Affirmative Action as a Magic Bullet
Why Republicans should run on opposition to race and sex preferences
A recurring theme in political discourse is the search for magic bullets. Very often, you will see an analyst say something along the lines of “I favor policy X. Look at polls, which show high levels of support for X. Just run on X and you can achieve overwhelming political victories!”
We can call this the “Magic Bullet Theory of Politics.” It’s the idea that there’s some low-hanging fruit out there, some very simple thing that can break the relative parity between our two major parties and give a candidate a clear electoral boost.
Some examples from recent years:
Bernie Sanders fans saying that Democrats can win elections by leaning hard towards economic redistribution.
Gun control activists arguing that overwhelming support for background checks and other reforms means that Democrats should run on their issue.
Immigration restrictionists arguing that opposition to illegal immigration means politicians should run as nativists.
In each case, Magic Bullet Theory does have some truth to it. Usually, the polls the activists are relying on are legitimate. And a limited version of Magic Bullet Theory is clearly correct; politicians can at the margins improve their chances of winning by adopting more popular positions. Usually, however, there are no magic bullets in politics, for the simple fact that if they existed they probably would have been used already. And while politicians can gain votes by ditching unpopular positions that are associated with their own party, they rarely do this because they are beholden to interest groups and activists on their side of the aisle.
Part of the problem activists have is that they get fooled by polling questions that do not give any counterarguments. But when a policy is being debated in the public sphere, people hear the case for both sides of an issue. A recent example I like is Biden’s student loan forgiveness plan. Ask about the policy in isolation, it has majority support. Name one tradeoff or cost, which opponents of a policy will inevitably do, and support goes negative.
For these reasons, you should always be wary of those claiming to have a Magic Bullet that a candidate can use to achieve political victory. With that disclaimer in mind, I think there is at least one plausible candidate.
The Public Really Opposes Affirmative Action
A few weeks ago, Blake Masters created a media firestorm when he brought up the issue of affirmative action.
I was pleasantly surprised to see this. For decades, polls have shown overwhelming opposition to affirmative action among the public. Every branch of conservatism – whether libertarian, populist, or establishment – should in theory be opposed to the government favoring some people over others on account of race and sex. Yet this has pretty much been a forgotten issue over the last few decades. While in 2000 Bush talked euphemistically of the “soft bigotry of low expectations,” after that we’ve gotten over two decades of Republican silence on the issue, as affirmative action was rebranded as “diversity” and became ever more prominent in American life.
Yet there is a strong case to be made that conservatives have dropped the ball here.
Between 1977 and 1989 Gallup asked Americans in four polls whether “women and minorities should be given preferential treatment in getting jobs and places in college” or if “ability, as determined by test scores, should be the main consideration.” Each time, Americans opposed race and sex preferences by an over 8-1 margin. The framing in Gallup led to lower support than in other surveys, but no matter how you ask the question, there is a long history of majority opposition to the pro-preference position. While most Americans may say they favor “affirmative action programs” in surveys where they’re not given any indication of what that means, support plummets as soon as you define affirmative action.
Unlike on other issues, where the public has moved left over the years, here we do not see much of a change. A 2014 Gallup poll asking about college admissions found 67% support for using only merit even if it meant “few minority students being admitted,” and 28% support for considering race. In March 2022, Pew found that only 7% of Americans thought race should be a major factor in college admissions, compared to 19% who thought it should be a minor factor, and 74% who said it should not be a factor at all. Support for sex preferences was even lower.
David Shor and his colleagues have a chart showing the reaction of voters to 113 Democratic positions, where they present the Democratic argument and the Republican response. Support for affirmative action was the second least popular liberal position, ahead of only lowering the voting age to 16.
This opposition to race and sex preferences reveals itself at the ballot box. The website Ballotpedia lists 9 times affirmative action has been voted on at the state level between 1996 and 2020. It has been rejected 8 of those times. The table below lists the initiatives.
We see a constant rejection of affirmative action across red, blue, and purple states. The margins aren’t as massive as polls recording 70% or more opposition would suggest. But one general feature of referendums is that voters are biased towards the status quo, more likely to want to vote “no” on an initiative than “yes.” One experiment showed that framing an abortion position as changing current law decreased support for the proposal by 6-7%. In that context, getting 55-60% of the public to vote “yes” on a ballot initiative is quite impressive, and not all that common. Of the nine referendums listed above, only the latest votes in Washington (2019) and California (2020) involved a “yes” vote that would change the status quo in order to implement affirmative action.
Another problem in interpreting these results is that, in ballot initiatives, people often don’t know what they are voting for. One study that looked at 1,211 ballot initiatives from 1997 to 2007 found that they were on average written at a grade level of 17, meaning less than a quarter of the public could even be expected to understand what they were reading.
With affirmative action initiatives in particular there is an added complication in that the language used to ban the practice is often extremely similar to the language that has been used to require it through civil rights law. Journalists and academics tend to highlight the confusion going in one direction: people who want to support affirmative action but end up voting in a way that opposes it. But the opposite must also happen. A survey of Colorado residents conducted after Amendment 46, the one anti-affirmative action initiative that failed, found that over a quarter of voters on each side of the referendum thought that they were voting for the opposite of the option they picked.
Imagine 70% of the public is opposed to affirmative action and 30% is in favor. At the same time, one-quarter of each side is confused by the ballot initiative and votes contrary to the way it intended. Instead of 70% opposed to affirmative action and 30% in favor, you will only win the referendum by a vote of 60-40, which is close to what we get in many of the initiatives listed above.
All of this is not to say that ballot initiatives are completely worthless for understanding public opinion. Usually, political activists through advertising campaigns make clear which side is which, so some voters at least know what they’re doing. But not everyone gets the memo, so there must be confusion at the ballot box, which introduces a degree of randomness into the process.
When a Republican increases the salience of affirmative action, there is none of the danger of confusion we see in complicated ballot initiatives written in legalese. Talking about the issue should present a stark contrast between the parties.
The votes to ban affirmative action look even more impressive again when we consider the resources that go into defending race and sex preferences.
In 2019, Washington rejected bringing back affirmative action. In support of the measure, top donors included SEIU ($500,000 cash), a state public sector union ($100,000), and the ACLU ($94,000 in-kind donation). The top donor against affirmative action was a man named F.K. Freeman ($50,000), whom I’ve been able to find nothing about online. The fourth largest was a Chinese takeout restaurant called Sizzling Pot King, which put up $18,000.
In the 2020 vote in California, supporters of affirmative action raised $31 million, compared to $1.6 million for opponents. Ballotpedia provides a list of individuals and institutions on each side of the question. In support of affirmative action, we see 4 government entities, 8 major corporations, 7 labor unions, and 13 activist organizations.
On the opposition side, we see no corporations or labor unions, and 5 small conservative organizations.
Supporters of affirmative action being much better funded and having institutional support is probably another reason why ballot initiatives tend to be closer than public opinion polls suggest that they should be. Yet the opposition to affirmative action is so strong and deeply rooted in public opinion that ballot initiatives to get rid of race and sex preferences still tend to win pretty comfortably.
Intensity of Belief
In politics, cardinal beliefs matter. If voters oppose affirmative action, but don’t care that much about the issue, then it may not make a difference in elections. In fact, talking about the topic may backfire if supporters of affirmative action are more likely to consider it a litmus test for gaining their vote. The issue of guns tends to be like this. The majority of the public might support some forms of stricter gun control, but supporters of the Second Amendment tend to be more intense in their beliefs and more likely to vote on that issue.
There’s little evidence, however, of deep support for affirmative action in any sector of the electorate. Liberal activists and politicians seem to put a lot less energy and effort towards defending affirmative action than they do into reforming policing, defending abortion rights, or preventing immigration restrictions. California in 2020 voting to keep the state prohibition on affirmative action appears to have inspired few liberal tears.
I suspect that this is in part because liberals are in denial about affirmative action and what it does. Faced with the fact that race and sex preferences are massively unpopular, institutions stopped talking about “affirmative action” and started talking about “diversity” to obscure what they were doing, and liberals were probably the only ones who took them seriously. I can’t prove this, but I suspect liberals are less aware than conservatives are of the reach and scope of affirmative action, and this will blunt any opposition to getting rid of it, as the left will assume that steps being taken are mostly symbolic.
If there is intensity on this issue, it appears to point in the other direction. In 1990, a Democratic pollster noted that “[w]hen we hold focus groups, if the issue of affirmative action comes up, you can forget the rest of the session. That’s all that’s . . . talked about.” Wokeness – particularly on the issue of race – appears to be what has permanently caused the split between Democrats and the white working class. While LGBT issues and abortion are also thought about in class terms, here public opinion is more nuanced, as can be seen in increasing public support for gay marriage and the voters of Kansas recently rejecting an initiative that would have codified a pro-life interpretation into the state constitution.
In a study I did with Eric Kaufmann and George Hawley, we found that when we showed voters Kirsten Gillibrand’s statements about having white privilege, opposition to her among white voters grew by a shift that was equal to about a standard deviation move to the right in ideology. In other words, all you have to do is make left-wing support for wokeness salient, and liberals will start behaving like moderates and moderates will start behaving like conservatives. Supporting reparations also had a statistically significant negative effect on willingness to vote for her. We found practically no effect of presenting her as more of an economic radical.
Matt Yglesias recently wrote up a summary of research showing that priming race at all across a wide range of domains, including economics and criminal justice reform, makes Americans more conservative.
This has led many to argue that the reason that the US hasn’t gone as far as other advanced countries in creating a social democratic state is that we are too diverse. While conservatives like Ann Coulter and Amy Wax worry about immigrants pushing the country to the left, the great legislative achievements that built the modern welfare state came under FDR and Johnson, before the era of mass migration. Look at the federal budget, and you’ll see that the programs passed under those two presidents are still responsible for most social welfare spending. Americans are so opposed to race targeted government initiatives that many of them reject redistributionist policies they would otherwise support.
In sum, we see massive opposition to affirmative action, whether measured through polls or ballot initiatives. We also find no evidence that liberals are more intense in their support for affirmative action, and a good deal of evidence using a wide variety of sources and methods – experimental, survey, and cross-national comparisons – indicating that highlighting liberal ideas and positions on race tends to be good for conservatives.
Why Haven’t Republicans Thought of This Before?
Given all this, why have conservatives not run on affirmative action? When I bring this question up to people, the standard response is that they’re afraid of being called racist.
But if the conservative politicians today are scared of Democrats calling them names, they sure don’t act like it. Doesn’t supporting Dobbs also anger the media? How about claiming fraud in the 2020 presidential election? Immigration tends to be a racially charged issue, and Republicans seem to have no problem talking about that. Same with Critical Race Theory.
The simplest explanation here I think is inertia. After Ronald Reagan failed to roll back civil rights law, Republican leaders decided to run from the issue. It might’ve made sense at the time, as before the rise of conservative media, liberals could frame stories of the day in the way they wanted with practically no pushback. Unlike on abortion and gun rights, where activist organizations formed to advocate for their issue, opponents of affirmative action never did the same. My model of Republican elites is that there is a general tendency to go along with liberals on the issues they care most about, except when the conservative base forces them to behave differently, in which case they do. Hence, Republicans do a lot on abortion and little on affirmative action. The key difference is not how liberals in the media react to the issue, but internal dynamics within the conservative movement.
Leaders can also be entrepreneurial and put issues on the agenda that elites would rather ignore. This is what Trump did with immigration in the 2016 primaries, tapping into a cause that put him on the same side of the base against the party elite. In the process, he transformed the elites by forcing them to be responsive to the base on this issue, and the days of party leaders like Bush and McCain pushing comprehensive immigration reform are long gone. Immigration worked as a kind of Magic Bullet in the 2016 primary, and was probably a net positive in the general election.
Republicans starting to run against affirmative action would likely be even more consequential for our politics.
A War on Data
Finally, it is important to consider what forms the fight against race and sex preferences should take.
It would be a mistake to simply focus on the narrow question of college admissions. In 1995, a conservative activist working in tandem with Bob Dole sent a letter to the Congressional Research Service “requesting a detailed account of every program on the federal books in which race or gender could be used as a factor in the selection process.” It came back with 160 programs. An enterprising politician could ask for a similar inventory today, whether at the state or federal level. This should be done with an eye towards publicly exposing and ultimately eliminating such programs.
Moreover, I have previously recommended resurrecting causes that Ronald Reagan fought for, namely attacking the Civil Rights Restoration Act and modifying the executive order requiring affirmative action in government contracting.
Finally, there should be an attack on the process of data generation itself in the form of a new “ban the box” movement. Race should be treated like religion, in that the general presumption is that it is none of the business of the government.
This helps solve the problem of government bureaucrats trying to get around affirmative action bans. For example, the University of California system is not allowed to use race as a factor in college admissions. But it recently filed an amicus brief with the Supreme Court arguing that Proposition 209 hurt its ability to recruit a diverse student body, and detailing the steps it is taking to admit more blacks and Hispanics. This may not violate the letter of the law, but it sure seems to violate its spirit, as it is still taking action in order to achieve a desired racial balance. If colleges and universities were prohibited from collecting the data necessary to know how diverse they are in the first place, it wouldn’t even be able to do this much. Preventing government and education bureaucrats from acquiring demographic information – with a few carefully carved out exceptions of course for things like medical research – is akin to taking keys away from a drunk, as if you give them any discretion on these matters, they’re going to use it.
Aiming at data collection makes it difficult for Democrats to obfuscate on what affirmative action is or how government uses race and sex to distribute benefits and impose costs. If such individual characteristics don’t already factor into government initiatives, then there is no harm in preventing the collection of data. In addition to making the argument for equal treatment, this can be framed as a war against bureaucracy and filling out paperwork, which everyone hates.
The way we collect data forms the entire underpinnings of wokeness as government policy. We worry about gaps between “whites” and “Hispanics,” but not “Catholics and Protestants” or “Arabs and Norwegians,” because of the ways in which government has decided to divide and classify its citizens.
Bureaucracy runs on numbers. While markets have the price system to direct behavior, government tries to meet benchmarks it sets for itself. What state officials are encouraged or allowed to measure shapes where they direct their efforts. Without the ability to collect data, the whole system of wokeness as law falls apart.
This is why France, which bans the collection of demographic statistics, cannot practice affirmative action or have policies targeted to increase the representation of certain groups in high status professors or reduce achievement gaps. According to the Harvard International Review, “the lack of racial data collection has left citizens of France in the dark about the rates at which people of color are stopped and searched by law enforcement, rates of workplace and housing discrimination, and rates of death due to COVID-19.” If only our government was similarly blind, it could do a lot less damage.
Race and sex preferences can be seen as a sort of Achilles heel for the left. Republicans have long ignored an issue that, if taken up, could see them act on their principles and satisfy their base, while also being on the right side of public opinion and making a major change in how the country is governed.
Richard Hanania's Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.