Last Updated on April 16, 2023
The purpose of this post is to cite the extent of black crime around the world. The first section is concerned with black crime in the United States. This section should not be terribly surprising to most people since I assume (hope) most people are aware that black people are disproportionately engaged in crime in the United States. I believe the latter two sections contain information that most people are not aware of. The second section covers misconduct among black children in the United States. The third and final section covers black crime and misconduct outside of the United States. The last two sections, especially the last section, will probably be educational for the vast majority of people reading this post. If not, the post at the very least serves as a useful fact sheet for those interested in this topic.
Black crime in the United States
The Bureau of Justice Statistics (BJS) has recently published a report “Prisoners in 2018” [archived] which shows that prisons are disproportionately composed of black prisoners:
- Despite being only 13% of the general population, blacks comprise 32% of the prison population (410k out of 1.3M, see Table 14).
- The imprisonment rate for black adults is over 5 times the rate for non-Hispanic white adults (see Table 6).
- The imprisonment rate for black 20-24 year-olds is 8 times the rate for their non-Hispanic white counterparts (Figure 2).
- The imprisonment rate for black 18-19 year-olds is over 12 times the rate for their non-Hispanic white counterparts (Figure 2).
- There are 75,700 blacks imprisoned for homicide (69,500 in state prisons, table 14; 6,200 in federal prisons, table 16) whereas there are only 45,200 non-Hispanic whites imprisoned for homicide (42,100 in state prisons, table 14; 3,100 in federal prisons, table 16). Thus, there are about 1.4 times as many blacks imprisoned for homicide as there are whites (74,700 vs 45,200). Since there about 4.6 times as many non-Hispanic whites as there are blacks (60% vs 13% of the population according to the U.S. Census [archived]), this suggests that blacks are about 6.4 times as likely as non-Hispanic whites to be incarcerated for homicide.
Many people believe that black disproportionate incarceration is largely the result of drug-related offenses. The data shows that this is not the case. The BJS report cited earlier shows that the vast majority of imprisoned blacks are not imprisoned due to drug-related offenses:
- The vast majority of black state prisoners are not imprisoned for drug-related crimes. The majority (61%) are imprisoned for violent crimes. Only 13% of black prisoners are imprisoned for drug-related offenses (including e.g. trafficking), and only 3.2% are imprisoned for drug possession. For comparison, 4.5% of white prisoners are imprisoned for drug possession (Table 13). Note that state prisons contain the vast majority (88%) of all prisoners (page 3).
- About half (45%) of black federal prisoners are imprisoned for drug-related crimes (Table 15) (federal prisoners constitute only 12% of all prisoners). But almost none of them were imprisoned for drug possession; 99% of federal drug offenders were sentenced for trafficking (see note g of Table 15).
Someone might object by saying that this data merely reports black imprisonment rather than black crime. This objection can be dismissed by using arrest statistics [archived] published by the FBI. The data shows that black people were arrested for a disproportionate amount of crime in 2018:
- Blacks commit 37% of violent crime (despite being 13% of the general population).
- Blacks commit 53% of murders and 54% of robberies.
- Among criminals under the age of 18, blacks commit over 58% of murders, 64% of robberies, and over 48% of violent crime.
- Blacks commit 20% more homicides than whites (including Hispanic whites) (4,778 vs 3,953). Since there are about 6 times as many whites as there are blacks (70% vs 13% of the population according to the U.S. Census [archived]), this implies that the black homicide rate is 6-7 times the white homicide rate.
Note that one problem with the FBI data is that it groups both non-Hispanic whites and Hispanic whites into the “white” category. If we had data to compare the black homicide rate to the non-Hispanic homicide rate, the racial disparity would be even greater. For example, Harris (2009) reported that, in Pennsylvania, the black homicide arrest rate was over 20 times the non-Hispanic white rate in Pennsylvania (Table 1). The racial disparities in crime are so great that, if every racial group constituted an equal share of the population, blacks would constitute 67% of arrests for homicides, Hispanics 30%, and non-Hispanic whites only 3% (Table 1).
To illustrate just how stark these disparities are, compare racial disparities in homicide to sex disparities in homicide using data [archived] published by the FBI. Among people aged 18-24 in 2005, the homicide offending rate for black males (203.3 per 100k pop) was about 9 times the rate for white males (22.4 per 100k pop) (page 79). For comparison, the homicide offending rate for white males of this age was about 11 times the rate for white females (2.0 per 100k) of the sage age. In other words, among 18-24 year olds, the magnitude of the homicide offending disparity between black and white males is comparable to the magnitude of the same disparity between male and female whites. Note, again, that these statistics include Hispanic whites under the white category; the racial disparity would be even larger if we compared blacks to non-Hispanic whites specifically.
Another point worth mentioning is that racial arrest statistics mostly mirror racial incarceration statistics. Beck and Blumstein (2017) analyzed FBI data on arrest suspects and BJC data on prisoners to estimate the degree to which disproportionate incarceration of black was the result of disproportionate arrests of black people. More specifically, the goal was to estimate the percentage of the black-white incarceration disparity that can be explained by the black-white arrest disparity. The researchers found that “approximately 70 to 75% of the racial disproportionality among state prisoners is accounted for by black-white differences in arrests” (page 11). In other words, the vast majority of black overrepresentation in prisons is the result of black overrepresentation in arrests, although the researchers do find that prison racial disparities are less accounted for by arrests for certain crimes (e.g., drug crimes) and certain regions (northeast). The researchers note that possible factors causing the unexplained 25-30% may be racial differences in criminal history, racial differences in the ability of defendants to hire effective counsel, and possibly bias by prosecutors, judges, etc. (page 11).
A similar earlier analysis by Tonry and Melewski (2008) posited that unexplained residual is primarily due to sentencing laws and policies that “specify punishments that are both absolutely and relatively severe for violent, drug, and gun crimes for which blacks are more likely than whites to be arrested and prosecuted” (page 31).
Murder arrests vs commissions
Someone might object that the FBI data merely reports high rates of black arrests, but they don’t show high rates of black crime. There are two possible ways to have high black arrest rates without corresponding high black crime rates. We have data to investigate these possibilities for homicides in particular. Neither of these possibilities is supported by the data on homicides:
- One possibility is that innocent black people are more likely to be arrested for homicide than innocent non-black people. To test this possibility, we can consider murder exonerations – cases in which a person was wrongly convicted of murder and later cleared of charges based on new evidence of innocence. This can help inform the degree to which murder convictions reflect murder commissions. While murder convictions are not identical to murder arrests, they are a close proxy. If we look at murder exonerations by race [archived], we find that blacks do not seem to be any more likely to be exonerated than, say, whites. The link shows that blacks constitute 551/1060 of people exonerated for homicide (as of August 3rd, 2020). So black people account for 52% of those who are exonerated for homicide. This is essentially equal to the percentage of black people who are arrested for homicide. This suggests that wrongful convictions do not play a significant role in explaining racial disparities in arrests/convictions for homicide.
- Another possibility is that black murderers are more likely to be arrested than guilty non-black murderers. However, evidence against this is the fact that homicides of black victims are more likely to be unsolved than homicides of non-black victims. An analysis by the Washington Post finds that an arrest was made in 63% of the killings of white victims, compared to only 46% of the killings of black victims. If anything, this suggests that racial disparities in homicide arrest rates may actually underestimate racial disparities in homicide commission rates.
The hypothesis that racial disparities in arrests do not reflect racial disparities in crime can be further dismissed by reviewing data on rates of crime victimizations rather than commissions.
One source of data for crime victimizations is the BJS’s National Crime Victimization Survey (NCVS), which is a survey that interviews individuals who were victims of crime. Data is limited to violent crimes of rape and sexual assault, robbery, aggravated assault, and simple assault. Beck and Blumstein (2017) performed an analysis of NCVS data among victims who identified the race of their assailant. They found that the share of assailants perceived to be black closely mirrored the share of arrestees that were black. Specifically, the ratio of the percent black among arrestees to the percent black among offenders for the aforementioned violent crimes was 0.97 (page 23), suggesting that the percent black among arrestees is nearly identical to the percent black among assailants.
For homicide victimization data, we have to use other sources. Because homicide victims are dead, we cannot rely on their perceptions to determine the race of the assailant. Instead, we can look at the race of the victims of homicide. This should give us information on the race of assailants because the vast majority of violent crime is intraracial rather than interracial. For example, according to data [archived] from the FBI, the vast majority (88%) of homicides against black victims are committed by black offenders. Therefore, rates of homicide victimizations of blacks should reflect rates of homicide offenses by blacks.
There are two main sources [archived] of data to measure homicide victimization rates by race – the Bureau of Justice Statistics and the CDC. Firstly, according to data [archived] published by the Bureau of Justice Statistics, blacks are disproportionately victims of homicides to a similar magnitude that they are disproportionately arrested for homicide. Specifically, despite making up just 12-13% of the population between 1980 and 2008, blacks constituted 47% of homicide victims, which is unsurprising given that blacks also constituted 52% of homicide offenders (Table 1).
Secondly, the CDC (2018) [archived] reported the leading cause of death for people broken out by race and age in 2016. The results show a homicide epidemic among young black males. For comparison, here are the leading causes of deaths for young white males (page 27):
And here are the leading causes of deaths for young black males (page 34):
A number of interesting findings stick out from this data:
- Homicide is the leading cause of death for black males aged 15-19, aged 20-24, and aged 25-35 (see page 34).
- Homicide is the cause of about half of the deaths black males aged 15-19 and black males aged 20-24. For comparison, homicide was the cause of only 3-5% of the deaths of non-Hispanic white males of this age.
- The homicide death rate for black males aged 20-24 (110.8 deaths per 100,000 population) is over 20 times the rate for similarly aged non-Hispanic white males (5.4 deaths per 100,000 population).
The CDC has also recently released a useful data tool that allows users to request more specific data on homicide deaths. I requested data on the number of homicides by race and ethnicity from 1999 to 2020. The data corroborates earlier findings, which is that there are far more homicide deaths for non-Hispanic blacks than for non-Hispanic whites:
This is the query criteria used to generate this table:
As you can see, the homicide death rate for non-Hispanic blacks (21.3 per 100k) is about 7.6 times greater than the homicide death rate for non-Hispanic whites (2.8 per 100k). To put this racial disparity into perspective, let’s consider the sex disparity in homicide death rate using the same data tool:
This is the query used to generate this table:
As you can see, the homicide death rate for males (9.4 per 100k) is about 3.9 times greater than the homicide death rate for females (2.5 per 100k). When analyzing the sex disparity within the races, the male:female homicide death ratio is 2.7 among whites, 6.1 among blacks, 2.3 among Asians, and 3.1 among Native Americans. Note two surprising findings from this data:
- The gap in homicides between blacks and whites is greater than the gap in homicides between males and females. In fact, the black to white homicide ratio (7.6) is about twice as large a the male to female homicide ratio (3.9).
- The homicide death rate for black females (5.9 per 100k) is greater than the homicide death rate for white males (5.2 per 100k) and Asian males (3.0 per 100k). Note also that the white male figure includes Hispanic males. The gap between black females and white males would be even greater if we only considered non-Hispanic white males.
The racial gaps in homicide deaths have been present since the collection of this data. For example, the following chart [archived] shows drastic racial differences in CDC-reported homicide death rates for persons aged 10-24 from 1990 to 2010:
See here [archived] for data on CDC homicide death rates that extend back to 1950. Even in 1950, the homicide death rate for blacks was several times the rate for whites. Here are some facts about homicide death rates in 1950:
- The age-adjusted homicide death rate for black males (47.0 per 100k) was about 12 times greater than the rate for white males (3.8 per 100k).
- The age-adjusted homicide death rate for black females (11.1 per 100k) was about 8 times greater than the rate for white females (1.4 per 100k). In fact, the homicide death rate for black females was 3 times greater than the rate for white males.
- The homicide death rate for black 25-34 year-old males (104.3 per 100k) was about 21 times greater than the rate for their white male counterparts (4.9 per 100k).
The most murderous U.S. cities
The racial disparities in crime are so great that cities with the largest percentage of black people tend to be the most dangerous cities. Conversely, the most dangerous cities tend to be the cities with the largest percentage of black people.
For example, consider the ten cities [archived] (Table 7) with the highest percentage of black people among cities with populations over 100,000 in 2010:
|Rank||City||Percent black||Homicide Rate||Population|
|3||Miami Gardens, Florida||78%||20.9||107,167|
|7||New Orleans, Louisiana||61%||37.1||343,829|
Homicide rates were calculated by using the 2018 FBI homicide data published here to determine the number of homicides per 100,000 population. Each of these cities have considerably high homicide rates. To see this, consider the following points:
- Data [archived] from the FBI indicates that the homicide rate for the entire country is 5 homicides per 100,000 population (see Overview section). The homicide rate for whites is around 3 homicides per 100,000 population, and the homicide rate for blacks is around 20 homicides per 100,000.
- Consider the major cities (population over 250k). The major cities from the above list are Detroit, Baltimore, Memphis, and New Orleans. As of 2019 [archived], these cities were 4 of the 5 most homicidal major cities in the country. The other city in the top 5 was St. Louis, Missouri, which has a black plurality of 47%.
- Now consider mid-sized cities (population over 100k and under 250k). The mid-sized cities from the above list are Jackson, Miami Gardens, Birmingham, Flint, Montgomery, and Savannah. As of 2012 [archived], these 6 cities were the 1st, 2nd, 3rd, 6th, 8th and 37th most homicidal mid-sized cities.
Now consider the 10 major cities [archived] (population over 250k) with the highest homicide rates in the country. Using demographic data from here [archived] and here [archived], one finds that these cities are disproportionately black to a significant degree:
- 4 of the 10 most murderous cities have black majorities (Baltimore, Detroit, New Orleans, and Memphis).
- 5 of the 10 most murderous cities have black pluralities and are at least 40% black (St. Louis, Newark, Washington D.C., Cleveland, and Philadelphia).
- The only city in the top 10 that did not have a black plurality was Kansas City, which had a relatively high black population of 29%.
- Keep in mind that only 13% of the United States population is black, and majority-black major cities are rather rare. In 2000, there were less than 20 cities with populations over 100k that were majority black.
In summary, blacks have disproportionate rates of imprisonment, arrests, deaths due to homicides, and crime. The racial disparity in crime is particularly striking for the most serious crimes, such as violent crimes and homicides.
The ubiquity of black crime
The previous section demonstrate that the crime rate for black Americans is much greater than the rate for white Americans. This section will show that this disparity is not isolated to particular contexts. I’ll focus on homicide since this is least likely to be influenced by reporting bias or differences in data coding. No matter how we slice the data, we consistently find higher rates of homicide for black Americans. Black Americans have higher rates of homicide for every state, every circumstance (e.g., gang related vs workplace homicides), every victim/offender relationship, every method of killing, etc.
Homicides by state
The extreme levels of black crime are not isolated to particular cities or regions of the country. Rampant black crime is found in every area of the country. A 2019 study [archived] by the Violence Policy Center analyzed homicides during 2016 and reported that the black homicide rate was extremely high in every state in the country. The study reports that the black homicide rate was 20.44 per 100,000, whereas the national homicide rate was 5.10 per 100,000 and the white homicide rate was 2.96 per 100,000. The black homicide rate in every state with more than 10 black homicide victims was over 3 times the average white homicide rate (see Appendix two: Connecticut had a homicide rate of 9.27 per 100,000, which was the lowest black homicide rate of all states with more than 10 black homicide victims).
Similar findings were reported by 2018 study [archived] relying on CDC data. In every state, the homicide death rate for blacks was at least 3 times that of whites, except for Hawaii where there was actually a slightly lower homicide rate for blacks. The authors note this disparity when describing the results of the study:
During the study period, the average ratio of Black to White firearm homicide rates across the 42 states in our sample was 8.7, and this ratio was greater than 3.2 in every state except Hawaii (Table 1). However, there was a substantial between-state variation in the Black-White disparity in firearm homicide rates, with a high of 21.2 in Wisconsin and a low of 0.8 in Hawaii. The five states with the highest Black-White firearm homicide rate ratio were Wisconsin (21.2), Minnesota (19.5), Michigan (19.3), Nebraska (18.8), and Pennsylvania (16.2). The five states with the lowest ratio were Hawaii (0.8), Arizona (3.2), Texas (3.6), New Mexico (3.8), and South Carolina (4.1).
Homicides by type
Furthermore, the extreme levels of black crime are not limited to particular types of homicide. The Bureau of Justice Statistics released report [archived] to describe homicide trends up until 2005. An updated version of this report was released [archived] showing homicide trends up until 2008. The data in the most recent report showed the percentage of homicide offenders by race broken out by circumstance (Table 7). The circumstances were “workplace”, “argument”, “gang related”, “drug related”, “sex related”, and “felony murder”. Black people were disproportionately associated with homicide offending for every circumstance. The data also showed the percentage of homicide offenders by race broken out by victim/offender relationship. The relationships were “intimate”, “family”, “infanticide”, and “eldercide”. Again, black people were disproportionately associated with homicide offending for every victim/offender relationship.
Homicides by method
Using the same CDC data tool mentioned earlier, we can review the number of homicides by method of killing (e.g., firearm vs cut/pierce). This is the number of homicides for non-Hispanic whites and blacks by method of killing from 1999 to 2020:
This is the query used to generate this table:
To avoid analyzing unreliable data, let’s consider only the methods of killing with at least 1,000 deaths for whites. As you can see, for almost every one of these methods of killing, the death rate for blacks is far greater than the rate for whites. The only exception to this is “Other transport”, which is likely the result of the 9/11 terror attacks (excluding the year 2001 from the data shows 0 deaths by “Other transport” for whites and just 1 death for blacks). For example, the black to white homicide death ratio is about 11.2 for firearm deaths (16.82 vs 1.58 per 100k), 5.8 for cut/pierce death (1.8 vs 0.31 per 100k), 3.1 for unspecified injury (1.16 vs 0.37 per 100k), 2.9 for suffocation (0.41 vs 0.14 per 100k), and about 2.5 for poisoning (0.05 vs 0.02 per 100k).
Perhaps the most surprising finding from this data is that the black homicide death rate from cuts and piercings (1.8 per 100k) is greater than the white homicide death rate from firearms (1.58 per 100k).
Homicides by age
Using the same CDC data tool, here are the number of homicides for non-Hispanic whites and blacks by age from 1999 to 2020:
This is the query used to generate this table:
As you can see, the homicide death rate for blacks is greater than the rate for whites for each age group. For example, the homicide death rate for blacks younger than 1 year is nearly 3 times greater than the same rate for whites of the same age (16.93 vs 6.07 per 100k). Likewise, the homicide death rate for those over 85 years old is over twice the rate for whites of the same age (4.19 vs 1.85 per 100k).
Perhaps the most surprising finding from this data is that the homicide death rate for elderly blacks (e.g., 75+ years) is comparable to, and sometimes greater than, the homicide death rate for young whites.
Misconduct among black youth
Black children exhibit misbehavior very early in life, as evidenced by reports of school suspension by race.
A 2010 report [archived] by the Southern Poverty Law Center was conducted to analyze suspension rates by middle schools in 18 of the nation’s largest school districts. Data on suspension was gathered from the Civil Rights Data Collection (CRDC). The study found that the suspension rate for black males was nearly 3 times the rate for white males (28.3% vs 10%). The suspension rate for black females was over 4 times the rate for white females (18% vs 4%). In fact, black females had higher suspension rates than white males.
Analyzing the suspension rates by race in each district showed that black children had greater suspension rates in every district except for Indianapolis, where the suspension rate for white children was 1 percentage point higher (Table 1a, Table 1c, Table 1e):
|Suspension rate for males by race|
A 2018 report [archived] by the U.S. Government Accountability Office on discipline disparities shows that the racial disparity in suspension rate is not accounted for by controlling for level of school poverty. At every level of school poverty, black students are significantly more likely to be suspended than any other race of students. In fact, the out-of-school suspension rate for black students attending schools with the lowest level of poverty (7.5%) is higher than the rate for white students attending schools with the highest level of poverty (7.3%) [Table 14].
The racial disparity in misconduct appears even at preschool. For example, a 2014 report [archived] by the U.S. Department of Education Office for Civil Rights shows that “Black students represent 18% of preschool enrollment, but 42% of preschool students suspended once, and 48% of students suspended more than once” (page 3).
The report [archived] by the U.S. GAO on discipline disparities shows black preschoolers are over 3 times as likely as white preschoolers to be suspended (1.1% vs 0.3%) [Table 17].
None of this data should be surprising given the statistics cited in the first section. Adult criminals tend to exhibit misbehavior in early childhood. People who engage in criminal activity usually display problematic signs very early in life.
Misbehavior or racism?
Now, some people might accept these facts but deny that it demonstrates early misconduct by black children. One might argue that this demonstrates only that black children are disproportionately punished for alleged misconduct, but it doesn’t show that they disproportionately engage in misconduct. Such a person might argue that racial disparities in suspensions are driven entirely or primarily, not by racial disparities in behavior, by a deep anti-black bias in schools across the country. There are a few reasons to believe that this hypothesis is not correct.
Firstly, while black teachers suspend black children at lower rates than white teachers do, they still suspend black children at a rate several times that of white children (Figure 2 [archived]). Consider the following data points:
- Whereas black male students are 2.96 times as likely as white male students to be suspended by a white male teacher (15.4% vs 5.2% suspension rate), they are 2.39 times as likely to be suspended by a black male teacher (12.7% vs 5.3% suspension rates). So the racial suspension disparity for male students when taught by black male teachers is about 70% of the disparity when taught by white male teacher (disparity of 10.2 percentage points vs 7.4 percentage points).
- A similar pattern emerges when comparing racial disparity in suspensions by female teachers. Black male students are about 2.93 times as likely as white male students to be suspended by a white female teacher (16.1% vs 5.5%), whereas black male students are about 2.69 times as likely as white male students to be suspended when taught by a black female teacher (13.7% vs 5.1%). So the racial suspension disparity for male students when taught by black female teachers is about 74% of the disparity when taught by a white female teacher (disparity of 11.6 percentage points vs 8.6 percentage points).
In other words, the vast majority of the racial disparity in suspension rates persists regardless of whether students have a black vs white teacher. Therefore, if racial differences in suspension rates are entirely or primarily due to an anti-black bias among teachers, then it appears that a similarly sized bias is found among black teachers. While it is certainly possible that black teachers exhibit similarly sized anti-black biases as white teachers, this hypothesis is certainly less immediately plausible than the hypothesis that bias by white teachers is driving the racial difference in suspension.
Secondly, large racial disparities in misbehavior are reported by teachers in low-stakes settings where no disciplinary action follows from the teacher’s report. Of course, this does not rule out the possibility of anti-black bias in the teacher’s memory of bias, but it at least shows that racial differences in disciplinary outcomes are driven by genuine beliefs in misbehavior rather than an active attempt by teachers to discipline black children more harshly. However, what’s more relevant is that several surveys show that black children are far more likely to report either perpetuating violence or being the victim of violence. For example, consider the following two reports by the National Center for Education Statistics:
- Data [archived] from the Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 (ECLS-K:2011) showed that teachers report that black third-graders are significantly more likely to assault other children. In the spring of 2014, teachers reported that the percentage of black third-graders who “pushed, shoved, slapped, hit, or kicked other students” (7.9%) was over 6 times greater than the percentage of white children who did the same (1.2%). Furthermore, black children were over 3 times as likely as white children to victimize another student in any way (14.8% vs 4.6%) (“victimize” also includes teasing other students, lying about other students, or excluding others from play on purpose).
- A similar report [archived] from the same dataset reported that black third-graders were more likely to report having been “pushed, shoved, slapped, hit, or kicked” by another student (20% vs 12.7%). Racial disparities in victimization of these actions likely reflects racial disparities in the commission of these actions because schools are highly segregated by race (Frankenberg 2019 [archived]) and because assaults tend to be intraracial rather than interracial.
- Data from the CDC (2020) [archived] show that black high schoolers consistently report having been in more physical fights than white high schoolers. For every year that data was collected, black high schoolers were more likely to have been in a physical fight during the 12 months preceding the survey. In the most recent year (2019), black students were about 2.4 times as likely as white students to report being in a fight within the past 12 months (15.5% vs 6.4%).
Thirdly, Zill (2019) [archived] used data from the National Household Education Survey (NHES) to show that the majority of the racial disparity in suspension rates are accounted for by differences in family background. Confirming prior reports, the data showed that the suspension rate for black children in elementary and secondary schools was about 3 times greater than the rate for white children in 2016 (24% vs 8%). However, adjusting for family structure (i.e. whether the family type was both married parents) alone reduced the racial disparity in suspensions by about 55%. Adjusting for both family structure and family SES (parental education and family income) reduced the disparity by about 60%.
Of course, it does not follow from this data that family structure and family SES are causally responsible for 60% of the racial disparity in suspension. It only shows that these family background variables are statistically associated with racial disparity in suspension. This data implies that racial disparities in suspension are driven primarily, not by anti-black bias in schools, but by either racial differences in these family background variables or racial differences in factors that correlate with these family background variables. Furthermore, these variables likely do not account for all of the variables that may explain racial differences in suspensions between blacks and whites. If one had a more robust set of controls, it is likely that the racial disparity in suspensions would be reduced even further.
Finally, a study by Wright et al. (2014) [archived] published in the Journal of Criminal Justice found that prior problem behavior accounts for the racial disparity in school suspensions. The researchers conducted their analysis using data from the Early Childhood Longitudinal Study – Kindergarten class, the “largest sample of school-aged children in the United States”. The authors collected data on suspensions during the 8th grade wave by asking parents whether the child had been suspended from school. Prior to including any controls, the suspension rate for black students was several times the rate for white students (OR = 3.78, page 5). In other words, the odds of suspension for black youth was 3.78 times the odds for white youth, consistent with prior data. The authors conducted a number of models to explain this disparity by controlling for a variety of covariates:
- In model 1, the authors controlled for variety of covariates (see Table 1). Some important control variables included gender, grades, poverty status, teacher race, parent-reported delinquency (based on the parent’s assessment of whether the child cheats, steals, and/or fights), percent of school that was black, and whether the parent judged the school to be a “bad” school (based on the parent’s assessment of whether the child’s school is good, emphasizes learning, has an alcohol/drug problem, and/or has problem with violence). After introducing these controls, the odds ratio for being black on suspensions reduced to OR = 1.89 (Table 2, model 1). In other words, prior to any controls, the odds for suspensions for blacks was 278% greater than the odds ratio for whites (because OR = 3.78). After introducing a variety of controls, the odds for blacks was only 89% greater than for whites (because OR = 1.89), implying a 68% reduction in the racial differential in suspension odds. In this model, the best predictors for suspension were parent-reported delinquency (OR = 6.83), being male (OR = 2.67), and being black (OR = 1.89).
- In model 2, the authors added a control for prior problem behavior. Prior problem behavior was measured using an average of teacher-reported problematic behavior from kindergarten, first-grade, and third-grade (page 4). After introducing this variable in the model, the odds ratio for being black on suspension reduced again to OR = 1.20, implying that blacks are only 20% more likely to be suspended than whites, a difference that was not statistically significant. Thus, the controlling for a robust set of covariates reduced the racial differential in suspension odds by about 93% (OR = 3.78 vs OR = 1.20). In this model, the best predictors for suspension were parent-reported delinquency (OR = 7.08), being male (OR = 2.14), being in a parent-reported “bad” school (OR = 1.68), grades (OR = 1.50), and percent black enrollment (OR = 1.28). Model 3 added an additional variable to represent the interaction of parent-reported delinquency and prior problem behavior. The results were similar to the findings here.
These results led the authors to conclude with the following:
Replicating prior studies, we first show a clear racial gap between black and white students in suspensions. However, in subsequent analyses the racial gap in suspensions was completely accounted for by a measure of the prior problem behavior of the student – a finding never before reported in the literature. These findings highlight the importance of early problem behaviors and suggest that the use of suspensions by teachers and administrators may not have been as racially biased as some scholars have argued.
In other words, the vast majority of the racial disparity in odds for suspension can be accounted for by controlling for a variety of confounders, with parent-reported delinquency being the best predictor of suspensions. This finding is completely unexpected if racial differences in suspension were explained by strong anti-black bias in the schools. After all, if schools possessed such an anti-black bias, why do black and white children with comparable backgrounds and histories have similar suspension rates? Does the anti-black bias magically disappear when comparing similar blacks and whites? No, the simplest explanation here is that the racial disparity in suspensions is not the result of race per se, but rather is the result of certain variables that correlate with race (such as e.g., delinquency, problem behavior, etc.).
Further evidence that disparities in discipline are real (i.e. not entirely driven by racial bias) is the fact that there is data suggesting that racial differences in misbehavior emerge as early as infancy. Bakermans-kranenburg et al. (2004) examined data from the NICHD Early Childcare Research Network to examine differences in attachment security between 142 black and 1,002 white infants. The children were around 24 months of age when differences in attachment security and compliance were assessed. The study found that black children displayed greater levels of misconduct (page 423):
A set of items indicating children’s compliance showed consistently lower mean scores for the African-American group. African-American children appeared to be less compliant to their mothers’ suggestions or requests, and were less inclined to ‘stop misbehavior when told no’. African-American children showed on average also more active and even rough behavior in the context of play. Play materials were more roughly handled and the children became more easily angry with toys.
It should be noted that the study by Wright (2014) has been recently criticized by Huang (2020) [archived]. The main criticism was that the effects of prior problem behavior (PPB) reported in Wright were primarily driven by the different samples used in model 1 vs model 2. Model 1 (which found an OR = 1.89 for black without controlling PPB) involved a full sample of 4,101 subjects, whereas model 2 (which found an OR = 1.20 for black after controlling for PPB) involved only 2,737 subjects due to missing data. Huang states that one should compare the two models using the same samples in order to make any conclusions about the effect of PBB. He performs just this analysis in his paper on the possible effects of PPB.
His initial sample included 4,360 students in his baseline. Prior to controlling for any covariates, black students were much more likely than white students to be suspended (OR = 3.71). After controlling for all covariates except PPB (e.g., gender, grades, parental education, poverty, delinquency, etc.), the effect of being black (OR = 1.92, model 1) was significantly reduced in the same sample, thus reducing the racial differential in suspension odds by about 66%, consistent with Wright’s results. Also consistent with Wright’s results, the effect of being black (OR = 1.18, model 2) was no longer statistically significant after including PPB, but this sample was significantly smaller than the original sample (n = 2,892 vs n = 4,360). Model 3 showed that the effect of being black (OR = 1.40) was already statistically insignificant for this reduced sample prior to including PPB. Thus, introducing PPB was not necessary to reduce the effect of being black into statistical insignificance. While this finding is interesting, the results from model 1 still indicate that the majority (about 66%) of racial disparity in suspensions can be explained by a variety of covariates, even without controlling for prior problem behavior.
The author then analyzed additional model specifications that provided estimates of the effect of being black using different analytic methods and different measures of PPB (page 2). These alternative model specifications were different in the following ways:
- Some models used multiple imputation to account for missing data. The sample after imputation was n = 4,918. Controlling for all covariates, the suspension odds for black students was 1.56 times the odds for white students (Table 4, model 6). On this model, the racial disparity in suspension odds was 79% lower than the original disparity (OR = 1.56 vs OR = 3.71).
- Some models used 5th-grade teacher-reported measures of prior problem behavior. This measure of PPB resulted in less missing data than averaging the data from kindergarten to third-grade. The sample was n = 3,946, 91% of the original sample. Controlling for all covariates, the suspension odds for black students was 1.67 times the odds for white students (Table 4, model 8). On this model, the racial disparity in suspension odds was 75% lower than the original disparity (OR = 1.67 vs OR = 3.71).
- Some models used parent-reported measures of prior problem behavior. The sample was n = 3,773, 87% of the original sample. Controlling for all covariates, the suspension odds for black students was 1.90 times the odds for white students (Table 5, model 10). On this model, the racial disparity in suspension odds was 67% lower than the original disparity (OR = 1.90 vs OR = 3.71).
- Some models used teacher-reported externalizing behavior as a measure of PPB. The sample was n = 3,128, 72% of the original sample. Controlling for all covariates, the suspension odds for black students was 1.28 times the odds for white students (Table 5, model 12), which was no longer statistically significant. On this model, the racial disparity in suspension odds was 90% lower than the original disparity (OR = 1.28 vs OR = 3.71).
Using a variety of different models, the results indicate that between 67% and 90% of the initial racial disparity in suspension odds is reduced once one controls for the covariates in the different models. I believe that model 8 is the most useful since it minimizes missing data and it relies on problematic behavior reporting by teachers, who are likely to have a more accurate understanding of student behavior in school. This model suggest that about 75% of the racial disparity is reduced, as the effect of being black decreased from OR = 3.71 to OR = 1.67.
Black crime around the world
There are similar relative levels of black crime in other countries.
Criminal activity is significantly more prevalent among black people in the U.K. than other ethnic groups.
Recent data [archived] published by the U.K. showed that, from the period of April 2018 to March 2019, the arrest rate for blacks was over 3 times the rate for whites (32 vs 10 per 1,000 population). This was greater than the arrest rate for all other major ethnic groups, including the arrest rate for Asians (12 per 1,000), mixed people (18 per 1,000), and those with “other” ethnicities (12 per 1,000). While the arrest rate for all ethnicities has decreased significantly within the past 15 years, the black arrest rate has consistently been about 3 times the white arrest rate. During the time period from 2006 to 2007, the black arrest rate was 57 per 1,000 and the white arrest rate was 24 per 1,000. When disaggregating the data by sex, we find that the disparity in arrest is even greater among men:
As one might expect from the arrest data, black people are also disproportionately represented in the prison population. 2018 data from the U.K. Ministry of Justice [archived] shows that black people account for just 13% of the prison population (page 39), despite representing just 3.3% of the general population (2020 U.K. Census data [archived]). Most shocking were the proportions of prisoners who were black among younger age groups. Black people comprised 21% of prisoners aged 18-24 years old and 29% of prisoners under 18.
The racial disparities were greatest among violent crime. For example, crime statistics [archived] by the U.K. government show for the three-year period year ending March 2017 to the year ending March 2019, blacks accounted for 18% of principal suspects for homicide. Again, this is despite the black people account for just 3.3% of the general population. For reference, whites account for 86% of the population and commit 70% of murders. This implies that the black homicide rate is over 6 times the white murder rate, a similar pattern that is found in the United States.
The prevalence of black crime is particularly stark in London, because blacks account for a much greater share of the London population than the U.K. population in general. In fact, black people account for 13.3% of the population in London (U.K. Census data), which is about the same as their share of the population in the United States. Interestingly, the relative share of violent crimes in London committed by black people almost perfectly matches the relative share of violent crimes in the United States committed by black people:
- A 2018 article [archived] by News Sky revealed that despite being just 13% of London’s population, blacks accounted for 48% of London murder suspects. These numbers are strikingly similar to the corresponding numbers in the United States population where, despite being only 13% of the general population, blacks account for about half of those arrested for murder.
- A 2010 article [archived] by the Sunday Telegraph released similar findings regarding violent crime in London. It reports that “the majority of men held responsible by police for gun crimes, robberies and street crimes are black.” Specifically, the findings were that “among those proceeded against for street crimes, 54 per cent were black; for robbery, 59 per cent; and for gun crimes, 67 per cent.”
- Data [archived] from the Metropolitan Police Service shows that blacks (or “Afro-Caribbeans”) committed 48% of knife crimes in London during the years 2008-2018 (14,527 out of 30,486).
The Metropolitan Police Service serves a homicide dashboard that contains real-time information on homicide victims and homicide offenders in London since 2003. As of October 2021, blacks constituted 1,685 of those accused of homicides since 2003, compared to a total of 4,019 persons accused of homicides since that time. Thus, black people accounted for 42% of those accused of homicide. In fact, there were more black people accused of homicide during this period than any other ethnic group during this period:
The racial disparity in homicides has become even more stark in recent years. For example, in 2020, there were 235 persons accused of homicide. There were 123 black people accused of committing homicide during this time, accounting for 52% of total homicides, a strikingly similar percentage that we see in the United States. The following figures shows the number of people accused of homicide in 2020 (as of October 2021) by ethnic group:
Thus, blacks were accused of twice the number of homicides as whites, despite the fact that the white population in London (59.8%) is nearly 5 times the black population (13.3%) (according to U.K. Census data as of 2020). This implies that the black homicide rate is over 9 times the homicide rate for whites in London, a disparity even greater than what is found in the United States.
Blacks are incarcerated at far higher rates than non-blacks in Canada. For example, a fact sheet [archived] by the Colour of Poverty reported “In 2016, black people comprise 3.5% of the general Canadian population, but made up 10% of the federally incarcerated population” (page 1). The same fact sheet reported that “In 2016, 40% of inmates in segregation at the Toronto South Detention Centre were black, while being 7.5% of the Toronto Census Metropolitan Area” (page 1).
An analysis by Owusu-Bempah (2014) [archived] published in The Oxford Handbook of Ethnicity, Crime, and Immigration reported data showing that blacks comprise 8.4% of all federally supervised offenders despite constituting only 2.5 percent of the Canadian population at the time of the study (Table 1).
This shows that the federal correctional supervision rate of blacks in Canada is over 4 times the federal correctional supervision rate for whites (245.6 per 100,000 vs 58.6 per 100,000). Data from the same study shows that blacks (and Aboriginals) are significantly overrepresented in provincial correctional institutions in every province for which data is available (Table 10.2):
The same data on racial disparities in incarceration by province were also reported by Khenti (2014) [archived].
Roberts and Doob (1997) published data showing that incarceration rates were particularly large in Ontario during the early 90s. The study revealed the following prison admissions rates for blacks in Ontario (page 481):
When the Ontario correctional admission data are converted to rates, the black-white comparisons become striking. Using 1991 census data from Ontario, and 1992-93 admissions data from Ontario correctional facilities, analyses reveal a prison admission rate of 705 per 100,000 residents for whites, compared to 3,686 for blacks. The rate for Aboriginals is 1,993 per 100,000. When the male data are analyzed separately, the black admission rate rises to 6,796 per 100,000, compared to 1,326 for whites and 3,600 for Aboriginals. These incarceration statistics show the same pattern as the data from the United States cited by Tonry (1994a, table 4).
In other words, the black incarceration rate was 5.2 times the white incarceration rate in Ontario during the early 1990s. Recent data published by Owusu-Bempah et al. (2021) [archived] indicates that the same basic pattern was present in 2010. The study indicated that the incarceration rate for black men in Ontario was 4,109 per 100,000 men, whereas the incarceration for white men was just 771 per 100,000 men (Table 2). In other words, black men were 5.3 times as likely as white men to be incarcerated in 2010. The disparity was greatest among young men, with over 7% of young black men being incarcerated compared to under 2% of young white men.
By comparison, in the United States during 2018, “the imprisonment rate of black males was 5.8 times that of white males” (page 1, Bureau of Justice Statistics 2020 [archived]), which is only slightly greater than the racial disparity found in Ontario.
As in the United States and in the United Kingdom, the disproportionate black imprisonment in Canada seems to be the result of disproportionate black crime. This can be seen by focusing on data regarding homicide victimization and offending:
- In a case study by Thompson and Gartner (2007) [archived] prepared for Enhancing Urban Safety and Security, researchers reported that racial differences in homicide victimization rates from 1992 to 2001 in Toronto. They report that “the average annual homicide rate per 100,000 blacks in Toronto was almost five times higher than the overall homicide rate (10.1 compared with 2.4 per 100,000 population)” (page 6). They note that “this pattern parallels that in the U.S.” Indeed, from 1980 to 2008 in the United States, the homicide “victimization rate for blacks (27.8 per 100,000) was 6 times higher than the rate for whites (4.5 per 100,000)” (page 3, Cooper and Smith 2011 [archived]).
- Similar numbers on Toronto homicide victimization rates over this same time period were reported in a study by Khenti (2013) [archived] in the Canadian Journal of Public Health. The author reported that young blacks accounted for 30% and 36% of homicide victims and offenders, respectively, in Toronto despite the fact that blacks only accounted for 10% of Toronto’s population (page 12).
- Large racial disparities in homicide rates were also found in Montreal. In a study by Khenti (2014) [archived] published in the International Journal of Drug Policy on the impact of Canadian drug law policies on black communities, it was revealed that “homicide rates among blacks in Montreal were estimated to be as high as 24 per 100,000, strikingly higher than their white counterparts at 3 per 100,000” (page 193). This implies that the black homicide rate is 8 times the white homicide rate, an even larger disparity than what is found in the United States.
As in the United States, black children in Canada exhibit misconduct very early. A fact sheet [archived] by the Colour of Poverty reported that “black students were 12% of the Toronto District School Board student population but represented 48% of all expulsions; Indigenous students were 0.3% of the student population and 1% of all expulsions; Eastern, Mediterranean and Southwest Asian students were 4% of the population but 8% of all expulsions.”
It is more difficult to find crime statistics by race in other countries. But all the data that I’ve found shows that persons of African ancestry are more likely to engage in crime:
- Switzerland: a 2010 article [archived] published in a Swiss newspaper reported that “Men from Angola, Nigeria and Algeria receive six times more criminal complaints in Switzerland than Swiss men. In return, the Swiss are almost twice as criminal as immigrated Germans.” (translate the page to English)
- Germany: in a 2016 article [archived] published to the Unz Review, immigrants from African and Middle Eastern countries have considerably higher rates of murder and sexual assault than European and East Asian immigrants. Note: the linked source in this article is no longer valid. You can access the original source here (Table 62) for 2018 data.
- Brazil: a 2019 article [archived] published in Think Brazil reported that “In 2017, the number of blacks killed was almost three times higher than among non-blacks (a group encompassing white, Asian, and indigenous people).” A 2013 study [archived] published in Aggression and Violent Behavior grouped the homicide rates by race and found that the homicide rate for blacks (34.6 per 100,000) was over 2 times the rate for whites (16.3), and over 5 times the rate for Asians (6.8) (see section 4.1). Trends in homicide rates can be found here [archived].
- South Africa: a 2008 article [archived] published in a South African newspaper reported that the incarceration rate for the “Coloured” (651 per 100,000) and African population (342 per 100,000) was over five times the rate for the white and Indian population (60 per 100,000). A 2004 study [archived] on female homicides in South Africa found that the rate of intimate femicide for “Coloureds” (18.3 per 100,000) was 6 times the rate for whites (2.8 per 100,000), and the rate for Africans (8.9 per 100,000) was over 3 times the rate for whites.
Around the world
Finally, we know that black people are more likely to engage in crime (compared to whites) around the world:
- When we look at the national homicide rates of countries, we find that countries with predominantly black populations (e.g. nations in the Caribbean and Subsaharan Africa) tend to have far higher homicide rates than European and Asian countries.
- Finally, consider The Handbook of Crime Correlates, a huge book published in 2009 that collects the summary analysis of crime research around the world. Table 2.3.2a summarized the results of a large number of studies that compared the rates of violent crime by blacks and whites in North America, the Middle East, Europe, and Oceania. Every study found that being black was significantly associated with crime. Also, every study that compared the rates of non-violent criminal and delinquent behavior found that being black was significantly associated with property offenses, drug offenses, and delinquency (Table 2.3.2b).
I searched for crime rates by race in every country that has sizable black and non-black populations. Many governments outright ban collecting official crime statistics broken out by race or ethnicity, so it was difficult to find this data for certain countries (e.g. France, Sweden, etc.). Based on my research, the same general pattern is found in every country that publishes crime statistics broken out by race: blacks are much more likely to engage in criminal activity than whites, who are less likely to engage in crime than Asians. The disparity seems to be most pronounced for more serious crimes. For example, in Canada, the United States, and the United Kingdom, the black homicide rate falls within 5-8 times the white homicide rate.
This post is agnostic with respect to the cause of the widespread black crime. I investigate the causes of black crime in separate posts here and here. This post is also agnostic with respect to prescriptions or recommendations. The point here is simply to collect data relevant to the topic of black crime. This is important to ensure that we are adequately informed when we investigate possible causes of and solutions to this problem.