Scott Sumner  

Crime and civil liberties

Craigslist Reduced Female Homi... Jonathan Weisman's Bias...

There's an interesting story in the National Review, discussing New York crime data:

Today in New York City, use of stop-and-frisk, which the department justified via the 1968 Terry v. Ohio Supreme Court ruling, has crashed. Yet the statistics are clear: Crime is lower than ever. It's possible that crime would be even lower had stop-and-frisk been retained, but that's moving the goalposts. I and others argued that crime would rise. Instead, it fell. We were wrong.
New York saw murders plunge to 290 in 2017, down 12% from the year before, and far below the 2245 murders committed in 1990. Chicago had 650 murders, despite having less than 1/3 NYC's population. While there are some plausible reasons that explain a part of the decline (improved health care for people who are shot, fudged data, demographic change, etc.,) there are no good explanations for the astounding scale of the decline in crime (not just murders):
It's possible there is some number-fudging going on with the crime statistics, but so far there is only scattered anecdotal evidence of that. And any serious effort to charge today's cops with distorting the numbers would have to consider whether yesterday's cops did the same. Moreover, if anything, the de Blasio-era police have been more antagonistic to the mayor than in any previous administration, taking such unprecedented steps as turning their backs to him en masse at a funeral and, at the nadir three years ago, launching a major slowdown in arrests and writing tickets. Police seem disinclined to do de Blasio any favors by giving him favorable crime statistics to brag about.
Chicago has a higher percentage of African American residents than NYC (32% vs. 25%), but in absolute numbers NYC has far more (over 2 million vs. less than 900,000. Indeed NYC has more of almost all ethnic groups than Chicago, so it's really hard to explain why NYC has 290 murders a year and Chicago has 650. This also suggests that the link between race and crime might be more complex than many people assume, (which is not to deny that race is strongly correlated with crime.)

The good news is that we don't need to curtail civil liberties to control crime. Kudos to the National Review for admitting they were wrong. I was not a fan of stop and frisk, but even I thought it would be mildly effective. The fact that it doesn't seem to have much effect on the crime rate is good news for civil libertarians.

Comments and Sharing

CATEGORIES: Economics of Crime , Liberty

COMMENTS (12 to date)
Matthew Waters writes:

For the nation-wide general drop in crime, the longitudinal data on lead exposure and abortion is pretty convincing. Health care for the murder rate in particular is a factor as well, but it's tough to measure.

Table 5 in first link has the three different correlations on state-year panel data. It shows effects for lead, abortion, number of police, "shall-issue" concealed weapon laws and beer consumption.

The p-values aren't very low except for abortion. The data probably will never show a smoking gun for what caused crime rates to decline. But the micro evidence for lead is really convincing.

As far as Chicago vs. New York crime rates, Chicago has had a far worse clearance rate for murders: 20% vs. 68% for NYC. As part of worse finances in general for Chicago vs New York, CPD has less resources. Both cities have very expensive pensions and health care for their municipal employees, but NYC has a much stronger tax base.

E. Harding writes:

"so it's really hard to explain why NYC has 290 murders a year and Chicago has 650"
Two explanations come to mind:
1. Worse/less policing in Chicago (highly probable)
2. Immigrant Blacks could contribute to the skew in the numbers.
3. Incomes being persistently higher in NYC than Chicago since the early 1980s (and especially since the Great Recession).

David R Henderson writes:

@Scott Sumner,
I was not a fan of stop and frisk, but even I thought it would be mildly effective. The fact that it doesn't seem to have much effect on the crime rate is good news for civil libertarians.
Well said, Scott. My thoughts exactly.

Dylan writes:

So with all the cop and detective shows set in New York and the continuing falling murder rate, I wonder if we're getting to the point where in any given year there are more fake murders set in New York than real ones?

Mark writes:

I suspect compositional changes explain a lot of the difference: in NYC income (and cost of living) seems to have gone up a lot more than in Chicago and there’s been a lot more gentrification, leaving fewer poor neighborhoods and less crime. Compare the south side of Chicago to Hell’s Kitchen or Harlem. Poorer , historically high crime neighborhoods in NYC are rapidly becoming middle class, reducing crime. The same thing hasn’t happened in Chicago, where if anything there’s some risk of ‘Detroitification’ where people move out to the suburbs leaving parts of the city in disrepair.

tOKEN writes:

Chicago has always had much more serious gang problems than New York, and in recent years, what used to be a few large gang organizations have broken down into smaller units, creating more borders between rival gangs, and therefore, more disputes that lead to violence.

It's an interesting question why the gang HHI in Chicago is greater than in New York.

maynardGkeynes writes:

It would be interesting to break out the Manhattan data separately, which is even more impressive. I believe there have been a few years recently where were there were virtually zero (or actually 0) "stranger to stranger" murders in Manhattan (ie., neither drug-related nor domestic/personal). The positive economic impact this has had on Manhattan is hard understate. I've often thought that the dead-weight economic loss of crime in cities like Chicago is far greater than people on the left are willing to acknowledge, at least publicly.

Brian writes:

"The good news is that we don't need to curtail civil liberties to control crime."


This may be true as a general principle (which I would certainly welcome), but this example only shows that stop-and-frisk has had no effect. You can't really draw any more general conclusions.

Murders in NYC fell dramatically in the 1990's, from 2245 in 1990 to 673 in 2000. Since then the decline has been slow and steady of -22 murders per year. This past year's result is well in line with that rate.

The steady decline indicates that it's basically independent of policy, and really more reflective of uncontrollable societal factors. That's really not surprising given the huge decline in the 1990's. Any low-hanging fruit readily addressed by policy likely would have been implemented then when the rates were higher (more motivation) and the results were potentially larger and more easily seen.

Scott Sumner writes:

Everyone, Thanks, lots of good comments. My general sense is that the scale of the decline in NYC is so large that multiple factors are probably required to explain it, including many that are discussed in this comment section.

James Pass writes:

Mr. Sumner's main point is that stop-and-frisk policies are not necessary to reduce crime in high-crime areas.

There aren't any stop-and-frisk policies in low-crime areas. There are all sorts of obvious ways to reduce crime that don't involve curtailing civil liberties. All these ways point in one direction - to a decent, middle-class standard of living. Policies like stop-and-frisk attack the symptoms of the disease but not the causes.

Chicago saw dramatic declines in murder since 2006, but starting in 2015 it jumped way up. New York City has seen a steady decline in murder since 1990. Here are some other cities that have seen murder rates recently jump UP:

St. Louis, Baltimore, Birmingham, Hartford, Washington DC, Kansas City, West Palm Beach, Savannah, Salinas, Waco.

Mr. Sumner says "there are no good explanations for the astounding scale of the decline in crime [in NYC]." I think with a little research one could find some good explanations for why murder rates have decreased in some cities and increased in other cities.

Sumner states "the fact that [stop-and-frisk] doesn't seem to have much effect on the crime rate is good news for civil libertarians." But the challenge for civil libertarians is to stand up for principles even when some polices are effective. For example, hypothetically, let's say that stop-and-frisk was somewhat effective at reducing crimes such as robberies or murders. Would Sumner then support the policy? How does a civil libertarian balance means and ends?

Lorenzo from Oz writes:

Folk in the US should stop talking about race. Seriously. And start talking about ethnicity.

If you say "blacks" who do you mean?

Ebonic-Americans (the descendants of slaves, a people who went through Jim Crow, have been in the US 2 centuries or more)?

Caribbean-Americans (the descendants of slaves freed a generation earlier, whose post-slavery experience is very different and who have migration effects)?

Recent sub-Saharan African immigrants who tend to be highly educated and are a notably successful group?

Because the things that causally matter differ among these groups.

Lorenzo from Oz writes:

I have a post expanding the points in my previous comment. (I have done it as a separate comment because linkage response is so variable here.)


Return to top