Afternoon all! I promised a thread on the myth of Black-on-Black crime, and here it is. Buckle up, because it's gonna be a long one. For those who have a short attention span tl;dr: it's racist bullshit deflection and I have sources to support that. Let's get started! 1/85

First off, shoutout to @mattomalley for his milquetoast response to the racist "blue lives matter" rally held in West Roxbury, and his utter failure to address the protest against a Juneteenth celebration which saw a lot of the same crowd in attendance. 2/85

Without that, I never would have had the interaction with a racist constituent who quoted "FBI STATS!" at me repeatedly and called me "donkey boy(?)" when I challenged his "Black-on-Black crime" racism, and which led to this thread. 3/85

I think most of the people who follow me know me in real life, so they know that I'm a horrible nerd and that, before I went to law school, I got my Master's in Criminology and my Bachelor's in Criminal Justice. Now seems like a good time to put that hat back on. 4/85

So let's talk about "Black-on-Black" crime. People just love to bring that up every time someone says "hello yes I think maybe the police should not murder Black people." There is a whole hell of a lot to unpack in that question, but let's start with what should be obvious. 5/85

What about White-on-White crime? According to a report by the Bureau of Justice Statistics (BJS), a branch of the U.S. Dept. of Justice (DOJ), 63% of violent crimes reported by Black victims were committed by other Black people. 6/85 https://permanent.fdlp.gov/gpo109667/rhovo1215.pdf

"Ah hah!" The racists say, "we were right!" Well, slow your roll there buddy, because I hate to break it to you but that same report found that 57% of violent crimes reported by White victims were committed by other White people. 7/85

Now, I lost the ability to do math when I got my JD, it's just a thing that happens, but those numbers do look awfully close to each other. Here's the real kicker though. White victimization is, in terms of raw numbers, way higher Black victimization. 8/85

In fact, it accounts for 63% of all violent crime in the United States. That being said, White, non-Hispanic or Latinx people make up 61% of the US population, so that's actually pretty close to where the number "should" be as far as representation. 9/85

If you add in Hispanic and Latinx Whites, that's 77% of the population. So, all-in-all, White people are about where we'd expect them to be in terms of victimization rates. But also, that's 3.7 million violent crimes. Compared to 850,720 in the Black population. 10/85

That 850k+ is actually about 14% of total victimizations, for the record, with 13% of the population being Black, so, again right about where statisticians would expect it to be. 11/85

The problem is that despite White people committing 2.1 million violent crimes against each other per year on average, and Black people only committing about a quarter of that, again at a roughly comparable rate, the narrative is "Black-on-Black." 12/85

Oh, and before we step away from this point, "well what about interracial crimes? White people being attacked by Black people?" Rate of White-to-Black interracial violence is 11% compared to 15% the other direction. The relationship there, though, is inverted. 13/85

Remember, Black people make up a substantially smaller portion of the population, whereas White people are the substantial majority. When you adjust for that it should really disturb you how often White people are attacking Black people. 14/85

This actually leads very well into the explanation of the phenomenon of intraracial crime: proximity. Generally speaking, there is an accepted idea in criminology of "means, motive, opportunity." 15/85

No matter what theoretical framework you hold to, these core tenets appear somewhere, somehow, within it. That's because they have been seen over and over again in reality. The vast majority of crimes occur where a person lives or works, or on the path between those places. 16/85

That's for both victim and offender, by the way. You just don't really see Johnny from the suburbs striking far afield to beat the piss out of someone. Why? Because it just doesn't make sense to. Violence is typically either heat of passion or calculated and intentional. 17/85

Both of those, except for certain select exceptions, involve a relationship with the victim. You don't usually plan to murder a random stranger just because, and even a serial killer who kills "randomly" typically does so within a set area that they frequently traverse. 18/85

In simple terms, you tend to punch the people who live around you because they are who is within arms reach. Nobody gets so mad that they're going to get in a killing mood hop on the train and go for a murder field trip. So, let's talk about racial segregation. 19/85

Now, I'll take off my criminologist hat for a second and tell you that I do not know as much as I should about redlining and other racist segregationist tactics that have been used to push Black people into economically depressed areas while facilitating White flight. 20/85

The good news is, there are literally entire websites devoted to redlining alone. Here's a resource to get you started, and a dirty little secret: a lot of research that gets done starts at wikipedia and uses the sources there to find other sources. 21/85 https://en.wikipedia.org/wiki/Redlining 

It's actually a really helpful research tool and it makes a lot of people's lives way easier. So take a look through there and look through the sources that are cited. 22/85

The tl;dr of it all, though, is that after about a century or so of housing segregation post-slavery, the feds decided to explicitly codify it with the National Housing Act of 1934, which established the Federal Housing Administration (FHA). 23/85

The FHA basically gave banks a legal excuse not to offer mortgages to Black people, or to people living in certain neighborhoods (spoiler, also Black people), because they were "high risk." 24/85

If you think that sounds like a valid reason, I'm honestly amazed you made it this far through my thread. But please do continue on. 25/85

Long story short, the government, banks, landlords, real estate agents, neighborhood and homeowners associations, and your everyday suburban people were racist as shit and violently forced Black people out of neighborhoods or prevented them from moving in. 26/85

Alright, criminologist hat, back on. So we now know that "Black-on-Black" crime isn't a "thing." It's certainly not something explicitly unique to Black people. The majority of crime is intraracial, not interracial. That's a function of our deeply segregated society. 27/85

So let's move on, then, to my absolute favorite subject: numbers. Specifically, why do the numbers you see most often quoted (by racists, the media, and racist media) look so much different from the ones I'm giving you? 28/85

Well, you may have already picked up on it if you're paying close attention. It's the word "victimization" that's key here. My new racist friend from last night got super shouty about "RAW NUMBERS" and "FBI STATS!" They love to think that "raw numbers" are somehow pure. 29/85

Oh my sweet summer children, they are not. In statistics, as in life, context is key. So, let's get some context on those "FBI STATS!" The information from BJS comes from the National Crime Victimization Survey (NCVS). It has its own issues, but it's pretty good stuff. 30/85

NCVS is largely regarded as a sort of gold standard data set. Depending on who uses it, it can be used for scaremongering (the "dark figure" of crime) or to point out the sizable racial injustices perpetrated by our punishment systems (Hint, this thread is the latter). 31/85

The reason I use the NCVS rather than "FBI STATS!" is pretty simple. It more accurately reflects the actual rate of criminal activity by tracking victimization rather than arrests and convictions. Which brings us NIBRS and the UCR. First, some names. 32/85

UCR is the older of the two systems, and it stands for Uniform Crime Report. Ironic, for reasons you'll understand shortly. The UCR basically created a list of 8 "Index" crimes that the FBI considers the worst of the worst, divided into "violent" and "property" crimes. 33/85

It then also tracks a list of other offenses that range from prostitution to simple assault to curfew offenses. The stuff that doesn't usually make for great television. The problems that arose from this were manifold. Big hitter no.1: the data is self-report. 34/85

The FBI doesn't collect this information, it relies on local law enforcement agencies to hand it over. Now, while most states did eventually make laws requiring reporting, it was not mandatory practice for a long, long time. On to Big hitter no.2: the ironic bit. 35/85

The UCR has set categories for crimes, which would be great if we had a universal statutory framework for criminal offenses, or data intake at arrest. But we don't. 36/85

A lot of the reporting relied on guesswork on the part of either the officer making the report or the analyst reviewing the information. Oh, also, the data is based on crimes reported. That's important. 37/85

The biggest problem there, as noted by basically every criminology professor ever, is that most crimes are unreported. Sometimes because the victims fear retaliation, frequently because they do not trust the police. 38/85 https://www.nap.edu/read/10581/chapter/2

Regardless of the reason, the UCR misses a lot of stuff. However, what the UCR does capture (relatively) reliably, are arrest and charge statistics. We will get into why that is a serious flaw in the data in a moment. Let's move on, for a moment, to the NIBRS. 39/85

The NIBRS is the National Incident Based Reporting System. It was a collaboration between the BJS and the FBI to create a more comprehensive, less pidgeon-holey system than the UCR. 40/85

If you want to know how that was going in 2015, 26 years after the system was created, check out this article. Tl;dr is "not well." 41/85 http://journals.sagepub.com.ezproxy.neu.edu/share/7CDDTBHKZSYMS4P9PGWY?target=10.1177/1525107115623943

There are severe coverage issues as law enforcement agencies struggle to adapt to the system, and like the UCR it still misses a whole lot of undiscovered or unreported crime. That's where the NCVS comes in. 42/85

The NCVS is actually conducted by an organization that knows how to do surveys and data gathering pretty well, the U.S. Census Bureau. Not saying it's perfect, but, y'know. 43/85

Setting aside many issues of racism associated with the Census Bureau for a moment, we can at least acknowledge they're better at collecting meaningful data than the FBI. That's part of why they're a gold standard. Importantly, it does not rely on reports to or by police. 44/85

So let's talk about that thing I put a pin in a few tweets back: police arrest statistics. This is the very core of why the "FBI STATS!" numbers are so problematic. Spoiler, it has to do with rampant racism. 45/85

Quick review of what we've covered so far: people tend to hurt others based on proximity. People tend to commit crimes and be victimized at a rate that is roughly equal to their representation in the population. FBI stats are based on arrests. 46/85

I'm betting you're seeing the problem already. If White people and Black people commit crime at about the same rate (if we're talking raw numbers though White people are WAY ahead), then why are so many Black people getting arrested? Well, let's talk about Broken Windows. 47/85

It's the impetus behind "zero-tolerance" policing. The idea is that failure to address minor disorder leads to people doing worse and worse stuff. If you leave graffiti up then before long people are murdering each other in the street en masse. We've all seen it. #sarcasm 48/85

As you may be able to gather from the way this thread is going, it is also complete bunk. Broken Windows is the brainchild of James Wilson and George Kelling. James Wilson and George Kelling cheated. 49/85