Let's talk briefly about the spike in shootings in NYC.

First of all, there is a spike in REPORTED shootings in New York City. In fact, June 2020 has the most shootings reported in NYC since 2006.

There are a few ways to approach this data. I want to start by asking questions that I will not have answers to.

So if you want clear answers about the data, this is not that thread.

Question 1: Why is the data lumpy?

First, let's remove the pandemic briefly. I promise we'll get back to it.

If you cut the data at 2013 and 2017, the shooting data radically shifts for each era, creating three distinct lumps.

Here's another view, with reported shootings at the yearly level.

But that doesn't align with overall crime trends in NYC.

Admittedly, shootings make up only a fraction of crime overall, but generally speaking I'd expect two things that I don't see.

Expectation 1: Shooting trends look similar to arrest trends.

Expectation 2: Shooting data does not have weird sudden drops or distinct eras; if it did, we'd also see that in arrest data. We don't.

Question 2: What could explain these lumps?

I have two possible explanations, but there may be more.

1. How shootings are recorded and reported has changed.
2. Everyone in NYC collectively decided to shoot their guns a little less.

Number two is a little far-fetched and I'd be surprised if the New Yorkers did that not once, but twice.

Option one on the other hand is a little more plausible. If there was a change in policy over how NYPD reported shootings, perhaps starting Jan 1 of 2013 and 2017, then the data would look lumpy.

Reported shootings would decrease.

And now we get to the crux of the issue. Notice that I've said REPORTED shootings. NYPD's data is based only on what is reported to them, and what they record.

Before I proceed, I want to note that NYPD has been caught manipulating data before. It may not be the case now, but we should pause and consider that our data source is not objective. https://www.reuters.com/article/us-crime-newyork-statistics/nypd-report-confirms-manipulation-of-crime-stats-idUSBRE82818620120309

Let's take a look at the spike in reported shootings now.

It's really important to note that there are typically fewer than six shootings reported in NYC on a normal day.

Even through the quarantine and shelter-in-place order, the data remains fairly constant.

Somewhere between May 15 and 29 the spike begins. Let's super-impose some events onto the timeline.

I'm obviously biased, but I can't help but notice a couple of things.

1. The shootings begin after the NYC budgeting process begins.
2. The shootings never return to one per day after May 28, the first day of George Floyd protests.

From here, we enter a land of speculation. So this is where twitter will have a reasoned discussion and we'll debate points civilly.

Right?

My guess: the shootings were always there and NYPD was not reporting them.

The lumpiness in historic reporting, lack of a dip during the shelter-in-place order, incongruity with arrest data and increase around budget time suggest the data may be manipulated.

All this data is available on the NYC Open Data portal, and anyone can analyze it.

I probably missed something. I hope I missed something. Because the other explanation is that NYPD is manipulating the data for their own ends.