Giving Compass' Take:

• The Marshall Project discusses the difficulties in filling in the gaps of crime measurement due to flaws in past data — new methodologies may help paint a more accurate picture.

• It's important for criminal justice advocates and policymakers to have reliable measurements. How can we support efforts to improve our survey-taking abilities?

• Here's how grassroots organizations are leading the way on criminal justice reform.

The government tries to measure crime in many ways, including the latest survey from the Bureau of Justice Statistics released recently. The National Crime Victimization Survey is meant to estimate how many people in the U.S. are victims of crime each year.

Like many crime studies, the survey is imperfect. But this year’s edition had more inconsistencies than most. In an attempt to provide a better picture of crime at the local level, the bureau changed its methodology. What it found when it surveyed new parts of the country, however, was that crime is even more difficult to measure than one might imagine.

The victims' survey is an important tool for researchers because a significant percentage of crimes are never reported to police and so are not counted in other measures like the FBI’s Uniform Crime Reporting program. By asking nearly 225,000 people if they’ve been victims of crime and whether they reported it to the police, the survey tries to fill in the gaps of other measures that can undercount crimes in communities where vulnerable populations do not go to the police for help as often.

The new victims survey finds that there were 5.7 million violent crimes in the U.S. last year — at a rate of 21.1 crimes per 1,000 people age 12 and older.

Read the full article about crime measurement by Yolanda Martinez at The Marshall Project.