2017 National Crime Victimization Survey Statistics: Did Violent Crime Go Up?

Hours before the government shutdown on December 21, the Bureau of Justice Statistics published the 2017 National Crime Victimization Survey (NCVS) estimates of crime.

Earlier this year, the 2017 FBI crime statistics showed a very very slight decrease in violent crime (homicide, rape, robbery, and aggravated assault), from 3.9 violent crimes per 1,000 persons in 2016 to 3.8 violent crimes per 1,000 persons in 2017.

But the FBI statistics are based only on crimes known to the police. The NCVS statistics come from more than 140,000 interviews with US households; household residents age 12 and older are asked about their experiences with crime. Unlike the FBI crime statistics, the NCVS statistics include crimes that are not reported to the police as well as those that are reported.

If the amount of crime stayed the same but more people reported crimes to the police in 2017 than in 2016, we would expect the FBI statistics to go up but the NCVS statistics to stay the same. Let’s look at serious violent crime in the NCVS. This includes rape and sexual assault, robbery, and aggravated assault — roughly the same set of crimes measured by the FBI, but excluding homicide and including sexual assault other than rape.

The NCVS report says that there were 7.3 serious violent crime victimizations per 1,000 persons age 12 and older in 2017. This estimated victimization rate is not statistically significantly different from the rates in 2015 (6.8 per 1,000) and 2016 (6.6 per 1,000). All of these statistics are, of course, higher than the FBI rates because in a typical year, only about half of serious violent crimes are reported to the police.

But be careful with the NCVS statistics from 2016! In October 2018, the Bureau of Justice Statistics issued revised numbers for 2016 that were lower than the estimates previously released in December 2017. The original report stated that there were 7 serious violent crimes per 1,000 persons age 12+ in 2016; the revised report decreased that rate to 6.6 serious violent crimes per 1,000 persons.

In my view, the original 2016 estimates are probably more accurate than the revised estimates. I’ll have more on this in a later post.

Whichever set of 2016 estimates you use, the main story from the NCVS statistics is that the level of serious violent crime was about the same in 2017 as in 2016.

Copyright (c) 2018 Sharon L. Lohr