In Part 1 of this evaluation of the report on Berkeley Police racial profiling issued by the Center for Policing Equity (CPE), we examined some of the considerations concerning racial profiling and racialization in the US that were absent from the report, either because of the CPE focus or because the data was withheld by the BPD. In particular, two categories of discussion were omitted in the report that seem important, statistics on arrests made by the police in the course of traffic stops, and the overall national situation with respect to police violence and race, in particular, the many uprisings and demonstrations that occurred across the countyr in 2014. Since that year occurs right in the middle of the period the report addresses, to leave it out decontextualizes what the report intends to study to a large degree. Part 1 of this evaluation can be found here []. And Part 3, if you wish to read ahead, can be found here].
Racial categories and the police “recognition factor”
A summary of some of the data
Berkeley is a very diverse community, which means there are people of every appearance. One would think that racial disparities in traffic stops would depend on clear visual differentiations between people, because profiling drivers depends on categorizing appearance. Yet the racial categories used by the BPD are far from well-defined (if that is even possible).
Though five racial categories are deployed, there is one "racial" term that plays an interesting role in this catgegorization, and that is the term “Hispanic.” It is one of the five categories. But it is also used to define the others, namely, non-Hispanic blacks and non-Hispanic Asians. It even uses this terminology with respect to whites (non-Hispanic whites), though the data for white people is used as a main base of comparison in the report’s commentaries. The report actually admits: “Subjects in the Hispanic category could be of any race.” Nevertheless, the report (and the BPD) not only use the term "Hispanic" to define the other racial categories, they use it as one of their "racial" categories.
In this evaluation, we will use the term Latino as more appropriate. "Hispanic" refers to language, and a driver’s language does not appear in a driver’s face (which is what an officer sees in making racially disparate traffic stops). "Hispanic" is also a generalization that eclipses the richness and differences in Latin American cultures (which include hundreds of indigenous languages). We should also not forget those who consider themselves "Chicano," namely the decendent residents of formerly Mexican territories absorbed into an expanding US during the 1840s.
The Asian-American population of Berkeley (listed as non-Hispanic Asian), for instance; is it a race, or simply a recognition of the vast and ungeneralizable linguistic variations among Asian peoples? How, then, are Philippinos to be considered, for many of whom Spanish is a primary language? These questions have been raised many times, and will continue to be as long as "race" is considered as an object to be defined. We’ll try to get beyond that in Part 3 of this evaluation.
What all this oddly suggests is not the undefinability of Latinos, but a blurring of the boundaries of recognition between all these groups. And that then poses the problem, given the vastness of the racial differentials that occur in police practices, of what the police are using as a means of recognizing people racially in order to produce those disparities. Indeed, it suggests that it is the police who are creating these groups by the overall manner in which they stop drivers, and it is their manner of recognizing that is what these racial categories represent.
Nevertheless, the racial breakdown of Berkeley is of whites (56%), blacks (8%), Latinos (11%), Asians (19%), and a category of "Other" that comes to 7%. The category of "Other" is a melange of indigenous, Alaskans, Pacific Islanders, and those listing multiple racial identifications on the census forms. And that simply exemplifies the vast diversity of the Berkeley population. There are Latinos, Asians, and African Americans who are as light as whites. And there are people from Latin America amd Asia who are as dark as African Americans. If appearance is problematic, then how are these categories used, and what is their meaning for traffic stop disparities? Yet, the data concerning racial disparities in traffic stops depends somehow on appearances, and strong disparities along "racial" lines are reported in the data on how different racial groups were treated by the police.
As if to complicate things, the report warns that not all traffic stops are of Berkeley residents, meaning care must be taken in relating the traffic stop data to population size. But what we have just said is cause to ignore that. The issue of police racial bias in their decisions to stop motorists occurs before learning of the driver’s residence. The officer can be assumed to be acting in terms of a consciousness of Berkeley’s resident population ratios in making those stops to the extent they result from racial profiling.
But let’s take a look at some of the data. In police tabulation, the Latino population of Berkeley is one fifth the size of the white population, but they are stopped twice as often as whites (on a per capita basis). [fig. 9, page 27] Assuming equal driving patterns, this indicates a recognition of Latinos as not white. The excess of Latino driver stops over those of white stops (per capita, that is, per 1000 of that group’s population) represents the level of profiling of Latinos. But it has to be based on visual factors since the officer, in stopping a driver on a racially profiled basis, only sees the driver’s face before making the stop. Black people are 8% of Berkeley’s population, or one seventh the size of its white population. Yet black people are stopped on average 6.5 times more often than whites on a per capita basis. The excess of black stops over white stops represents the level of profiling of black people, which is an extraordinary level of excess. Asians are the only group whose per capita traffic stops are less than those of whites. And this variation in rates of stops indicates that the police are picking and choosing.
It gets worse if we look at the actual numbers. The graph (fig. 9) states that, in 2016, white stops occurred at a rate of 51 per 1000 of white population, while black stops occurred at a rate of 330 per 1000 of black population. Hence, the 6.5 rate at which black drivers are more likely to be stopped than white drivers. The “per 1000” figure signifies that white stops represent 5% of the white population, while black stops represent 33% of the black population. (Latino stops, at 100 per 1000, represent 10% of the Latino population.) This figure indicates the extent to which drivers from each group are singled out (with duplications, of course, but there is no data in the report on multiple stops of particular individuals, so the figure remains representational). For black people, this is not new. It is satirized as “driving while black.” Despite the vast variety of appearnces of black people, somehow the police find enough of them to represent fully a third of the black population.
Acting on racial bias is not a mechanical phenomenon. It is always intentional. That is, a decision is made on the basis of noticing and recognizing. Especially in the case of “implicit bias,” for instance, it is that implicit bias that then drives the decision (to stop) made in the wake of observation. And the degree of disparity in these figures suggest a high degree of decision to stop black drivers, even to the point of seeking them out from the vast variety of appearances that drivers present. That implies that there is a "search" component responsible for this degree of excess. And that makes the degree to which black drivers are stopped a deliberate project of the department. The police not only notice the race of the driver; they have to have been looking for drivers of that particular racial group to stop at the rate differential that actually occurs.
Indeed, in Berkeley, with people of every appearance, the frequency of black stops is excessive not only with resepct to the group’s population size, but also with respect to their percentage of all stops of people of color. Black stops account for 50% (approximately) of all stops of people of color, while constituting 22% of the whole (POC). That additionally implies that, for the police, the appearance of racial difference is well-defined. Yet physically that cannot be the case.
To search out members of a group is no longer "profiling." To profile means to impose by attribution a characteristic or value (such as a propensity to be violent) on a person upon encountering them, simply by association with how they comport themselves or how they look. But “to search” means to have that attribution in mind prior to encounter with the person on whom it will be imposed. It is the fact of imposition that constitutes the generalizations inhabiting racialization. To search for black drivers to stop represents an overall process of racializing them as black.
To say that “this is not new” is not to make a glib observation. To not be "new" means there is a history at work here. It is a history that contextualizes the fact that Berkeley PD traffic stops represent 33% of the black population, or that every black driver has a 1 in 3 chance of being stopped. It recalls the fact, imminent in the mass incarceration of POC in the US, that black men between the ages of 20 and 40 have had a 1 in 3 chance of being thrown in prison at some point in their life. It is that campaign, politically known as the “war on drugs” (more appropriately identified as a “new Jim Crow” by Michelle Alexander) that has made the US prison system the largest in the world.
If traffic stop ratios result from searches rather than law enforcement encounters, then the police are not innocent in the results. The fact that the number increases from year to year takes on a different meaning. In 2012, black traffic stops represented 21% of the black population (213 stops per 1000 of population). It rose to 28% in 2013, and stood at 33.8% in 2014 (a little bit higher than for 2016). With respect to white drivers, whose traffic stop rate was 40 per 1000 in 2014 (or 4% of the population, up from 3.2% in 2012), black drivers had an 8.45 greater chance of being stopped than a white driver, still as only 8% of the Berkeley population. (For Latinos, the 2014 rate was 10%, up from 5.5% in 2012.)
That steady increase parallels a different historical factor, that of the US police kill rate. In 2012, the Malcolm X Grassroots Movement estimated that an unarmed black person was shot and killed in the US every 28 hours. That signifies the killing of 320 such people by the police that year. By 2014, that number was up over 800. And by 2015, it had climbed to more than 1100 unarmed black people killed by the police. That is more than 3 a day – a form of mass murder committed daily by the government. Berkeley police data are not innocent either.
There is an odd twist in the BPD data, however, as we move from 2014 to 2015, a "dip" in incidence (mentioned earlier in this evaluation). While the number of black traffic stops increased linearly from 2012 to 2014, reaching a ratio of 8.45 to white stops, it fell to 6.1 in its ratio to white traffic stops in 2015, a decline of 28%. The yearly rate (per 1000) of white traffic stops continued to rise linearly from 2012 to 2016. There was, however, a dip in the rate of Latino traffic stops which dropped from 2.1 in 2014 to 1.75 in 2015, a decline of 19%.
This "dip" in traffic stops appears quite markedly in figure 5 (page 23). It begins at the end of the summer, 2014, hits bottom in January, and rises to former hieghts in March 2015.
Two questions (at least) emerge from this. Why did the ratio became so large in 2014, and what would explain the precipitous drop in January, 2015? After all, the driving capabilities of Berkeley residents had not changed from 2014 to 2015.
Throughout the second half of 2014, there were uprisings against police violence and militarism in many parts of the US, significantly in Ferguson and in Baltimore. These uprisings were not spur-of-the-moment events. Police violence had been increasing for years. What pushed some communities of color past their breaking point were the killings of Michael Brown in Ferguson, Freddie Gray in Baltimore, and Eric Garner in NY. And massive demonstrations erupted throughout the country in solidarity with and support for the many movements demanding justice for the victims of police killings, including Berkeley and Oakland.
Did those demonstrations have an effect? Should we hypothesize that the Berkeley PD, on its own, responded to national events by curtailing its excessive treatment of black and Latino people? Did the police undergo a change of heart, and back off from singling them out for excessive traffic stops? Or did that dip in black and Latino traffic stops represent something else?
Insofar as the increase in black traffic stops leading up to late 2014 paralleled a national trend of increasing violence, it would not be farfetched to assume that the dip in ordinary civilian policing was also in coordination with a national trend. And we know that policing in major cities, because of these uprisings, shifted to strategies focused on counteracting social unrest and strengthening crowd control logistics. That would imply, at least temporarily, a reduction in ordinary civilian patrolling. And the "dip" would suggest that Berkeley PD was put on similar alert – against social unrest.
In short, it would probably be a mistake to interpret this dip in traffic stops as a decrease in police activity itself. Rather, because of its historical context, namely, the surge of movement activity across the nation around the issues of police violence, one could more reasonably speculate that it marks a shift in strategy from ordinary responsiveness to one prioritizing crowd control readiness. And this would suggest a more conceptual connection between the Berkeley police and the national situation. In the wake of the Berkeley demonstrations of December, 2014, when many civilian complaints of police excessive violence emerged, the BPD issued a report in 2015, evaluating their actions and responses. Though they suggest that they were perhaps overly zealous, their argument parallels that of the police elsewhere in the country that they were placed on the defensive by the demonstrations or uprisings, and that their violence was essentially defensive violence. Video evidence, however, clearly shows that the BPD initiated violence against the demonstrators. And this was clearly the case as well in Ferguson.
In other words, rather than see the BPD acting autonomously, it would be more realistic to conclude that its strategies were linked to federal policy directives and coordination through US fusion centers – in Berkeley’s case, NCRIC (Northern California Regional Intelligence Center) – which would communicate advisories concerning such policy shifts nationwide. That would in fact explain the adamance with which the BPD argued for renewal of contracts with NCRIC in 2015 and 2017. Indeed, it was an adamance wholly out of character with what the BPD claimed they would receive. In pressing its requests, the BPD misrepresented its needs and misinformed the City Council about the benefits it sought (see the May 14 and June 20, 2017, hearings on the issue). In part, the police were untruthful in claiming they needed those connections for facilities that they already possessed, and for reasons that were for the most part irrelevant. In acceeding to BPD’s demands, the City Council ignored an enormous outpouring of popular sentiment against those contracts by the people of Berkeley.
What this represents is a process of federalization of urban police departments, beyond the militarization that has been debated over the last 25 years (involving such programs as Urban Shield, for instance). Federalization would amount to external influence on local police strategies in the name of coordination. Supporters claim it will be a way of attenuating racial bias in policing, while critics suggest it will erode local civilian control of the police.
The recognition factor
Let us turn to the fact that the police stop so many black drivers that it represents what amounts to stopping 33% of the black population of Berkeley in one year. The fact that Latinos are stopped twice as often as whites, and black drivers stopped 6.5 times as often as whites reflects something more profound than simple racial bias. In light of the wide variety of people of color who drive in Berkeley, what are the police using as a “recognition factor”? How are the police able to notice and stop those who are actually African American to the extent that they do? Many Latinos are dark, and many African Americans are light. What enables the police to pick out members of the smallest racial group for the highest number of stops? What do they look for in their search for drivers to stop?
Indeed, the number of African Americans stopped is roughly equal to the number of all other POC stopped. In 2016, there were 13,469 white stops, 13,351 stops of non-black people of color, and 13,594 black stops (multiplying the “per 1000” rate for that year times the number of thousands in the group’s population). They are all within a few tenths of a percentage point of each other. Yet black people are only 8% of the population, while non-black POC are 36%. How are the police able to find black people fully a third of the time in order to stop them? What are the police looking for when they stop a driver they identify as black so often that it equals the rate at which all other POC are stopped? Hanging out in black neighborhoods is obviously one way. Are the police operating on some kind of quota system? Or is their "occupation" of black communities part of their campaign to criminalize those communities? Is this too part of a policy handed down to them by federalization of urban policing?
However it is done, it can’t be easy, given the diversity of the city. Only people who “love their work” would be able to do it. Whatever kind of search function the BPD are using, it would have to be powered by an intentionality, a consciousness driven by a desire, a desire to recognize African Americans. To postulate such a desire goes beyond racial profiling, and the inequity of racial discrimination. It suggests instead a kind of obsession, though we would have to understand that term in relation to institutionality. That is, it is a political project.
The fact that this high rate of searching out, recognizing, and stopping African Americans indicates that black people are not just another sector of POC, one among many. What this “recognition factor” indicates is that black people play a special role in this society, one which requires that they be subjected to special treatment. That will not be news to most black people. The important question is its importance for white people, and for the white institutions that do it. That goes beyond mere recognition,
Are we back to some form of what had been (a century ago) called “scientific racism,” by which a special sense of presence was imposed on African Americans to immerse them in a prior dehumanization? Is this part of the ideology of the Berkeley police? Or is this an alternate form of objectification designed to deprive people of their sovereignty and autonomy as people?
In the third section of this critique of the CPE report, we will look a little more deeply into this “recognition factor.”
s