About . . . . . . Classes . . . . . . Books . . . . . . Vita . . . . . . . Links. . . . . . Blog

by Peter Moskos

September 23, 2016

Brennan Center: No need for "most Americans" to worry about more murders

The good people at the Brennan Center for Justice at NYU School of Law have assured us (pdf link to report):
Reports of a national crime wave were premature and unfounded, and that "the average person in a large urban area is safer walking on the street today than he or she would have been at almost any time in the past 30 years."
The authors conclude there is no evidence of a national murder wave.... Most Americans will continue to experience low rates of crime.... There is not a nationwide crime wave, or rising violence across American cities.
Ah, yes.

Cause for a moment there, I was kinda worried that more people were getting killed. But it turns out, I guess, that it was illiberal of me to care about people who are particularly at risk of being killed. Also, did you know:
Homicides are concentrated in the most segregated and poorest areas of the city.
I hadn't thought of that. And since that's not where the "average person" lives, I guess we don't need to worry.

Maybe I should just jump on this illogic ideological bandwagon of denial to see where it goes:
By "historic standards," racism is pretty low in America. QED: Not a problem.

#BlackLivesMatter can close up shop because "most Americans" don't have to worry about being shot by police.

Enough with all those new letters, the "average American" doesn't face any LGBTQ discrimination.
Check, check, and check. Problems solved!

Oh, but while you're here. Not that it's any cause for concern. But there is this little issue...:
The murder rate is projected to rise 13.1 percent this year.... [and] 31.5 percent from 2014 to 2016.
Say what?!

[Update: 2015 stats are out. The rate, based on recorded homicide, increased from 4.26 to 4.75 per 100,000. The rate, based on estimated homicides, increased from 4.44 to 4.90. Recorded 13,594 homicides in 2014 (estimated 14,164). 15,192 in 2015 (estimated 15,696). 2014 estimated population 318,857,056. 2015: 320,090,857.]

If these numbers are correct -- and they may not be (there is some odd math in this report*; and keep in mind 2015's national UCR stats haven't yet come out) -- but if these Brennen Centers estimates are correct, that would mean 2015 saw a 16.3 percent increase in the homicide rate.

So all we've got is just your average 16.3 percent annual increase in murder. I mean, we had one of those, well, uh, actually, never. This would be the largest increase since the government has been keeping track. (An increase in 1921 might have been greater, but we don't really know.) The last time the UCR recorded a 31.5 percent increase in two years was, oh, never.

[In raw numbers the homicide increase is the greatest in 25 years. But it's standard industry practice to use rates and percentages.]

[Update: I've been informed over in the twitter world that when they say "nationwide" they don't mean "nationwide" but "in the top cities." I would expect the national increase to be less than what is found in the top cities. But I don't know. Anyway... the 2015 UCR data will be out this week. And then, at least when it comes to last year, we can all stop speculating and know how big the increase in homicide was.]

As to their overall point that homicide may be up but "crime" is little changed? I just call bullshit. Not on their analysis, per se. It's just that crime numbers are not as reliable as homicide numbers. Trust homicide. Crime numbers are heavily influenced A) by proactive police and arrests (which are both down) and B) non-reporting (probably up). I trust the strength of the correlation between homicide and other violent crimes more than I trust the data on other violent crimes. If homicide is up, violent crime is up. Trust me on that one.

*They've got some weird math here I can't figure out:
The national murder rate is projected to increase by 13.1 percent. Nearly half of the increase (234 out of 496 new homicides) will occur in Chicago. (page 1)
But if the national rate goes up 13 percent this year, we'd see something closer to 1,500 more homicides. (Based on 2014 rate of 4.5 and 13,472 homicides.) And Chicago's numbers will be up by about 200 this year. This is closer to 15 percent. What gives?
Baltimore, Chicago, and Houston are projected to account for 50 percent (517 of 1041) of new homicides between 2014 and 2016. (page 8)
But if the murder rate is up 30 percent, we'll have closer to 2,500 new murders. I do not understand.

Also, these semi-annual "crime isn't up" reports from the Brennan Center have this odd habit of saying, "if we remove the cities where the increase is the greatest, the increase really isn't so great. (An odd statistical proposition, to say the least.) But let's play along and "pull a Brennan." Let's remove Chicago, Houston, and Baltimore because (I think) in terms of raw numbers, those cities have the greatest increase in homicides, 2014 - 2016 (roughly 240, 165 and 115, respectively). After we "pull a Brennan" we lose about 520 murders. That's a lot, but we'd still have close to 15,500 homicides this year, which would be a 2-year increase of 15 percent. And even that should be cause for alarm.


Peter said...

So what could explain such an unprecedented rise in the homicide rate? Could it be evidence of the Ferguson effect? Really enjoy this blog btw.

Bill Harshaw said...

Okay, say we have a 15 percent increase in homicides, meaning also a 15 percent increase in violent crime. So what...what do we do about it? Do we dust off Clinton's 100,000 police initiate from the 1990's? Do we react state by state or city by city? Do we legalize pot nationally and hope it diverts the market away from hard drugs? Do we mandate nationwide stop and frisk? Do we reverse the trend towards emptying prisons? More three strikes and you're in for life? Do we increase police pay by x percent across the board? Require all local forces in cities over 50,000 to have their trainers trained by the FBI?

Does anyone have a set of measures proven to reduce violent crime?

I'm too old to get alarmed, unless someone can point me to what our politicians should do.

Peter Moskos said...

Well the first thing we do is stop digging. For starters we need to defend police officers who shoot armed suspects who are an imminent threat. (And yes, punish those officers who commit crimes.) We need to fight the narrative of those who really are anti-police, those who believe that cops are malicious and the greatest problem facing America.

We're moving back to de-policing. That's wrong. People in high-crime areas, blacks in particular, want *more* police (better police, too). We need to heed their desires. We need to incentivise good policing, and we can't if the only way we judge officers is the ability to not make the evening news.

And once we can get back to seeing police as part of the solution, we figure out strategies and use tactics that work. Of course that is easier said than done, but we were doing it. In bits and piece and fits and starts, we were on the right track. Violent crime had been going down for 25 years! This wasn't magic or ordained by God. It was, in part, because of good policing. And crime was going down even in the most stubborn cities, like Chicago and Baltimore. It was going down until we told police that they would be judged only by their ability to do no wrong.

We analyze problems at a local level and act appropriately, be that hot spots policing, more use of foot patrol, and shifting the drug trade off public streets. We need police to can address the quality-of-life issues that residents care about. We should encourage cops to make discretionary stops based on reasonable suspicion (without overly onerous paperwork).

There is no silver bullet to crime reduction. Good policing is a lot of hard work. But it was working. We need to avoid a reactive system where cops become nothing more than report writers, sitting in their car till somebody reports a crime. The worst thing we could do right now is throw our hands in the air and tolerate thousands of more murders (and multiple times as many other violent crimes) by pretending we have no clue.

In cities like Chicago and Baltimore, the rise in crime is so obviously linked to changing in policing, particularly the end of discretionary order-maintaining policing that has defined good police work for 187 years.

Thos Wallace said...

In Chicago, Eddie Johnson speaks regularly about the SSL or strategic subject list. This is a list of 1,400 people who have a numeric score based on prior criminal record, gang membership, interactions with police, etc. Johnson says that the majority (if not 3/4) of homicides are people on this list. That is victims. Also, most suspects are on the list.
This is another way of saying roughly the same thing. The average person isn't facing a lot of risk of homicide in Chicago.

Johnson is also a believer in strict and lengthy prison sentences -- especially for gun violations and other crimes committed by these 'known' criminals.

At the same time, there is movement to reduce incarceration rates in the US.

Do you think it is reasonable to have policy that targets individuals on the SSL for strict enforcement of laws and more lengthy sentences? After the fact, when someone is apprehended and charged with murder, it seems obvious that they could have been in prison for earlier crimes if they had been more aggressively prosecuted, charged, and sentenced.

No one seems to want to go to a strict 'operation exile' approach where all illegal gun charges are brought in Federal Court and receive a minimum sentence of 5 years (or more depending on circumstances).

My personal thinking is that it sounds good in theory. But perhaps the SSL will just fill up with replacements. Or maybe Johnson is overstating the degree of concentration of likely criminals.

Here is the guy who was involved in Dwyane Wade's Cousin:

Derren Sorells was just released from prison two weeks before Friday's shooting. He was convicted in 2012 for a motor vehicle theft and escape from custody, but was released early. He had six felony arrests in his background. His brother Darwin Sorells was also on parole for a felony gun charge. He was sentenced to six years in prison in January 2013 and was released from prison earlier this year.


According to police, Derren Sorrells was on an ankle bracelet when Aldridge was killed, but it was “inactive,” said Fox News.
Fox News said, according to police, that “Derren’s ankle monitor was inactive from 8 a.m. to 4 p.m., ostensibly so he could look for work.”

The last time Chicago tried to target gang leaders and break up gangs, it was later a matter of speculation that homicides were up because gangs had fragmented and there was no gang discipline and teenage, neighborhood, 'Beefs' ... frequently on social media ... was the new 'cause' of increased homicides.

So, it is hard to have much faith in a silver bullet. But predictive modeling of likely perpetrators and suspects seems promising.

EA5 said...

But in Chicago and Baltimore, its also obvious that the police departments need to be overhauled. Even ignoring the sections on stop and frisk and the disputed anecdotes, the DOJ report on Baltimore paints a picture of a highly dysfunctional department, where policies are poorly communicated, training facilities are decrepit, and accountability processes are deficient.

And its pretty clear that the Chicago PD has been rotten to the core for a while now. The question is, how do you retrain an entire department, implement new processes for accountability, identify and remove the worst officers (and probably most of the supervisors), all while policing a very high crime city and maintaining officer morale in the face of massive resistance from the officers?

But it is clear that the DOJ and police reform advocates don't believe its possible for a stop and frisk policy targeted at only one race, especially if it exists in both high and low crime neighborhoods, to be consistent with constitutional policing. If stop and frisk actually is necessary and should be continued after departments are overhauled, then there needs to be a discussion of how it works within a reformed system. But maintaining stop and frisk can't be used as an excuse to not address the major problems that do exist within many major police departments or else it won't survive.

Peter Moskos said...

The DOJ report was more political excuse than accurate description. I wrote a bit about it.


Corners and stoops do need to be cleared. And it can be done legally and constitutionally. The DOJ disagrees. And they are getting their way. It's not working out too well. I'm not certain what they want residents to do when drug dealers are camped outside their front door. Or maybe they simply do not care.

The DOJ, I believe, wants to end proactive and quality of life policing. That comes from an ideological belief that goes back to the Kerner Commission, namely police don't play a major role in crime prevention. This philosophy is dangerously wrong.

I can't speak about Chicago (though it does appear from afar that Chicago has more serious problems. God knows the BPD has its problem, but no, I don't think it needed a complete overhaul. Before Freddie Gray, arrests were down, crime was down, complaints were down. Things were getting better! That all changed.

Baltimore needs to overhaul its *political* leadership, particularly at the level of mayor and state's attorney. The DOJ report was a way for these incompetent elected officials to say, "not our problem." But it *is* their problem. And passing the buck isn't the solution. The problems start at the top and filters down. Even if one does believe the DOJ's report (and some of it was true), they offer no solutions. Indeed, many problems are systemic. So overhaul will lead back to where we are. What needs to change is the political system, the one that scapegoats police officers for their failures.

Despite the DOJ's odd fixation on long-gone Mayor O'Malley, the problems of today are not rooted in "the early 2000's." If too many arrests were the problem, the city would have rioted in 2003, not in 2016.

So yes, if I could, I would simply turn back the clock 3 years and keep enjoying gradual but steady progress. And I'd buy better police transport vans.

Peter Moskos said...

Thos, I'm strongly against more incarceration. I am for more consistent incarceration. We need to lock up people for much shorter amounts of time. Too many people get probation before judgement, basically no punishment. Other get years or decades when months or years would do a better job.

My main problem with Chicago's "predictive modeling" is that I can't help but notice it doesn's seem to be working. Cops already know who the problem groups are. You shouldn't need fancy modeling to figure that out (though it may help). What do you do with those lists of names after you have them? Some non-police intervention is certainly in order. But we need to tell police what they should be doing with the gang on the corner, holding their turf, armed and ready to die. I hear no answers. And pulling back has caused more death.

EA5 said...

Yeah, I've read everything you've written about the BPD DOJ report. I guess I don't see it as a way of shifting blame from the politicians but a description of all the things the politicians need to fix. I do generally think about it as a problem with a government agency though, so by definition it implicates the politicians and the voters. They do talk a lot about O'Malley but they also have sections on staffing and training issues, which are clearly recent political decisions. The sections on accountability and investigating misconduct though are pretty damning if they do accurately describe misconduct investigations, and that seems to be more of an issue that exists mostly within the department. And the section on sexual assault investigations is just stomach churning. But given the overall levels of untested rape kits in Maryland, that also reflects statewide budget priorities.

Thos Wallace said...

This is more than a lot of people may want to know, and I'm sure there is nothing in this that will surprise Peter. However, Chicago is constantly used as an example. This is some general information that I have come across trying to understand more about the city.

For people not familiar with Chicago, in spite of the headline homicide numbers, it fits the Brennen narrative ... "Most Americans will continue to experience low rates of crime"
The city is divided into 77 'Community Areas'. https://www.cityofchicago.org/content/dam/city/depts/doit/general/GIS/Chicago_Maps/Community_Areas/Community_Areas_W_Numbers.pdf

In the 1920's, the University of Chicago's Social Science Research Committee set up 'community areas' which, at the time corresponded to neighborhoods. These have not changed except for adding 2 areas associated with O'hare Airport. It is a good, 'clean' way to compare statistics over time, unlike neighborhoods or other commonly used divisions which evolve and change. However, they don't correspond exactly to current neighborhoods or police districts or wards or anything else. They are still largely recognizable as neighborhoods.

Out of these Community Areas, two of them account for between 20% and 25% of the city's homicides this year. Austin on the West Side and Englewood on the South Side. Englewood is actually two community areas (Englewood and West Englewood), although I have never heard of anyone making the distinction. So far this year, Englewood has the most homicides, 61 followed by Austin with 60.

Englewood's population has declined from 101,000 in 1990 to 61,000 in 2010 based on census data. Austin's population has declined from 114,000 to 99,000 during the same period. Chicago's total population has remained about 2.7 million during this period. So you can get a good idea of homicides per 100,000 for these areas without using a calculator -- Englewood's being over 100 and Austin around 100.

A lot of the remainder of homicide concentration isn't as obvious, since the Community Areas are either small or not homogeneous. For example, directly north of Englewood is Community Area 'New City' which is primarily the neighborhood, 'Back of the Yard'. It has a population of 44,000 at 2010, with the ethnic breakdown about 60% Hispanic, 30% black and 10% white. Its 24 homicides will end up with a rate per 100,000 lower than Englewood, but roughly similar to an average of the adjacent areas. All the homicide figures are year to date and from http://heyjackass.com on Sept 24, 2016.

The main thing that people aren't likely to understand if they just read the main stream media is the extent to which the problem is concentrated. Which means that if you don't live in those areas, you don't notice.

Another way to get a sense of what is going on in Chicago is to simply read the homicide news reports. http://homicides.suntimes.com The Sun Times does a pretty good job with this now. If you want to get a feel for what's going on, there is nothing better than to read through 100 of these. You can do it in a couple of hours and I promise you will know more than Watching talking heads on CNN bloviate about crime.

Adrian said...

regarding the "well what do we do about it" question, check out George Mason University's compilation of studies on what sort of police interventions are effective: http://cebcp.org/evidence-based-policing/the-matrix/

Peter Moskos said...

That's for that link!

And Theo, I still read the Sun-Times every day (well, online I glance at the Metro section).