Another way to get crime stats to look good is to ignore crimes. Even from the article we are discussing here it sounds like there have been dozens of instances of assault and robbery that were ignored. Ignore enough and people stop wasting their time reporting crimes. Stats look great, the streets tell a different story.
Oops, didn’t notice the first link was for New York State. The CompStat link includes longer-term trends, though, which is what I was referring to.
Crime is down overall, and I suspect that if you controlled for the murder rate increase in all US cities over the past two years it might be flat or down in NYC as well.
Murder rate is arguably the most useful crime metric because it's the hardest one for politicians/police to game. Changes in enforcement can change trends for assaults and property crime, but you can't really hide bodies.
Well, you'd have to show that murder rates correlate with rates for other crime. But more important is digging into the thorny issue of what — and who — defines "crime". Because there are a ton of aspects to it: what laws are on the books, what laws are actually enforced, what various special interest groups want to legislate, what people analyzing data choose to include.
For example, as of 2017 feeding pigeons was a misdemeanor in Las Vegas [1]. That will show up in crime data, but you'd be hard-pressed to convince me that it correlates with the murder rate. And if you wanted me to compile a report on crime data, I'd probably ignore it altogether (which is essentially what people are doing when they refer to "violent crime" states).
For a slightly more charged example, let's say property owners push a law against sleeping in public. If they succeed and police don't enforce it, are they "gaming" the metrics? What if most other locals actually oppose the law?
Are you saying that we shouldn’t control for a coordinated national change in murder rate when evaluating the impact a mayor’s specific policies have had on their city?
Anti-police policies are not exclusive to the NYC though. If you want to control you need to show that murders also rise in the cities that don't have the same policies. Chicago, SF, Minneapolis don't appear a sensible choice for a control group in this regard.
NYC is subject to the same national economic, demographic, and cultural trends. For example, recently there has been a respiratory pandemic that affected the entire nation.
"Safety" in a city is an outcome of a very large number of interconnected factors. It is not under the control of any one mayor or police commissioner. More importantly, perception plays a large role in how people judge the safety a city or a neighborhood.
"Safe" is also very hard to define. Safe for who? Safe in what sense? It's a giant grey area and no one has ever been able to "solve" the crime problem. Getting tough on crime doesn't work, nor does endless social programs, nor does sweeping it under the rug.
Hopefully Adams will take a pragmatic approach and do what he can without making it worse. I wish him well.
There are plenty of non-exogenous causes, they just don't have the same cool-factor for publication as things like leaded gasoline. For example many people were released from institutions during the massive de-institutionalization push that started in the mid 50s [1], going from ~340 institutionalized people per 100,000 in 1955 to about 60 per 100,000 in 1980. One can even argue that the sum of incarcerated and institutionalized has averaged roughly constant (about 450 per 100,000) with a dip during the period of increasing criminality from the 1960s to the 1990s as the incarceration rate had to climb from ~100 to ~450 per 100,000 to make up for the mass de-institutionalization.
The cited study is pretty interesting:
"The juxtaposition of these trends and the current high incidence of severe mental illness among those behind bars begs the question of whether the mentally ill have simply been transinstitutionalized from mental hospitals to prisons and jails. A related question concerns the extent to which the unprecedented growth in incarceration since the late 1970s is driven by a reduction in public investment in inpatient mental health services. Past changes in sentencing and corrections policies are currently under heightened scrutiny as state prison populations are at record levels and many states are seeking to scale back correctional populations with an eye on the fiscal benefits of doing so. To the extent that the run-up in state prison populations was driven by deinstitutionalization, the current focus on sentence enhancements and the evolution of the U.S. sentencing regime may be misplaced."
If crack was a big driver for the previous crime wave, it heroin/fentanyl is likely to be a driver for this one. Or, on the West Coast, meth. It’s made a pretty big comeback since 2017, but it’s been mostly overshadowed by the opioid crisis.
(That link only covers through 2019, but if you look at the official CompStat data the two year trend for 2019–2021 is also negative: https://www1.nyc.gov/assets/nypd/downloads/pdf/crime_statist...)