Thread by @timjacobwise

1/ This will be a long thread, but necessary, so as 2 counter some of the right wing & white nationalist propaganda concerning crime in America: specifically crime by black folks, and especially crime by black people against white people, both of which they claim are “epidemics.”

2/ These claims have been around a long time, and those who make them often marshal data to support their arguments. While the data is mostly accurate (at least in its raw presentation), the interpretation and understanding of the data is fatally flawed...

3/ I have written long form rebuttals to this stuff before, but since lots of folks have grown accustomed to twitter-debates and won’t read long form pieces (or merely would like to have easily re-tweetable material here), here goes…

4/ I will probably have to do multiple threads on this bc it usually cuts me off after 25 or so tweets. First, let’s explore the generic claim about crime (especially violent crime) by black people being “out of control,” or as Trump would proclaim, evidence of “mass carnage..”

5/ Facts matter. In truth, crime in general and by black folks in particular is not rising or high by historical standards. Quite the opposite. According to FBI data, crime-data-explorer.fr.cloud.gov/explorer/natio…
Violent crime is down 42 percent since 1995 and homicide is down 35 percent...

6/ Using a slightly different data set bjs.gov/content/pub/pd…
which includes crimes not reported 2 police (but chronicled in victim surveys), from 1993-2017, the violent victimization rate fell by 74% (79.8 crimes per 1000 people to 20.6). So we’re getting safer, not less so…

7/ According to additional DOJ data, bjs.gov/content/pub/pd…
black crime has fallen especially far. From 1994-2015, so called “black on black” crime fell 78%, from nearly 67 victimizations per 1000 blacks to only 14.5...

8/ ...meaning that by 2015, fewer than one-and-a-half percent of African Americans were being violently victimized by another black person in a given year...

9/ Similarly, black crime against white people – the thing racists are really worried about – has plummeted as well. From ‘94 to ’15, black-on-white violence dropped by 80%, from 14.9 victimizations per 1,000 white persons to only 3 victimizations per 1,000 whites...

10/ In 1994, the black-on-white victimization rate was almost 50 percent higher than the white-on-back victimization rate, but by 2015 the rates were virtually identical, mostly due to the drop in black-on-white victimization...

11/ Now, is it true there are far more “black on white” crimes, numerically speaking, than “white on black” ones? Well yes. But this is far less meaningful than racists would like u to believe. First, let’s just look at the data and then explore what it does and does not mean..

12/ According to the DOJ report referenced above, there are about 540,000 “black on white” violent victimizations (mostly simple assault, no weapon, no injury) annually, and about 91,000 “white on black” victimizations (also mostly simple assault, no weapon, no injury)...

13/ At 1st glance this might seem scary…My God! The black folks are on an anti-white crime spree! After all since black folks are 12.5% of the population, and whites 62%, these numbers mean any given black person is 31x more likely to victimize a white person than vice versa!..

14/ Sounds like a bombshell revelation…but it’s not. Here’s why…1st, just consider the numbers themselves...

15/ If blacks victimize whites 540k times a year, as a % of 172.6 million potential victims (the avg number of whites 12+ & and thus eligible for consideration in crime data during the years under review in this report) this is a victimization rate of 31 per 10,000 white people…

16/ This comes to 0.31 percent of whites, at most, who will be victimized in a given year by a black person. Meaning, one white victim of a black offender for every 322 white people…

17/ Meanwhile, If whites victimize blacks 91.5k times a year, as a % of the black pop 12+ (32.6 million people) this is a victimization rate of 28 victimizations per 10,000 blacks, or 0.28% of blacks who'll be victimized by whites...

18/ This means, 1 black victim of a white offender for every 357 black people. In other words, the rates at which whites and blacks are violently victimized by the racial “other” are basically identical here...

19/ So, whence the figure than any given black person is 31x more likely to victimize a white person than the reverse? Simple…and if Nazis new how to do data interpretation (or were honest) they would get this…

20/ First, there mere fact that the general violent crime rate is higher for blacks than whites (due to a range of socioeconomic factors that highly correlate with crime) explains a huge chunk of the disparity...

21/ So, for instance, and according to the DOJ data in this report, although black offending has dropped a lot over the years, so has white offending, and so the black violent crime rate is still about 2.7x higher than the white rate...

22/ The data shows a black rate of about 43 violent crimes per 1000 black people, and 16 violent crimes committed per 1000 whites (and yes, these are 'real whites' for those Nazis who think this number includes Hispanics who are folded in w/whites…it does not. Read the report)..

23/ So that’s a 2.7 to 1 ratio. Why does this matter? Obvious…If one group has a general offending rate 2.7x higher than another group, we should expect the first group to victimize members of the 2nd group more often than the reverse, even w/o deliberate racial targeting...

24/ Adjusting the 31:1 BoW to WoB ratio for general offending rates alone brings the disparity down to 11.5 to 1. In short it cuts it by almost two thirds...A 2nd factor that will explain part of the remaining disproportion will be relative population sizes.

25/ If 1 group is more plentiful in population than 2nd group, members of group 1 are more available as victims of offenders in group 2 than vice versa...(MORE ON THIS IN A FEW MINUTES...TWITTER'S MAKING ME SEND THESE 25 AS A BATCH...) gimme a sec (in the air right now)...

26/ In this case, black offenders (already more plentiful per capita) will have more potential white victims than white offenders (less plentiful per capita) will have potential black victims....

27/ In short, any black person inclined to commit a violent crime will have an easier time finding a white victim than vice-versa...

28/ During the period under review, there were 5.3x as many whites as blacks in the US, meaning the 11.5: 1 ratio of BoW to WoB offending has 2b adjusted for the factor of population availability...

29/ This adjustment would bring the unexplained disproportion down to only 2.2 to 1. Other adjustments would bring it down even further...

30/ For instance, robbery (motivated by pecuniary interest & rarely about ‘racial targeting’) is hugely lopsided, BoW as opposed to WoB. Of course, bc if one is looking to rob someone, given economic disparity, BoW robbery makes more pecuniary logic than the reverse....

31/ The major categories of crime (numerically) like assault tend to be more even...So if we removed robbery from the equation due to its non-racial motivation (or merely adjusted rates to account for this important difference) the 2.2 to 1 ratio would likely shrink even more...

32/ But for the sake of argument, forget that one. Let’s go back to the population issue, and the matter of victim availability...

33/ Above, I was merely adjusting the interracial crime disparities to account for aggregate population differences (i.e. the 5.3 to 1 greater availability of white victims, relative to black ones). ..

34/ But the adjustment needs to be greater than that. Bc whites/blacks aren't evenly spread throughout the population, such that the 5.3: 1 population ratio is the relevant number. The real issue is: what are odds of a given white person encountering black person, or vice versa?

35/ ...…and thereby having a realistic opportunity to victimize them? Sadly not much research has been done on this lately. But in the 80s, a scholar at the U of Oregon crunched the numbers. jstor.org/stable/2780040…...

36/ What he discovered was that bc of spatial isolation – especially white isolation from blacks – the odds of a white person encountering a black person were only about 3%. That’s another way of saying that about 3% of people whites encounter on avg are black...

37/ (and this was in Metro statistical areas: it would be even lower in small towns, rural areas, etc)...Black folks are not as isolated from us however....

38/ Bc whites often work in, commute through, or recreate in urban areas where POC live or around where they live, black folk encounter us at a higher rate. The data shows the odds of a black person encountering a white person (thus having opportunity to victimize them) is 57%...

39/ This means black folks are 19x more likely to encounter whites than the reverse. So now Nazis ”has math problem…”

40/ Rather than adjusting the interracial crime disproportion to account for the 5.3x greater availability of whites as possible victims – which brought the gap down from 11.5:1 to 2.2: 1, you have to adjust it for the 19:1 gap in interracial encounters…

41/ Bc whites are not merely 5.3x more available as victims of black folks, but 19x more available. So having already reduced the interracial victimization disparity from 31:1 down to 11.5:1 (by adjusting for general crime rates), we must further reduce it...

42/ by controlling for the 19:1 encounter rate differences...And when u make that adjustment, which you can do by multiplying the 2.7: 1 general offending gaps by 19, the predicted ratio of BoW to WoB offenses—based solely on random chance—would be 51: 1, rather than 31: 1...

43/ In other words, BoW offending relative to WoB offending is 40% below what random chance would predict…There is no widespread racial targeting going on. Period.

44/ Of course there are other problems w/the racist argument. Using black crime data to scare whites also ignores that we are 4-5x more likely to be victimized by other whites than by blacks or Latinx folk...

45/ But the bottom line is, those who use data to prove black criminality or a crisis of black folks targeting white folks are liars, fools or both...

46/ And I now invite them to pick up my damn mic off the floor…you’re welcome. END