There is a growing consensus that race-based affirmative action is both legally problematic and ineffective at promoting racial justice and social harmony, not to mention good education and competence in employees. There seems little doubt that the rules will be rewritten in the next few years and that the result will neither be pure merit-based selection on objective criteria, nor for diversity and racial justice to be the dominant considerations in selection. There seems to be a long-standing, solid majority opinion in favor of tipping the scales here and there without abandoning ability as the major determining factor. Statistical discrepancies among groups can be tolerated if they favor historically oppressed groups, but no individual should be completely blocked from advancement to make room for others.
A case in point is a recent New York Times opinion piece, “It's Time to End Race-Based Affirmative Action” by John McWhorter. McWhorter is a smart guy and one of the world’s great lecturers and popular writers. I recommend his books and recorded lectures highly. But he suffers from a common and dangerous confusion with historical roots. He accepts the theory and practice of modern affirmative action, but wants to substitute some kind of index of social disadvantage for race. He accepts that race-based affirmative action was the right medicine in the 1960s and early 70s, but thinks it has passed its useful life.
For over a century, since its inception until the mid-1960s, the national Republican party stood four-square for equal treatment under the law for all races. Prior to the mid-1960s the main problem for Black people was explicit government repression. Race was not a social construct, it was a legal one. “Passing,” a person with some Black ancestry pretending to be white, was often a crime.
Only Whites could become naturalized citizens from 1790 to 1954, and elimination of that provision was fought by Democrats, including a veto by President Truman. Non-whites were prohibited from many professions and government jobs, often including military service, by law, not custom. In many places they could not own land, nor live in certain places, nor use public facilities. During World War II, racial concentration camps were established, mainly for people with at least 1/16th Japanese descent, but for tens of thousands of non-Japanese as well. School and other segregation was enforced by law in many places.
Of course layered onto this legal discrimination was extra-legal repression, up to and including lynching. It’s important to remember that when I say the national Democratic party before the Civil Rights era was racist, I don’t mean it tolerated desperate impacts on Blacks, I mean it actually favored lynching as a means of controlling the Black population.
In the early 1960s, the national Democratic party made a seismic shift and decided to go after Black votes at the expense of most Southern and many Northern whites. The national Republican party moved in the racist direction to pick up the latter voters. It took about 20 years, with lots of confusion and turmoil for this to play out.
The distressing point to realize is the remedies for historic anti-Black oppression were written by the same people who had legislated and enforced that oppression. Democrats like Lyndon Johnson did not have Road to Damascus conversions to justice for all, they made a cynical calculation that promising things to Blacks would win more elections than promising things to white racists. They did not reject their bedrock views that politics was handing out favors in return for votes, nor that racial discrimination was wrong, they just shifted their target recipients.
Politicians who favored, and even participated in, lynchings remained in office. Sheriffs who oversaw lynchings or did not investigate them, remained sheriffs. Judges who convicted Blacks out of pure racism and acquitted whites for crimes against Blacks, remained on the bench. Administrators who had enforced explicitly racist rules continued in their jobs and got promoted.
In the early Civil Rights era, Republicans won victories for equal, color-blind treatment, reversing oppressive government laws. In the later Civil Rights era, Democrats imposed new racial-based rules in compensation for past racial-based rules. Affirmative action is a prime example. That doesn’t mean affirmative action was a bad thing to do, but it has a suspect ancestry.
The original concept of affirmative action, the one that explains its name, is clearly good. Institutions should not shrug off disparate impact with excuses like, “Not many women apply to engineering programs,” or, “We don’t see many qualified Blacks to admit.” Rather than waiting for women and Blacks to show up, you should take affirmative actions to find out why you’re not getting enough, and to reduce barriers. That might mean outreach campaigns to targeted groups, remedial programs to help underprepared applicants, policies to make the institution more welcoming to underrepresented groups or other policies.
But the political push for affirmative action was simpler. Push up acceptance rates for Blacks, and later other groups, so Blacks will vote Democratic. That meant quotas. Moreover, quotas are much easier for administrators than the difficult challenge of chipping away at racism and its legacies. Republican objections to quotas left us with the current incoherent system of mixed rationales and hair-splitting policy distinctions. The only people who like the current system are administrators and lawyers who have built up specialized knowledge that could be made obsolete if a principled system is adopted.
When the Democratic party decided to court Blacks it brought two streams of thought together. The first is political horse-trading, the idea that politicians should negotiate favors among groups to build majority support, with no thought to consistency, principle nor long-term wisdom. If workers are unhappy, cut payroll taxes or increase unemployment benefits. If Blacks are unhappy, give them more government jobs and places at Harvard. If tenants are unhappy, limit rent increases.
The second idea is technocratic top-down planning by experts. To fix racial oppression you get the smartest people you can find who have studied the issue and design consistent, principled programs to solve the problem, root and branch.
Both of these ideas have some merit on their own (along with obvious disadvantages and historical disasters), but the combination is always massive failure, often with horrific consequences. McWhorter is all for technocratic top-down planning, a.k.a. social engineering, he just thinks there’s a better criterion than race to use. While that’s a logical position, it’s naïve to promote it when you know it’s strongly linked to horse-trading. The result will obviously be more micromanagement by horse traders, who can justify any payout or punishment as an adjustment to the index of social disadvantage. Moreover it’s irresponsible to promote social engineering without mentioning its horrific and spectacular failures. It may work in some cases, but it’s a dangerous tack to take.
That leaves two issues: what should have been done in the 1960s, and what we should do today. I think there’s a clear answer to the first. After removing the explicit legal oppression against Blacks, there should have been compensation and some sort of Truth and Reconciliation Commission like those done in South Africa and Argentina. Of course, no amount of compensation could have made up for all the horrors of slavery and Jim Crow, and the vast majority of individual criminals and victims would have been overlooked. But a serious effort to give monetary compensation to explicit legal victims, and to punish or at least fire some law enforcement personnel and judges, would have gone a long way toward confronting the evils of the past and allowing people to move on to the future with positive rather than negative focuses.
But that ideal is silly as it was not remotely politically possible at the time. So if someone wants to argue that race-based quotas were the best we could have done, I have no strong counterargument. Quotas are a form of compensation, and perhaps they empower individuals to find their own truth and reconciliation.
The answer to what we should do now is also clear: true affirmative action. Bottom-up efforts to chip away at the legacies of oppression. Figure out why some groups are under-represented in some areas. Don’t assume that differences are bad, after all we don’t want women to achieve the same homicide rates as men, or get more Blacks into the Ku Klux Klan. Don’t insist all programs help all groups. For example, a program that trains more women engineers is a success, even if the women are mostly Asian and Jewish. Don’t mandate top-down policies designed by mandarins in Washington, let institutions experiment on their own. Drop the failures and expand the successes.
The role of the government should be to insist institutions receiving federal funds make credible efforts to improve equity, not to sit in judgment on every selection decision. Courts should also back away from micromanaging university admission decisions, along with employment, promotion and other matters. Discrimination case decisions should be based on intention and individual harm, not detailed review of processes or aggregate statistics.
Unlike the 1960s, I think the clearly correct policy today is politically feasible. In fact, I think it would be more popular than the current system as well as both extremes of total color-blindness or quotas. I don’t expect it to be adopted. Not because it’s unpopular, but because it’s hard. It requires a lot of work by administrators at a frustrating task where even success comes with a lot of complaint and criticism. It means confronting the legacy of oppression, and thinking hard about principle.
My prediction is that courts will make it harder for selection to clearly disfavor groups like Asians and Jews, leading to a squeezing out of the middle group of applicants without either top quantitative qualifications or documented disadvantage. I further predict this will be a politically and administratively acceptable result. No cohesive group will be harmed sufficiently to generate effective political protest, and the system will be easy to run. It will work tolerably well in training and employing competent people, while offering reasonable opportunity to some members of disadvantaged groups. It will support large armies of bureaucrats, lawyers and administrators.
But we could do better, much better. And beyond affirmative action, smart technocratic progressives would do well to jettison their historical alliance with cynical horse-traders. Technocratic solutions only work when administered by widely respected experts, who are considered to be apolitical. When science and politics ally, the result is Big Science, and usually disaster. An index of social disadvantage would lead to just such a disaster.
