You Cannot Cite Data That Does Not Exist

This past week, Bill O’Reilly cited, for the second time, statistics that supposedly show that more Whites than African-Americans are killed by police. Although he framed the discussion of these statistics by stating that he was in no way condoning the recent police shooting of an unarmed African-American man in North Charleston, South Carolina, he was nonetheless arguing that the assertions that Black men are disproportionately victims of police brutality is some sort of lie being disseminated by Liberal activists and Liberal media.

Putting aside for a moment that the study that O’Reilly cited has been described as limited, somewhat dated, and something of a statistical outlier, one does not need a degree in mathematics to challenge O’Reilly’s assertions on several levels.

First, since White Americans outnumber African-Americans by a four-to-one to five-to-one margin, it is possible that more Whites than African-Americans are killed by police. But unless Whites are being killed at a rate proportionate to their percentage of the population–that is, at a rate four to five times the rate at which African-Americans are being killed by police—then African-Americans are disproportionately the victims in police shootings. Moreover, if the statistics for the two races are anything close to equal, then African-Americans are very disproportionately the victims of police violence.

Second, the total number of people killed by police is not really the issue. Presumably, large numbers of those killed by police have just committed violent crimes and have then been killed in confrontations with the police. The issue here is the seemingly disproportionate frequency of police killings of African-American males (1) who are stopped by police on “routine” matters, such as the broken brake light that was the ostensible cause of the traffic stop that preceded to the police shooting in North Charleston, (2) who are unarmed at the time that they are killed by police, (3) who may have resisted being restrained by police but did nothing that so physically marked the police officers that the claim that the officers were in fear for their lives can be justified, and (4) who, in many instances, have actually been running away from police when they have been gunned down.

And, third, since cell phone videos of these police shootings are a relatively recent phenomenon and since, in the absence of such videos, the police officers’ own reports on what occurred generally go unchallenged or are given the benefit of the doubt, it would be very difficult to statistically isolate the data that is actually germane to the argument—if, in fact, some sort of adequate data pool actually existed.

And that is the biggest flaw in Bill O’Reilly’s assertions. In 2000, federal legislation mandated that a database should be created to track police shootings. It took five years for the database to be developed and for the more than 17,000 municipal police forces to be contacted about the need to provide their data. An initial but very incomplete year’s worth of data was collected. Then the funding for the database was not renewed by the Congress—for eight years—until the recent attention to the police killings of unarmed African-American males created pressure for the restoration of the funding for the database.

In other words, the F.B.I. can provide very precise data on American criminals and the crimes that they have committed. But there is almost no data on how often American police discharge their weapons, how often those shootings result in deaths of suspects or bystanders, or what percentage of those shootings are clearly justifiable, ambiguous, or clearly criminal.

In an era in which even personal computing devices allow unprecedented data collection and analysis, such a lack of data on a matter of such great public concern seems deliberate and ideologically driven. As a society, we have decided that we don’t want to know the facts related to these issues—or, rather, our legislators have decided that they don’t want to know and/or us to know those facts.

In several states, the legislatures have actually mandated that state funding not be used for studies of social and economic issues—legislation that reaches well beyond even the recent elimination of the funding for several centers devoted to the study of such issues in North Carolina. Those promoting such legislation have argued, with almost incredible hutzpah, that such research is too often skewed by the ideological predispositions of those conducting it and that such research is a very limited public benefit. Without attempting to justify every study that has been supported by public funding, because there are always a few of very dubious value that seem always to get funded, I think that we clearly need to argue much more vehemently and persuasively that the elimination of such funding is ideologically driven and undermining the public interest.


One thought on “You Cannot Cite Data That Does Not Exist

Your comments are welcome. They must be relevant to the topic at hand and must not contain advertisements, degrade others, or violate laws or considerations of privacy. We encourage the use of your real name, but do not prohibit pseudonyms as long as you don’t impersonate a real person.