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

by Peter Moskos

March 7, 2016

Remain Calm!

Apparently there's still no need to worry.

Here is what I think matters: 2015 will almost assuredly see (we don't the numbers for sure yet) a double digit in increase in homicide. See this Washington Post piece for a clue.

And yet, if you listen to Ames Grawert and James Cullen, there's no need to worry:
Rather than stoking unfounded fears of a new crime wave, always just beyond the horizon, we should take this opportunity to ask how we can expand on the public safety gains of the past 25 years.
While there were 471 more murders in large cities in 2015 than 2014, more than half (260) of that increase occurred in just three cities: Baltimore, Washington and Chicago.
My favorite ideological statistical shenanigans: if you ignore places where crime is up, crime isn't up!

America has not seen a double digit increase in homicide since 1971. (1986 and 1990 came close.) Since 1971 is my entire lifetime. So, yeah, we probably saw the biggest annual increase in murder in my lifetime and perhaps ever. Seems like something to worry about.

But no. We who care about these dead people are just stoking public fear, as if police have anything to do with confronting murderers, and perhaps even preventing a few shootings.

In Baltimore, mayoral candidates are talking about how best to reform police. Very little on how to prevent shootings. They should be talking about how to get back to how they were exactly one year ago, before police were seen as the problem and violent crime doubled.

Just remember, no matter what happens, if it's not ideologically expedient to worry about rising homicide, just repeat this mantra: Remain Calm. All is well.

Related, at this is an interesting piece of the jigsaw puzzel. Homicides are down thirty-some percent in NYC this year, which seems to negate last year's increase in NYC. At least here in New York, the sky is not falling.
Mac Donald predicted in 2013 that if New York City ended its controversial stop-and-frisk program, crime would skyrocket back to pre-1990 levels.

Well, stop-and-frisk formally ended in 2014, and the lights still haven’t gone out on Broadway. In fact, as the number of stops by police tapered off, so did the city’s murder rate, hitting a historic low the same year the program ended. Despite a small increase, the murder rate remained low in 2015, while shootings, major crime and arrests all fell in tandem.
NYC is OK. But elsewhere, I'm not so sure.

[Thanks to EyeRishPirate for bringing this to my attention.]


Andrew Laurence said...

If more than half of the increase is in just three cities, perhaps it's not a nationwide problem, but that's not the same thing as saying it should be ignored.

Peter Moskos said...

Indeed. But that's exactly what crime-rise apologists are doing, saying it should be ignored until the problem is universal. It makes no sense.

Pragmatic Liberaltarian said...

" as if police have anything to do with confronting murders, and perhaps even preventing a few shootings."

You're written a lot about those downplaying any link between police criticism and increased murders. And, I think pretty persuasively make the case that the increased murder rate in 2015 is concerning and significant and maybe even alarming. But, I've seen very little in explaining how police criticism is actually causing increased murders; you suggest/assert it often as in the quote above, but don't explain it. As you often note, correlation is not causation. To make the case for causation, there needs to be some reasonable causal mechanism between the two. This leads me to a question that may seem obvious to you, but is not to someone like me living in a nice safe neighborhood where I pretty much never see the police except when responding after the fact to the rare call. So, the question: what is the causal mechanism is between proactive police work and prevention of murder? I think I once saw MacDonald suggest that stop and frisk finds people possessing illegal drugs or weapons and prevents murders by arrests and imprisonment of such persons before they might otherwise murder someone, and presumably by also deterring others from possessing illegal weapons while out in public for fear of police stopping and questioning them. Is that basically the theory or is there more to it?

john mosby said...

Prag, that and this: proactive investigation post-murder helps (in theory) to keep down murders by solving them and convicting the murderers, thus giving you specific deterrence (of the convicted murderer while she's in jail) and general deterrence (of other potential murderers who see the likelihood of getting caught).

Proactive policing after a crime has occurred includes things like: sealing off the crime scene for a thorough analysis; separating and interviewing potential witnesses; canvassing the neighborhood for other potential witnesses; maintaining a presence in the days after to pick up on disrupted patterns that may point to potential suspects (eg, "where Lefty at?"). Done well, these techniques tend to solve homicides. And they have the by-product, through the spike of police presence, of deterring revenge killings.

But if people are poking cellphone cameras in your face, running in and out of the street simultaneously screaming "Why weren't you here?" and "Stop Snitching!", complaining about why you left the body in place for so long, threatening you with complaints and lawsuits for bracing witnesses/suspects, etc., then you will probably not do these techniques well or at all.

And the murder won't get solved. And Lefty and his mates will realize they can get away with another one.


Peter Moskos said...

Prag, I think I've written the answer to your (good) question here:
and here:

Matt Ashby said...

Excluding three cities from the national picture is obviously inappropriate (and statistically dishonest), but I'm not sure it's helpful to talk about a single national picture if it's known that there are different places with substantially different trajectories. What would be helpful here would be to talk about homicide trends in particular cities (and potentially areas within cities) rather than trying to divine anything (is there a problem? is it getting worse?) from a national average that may not accurately represent any group of actual places. Doing this allows you to talk about a big increase in homicides in Baltimore while at the same time acknowledging that the trend in homicides may be down in some places and static in most. I'm not sure national totals or averages are at all helpful here.

I'm also not sure it's valid to compare percentage changes now to those in the past, given that we know there is a clear trend in the numbers. As the number of incidents goes down, the chance of a double-digit percentage change goes up, because it takes fewer incidents to make up at least 10% of the total. Imagine if there were only one homicide in America one year, but two the next: that's a 100% increase, the sort of percentage change that is very unlikely if you have 13,000+ incidents.

I totally understand your frustration with people who want to fit the evidence to their own point of view, by the way!

Peter Moskos said...

I could make the same point in absolute numbers. The last time we saw more than 1,000 more murders in any given was 1990, 26 years ago. But if I spoke in absolute numbers -- and I guarantee this because this isn't honest statistical criticism as much as ideological claptrap -- people would say "It's not fair to use absolute numbers since America is larger now." (And to some extent, that's true... which is why we use rate).

Until this recent homicide increase, I have never heard anybody even whisper that a percentage change is an unacceptable statistic method to describe changes in crime. (I understand year-to-year and city-to-city can be problematic.) Of course the n has to be of a certain size. But we're talking nationwide. And luckily, statistically speaking, we here in the US have no problem with too few homicides.

And here's the other thing. Homicides in America have always been geographical isolated. The entire state of New Hampshire has something like 20 murders. When it comes to murder, we're never talking about a universal increase to all people.

It's the geographic and social and economic and racial isolation that makes these increases so terrible, and yet so easy for others to ignore.

JPP said...

@John Mosby,

Aren't your examples of proactive policing an oxymoron? How can post-murder investigation be proactive?

I understood proactive policing to mean not being reactive to crime. It looks like your examples are simply "policing", and not proactive.

I think doing a self directed search on a guy who is frequently around or involved in shenanigans and finding a weapon could be considered proactively preventing an act of violence. On the other side of the issue, 5 searches with no result could be called harassment.

john mosby said...

If you don't want to call it proactive, how about just active? My point is that, even after a crime has taken place, there's a spectrum of policing available, from just hosing down the sidewalk to actively trying to solve the crime by engaging with the community. The Ferguson effect is pushing the needle toward the sidewalk-hosing end.


Jay said...

It seems to fair to focus on cities where the increase is greatest, especially to ask whether they are different in kind or in degree from other cities, and from one another. Are the reasons for the increase in Chicago the same as those for Baltimore and DC? Changes in Baltimore policing, as Peter argues, may have been an important factor. But is anyone claiming a “Ferguson effect” in Chicago? My impression is that until the VanDyke video, the Chicago police have been operating with the same latitude they’ve had for decades. (The VanSlyke video became public far too late in the year to have any impact on the 2105 crime statistics.)

Peter Moskos said...

Those are good questions.

I think Chicago has seen a huge increase since the video became public. In 2015 there was only a small increase in crime (15 murder increase, I think).

So yes, my hypothesis is that each city sees has it's own "Ferguson" event that matters. And New York is still the happy outlier that makes it hard to generalize.

IrishPirate said...

I suspect the recent policy that cops have to fill out a form every time they have an interaction with a citizen has had more of an impact than the Van Dyke/McDonald shooting response. If the bangers are less afraid of being stopped and "harassed" by the evil cops they're more likely to carry guns. If they're carrying guns they're more likely to shoot someone. If they have to hide the gun in the bushes or somewhere where it isn't immediately available some shootings won't occur.

Here's a take on it: https://secondcitycop.blogspot.com/2016/02/saw-this-comment-in-yesterdays-post.html

Plus the constant protests and the sometimes lacking police response has negatively impacted morale.

I watched a video taken just before XMAS when protesters blocked stores on Michigan Avenue and literally fought cops as they tried to make an arrest. Cops backed off. That's a BOZO NO NO. You fight the police you have to be arrested. If that means calling in reinforcements to do it then you fucking do it. If it means negative videos hit the airwaves/internet that annoy "Da Mare" then so be it.

Here's some video. http://wgntv.com/2015/12/24/black-christmas-chicago-mag-mile-protest/#

I can't find it, but there's other video out there of a large group wrestling with the big older cop wearing the Silver Oak Leaf insignia in an alley. Bike cop jumped in, but it was ugly and if he had drawn his weapon it would have been appropriate.

At the moment the few hundred regular protesters seem to believe that they can use force or violence and get away with no repercussions. Too often they're right.

People have a right to protest, but once they get physical the cops need to clamp down.

The Van Dyke shooting was horrible, but I wouldn't put police involved shootings on the top ten list of problems facing black Chicagoans. I might put the 99% of shootings that don't involve cops as number 1-10 on that list, but then again I'm a pasty white guy who is seldom stopped by cops.

I also know that to paraphrase the first "Mare Daley" cops don't create disorder...they preserve da disorder. There's some twisted wisdom dere.

Pragmatic Liberaltarian said...

Has there been any analysis of the types of murders taking place in these cities? By that, I mean who's getting killed, how, by whom, etc.? Is the increase gang members killing one another? Is it an increase in domestic violence? Is it victims of armed robbery? Id think we'd have a much better idea if there's a Ferguson Effect and exactly how it is leading to increased murders if there were better answers.

Peter Moskos said...

I have no reason to believe the increase isn't just more of the same: mostly young black men with stupid beefs, access to guns, and a strong link to public drug dealing. Where gang influence is strong, it's linked to gangs; where it isn't, it isn't. I think it's just our "normal" violence ramped up. I don't think the nature of the killings has changed at all.

campbell said...

http://heyjackass.com/ keeps great stats on Chicago shootings. When you have 100 dead men and only 9 women, I'm pretty sure it's safe to say it's not a wave of domestic killings.