As of mid-December, the homicide rate is up in 16 of America’s 20 largest cities. That continues a trend of increasing murder rates which began last year. The Wall Street Journal reports:

Sixteen of the 20 largest police departments reported a year-over-year rise in homicides as of mid-December, a Wall Street Journal survey found. Some notched minor increases, while Chicago saw one of the most dramatic jumps, with more than 720 murders—up 56% from 2015.

Nationally, the murder rate rose in 2015 for the first time in nearly a decade, though it remains well below where it stood during the 1990s…

Nationally, 37 of the 65 largest police agencies, including ones in San Antonio, Las Vegas and Memphis, Tenn., reported year-over-year homicide increases as of Sept. 30, the Major Cities Chiefs Association said. In 2015, 44 police departments reported increases, many for the first time in years.

“I have a lot of concern about that many cities experiencing those increases in violence,” said Darrel Stephens, executive director of the association, which represents chiefs from the largest departments.

The WSJ published this chart showing where the increase is taking place:

wsj-homicide-chart

There is disagreement about why violent crime is on the rise after two decades of decline since the early 1990s. One suggestion is the so-called Ferguson Effect, which suggests that police are holding back on intervening in some situations that could lead to them winding up on YouTube.

Last year FBI Director Comey said the Ferguson Effect might explain the timing and location of some of the increase in crime. He was criticized by the White House and progressive outlets at the time for taking that stance.

Earlier this year one of the researchers who had previously discounted the Ferguson Effect reversed course after taking a close look at the FBI’s data. Criminologist Richard Rosenfeld, whose research had previously been cited as proof the Ferguson Effect wasn’t real, said in May, “The only explanation that gets the timing right is a version of the Ferguson effect.” He added that it was now his “leading hypothesis.”