Behind the epidemic of police killings in America: Class, poverty and race

Part one

By Benjamin Mateus
20 December 2018



The steady rise in police killings in the United States is the manifestation of an ongoing civil war between the ruling elite, the top one-tenth of one percent, and the working class. It is not “white cops vs. black youth,” as portrayed by the media and groups like Black Lives Matter and the pseudo-left, anxious to elevate race over class. It is the armed representatives of the capitalist state (frequently black and Hispanic, as well as white) against the most impoverished sections of the working class, white, black, Hispanic and Native American.

This study reviews all the data available on police shootings for the year 2017, and analyzes it based on geography, income, and poverty levels, as well as race. It identifies a major omission in all the published accounts: the vast and rising death toll among working-class white men in rural and small-town America, who are being killed by police at rates that approach those of black men in urban areas.

Police violence is focused overwhelmingly on men lowest on the socio-economic ladder: in rural areas outside the South, predominately white men; in the Southwest, disproportionately Hispanic men; in mid-size and major cities, disproportionately black men. Significantly, in the rural South, where the population is racially mixed, white men and black men are killed by police at nearly identical rates. What unites these victims of police violence is not their race, but their class status (as well as, of course, their gender).

The wave of killings by police officers that occurred in the period from 2014 to 2016, many caught on personal smartphones and released through social media channels, led to an outpouring of popular rage against these crimes. Protests against these horrific killings perpetrated by the police officers, who for the most part face little or no consequences, have become commonplace. Some of the best-known cases include:

All of these victims were black and poor. All the killings took place in urban areas or their suburbs. In all but the Castile case, the officers involved were white. These demographic characteristics were well publicized and became the media template for all such police killings. But a more thorough investigation shows that the killing of poor blacks by white cops is only one aspect of the reign of terror by American police against the working class, and not even the most common form of such killings.

Enormous resources have been mobilized to spread the “race, not class” mythology of police killings. Black Lives Matter (BLM) was formed in the aftermath of Trayvon Martin’s killing by neighborhood watchman George Zimmerman in 2012. As a response to the growing tension within these communities associated with these killings, BLM was used to channel the anger along racialist lines and, ultimately and politically, in support of the Democratic Party.

As we wrote in 2017, “From the beginning, the ‘mothers of the movement’ Alicia Garza, Patrisse Cullors and Opal Tometi—who collectively adopted the famous hashtag—specifically opposed uniting blacks, whites and immigrants against the brutal class-war policies of the capitalist state. Instead, the group did its best to confine anti-police violence protests within the framework of the capitalist system and push a racialist and pro-capitalist agenda.”

Big money donors lined up behind BLM. The Ford Foundation made a six-year $100 million investment in the organization. BLM garnered endorsements from companies such as Facebook, Nike and Spotify. BLM has gone on to partner with Fortune 500 New York ad agency J. Walter Thompson (JWT) to create “the biggest and most easily accessible black business database in the country.”

The data on police killings

Before 2015, two federal databases, from the FBI and Bureau of Justice Statistics (BJS), were the main sources used to track police shootings. However, these figures, self-reported by the police, have been found to be incomplete and inaccurate. Accordingly, several private groups and media outlets began to systematically track police killings through news reports, publishing the data on their websites.

The Washington Post tracks people “shot dead” by police, while the Guardian records “all people killed by police, regardless of the means” and Fatal Encounters attempts to count all police killings as far back as the year 2000. Fatal Encounters, besides using news sources, also uses several other public databases and research methods. In 2015, the Washington Post counted 990 shot dead while the Guardian counted 1,146 killed. Fatal Encounters counted 1,357 killed.

Academicians and researchers, to provide a veneer of scientific corroboration to the racialist perspective promulgated by mainstream media, have been using the information in these databases to publish peer-reviewed articles to substantiate their perspective on identity politics, push through racialist agendas, and direct the national discussion on these issues. However, there is little substance to their analytical approach, which generally avoids the socio-economic dimension to the phenomenon of police killings, or at best, treats it as one more (usually lesser) factor, in the name of “intersectionality,” in which class is blended in to the basic framework of identity politics.

By example, a recently published paper in the American Journal of Public Health titled, “Risk of Police-Involved Death by Race/Ethnicity and Place, United States, 2012–2018,” toes the line that “race plays a powerful role in explanations of police-involved killings in the United States.” Their results note that in the US police kill on average 2.8 people per day. Blacks are killed at 1.9 to 2.4 deaths per 100,000 per year, a threefold higher rate than whites, who are killed by police at a rate of 0.6 to 0.7 per 100,000. Beyond these rudimentary statistics there is little else to take from the study than to accept the authors’ conclusions unquestioningly, “Indeed, our results show that—like other police-related outcomes, which vary across the nation according to local political and social forces—police-involved deaths are contingent upon local contextual environments. Structural racism, racialized criminal-legal systems, anti-immigrant mobilizations, and racial politics all likely play a role in explaining where police killings are most frequent and who is most likely to be a victim.”

The method of this study

The present study was undertaken to better understand the demographics and economic aspects of police killings in the US, treating police killings as though we were observing an epidemic and seeking to understand the main risk factors that made sections of the population more or less vulnerable. We used the 2017 data from the Washington Post database for our analysis. These were cross-referenced with the website to ensure accuracy and attempt to fill in missing information such as race, name, and age of some of the victims.

News reports from approximately 15 to 20 percent of these events were reviewed to glean the context of some of these killings as well as help locate the site of shooting for further in-depth analysis. We used information on the location of the shooting to perform a ZIP Code analysis for some of the metropolitan centers and medium-sized cities where blacks were killed at a much higher proportion than in the population. The US Census Bureau website was used to obtain demographic and economic data on states, cities and towns where police killings took place. Data USA and websites were used for population centers under 5,000 people. Economic data used included Median Household Incomes (MHI), Percentage in Poverty (PP), and employment and “not in the workforce” categories to determine the role of these factors in police killings.

Academic studies and news reports use national demographics to compare the rate of those killed by their race. For instance, blacks make up approximately 12 percent of the US population, but they make up about 25 percent of those killed by police. Therefore, blacks are killed by police at more than twice their representation. These numbers are standardized to a rate of numbers killed per 100,000 to form a consistent unit to compare across racial groups.

It should be understood that we had to use the category of “race” despite its completely unscientific character, because virtually all data on police killings describes the victims in such terms. More than three percent of the adult US population, and ten percent of all children, are officially categorized as “mixed race,” and from a historical standpoint, the proportion of Americans whose ancestry combines white, black and Native American is even larger. But victims of police killings, and their killers, are not usually categorized in media or government reports in that fashion.

Our statistics also used this methodology to conform to the published literature. However, our hypothesis differs from the published studies. The United States is not homogenously diverse. There are significant variations from state to state and from population centers like metropolises to small rural communities in how the demographics are configured. The Southeast states have large black populations in both urban and rural areas, the Midwest is predominately white, particularly outside city centers, and the Southwest has a very high Hispanic population. Metropolitan centers have higher minority populations while rural communities have a preponderance of white people. There are also considerable socioeconomic variations within these regions.

We tabulated the population estimates and demographics based on the locations where police killings took place which allowed us to compare these regions against the nation as a whole. This also provided the ability to compare economic data for these regions and provide a more accurate estimate of the real picture.

We used the Excel spreadsheet for performing and tabulating the basic statistics. The entire Washington Post data set was imported for this analysis. Web-based chi-squared tests were performed using analytic software to estimate confidence intervals and determine the significance of the findings, a statistical method used to denote a difference that has a low probability of being a chance occurrence. The P-value of 0.05 or less connotes that there is less than a 5 percent probability that the significance in the finding was due to chance. A P-value closer to 1.00 suggests that differences are random or minor. We also employed statistics that looked at observed vs. expected outcomes in each state. The observed outcomes are the reported racial distribution of killings in each state. The expected numbers are derived from the states’ actual demographics.

For example, in the state of Alabama, there were 25 people killed. Fourteen (56 percent) were identified as white, seven (28 percent) were black, and 1 (4 percent) was Hispanic. The actual state demographics are 65.8 percent white, 26.8 percent black and 4.2 percent Hispanic. The observed vs. expected outcomes P-value was 0.892 for blacks, meaning that the racial difference was not significant. This has enormous political importance: in Alabama, a state which historically is a byword for racially motivated police violence, there was no “preference” by the police for blacks as targets over whites. Similar “neutrality” by the police was found in Mississippi, an equally backward state from the standpoint of its history of racism. The racial explanation of police violence falls apart in precisely the locations where it should be most blatant.

To be continued