Racial inequality and discrimination is a big topic, and one that contains many subsets of subtle, and not so subtle issues. Today I want to go through discrimination as a concept, and look at the causes and types that lead to lower incomes for many groups of society. I will note here that this essentially takes discrimination at face value, and does not examine the moral and ethical questions underpinning their development. That historical context is valuable and important, but it is not the method of examination I will be using today. I will be looking at these in a manner I find conducive to setting up what needs to be examined for an incentives based solution. Essentially this approach is meant to say, we re at point A, and we need to get to point B, here’s how, but it is not my intent to dismiss any of the historical, ethical, or humanitarian aspects of this issue.
I can identify three distinct causes of discrimination, and will deal with them one by one. The first can take place when there is no racism in a society, but was in the past. The second involves information asymmetries, and how those may be developed. The third is probably the simplest to understand, preference based discrimination. This is the desire of an agent to support and be with those belonging to her group more than those of other groups. Preferences are almost always exogenous to models of human behavior, so dealing with this has no elegant solution. In firms, preference based discrimination takes the form of a principal-agent problem, where a manager in a firm may take actions that are optimal for herself, but not the firm. This problem already has some solutions, usually taking the form of stock options and bonuses tied to performance. These types of compensation however, are vulnerable to wide fluctuation due to random shocks in the economy, so any solution to preference based discrimination would either need to act on managers directly, or do a better job at tying firm performance to inequality. I will go into more detail on why exactly preference based discrimination is sub-optimal for the firm in a later post, but the logic follows closely to this post.
The second issue here is information asymmetries. This has more to do with cultural characteristics than specifically race, but the two can be closely intertwined. This problem comes about when one associates particular characteristics of behavior(speech, mannerisms, etc.) with success. Note, that these characteristics have nothing to do with productivity. How you wear your hair, what words you prefer, things like these have nothing to do with your actual potential economic output. If one assumes, however, that these characteristics are well correlated with traits such as intelligence, or work ethic, then we will find an inefficient disparity among workers, particularly when setting wage, and making a hiring decisions. Here is an article that makes a similar point. It finds that conflict comes not when there are different cultures, or different races, but when different races correspond with different cultures. The question remains of how these trait associations come to be. The simplest example I can give here is the school system. Let’s take the idea of AP tests and classes. From here, we can see that black/African American students, made up 14.5% of students, yet only took 9.2% of AP tests. A similar trend can be found in honors courses, and this isn’t even taking account societal forces that pressure disenfranchised students out of school systems, and into prisons. If we have more white people in AP and honors courses, then more traditionally white mannerisms may be associated with successful traits, which can lead to this asymmetry. In order to deal with this problem, we must focus on providing equal access to education, and preventing discriminatory processes of guidance concealer lead class selection in our school systems. There are other examples of how these information asymmetries can come about, which I will go into more detail in a later post, but the point here is that this is a problem, and one that discrimination and inequality must address.
The last issue I want to go in to here, is to show that even if no information base discrimination, and no preference based discrimination exist, we will still have inequality in society, if they existed at some point. I won’t bother to show that it existed at some point, you do not have to go far back in history to find overt and specific language of racial bias in public documents and laws. Let’s assume now, that this discrimination had the effects of limiting access to human capital, and household wealth. Let’s now assume that human capital grows exponentially across generations. This is, a family passes on its human capital to their descendants, and those descendants build on that base. Even if we assume equal access to education, and therefore equal growth rates of human capital, we will have that while the ratio between groups will remain constant, the difference between groups will continue to grow. This is essentially to say, it is not enough to simply stop racism, one must actively seek to repair the wounds it caused. I will also note, that even if one did not start with an information asymmetry problem, this sort of dynamic can cause it, which would result in decreasing the growth rates for the historically marginalized group, and cause the ratio to diverge.
Any potential solution to discrimination must(at the very least) address all of these problems. These were treated here as mostly separate issues, but there is evidence that they all exist in our society. It is not enough to simply end preference based discrimination, we must actively work to change the way expectations are made between groups, and to systematically repair disenfranchisements of the past.