“I was the first person who asked him why he killed someone,” recalled Amy Brittain, an investigative reporter for The Washington Post. As part of a Thursday session on prison and jails, Brittain discussed one of the subjects of her investigation into an old Washington, D.C. law that allowed violent offenders to remain on the streets.

The speakers covered policy shifts in California, the lack of transparency and data in the criminal justice system, and the special interests influencing the system from behind the scenes. Here are a few takeaways from the panel:

1. The data can lie — if it exists at all.

It’s hard to get data from the criminal justice system, often because it just isn’t being collected, said Paige St. John, a writer for the Los Angeles Times who investigates the justice system in California. For instance, the state of California once hired the Rand Corporation to assess the effects of a number of criminal justice reforms in the state. The research institution returned with this: We can’t. The data doesn’t exist.

“That was astonishing to me,” St. John said.

Comparisons can also be difficult. For example, it was impossible to compare California’s post-reform crime rate to other states because no one else had implemented similar initiatives. To determine if some of the reforms, which led to the release or diversion of 18,000 inmates, had caused a spike in crime in California, Magnus Lofstrom, a senior research fellow at the Public Policy Initiative of California, looked at other states with similar crime patterns to see if they had a comparable spike. What they found was that there had been some increases in the crime rate that could be attributed to the reforms — but they were isolated to auto thefts.

2. Anecdotal evidence might be all you have — and it can be misleading.

Brittain had been collecting the names of men who had been convicted for assault during a spike in violence on D.C. public transit. When she looked up the names in the D.C. court system, she came across something she’d never heard of, tied to a 1980s law called the Youth Rehabilitation Act. She found that offenders, some of them as old as 21, were being diverted from prison and not receiving substantive rehabilitation services, thanks to a law considered by many in D.C. to be a joke. No one was collecting data on the program, so she had to rely on the interviews with people who had been touched by it. That went a huge way towards understanding how the law was failing offenders and victims alike.

St. John cautioned, however, that it can be easy to repeat advocacy rhetoric, especially when it shows up in congressional hearings. But when data is lacking, you can follow people through the system to see where it falls short — St. John called this her “radioactive tracer.”

3. Ballot initiative are sometimes much more than they seem.

While St. John was reporting on the California ballot measure Proposition 47, also known as the Reduced Penalties for Some Crimes Initiative, the Los Angeles Times was funding her position from a Ford Foundation grant. She later learned that the same foundation was funding research to fuel Prop 47, as well as a number of other avenues pushing for its acceptance. The foundation was funding St. John’s position as well as a number of advocacy efforts and research initiatives on criminal justice reform, not all of which were apparent to the public or to St. John herself. St. John, who is no longer funded by the Ford Foundation, followed the money and went on to publish a story on the massive web of foundations pouring money into the proposition.

Emily Hopkins in the 2017 Google News Lab Fellow for IRE & NICAR.