ObserveandReport Posted April 6, 2016 Posted April 6, 2016 The topic came up in one of my classes. I assembled some stats and reflected on them. Love to see what you make of the work. Race, Violence, and Other Lighthearted Topics So I thought a lengthy post about everyone’s two favorite things, race and math, would be a great way to alienate friends. All kidding aside, I realize the topic of race as it relates to crime is a contentious one. All the more reason it should be evaluated with rigorous objectivity. Before we dive into the numbers, a caveat from Wikipedia 4 about a source I rely upon, the Nation Crime Victimization Survey: “NCVS statistics do not represent verified or evidenced instances of victimization. As it depends upon the recollection of the individuals surveyed, the NCVS cannot distinguish between true and fabricated claims of victimization, nor can it verify the truth of the severity of the reported incidents. Further, the NCVS cannot detect cases of victimization where the victim is too traumatized to report. These factors can contribute to deficits in the reliability of NCVS statistics.” Further… “The NCVS program is focused upon metropolitan and urban areas, and does not adequately cover suburban and rural regions. This can lead to misrepresentations regarding the nature and extent of victimization in the United States.” That being said I’ve chosen to rely on this source because unlike FBI Uniform Crime Reports (which I reference for other reasons later), the NCVS is based on surveys which may include unreported crimes and be more representative of crime statistics on the whole. 5 What The Numbers Say Now when it comes to the breakdown of who is hurting who, there are two major issues. First, I only found extensive tables from 2004. Just to give you an idea, the U.S population was 293 million in 2004 with approximately 12.8% African-American, 14.1% Hispanic/Latino, and 80.4% white. 6 The second problem is that at this time, the NCVS didn’t set out a category for Hispanics or Latinos in the data below, so numbers may be skewed somewhat, again. On to the numbers (I know your chomping at the bit). 52% of all violent crimes against whites were committed by whites, whereas a mere 9.8% of the total violent crimes against whites were committed by blacks. Both blacks and whites committed crimes at levels lower than their perspective population percentages. Conversely whites committed 10.8% of violent crimes against blacks, whereas blacks committed 73.6% of all violent crimes committed against blacks 7 (see table 42). This reflects the oft repeated “black on black” talking point. That is, blacks are victimized by other blacks at a far greater rate than by any other demographic. But what about as a whole? Well about 17.7% of all violent crimes are committed by blacks (see tables 42 and 43) 8 , which is disproportionate, but while not statistically insignificant how this 4.9% overrepresentation translates to the average person’s risk assessment is far from clear. This is shown by the previous numbers, showing white’s to be more likely to be attacked by whites and blacks by blacks. Ought one focus more on their own race? Perhaps. Consider that when it comes to what kills us (other than disease and other natural causes) the usual culprit is medical errors, not violent crimes. 9 10 A good question is, how could you effectuate a 1.4 increase of normal attention paid to any individual racial demographic? Is such a thing plausible, or cost effective as far as time spent versus resulting safety and other means of threat prevention? How could these questions be measured? These aren’t rhetorical questions… any sociology majors feel free to chime in. Does this mean that there is not a problem of where violence is occurring across racial lines? No. Unfortunately, a disproportionate (not a majority or anything close) amount of violence is occurring in black communities, the root cause of which is probably an entire field of sociology unto itself. The Criminal Justice Response While to many of you this may not be surprising, it bears substantiation: Blacks and minorities are overrepresented in arrests and charges filed, even when compared with the actual rate of committed crime. For instance when it comes to non-negligent homicide blacks were charged at a rate of 27.5% of the total charges filed (see table 43). 11 This effect is not limited to violent crime however. When it comes to drugs, whites, blacks, and Hispanics use drugs at about the same rates, yet 20% of inmates are Hispanic and 45% are black. 12 If one were to look at the actions of our officers and DA’s, it would be easy to think that minority communities pose a greater threat to the peace, but the actual rates of crime show a far murkier picture of crime and those who commit it. Final Thoughts Personally, I don’t think crafting risk assessments on racial lines is effective for the average person. Highly trained professionals may be able to use these methods to great effect, 13 14 but in general humans are bad at dealing with numbers15 on the fly (not to mention high-stress situations tossing emotion into the mix). Also, not mentioned here is the disproportionate representation of minorities below the poverty line. I’ll look for my sources if any are interested, but from what I remember it correlates closely with the arrest records, which is suspicious to say the least (and raises a host of other questions.) Here's a quick reference point: http://www.npc.umich.edu/poverty/ We should also remember that we suck at empathizing with others, especially as the numbers go up. 15 Next time you hear your uncle talking about how the (insert racial group here) is ruining America, you can bore them to pieces with all the information contained here. To quote the late, great Christopher Hitchens, “Evolution has meant that our prefrontal lobes are too small, our adrenal glands are too big, and our reproductive organs apparently designed by committee; a recipe which, alone or in combination, is very certain to lead to some unhappiness and disorder.” Sources 1. http://www.crimeinamerica.net/2010/12/13/what-are-my-chances-of-being-a-victim-of-violent-crime/ 2. http://time.com/4117442/us-terror-attacks-infrastructure-paris/ 3. https://www.fbi.gov/news/stories/2014/september/fbi-releases-study-on-active-shooter-incidents/pdfs/a-study-of-active-shooter-incidents-in-the-u.s.-between-2000-and-2013 4. https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States#Victim_Surveys 5. http://www.bjs.gov/content/pub/pdf/ntcm_2014.pdf 6. http://www.census.gov/population/pop-profile/dynamic/RACEHO.pdf 7. http://www.bjs.gov/content/pub/pdf/cvus0402.pdf 8. http://www.bjs.gov/content/pub/pdf/cvus0402.pdf 9. https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2011/crime-in-the-u.s.-2011/violent-crime/murder 10. http://journals.lww.com/journalpatientsafety/Fulltext/2013/09000/A_New,_Evidence_based_Estimate_of_Patient_Harms.2.aspx 11. https://www2.fbi.gov/ucr/cius_04/documents/CIUS2004.pdf 12. http://www.politifact.com/truth-o-meter/statements/2009/jul/28/arianna-huffington/arianna-huffington-calls-drug-imprisonment-rates-b/ 13. https://www.samharris.org/blog/item/in-defense-of-profiling 14. http://www.huffingtonpost.com/daniel-wagner/what-israeli-airport-secu_b_4978149.html 15. https://www.youtube.com/watch?v=Y0Z0raWIHXk 3 1
labmath2 Posted April 7, 2016 Posted April 7, 2016 The problem with facts that do not show causation is that you cannot tell exactly what they mean. Even events like the Trump manager grabbing a reporter is so muddled with conflicting data and the only conclusive evidence is a discontinuous video that makes it hard to tell (for me) if the reporter actually grabbed Trump. Now think of the data about race and crime. How much crime do each race actually commit? How do they commit the crime? Then there is what counts as crime and how it is investigated and prosecuted (this changes by location). Who is more likely to be prosecuted? This includes all other mitigating circumstances that go with those like false conviction or false accusation (some think false conviction rarely happens, but given that overturning a conviction seems to be harder than getting one, i'll let you be the judge). Based on the current facts, it does appear blacks commit more violent crime.
ObserveandReport Posted April 7, 2016 Author Posted April 7, 2016 Based on the current facts, it does appear blacks commit more violent crime. Indeed they do! About 1.4 times their percentage of the population. This is a far cry from what the FBI UCR reports say, where blacks commit crimes at 3 or 4 times the rate of whites. This number is far from alarming, though less so for the reasons you mentioned and then EVEN less so when you factor in the notion that the NCVS is done in urban settings, automatically over representing blacks. That being said, the NCVS is probably the best thing we have due to the uniformity of how crime is defined in all the jurisdictions where it is distributed. More than that, the sample is quite likely more inclusive than UCR because not only would people who reported crimes answer, but also many people who are considered "dark horse crime" victims. 2 1
ObserveandReport Posted April 7, 2016 Author Posted April 7, 2016 Why do I have little "-1" on each of my comments? Both of my statements are well sourced conclusions made in good faith. Curious.
shirgall Posted April 7, 2016 Posted April 7, 2016 Why do I have little "-1" on each of my comments? Both of my statements are well sourced conclusions made in good faith. Curious. Ignore the popularity contest. As for the NCVS, as I posted in another thread: Possible problems with the National Crime Victimization Survey and other Census Bureau surveys? Quote During a two year investigation of the Census Bureau my whistleblowers and I have: — Caused the Justice Department to look into whether crime data is being falsified and whether the budget for that survey is being misused. That investigation is still ongoing. — Found one verified, massive case of data falsification in the Philadelphia region of Census that could have changed national economic statistics. This resulted in an Inspector General and Congressional investigations. http://nypost.com/20...-vet-weighs-in/
ObserveandReport Posted April 8, 2016 Author Posted April 8, 2016 Ignore the popularity contest. As for the NCVS, as I posted in another thread: Possible problems with the National Crime Victimization Survey and other Census Bureau surveys? Quote http://nypost.com/20...-vet-weighs-in/ Then do we ignore statistics linking race and crime? The FBI UCR are just as plagued by problems.
shirgall Posted April 8, 2016 Posted April 8, 2016 Then do we ignore statistics linking race and crime? The FBI UCR are just as plagued by problems. We don't ignore them, but we have to look for alternative sources data if we can. I'll take the NCVS over the SPLC any day, for example.
Guest Gee Posted April 8, 2016 Posted April 8, 2016 You've acknowledged that there exists no category for hispanics and latinos, that they're lumped in with whites and this will skew the results, but then you've talked exclusively about white crime? Didn't say, I don't know, write the category as white, latino and hispanic? Why did you do that? You've put alot of care and attention into this, even referencing it and that is great, so why?
ObserveandReport Posted April 11, 2016 Author Posted April 11, 2016 You've acknowledged that there exists no category for hispanics and latinos, that they're lumped in with whites and this will skew the results, but then you've talked exclusively about white crime? Didn't say, I don't know, write the category as white, latino and hispanic? Why did you do that? You've put alot of care and attention into this, even referencing it and that is great, so why? To be frank, I'm not sure if they are actually lumped in. I didn't see anything in the data set saying so. However Wikipedia mentions that the UCR have this weakness. I assumed it was a possibility because of the UCR having this error and so put it as a caveat. Ill see what I can scrounge up.
Guest Gee Posted April 11, 2016 Posted April 11, 2016 To be frank, I'm not sure if they are actually lumped in. I didn't see anything in the data set saying so. However Wikipedia mentions that the UCR have this weakness. I assumed it was a possibility because of the UCR having this error and so put it as a caveat. Ill see what I can scrounge up. Both the UCR and the NCVS surveys fail to separate Hispanics and Whites thus White crime figures include all Hispanic crime figures. Of course this misleading inflates the statistics for "White" crime. Once you've confirmed your mistake are you going to apologise for insinuating that my white uncle is a hysterical racist?
ObserveandReport Posted April 11, 2016 Author Posted April 11, 2016 Both the UCR and the NCVS surveys fail to separate Hispanics and Whites thus White crime figures include all Hispanic crime figures. Of course this misleading inflates the statistics for "White" crime. Once you've confirmed your mistake are you going to apologise for insinuating that my white uncle is a hysterical racist? I'll be happy to admit my error when you provide the source confirming your assertion. I'll even thank you for correcting me and doing the research that I struggled to do. I can't apologize however, because my statistical error doesn't make your uncle not a racist. That burden is much higher, as we know incidence of racism in uncles is several standard deviations from the mean. As a quick follow up, the NCVS did record victim data about Hispanics vs. non-Hispanics prior to 2004: http://bjs.gov/content/pub/pdf/hvvc00.pdfThis doesn't prove that offender data did not lump Hispanic whites and non-Hispanic whites together. However, it is also more than possible that the main source of my original post was merely a reflection of black and non-Hispanic white crime. Let my know if you all find something definitive.
Guest Gee Posted April 22, 2016 Posted April 22, 2016 I'll be happy to admit my error when you provide the source confirming your assertion. I'll even thank you for correcting me and doing the research that I struggled to do. I can't apologize however, because my statistical error doesn't make your uncle not a racist. That burden is much higher, as we know incidence of racism in uncles is several standard deviations from the mean. As a quick follow up, the NCVS did record victim data about Hispanics vs. non-Hispanics prior to 2004: http://bjs.gov/content/pub/pdf/hvvc00.pdfThis doesn't prove that offender data did not lump Hispanic whites and non-Hispanic whites together. However, it is also more than possible that the main source of my original post was merely a reflection of black and non-Hispanic white crime. Let my know if you all find something definitive. I've forgotten the acronyms but the FBI one counts hispanics as white and there is a high degree to correspondence between the FBI one and the other one which logically implies that the other one must count a great deal of hispanics as whites. Classification of HispanicsThe UCR classifies most Hispanics into the "white" category. The NCVS classifies some Hispanic criminals as "white" and some as "other race". The victim categories for the NCVS are more distinct. According to a report by the National Council of La Raza, research obstacles undermine the census of Latinos in prison, and “Latinos in the criminal justice system are seriously undercounted. The true extent of the overrepresentation of Latinos in the system probably is significantly greater than researchers have been able to document. Lack of empirical data on Latinos is partially due to prisons’ failures to document race at intake, or recording practices that historically have classified Latinos as white.[18] Overall the FBI didn't include a 'Latino' or 'Hispanic' category until recently and 93% of Hispanics are classified as "white" by law enforcement officers (irrespective of their ancestry) often inflating actual white criminal statistics.[19][20] https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States You can find the links on the page. But the uncle thing....... Next time you hear your uncle talking about how the (insert racial group here) is ruining America, you can bore them to pieces with all the information contained here. So if I accept your words, you've been talking about Blacks and Whites. And you're making a case that Black crime statistics do not reflect reality. So the uncle is white. And the hysterical because his reaction to Black crime is disproportionate to reality. And racist because in the reaction is implied a racial bias against Blacks. Are you going to apologies for crafting an argument whose only conclusion is that white people are racist? Because I'm white, and I don't like it.
ObserveandReport Posted April 22, 2016 Author Posted April 22, 2016 I've forgotten the acronyms but the FBI one counts hispanics as white and there is a high degree to correspondence between the FBI one and the other one which logically implies that the other one must count a great deal of hispanics as whites. https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States You can find the links on the page. Thank you for restating the same thing I said except without adding evidence and drawing an inference whose likelihood is unknown. Cantelope and honeydew have a high degree of correspondence, but they taste quite different. I can see why one might draw that inference, but it isn't certain, nor is the likelihood quantifiable. Thus, I left it out of the equation. I've forgotten the acronyms but the FBI one counts hispanics as white and there is a high degree to correspondence between the FBI one and the other one which logically implies that the other one must count a great deal of hispanics as whites. https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States You can find the links on the page. But the uncle thing....... So if I accept your words, you've been talking about Blacks and Whites. And you're making a case that Black crime statistics do not reflect reality. So the uncle is white. And the hysterical because his reaction to Black crime is disproportionate to reality. And racist because in the reaction is implied a racial bias against Blacks. Are you going to apologies for crafting an argument whose only conclusion is that white people are racist? Because I'm white, and I don't like it. You are humorless. There are plenty of racist people of all colors. White racism against blacks was the specific target of the article which you so astutely grasped. The comment was a self-deprecating nod to the fact that everyone's "got that crazy uncle." It's a common trope especially brought up by comedians involving the topic of Thanksgiving. Thank you for applying rigor to sufficiently stifle all humor and induce this long explanation. If what you grasped was "all white people are racist" you are as sensitive and loose with logical inferences (and quite frankly dishonest) as any social justice warrior. I genuinely dislike your style of argument. If you want to have the fight "ALL whites are racist" "NO not ALL whites are racist!" please go elsewhere. What I presented was a well sourced argument with some conversational tones at times.I think you are either arguing in bad faith or simply trying to white knight for whites (white-white knighting?). You have the right to be offended and I have the right to give offense. Grow thicker skin or make me believe what I said would be hurtful to a reasonable person. You have clearly done neither.
ObserveandReport Posted April 22, 2016 Author Posted April 22, 2016 My longer response will post soon...but note that I asked for "definitive" evidence. Either you deliberately ignored that, or don't understand the word. In response, I have taken it upon myself to actually e-mail the BJS.
ObserveandReport Posted April 22, 2016 Author Posted April 22, 2016 Update: The response email was this "The NCVS has been collecting data on both race and ethnicity since its inception in the 1970s. You can use our online data analysis tool to examine crime rates for non-Hispanic whites and Hispanics going back to 1993 (http://www.bjs.gov/index.cfm?ty=nvat). Best, Lynn Langton, Ph.D. Chief, Victimization Statistics So you were wrong in your conjecture. Update: before 2012 there was no specificity to the offender race data. You were right. Thanks for motivating me to do the work. That may be inflating the black data as well...so it's not as though either conclusions are accurate. I am searching for the post 2012 data. The other thing was a joke.
ObserveandReport Posted April 23, 2016 Author Posted April 23, 2016 Update: I was wrong. I emailed the PhD. back about race on race variable statistics, and she confirmed that before 2012 the only questions asked were "white" "black" or "other" when it came to the race of the offender. I was wrong in my assessment of your conjecture. I'm going to try and get the post 2012 offender stats. That being said, I take it the vast majority of people who comment on race, which includes many racists of various backgrounds, haven't done a modicum of research into the matter. That was my point with the joke. It is nearly inconsequential to the conclusions made in the OP. I've forgotten the acronyms but the FBI one counts hispanics as white and there is a high degree to correspondence between the FBI one and the other one which logically implies that the other one must count a great deal of hispanics as whites. https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States You can find the links on the page. But the uncle thing....... So if I accept your words, you've been talking about Blacks and Whites. And you're making a case that Black crime statistics do not reflect reality. So the uncle is white. And the hysterical because his reaction to Black crime is disproportionate to reality. And racist because in the reaction is implied a racial bias against Blacks. Are you going to apologies for crafting an argument whose only conclusion is that white people are racist? Because I'm white, and I don't like it. So there you go. My conclusion was that some people were racist, not all. If your assessment of a race is not based on facts, but rather the preconceived notions you have of that race, you're by definition a racist. I have no opinion as to whether or not you are a racist, other than most white people are not racists and you are most likely to be in that majority. How you gathered otherwise is a mystery, but thank you for harping on my statistical error. I will rectify it at the earliest possible time. I actually owe you a bit more than that because you forced me to do the research I was too lazy to do. Thanks. Oh, and I still won't apologize about your uncle because that was a joke.
Recommended Posts