The central questions of this course are: What causal role, if any, does race play? How should the social sciences, such as sociology and economics, study the causal role of race1 The standard approach is to identify a racial disparity in the raw data, e.g., the raw data shows that the rate of police stops differs across races. If, after controlling for several variables other than race, the disparity does not disappear, race is thought to be a plausible (causal) variable that explains the disparity. The Fryer’s paper on racial disparities in police shootings discussed below is an example of this approach.? In what way can race be considered a cause? In order to address these questions, we will examine the philosophical and statistical literature on causality, as well as the literature on theories of race.
Talks of race as a cause are ubiquitous today. Claims of racial discrimination in hiring, for example, are allegations that race played a causal role in determining the hiring decision, and that this causal influence was inappropriate and unjustified. But again, in what way does race play a causal role in hiring decisions? Understanding the causal role of race can also help to clarify discussions about structural racism. How can social structures be causes and what role does race play within these social structures?
Ronald Fryer, a Harvard economist, wrote a paper on racial differences in police use of force. The paper came out in 2017 (though it was only published in 2019) and was widely discussed, even in mainstream media such as the New York Times.2 Fryer (2019), An Empirical Analysis of Racial Differences in Police Use of Force, Journal of Political Economy, 127(3). See the 2016 New York Times piece titled ‘Surprising new evidence shows bias in police use of force but not in shootings’. Based on empirical data, the paper made two central claims.
The first claim: there is racial bias in non-lethal use of force by the police. More specifically:
Using data on police interactions from NYC’s Stop and Frisk program, we demonstrate that on non-lethal uses of force – putting hands on civilians (which includes slapping or grabbing) or pushing individuals into a wall or onto the ground - there are large racial differences. (p. 3)
With all controls, blacks are 21 percent more likely than whites to be involved in an interaction with police in which at least a weapon is drawn and the difference is statistically significant. Across all non-lethal uses of force, the odds-ratio of the black coecient ranges from 1.175 … to 1.275… (p. 4)
Data from the Police-Public Contact Survey are qualitatively similar to the results from Stop and Frisk data, both in terms of whether or not any force is used and the intensity of force, though the estimated racial differences are significantly larger … the odds ratio is 2.769 for blacks and 1.818 for Hispanics. (p. 4)
These findings are, in many ways, are unsurprising.3 Note that different databases show different degrees of racial bias. Fryer’s paper considers a number of different explanations, one of them being that one dataset supplied the police perspective while the other supplied the civilian perspective.
The second claim of the paper, however, is more surprising: there is no racial bias in lethal use of force by the police. The data revealed no racial bias in police shootings against civilians:
Using data from Houston, Texas – where we have both officer-involved shootings and a randomly chosen set of potential interactions with police where lethal force may have been justified - we find, after controlling for suspect demographics, ocer demographics, encounter characteristics, suspect weapon and year fixed effects, that blacks are 27.4 percent less likely to be shot at by police relative to non-black, non-Hispanics. This coecient is measured with considerable error and not statistically significant. This result is remarkably robust across alternative empirical specifications and subsets of the data. Partitioning the data in myriad ways, we find no evidence of racial discrimination in officer-involved shootings. Investigating the intensive margin – the timing of shootings or how many bullets were discharged in the endeavor – there are no detectable racial differences. (p. 5)
Odds ratio non-lethal use of force,
stop and frisk datasets.
What are we to make of these claims? How should we interpret them? In particular, what does it mean to say that there was no racial bias in police lethal use of force? It might mean something like this: the race of the person interacting with the police officer had no causal role in affecting the officer’s decision to use lethal force. But what does it mean to say that ‘race’ plays (or does not play) a causal role?
The question of what causal role race plays is not merely a question about the social sciences. It is also a question that matters a great deal in policy and law. What does it mean, for example, that a candidate who applied for a job was discriminated on the basis of race? A plausible answer is in the form of a counterfactual test:
Had the candidate been of a difference race, the hiring decision would have been different.
If the counterfactual is true, then there is racial discrimination. If the counterfactual is false, there is no racial discrimination. It is difficult to operationalize the counterfactual empirically. Instead of race itself, studies will use indicators of race that can be manipulated empirically. Audit studies, for example, have shown that by changing the last name on otherwise identical resumes the probability of a callback changes.8 See, for example, the paper Bertrand and Mullainathan (2004), Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination, The American Economic Review, 94(4). We will discuss audit studies more in detail later in the course. This result can be interpreted as evidence of racial discrimination in hiring decisions, so long as a candidate’s last name tracks the candidate’s race.
This counterfactual approach, however, faces a difficulty. It is incompatible with a popular view about race, that is, race as a social construct. We will discuss this critique in detail later in the semester. For the time being, here is a succinct statement of the problem:
If, as the social constructivist account suggests, race is a social category constituted by a set of social practices, institutions, norms, expectations, and so on, then to speak of a decision as being “directly caused” by race white or race black is incoherent. In other words, … [to think of] race as a singular node that independently causes downstream effects, or … [to see] race as something that can be isolated from confounders and mediators such as “socioeconomic status” or “high school performance” rests upon an incorrect theory of what race is–and why it matters in our society.9 Hu (2019), Disparate Causes, pt. I, Phenomenal World.
So, again, in order to understand (and adjudicate) claims of racial discrimination in hiring, we should first understand what causal role (if any) race plays. This question will accompany us throughout the semester.