a representative jury pool, then defendants haven’t been given a fair crack at justice. “Halting jury trials for a short time in order to figure out why an entire generation is absent from the jury pool is a small price to pay,” George says. “Especially considering the incredibly high stakes of a felony prosecution.”

After the appellate court decision, the trial judge granted George a continuance and postponed the trial until Aug. 26. George and her colleagues will now collect more information. They recently obtained jury questionnaires going back to 2015, and are now collecting data from those forms to see just how long Calcasieu Parish juries have been skewed. As in East Baton Rouge Parish, if they can show that some flaw in the jury selection process has been making people under 30 systematically less likely to be chosen, it could cast doubt on every guilty verdict in the parish since the selection system was put in place.

Read more: