Last week I read an interesting blog post from Kevin Drum based on two sets of U.S. crime data collected two different ways by separate parts of the Justice Department.
Data derived both from large-scale population surveys and from police reports to the FBI show that property crime was down around 6 percent in 2015 relative to the previous year, in keeping with long-term trends since crime peaked in the early 1990s. (Data for 2016 won't be available for a number of months, so this in the most up-to-date information we have.)
Curiously, however, while survey data showed a drop of over 11 percent in violent crime, police reports indicated an increase of 3 percent. Drum points out that a closer look at the numbers by Rick Nevin (link) suggests that this may reflect a change in how police departments are reporting crime statistics to the FBI rather than an actual increase in criminal violence in 2015 over 2014.
On the other hand, the murder rate did shoot up in 2015, mainly the result of a spike in murders in about a dozen large cities (some other large cities experienced a decline). In percentage terms, the jump in the murder rate from 2014 to 2015 was the largest in 45 years. But the murder rate for 2014 was at a record low, and as Rick Nevin pointed out in the piece mentioned above
The 2015 murder rate increase was the largest percent increase since 1968, but that percent change is relative to the record low in 2014. Larger absolute increases in the murder rate in 1990, 1986, 1980, 1979, 1971, and 1970 look smaller as year-over-year percent changes because the prior year murder rates were much higher than the record low 2014 rate. Even after the 2015 increase, the 2015 murder rate was still lower than it was in 1960, lower than it was in any year from 1964 through 2012, and just about half of the average murder rate from 1969-1995.
Back in the second presidential debate Donald Trump accurately referenced the large percentage increase, but he later started falsely claiming that the murder rate itself was the highest in 45 years, which obviously is not remotely true. This was pointed out by Factcheck.org and The Washington Post's Fact Checker among others. Trump also claimed that the murder rate was up in New York City, which was also incorrect.
Of course, it's likely far more people heard and believed Trump's false assertions about a record-high murder rate than read the fact checks (let alone this blog post!).
(For what it's worth, I wrote a similar item about this back in April based on FBI crime statistics through the first half of 2015, full-year data for 2015 not having been available then. The conclusions were basically the same.)