Stop-and-Frisks, Crime Spikes and a Lesson in Statistics


A few months back, we reported on the outrageous statements made by the Boys in Blue that praised the low murder rate as a sign of victory for stop-and-frisk. And our argument was simple: if you stop-and-frisk everyone, then, yes, the murder rate will go down. But this is the kind of upside-down thought pattern that had our head spinning, too, and it continued today in the media.

This morning, the New York Post had a breakthrough: its reporters came to the conclusion that the widespread practice of stop-and-frisk by the NYPD is a success. We solved it, guys. No more complaining about the infringement on constitutional rights, racial profiling or any of that other left-leaning drivel that the NYCLU loves to cook up.

But how did the Post come to this conclusion? Well, the media organization simply just looked at the numbers. According to Rebecca Harshbarger and David Seifman, the reason why we can lay out the ‘Mission Accomplished’ banner for the controversial means of searching citizens, notably minorities, is because of the corresponding crime statistics that go with it.

Exhibit A: Between January 1st and March 31st, there were 203,500 individuals who were stop and frisked, 881 guns seized and 24,571 crimes recorded in the books.
Exhibit B: Between April 1st and June 30th, there were 133,934 stop-and-frisk targets with 732 guns seized and 27,832 major crimes reported.
Notice the drop in stop-and-frisks and the 12% spike in crime. Well, there you have it, people – it’s a wrap… unless you’ve ever taken a class in statistics. Then, this conclusion might seem a little far-fetched to believe.
Okay, how do we say this to the Post and its conclusion without coming off as pretentious, finger-wagging elitists that have no sense of a reality that doesn’t include being judgmental. This is going to be hard; take a deep breath, John. Well, here goes nothing:
You can’t do that. You just can’t. Although this is hardly mentioned anywhere in the media besides maybe in Ezra Klein or Nate Silver’s writing, what the New York Post did with the stop-and-frisk and crime numbers is against mathematical axioms.
And the rule-breaking is simple: you cannot make a wholehearted declaration of something based on a correlation. It is statistical logic that just because X spiked and Y fell flat doesn’t mean that X and Y are in bed together. A correlation is basically a relationship between two factors that may have some sort effect on each other; when you write research papers smothered in statistics, every student knows to stress that maybe. In other words, one doesn’t necessarily explain the other.
What the Post is doing underlies the main problem with basing a conclusion off of a correlation: there are other factors we must all take into account. To say that the period between April 1st and June 30th had seen a spike in crime because of the stop-and-frisk deficit, we must also say that this period had the exact same conditions as the period before it.
Which it did not: one could argue that the warm weather is an enormous variable – DNAInfo columnist Murray Weiss made note of this earlier on in the year, when New York was stuck in that strange autumn-that-never-ended period and saw higher-than-average crime stats as compared to the Snowpocalypse winter of 2011. This year, it goes without saying that temperatures were much higher in the period with more crime recorded; this is a huge factor for the Post to simply skim over.
But, in regards to media narrative purposes, the declaratory, Big Picture thinking and writing style is more suitable to explain analytical crime statistics to a mass audience. It is much easier for a reader to understand that, hey, if crime rates spike and stop-and-frisks are down, the latter must explain the former. So, with that, our ‘Mission Accomplished’ banner analogy actually makes much more sense: when Dubya hung up that claim of victory at the end of the Iraqi invasion in 2003, he was signing off on a situation that had only started, not taking into account the countless other factors that would lead to many more years of civil strife and bloodshed.
It’s when you dig a little deeper into that overreaching thought pattern that you start to see what’s really going on here.