Fifteen Questions About Prop. 16 and Prop. 209

The extraordinary protests and marches that swept the United States during the late spring, in the wake of the death of George Floyd, centered on calls for racial justice, but specific proposals to define and achieve racial justice were scarce.  Aside from calls to “defund” or at least “remake” police departments, the demands of many protesters seemed to boil down to an insistence that American institutions achieve something like proportional racial representation. One of the most tangible legislative proposals to flow from this, and one that is now playing out in California’s November election, is Proposition 16 (Prop. 16).  In the Essay below, I imagine a sort of conversation between myself and someone trying to understand why Prop. 16 is on the ballot, and whether it is a good and useful idea.

1. What are Prop. 16 and Prop. 209?

California is one of many states where amendments to the state constitution must be approved by popular vote.1   In 1996, California voters approved, by a nearly 55-to-45 margin, Proposition 209 (Prop. 209), which amended the state constitution to provide, “the state shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.”  The measure followed the Regents of the University of California (UC Regents) passing in 1995 two measures, SP-1 and SP-2, that placed similar limitations on the university system.  These measures ignited what had been a smoldering national debate over the proper limits of affirmative action.  Both the UC measures and Prop. 209 were strongly supported by Pete Wilson, California’s popular Republican governor who, for a while, was a leading contender for the 1996 Republican presidential nomination.  The central organizing force behind both SP-1/SP-2 and Prop. 209 was Ward Connerly, an eloquent African-American businessman whom Governor Wilson had appointed to the UC Regents in 1993.2   After the Prop. 209 victory, Connerly started a national movement to ban the use of racial preferences, leading several other referenda campaigns.  By 2010, ten states or statewide university systems had adopted similar bans, though none have since that time.3

UC administrators almost universally opposed both SP-1/SP-2 and Prop. 209, and their opposition did not noticeably decline after the measures became law.  In 1998, California elected its first Democratic governor in sixteen years, and the composition of the UC Regents (who were appointed by the governor and the state legislature) gradually became more liberal.  Governor Arnold Schwarzenegger (2003–2010) and Governor Jerry Brown (2011–2018) continued to appoint moderate-to-liberal regents.  In 2001, at the urging of UC’s administrative leadership, the Regents revoked SP-1 and SP-2, though this had no immediate practical effect because Prop. 209 had superseded both measures. 

In 2012, again with the support of UC administrators, State Senator Edward Hernandez introduced “California Senate Constitutional Amendment 5,” which proposed to repeal Prop. 209.  When the Senate passed the measure in January 2014, it precipitated widespread organizing and mass meetings within California’s Asian-American (and, particularly, its Chinese-American) community.  Backers of SCA-5, taken aback by this opposition (and by the state Republicans’ emerging alliance with the opponents) withdrew the amendment before it came to a vote in the State Assembly.

In January 2019, opponents of Prop. 209 tried again, this time with much more preparation.  Three Assembly members introduced “ACA-5,” a measure very similar to SCA-5, that also proposed to amend the California Constitution and eliminate Prop. 209’s anti-preference and antidiscrimination provisions. ACA-5’s backers had lined up a wide range of supporters, including, notably, some Asian-American civil-rights organizations.  Defenders of Prop. 209 did not notice the new measure until March 2020, when it sailed through its first committee hearing.  COVID-19 shutdowns substantially inhibited the ability of ACA-5 opponents to follow their 2014 playbook, so opposition was muted.  ACA-5 was adopted by both the Assembly and the State Senate in June 2020 by more than the required two-thirds vote, with Democrats giving unanimous support.  That put the measure on the November 2020 ballot, and the Secretary of State, within a few weeks, created the specific wording of Prop. 16.

2. Didn’t something like this happen in Washington State last year?

Yes.  The State of Washington was the second, after California, to adopt a ban on racial preferences.  Its “Initiative 200” passed in 1998 with a 58-percent majority and, like Prop. 209, faced significant opposition and resistance from leaders at the University of Washington and in state government.  In 2019, the Washington state legislature approved “Initiative 1000,” which would have repealed the ban on racial preferences.  Opponents to Initiative 1000—again, led by Asian Americans—gathered enough signatures to put a referendum on “Initiative 1000” on the November 2019 ballot.  By a narrow margin (50.5 percent to 49.5 percent), voters rejected Initiative 1000, keeping the racial-preference ban in place.

3. Why has the Prop. 16 debate centered on the University of California?

This is certainly somewhat puzzling.  Prop. 16 would affect all state and local government activities and programs in California, and UC accounts for only about 3 percent of state expenditures.  On the other hand, debates about public-sector affirmative action have always been centered on elite higher education (witness the Supreme Court cases involving the University of Texas (2013, 2016) and the University of Michigan (2003), and the current Harvard lawsuit before the First Circuit).4 The University of California is in many ways the jewel of public universities in America.  Moreover, these debates are dominated by California “elites”—legislators, journalists, lawyers, professors—who often have ties to the University of California or hope to send their children there.

4. Why is the “short-form” wording of Prop. 16 controversial?

Those involved in these debates have long known a basic fact:  the public broadly supports “affirmative action” efforts aimed at helping disadvantaged groups but opposes “racial preferences.”  A 2018 Gallup poll asked Americans whether they “generally favor or oppose affirmative action programs for racial minorities.” Sixty-one percent of respondents said they were in favor, while 30 percent were opposed.  A 2019 Pew survey found that 75 percent of Americans thought it was “very or somewhat important” for “companies and organizations to promote racial and ethnic diversity in their workplace.”  But consider, in contrast, this question posed by Pew:  “When it comes to decisions about hiring and promotions, do you think companies and organizations should take a person’s race and ethnicity into account, in addition to their qualifications, in order to increase diversity in the workplace, [or should they] only take a person’s qualifications into account, even if it results in less diversity in the workplace?”  Seventy-four percent of respondents said only qualifications should play a role.

Are people just confused, or is there a real difference between “racial affirmative action” and “racial preferences”?  There is a difference.  “Affirmative action,” as originally conceived and implemented by various federal agencies in the early 1960s, focused on trying to improve the pool from which people were chosen for a job, a contract, or a university slot.  This meant bypassing the “old boy network” through outreach, making sure that every potential applicant had the information and tools to understand potential opportunities and compete for them.  Preferences entered the picture at the end of the 1960s and early 1970s, when many critics said that traditional affirmative action was not producing adequate results.

The difference in public support for “affirmative action” versus “preferences” is all-important in the Prop. 16 campaign.  If proponents can persuade voters that Prop. 16 is a measure to permit “affirmative action,” then they are likely to win.  If opponents can persuade voters that Prop. 16 will institutionalize racial preferences, then the measure will likely fail.  Naturally, both sides had their own view of how the Secretary of State should summarize Prop. 16 on the ballot.

The “ballot title” of Prop. 16 was not a complete win for either side.  It states that the measure would “allow diversity as a factor in public employment, education, and contracting decisions.” Someone reading the more detailed language that follows can figure out that Prop. 16 would allow the government to use racial and gender preferences, but it is not obvious.

5. Isn’t the difference between “affirmative action” and “racial preferences” somewhat academic?  Didn’t Prop. 209, so far as the University of California is concerned, effectively get rid of both?

Absolutely not—and that is key to understanding this entire issue.  In the years before Prop. 209, UC engaged in relatively little outreach to disadvantaged communities.  It simply augmented its normal admissions pool (under which the top eighth of California high school students are automatically eligible for admission to some UC school) with a race-conscious “special admissions” procedure.  The result was that the pool of Black and Hispanic applicants, and the number of minority students admitted to UC, was stagnant or, in the case of Black admits, actually declining from 1989 through 1997 (see Tables 1 and 2, below).

In 1997, faced with the loss of the ability to use racial preferences, UC embarked on a massive, multiyear effort to build bridges to disadvantaged schools across California.5   It started counseling low- and moderate-income students on how to complete the necessary high school course requirements to be UC-eligible; many UC colleges formed partnerships with underperforming high schools; and the university began to systematically take individual disadvantage (e.g., low socioeconomic status) into account in admissions decisions.6   In other words, ending racial preferences ushered in the beginning of traditional “affirmative action” strategies at UC—which, by the way, were perfectly legal under Prop. 209.  Of course, these outreach efforts could not be racially discriminatory—Asian-American students, for example, could not be barred from counseling sessions—but they were sufficiently targeted so that the overwhelming majority of beneficiaries of the outreach efforts were Black and Hispanic.

The university reformed admissions procedures as well.  Some UC campuses had probably used socioeconomic preferences to some degree before Prop. 209, but post-Prop. 209, these preferences were somewhat expanded and systematized.  Such preferences were, of course, completely legal, and although these preferences disproportionately benefited Blacks and Hispanics, they changed the focus of admissions officers from “group representation” to “individual hardship” and “obstacles overcome.”  By the early 2000s, over a third of UCLA freshmen had received Pell Grants, which are generally available to persons in the bottom half of the income distribution, compared to around 10 percent of freshmen at elite colleges more broadly.

6. Did the new outreach efforts make a tangible difference?

Yes, indeed.  They produced a dramatic increase in applications from low- and moderate-income students, especially from Blacks and Hispanics.7   While Black applications to UC were 2,191 in 1989 and 2,141 in 1997 (the last year before Prop. 209 went into effect), they had risen to 3,307 by 2006 and 5,978 by 2013—a stunning turnaround (see Table 1).  Hispanic applications rose at a still faster rate, even when one takes into account the growing share of the population that was Hispanic. 

Unique applications from California high school students to UC in year by race.  Source:  UC Application, Admission and Enrollment of California Residential Freshmen for Fall 1989 through 2013 (UCOP, 1/14)

In short order, these changes produced similar, though less dramatic, increases in underrepresented minority (URM) enrollment.  There was a short-term drop in both Black and Hispanic enrollment (from 1997 to 1998), as UC got its new outreach and socioeconomic-status (SES) initiatives underway. But by 2000, enrollment levels for these groups had recovered and, within a few years, reached new highs.

Total UC freshman enrollment from California high schools in year by race. Source: UC Application, Admission and Enrollment of California Residential Freshmen for Fall 1989 through 2013 (UCOP, 1/14).  Columns do not sum to total because some smaller racial categories are omitted.

7. Yet one regularly hears that minority enrollment at Berkeley and UCLA crashed after Prop. 209.

First-year URM enrollment had been declining at Berkeley and UCLA even before Prop. 209, and declined further afterwards (see Table 3).  Prior to Prop. 209, those two campuses used much heavier racial preferences than the other six campuses did.  The Black-White gap in SAT scores at Berkeley in the early 1990s was 288 points, according to an analysis published in the Journal of Blacks in Higher Education.  For URMs as a whole in the mid-1990s at Berkeley and UCLA, about 90 percent of Asian-American applicants had stronger academic credentials than the median URM applicant, but the URM admissions rate was double the rate for Asian Americans.  Given these differentials, it was inevitable that Black and Hispanic numbers at these two campuses would decline under a race-neutral regime. 

Total UC freshman enrollment from California high schools at Berkeley and UCLA by URM race and year.

Total UC freshman enrollment from California high schools at all other UC campuses by URM race and year. Source: UC Application, Admission, and Enrollment of California Residential Freshmen for Fall 1989 through 2013 (UCOP, 1/14)

But I believe this was a good thing.  In the first place, the ultra-aggressive preferences at Berkeley and UCLA in the early- and mid-1990s had the effect of concentrating half of all Blacks and a third of all Hispanics in the UC system at these two campuses.  Relative to Whites and Asians, Blacks and Hispanics were more integrated across the UC system after Prop. 209 than before it.  Second, huge racial preferences created a big academic gap between URMs and non-URMs at Berkeley and UCLA.  From 1992 to 1994, Blacks had a 13.5-percent four-year graduation rate at UCLA—an appallingly low rate.8   In this environment, Blacks and Hispanics had much lower grades, were far more likely to drop out, and were much less likely to graduate on time (or to graduate at all) compared to Whites and Asian Americans.  All those gaps greatly narrowed after Prop. 209.

8. I think you’re now referring to the “mismatch” issue.

Yes.  The mismatch hypothesis suggests that students are academically harmed when they attend classes or schools where nearly all the other students have much stronger levels of academic preparation.9 Imagine, for example, a student who wants to become a chemist and maybe teach chemistry at a university.  Her high school grades and test scores are solid, but not stellar, and her high school offered Honors Chemistry but not Advanced Placement Chemistry.  She nonetheless receives a preference to UC Berkeley and is surrounded by classmates who do have stellar grades and test scores and scored “5s” in AP Chemistry.  She might be inspired by the competition and rise to the challenge, but she is much more likely to be flummoxed.  She has difficulty keeping up in the first-year chemistry course, gets a mediocre grade the first quarter, and falls even further behind in the second quarter.  This student is, at a minimum, likely to be dissuaded from the idea of an academic career and switch from chemistry to a less difficult field. At worst, she may drop out of college, or take six years to attain a degree in a new field, and have a mediocre transcript.  This same student at a somewhat less competitive but still excellent school, such as UC Santa Barbara, is much more likely to thrive, stick with chemistry, and achieve her academic aspirations.10

Mismatch has been documented in several careful studies by eminent social scientists whose work remains unrebutted.11   Prop. 209, as you can imagine, provided a great “natural experiment” for assessing the mismatch hypothesis.  Unfortunately, UC has played politics with the issue, giving only limited data to independent scholars while giving access to “secret” data to in-house researchers.  The independent scholars have nonetheless produced high-quality research that has been published in top-flight social science journals, like the American Economic Review, and their findings come down heavily in favor of the mismatch hypothesis.12   Some of the in-house scholars have disagreed, but they have not made their data, or the basis for their conclusions, public and available for replication.

Academic findings aside, the numbers tell such an overwhelming story that they largely speak for themselves.13   The four-year graduation rate of URMs in the UC system nearly doubled in the ten years after Prop. 209’s adoption.  The number of African-American graduates slightly dipped initially (again, before UC’s new outreach efforts kicked in), but had risen to a record level by 2006 (roughly corresponding to students entering in 2001), and by 2017 the number of Black UC graduates had risen 70 percent above pre-Prop. 209 levels.  A higher proportion of these graduates were in the sciences, so that the number of Black STEM UC graduates rose from an annual average of around 200 before Prop. 209 to 510 in 2017 and 558 in 2018.  The numbers for Hispanics were even more impressive: the number of Hispanic UC graduates quadrupled over the twenty years after Prop. 209, and the number of Hispanic UC graduates in STEM fields quintupled.

In short, within a few years of Prop. 209’s implementation, URMs were flourishing at UC, in both absolute and relative terms, to a degree they never had before.

9. If these numbers are so impressive, why does one continually hear that URMs are “underrepresented” at the university?

Much depends on what you mean by “representation” and why you are asking the question.  UC is, by design, intended for the most academically successful high school students.  Under the California Master Plan adopted in the 1960s, UC was required to admit students from the “top eighth” of California high school graduates.  The Cal State system, an equally large system with many outstanding programs, is required to admit students from the “top third.”  And the California community college system provides even broader access, admitting anyone who meets fairly basic academic requirements. Students who do well at the community-college level transfer in large numbers to UC to finish their degrees.

Now, this top-eighth requirement for UC admission had been broadened in various ways over the years, often to make sure that the “elite” schools are not unfairly excluding people who have not had the same access to AP courses or other advantages.  But the UCs remain selective campuses—in some cases, highly selective—and if one is evaluating the pool from which UC should admit, it only makes sense to apply some academic criteria to the definition of that pool. 

The tables below illustrate what I am talking about.  Table 4 shows the racial makeup of California high school graduates, and UC freshmen admitted from California, in 1997 and 2017.  The ratios in columns (c) and (f) show “relative representation” when one considers this whole universe of high school graduates.  (For example, the first number in column (c), .51, is the ratio of column (b) to column (a), and means that, in the late 1990s, the average Black high school graduate in California had 51 percent as great a chance as a random high school graduate of attending UC as a freshman.)  By this measure, the representation of URMs went up sharply between 1997 and 2017, though their level of representation was still far below that of Asian Americans.

Tables 5 and 6 add academic context to the demographic data.  Using data from the College Board, I compare UC students with the pool of high school graduates who are in the top half, top third, and top eighth of California high school students, for both 1997 (Table 4) and 2017 (Table 5).  Several striking patterns stand out.  First, representation of URMs as a whole at UC improves markedly from 1997 to 2017, by any measure.  Second, Black and Hispanic representation numbers converge between 1997 and 2017—tangible evidence that UC moved from primarily racial preferences (which favored Blacks) to socioeconomic preferences (which favored both Blacks and Hispanics, more or less equally).  Third, by any academic measure, Whites emerge as the most underrepresented group in 2017.  And fourth, when we use the top-eighth measure (in red in both tables), which is the measure most in tune with UC’s avowed academic standard for admission, Asian Americans are no longer overrepresented relative to URMs.15

The reason why the “representation” numbers change so much when we consider academic performance is that there are very large performance and preparation gaps along racial lines—not just in California, but in America generally.  It is fashionable to contend that these are simply reflections of poor schools in nonwhite communities, but very careful research has shown, for example, that two-thirds of the high school Black-White test score gap exists by the age of five.  Race itself does not explain the gap, but many factors that correlate with race create it, such as low birthweight, fewer books at home, and parenting practices.16

The apparent view of many advocates of Prop. 16 is that the problem is not with the academic preparation of students, but with the flawed measures that UC and other universities use to measure preparation.  This was a key sentiment behind the UC Regents’ decision, in April 2020, to forbid UC schools from considering SAT scores in the undergraduate admissions process.  This is a sadly misconceived argument, however, since many measures of high school achievement, including the National Assessment of Educational Progress, show the same racial gaps as the SAT.  Ignoring credentials ultimately means eliminating the academic selectivity of the UC admissions process, or creating serious mismatch problem for URMs, or both.

Before Prop. 209, I believe that people from low socioeconomic backgrounds were significantly underrepresented, even after adjusting for academic preparation.  UC was not doing enough outreach to find very promising students attending weak high schools.  Prop. 209, as I have explained, did a lot to solve that problem. Today, by any meaningful standard, it is, ironically, the Whites who are the most underrepresented group.  I daresay that very affluent Whites are still managing fine, but low-SES Whites—of whom there are many in California—are disproportionately overlooked by the current system.

10. One also hears that URMs—Blacks in particular—feel isolated on UC campuses, and that increased diversity would help.

Before Prop. 209 passed, one of the widespread predictions in circulation was that eliminating racial preferences would have a “chilling effect” upon URMs, particularly Blacks.  The idea was that they would perceive UC as less welcoming, and therefore not apply at all, or turn down admissions offers if they were accepted.  In fact, however, just the opposite happened, and in a big way.  As I’ve already discussed, the sheer numbers of Black and Hispanic applicants jumped dramatically after Prop. 209.  But even more surprising is that URMs became substantially more likely to accept admissions offers after race neutrality went into effect in 1998.  Economist Kate Antonovics and I investigated this some years ago, and we called what we found a “warming effect.”  URMs were more eager to attend UC after preferences were eliminated, and the size of the warming effect was directly proportional to how large the racial preferences had been before Prop. 209.

The obvious explanation of the warming effect—and one supported by lots of indirect evidence—is that URMs welcomed the idea of attending a campus where there would be no stigma attached to them as “only getting in through a preference.”  This makes common sense, of course—who would want to deal with the suspicion of getting in through the back door?  URMs probably also recognized that the value of a UC degree for Blacks and Hispanics went up after Prop. 209.  Employers often take one’s university as a proxy for ability—but since employers are aware that most elite universities use large racial preferences, they presumably discount the “credential effect” for Blacks and Hispanics graduating from elite schools.  After Prop 209, such discounting of a UC degree would make no sense.  Even high school students can appreciate this logic.  In any case, the warming effect was large and very real.  

11. You observed earlier that Blacks and Hispanics, at the group level, lag behind in high school academic achievement, and this impairs their numbers at UC.  How do we deal with that problem?

We start by acknowledging that it is a problem.  Many of the Prop. 16 proponents who demand proportionate representation at the university are pretending that the achievement gap does not exist.  Getting rid of SAT scores, for example, is a way of papering over the problem, not dealing with it.

One real solution is investing more in underperforming high schools—and California has actually done that over the past generation.  Consider this important but generally unheralded fact: the proportion of Hispanic high school students in California who graduate has risen from 66 percent to 88 percent over the past twenty years—a very big achievement unequalled anywhere else in America.  The proportion of Blacks achieving grades and test scores near the top of their high school classes has also increased steadily in California—and here again, that achievement has not been duplicated in most of the rest of the nation.  There is still a substantial achievement gap, but California—arguably because of Prop. 209, rather than in spite of it—has made very substantial progress.

12. Much of the frustration that has led to proposals like Proposition 16 stems from a feeling that levels of racial inequality in our society are simply not acceptable.  What would you do about that?

Thank you for asking.  There actually is an important solution that people are not talking about nearly enough, and that is reducing housing segregation.  Out of the nation’s sixty largest metropolitan areas, about ten have made substantial progress in lowering levels of Black/White housing segregation.  And in every one of those metro areas, Blacks have made dramatic gains in employment levels, income levels, health outcomes, life expectancy—and yes, test scores.  Several of these metro areas with lower segregation are in California, including San Diego and Sacramento.  Housing segregation in most of the United States is still very high, and it is as close to a genuine “root cause” of inequality as anything we are likely to identify.  It of course is the main source of school segregation; it separates minorities from suburban jobs; it concentrates them in areas with weaker health care systems; and it also seems to influence behavior and norms in counter-productive ways.  Moreover, there are a number of straightforward ways to reduce segregation, like providing mobility counseling that encourages people (of all races) to consider integrated neighborhoods, or helping to stabilize moderate-income housing in gentrifying neighborhoods.

So there are solutions to racial inequality.  But they do not lie in legalizing discrimination to create racially proportional university admissions.  That not only papers over the real problems, but also, as I have pointed out, hurts the people it is intended to help.

13. If Prop. 16 would be as counterproductive as you suggest, why are UC leaders supporting it?  

I can think of four reasons.  First is the problem of cancel culture—the fear of being ostracized if one says something that can be portrayed as racially insensitive.  This fear is greater in academia than in American society generally, and it is greatest in the ranks of university administrators.  Literally two hours before I started writing my answer to this question, a senior faculty member at UCLA—a nonwhite with liberal views—told me how much he appreciated that I spoke publicly about “things that most of us believe privately.”  Dozens of my colleagues have said similar things, and also often note that it is simply too difficult and too risky to express those views in public discussions.

Second, there is an unwillingness to admit a mistake.  Many UC administrators genuinely feared in 1996 that, without preferences, URM numbers at the university would tank and never recover.  They were very public about these fears, and continued to predict doom even after Prop. 209 passed.  They therefore became invested in confirming their own biases, ignoring the growing evidence that Prop. 209 was actually a boon for URMs at the university, and that large racial preferences had done great harm. The path of least resistance—especially for administrators—is to cling to the powerful “groupthink” that views Prop. 209 as a reactionary measure passed by social deplorables.

Third, UC administrators face relentless pressure from state legislators—especially a very powerful Latinx caucus—to increase URM numbers.  By 2010 or so, UC had done pretty much everything that could be done legally to raise those numbers.  With a little courage, those administrators would have confronted legislators and said, “Look, URMs are now substantially overrepresented by any reasonable measure of the qualified pool.  Doing more would violate the law and reintroduce the mismatch problem to the university.” But instead, UC officials have increasingly connived at reintroducing race in a variety of ways.  Now they are trapped, and Prop. 16 would give them a legal path forward.

And fourth and finally, there is a lack of imagination and courage in presenting proactive, forward-looking strategies to attack problems like housing segregation.  Both university leaders and political leaders have suffered from simple intellectual laziness in not thinking hard about root causes.

14. Do you think Prop. 16 will pass?

If you visit Ballotpedia and look at the list of Prop. 16 supporters, versus the opponents, it is quite overwhelming.  Dozens of organizations have endorsed Prop. 16, as have California’s two leading newspapers.  The supporters have raised millions of dollars; as of this writing, they have something like a 20:1 fundraising advantage over the opponents.  California is one of the most liberal states in the country, and the Floyd protests have given a new urgency to diversity goals.  So, many observers thought that Prop. 16 would be a shoo-in.

But in September 2020, during the writing of this piece, the Public Policy Institute of California released the first public opinion poll on Prop. 16.  In their results, only 31 percent of likely voters supported Prop. 16, while 47 percent opposed it and 22 percent were undecided.  Strikingly, a plurality of Hispanic voters opposed Prop. 16 as well. This suggests that a fair number of voters are aware that Prop. 16 is about racial preferences, not just about “affirmative action,” and that the long-standing popular skepticism about racial preferences still persists, even in California.

15. Does University of California’s experience with race-neutrality provide lessons that are relevant to the arguments of other contributors to this series?

I think so.  For example, the key point of Coleman Hughes’ essay is that there is a fundamental trade-off between “pro-equality” policies—treating everyone the same—versus “pro-equity” policies, which aim to achieve equal outcomes across groups.  Hughes is quite right that this distinction is important and is often lost (or deliberately obscured) in affirmative-action debates.  But UC’s experience under Prop. 209 shows that the two are not necessarily in conflict.  Implementing race-neutral admissions had three big effects: (1) it reduced mismatch for Blacks and Hispanics, thus greatly improving their outcomes at UC; (2) it made UC more attractive to them (the “warming effect”); and (3) it prodded UC to launch much better-funded and better-conceived outreach and remedial efforts, which were both lawful and disproportionately beneficial to Blacks and Hispanics.  UC’s graduation numbers before and after Prop. 209 leave no doubt that race-neutral admissions were a win-win for both equality and equity.

Daniel Ho and Alice Xiang argue that outcome-predicting algorithms that take race into account can improve fairness but are overly restricted by current interpretations of antidiscrimination law.  Here, too, the Prop. 209 experience—and research on college and professional school outcomes—is helpful in suggesting how their ideas can be constructively applied.  Careful research on college admissions shows that taking race into account in an algorithm does not improve our prediction of how students will perform in college or professional school, if we carefully control for their prior level of academic achievement and (sometimes) their socioeconomic status.  Since college admissions officers almost uniformly desire to use racial preferences (it makes their diversity efforts easier), the discipline of strictly race-neutral algorithms is more likely to achieve fairness and good results.

In contrast, there is no doubt that UC’s post-Prop. 209 outreach programs took race into account in coming up with “algorithms” (formalized or not) about where to target efforts; UC was particularly eager to make sure that it built the pipeline from Black and Hispanic communities to UC.  In a strict sense, this may have violated Prop. 209, but not in a way that ever did, or ever would be likely to, prompt litigation.  Indeed, creating a more formal algorithm to predict, today, where outreach is most likely to be effective (say, by measuring the increase in competitive applicants from a high school following a UC intervention) would be beneficial, and might actually reveal that UC is underinvesting in low-income White communities, whose members, as I showed earlier, are now most underrepresented at UC.

All this is to say that carefully studying and learning from UC’s actual experience with Prop. 209 provides rich lessons applicable to the full spectrum of discussions that aim to think clearly about, and apply fairly, the lessons of affirmative action. 

• 1See Cal. Const., art. II. See also The Constitutions of California and The United States, with Related Documents (2017–18 ed.).
• 2The story of Prop. 209 is recounted in Richard H. Sander & Stuart Taylor, Jr., Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It, ch. 7 (2012).
• 3Arizona was the last, when voters adopted Proposition 107 in 2010 by a 59–41 margin.
• 4These cases are Fisher v. University of Texas at Austin (Fisher I and Fisher II); Gratz v. Bollinger; Grutter v. Bollinger; and Students for Fair Admissions v. Harvard.
• 5See Thomas B. Timar, Rodney Ogawa & Marie Orillion, New Directions for Outreach: Report of the University of California Outreach Task Force (July 1997) (link); Karl Pister, UC Outreach: Systemwide Perspective and Strategic Plan (Sep. 1998); Denise D. Quigley & Seth Leon, The Early Academic Outreach Program (EAOP) and Its Impact on High School Students’ Completion of the UC Preparatory Coursework (Mar. 2002) (link).
• 6Kate Antonovics & Ben Backes, The Effect of Banning Affirmative Action on College Admissions Policies and Student Quality, 49 J. Hum. Res. 2 (2014).
• 7The Quigley and Leon report, for example, found that the Early Academic Outreach Program, launched by UC after Prop. 209, doubled the likelihood that disadvantaged high school students would complete the course requirements for UC admission.
• 8Data compiled by the author from UC Statfinder in 2010.  That website is now defunct; current data posted by the university does not extend back this far.
• 9Rigorous experimental demonstrations of the mismatch effect include:  Esther Duflo, Pascaline Dupas & Michael Kremer, Peer Effects, Teacher Incentives, and the Impact of Tracking: Evidence from a Randomized Evaluation in Kenya, 101 Am. Econ. Rev. 1739 (2011); and Scott Carrell, Bruce Sacerdote & James West, From Natural Variation to Optimal Policy? The Importance of Endogenous Peer Group Formation, 81 Econometrica 855 (2013).
• 10One of the most convincing demonstrations of the mismatch effect in the sciences is Frederick L. Smyth & John J. McArdle, Ethnic and Gender Differences in Science Graduation at Selective Colleges with Implications for Admission Policy and College Choice, 45 Research in Higher Ed. 353 (2004).  The best work on how mismatch erodes the academic aspirations of students receiving large preferences is Stephen Cole & Elinor Barber, Increasing Faculty Diversity, ch. 8 (2003).
• 11None of the studies in the preceding footnotes have been rebutted or even seriously critiqued, to my knowledge.
• 12Peter Arcidiacono, Esteban M. Aucejo & V. Joseph Hotz, University Differences in the Graduation of Minorities in STEM Fields: Evidence from California, 106 Am. Econ. Rev. 525 (2016); Peter Arcidiacono, Esteban M. Aucejo, Patrick Coate & V. Joseph Hotz, Affirmative Action and University Fit: Evidence from Proposition 209, 3 IZA J. Labor Econ. (2012).
• 13I summarize and document a number of these trends in a working paper on file with the University of Chicago Law Review. See Richard Sander, A Brief Commentary on Zachary Bleemer’s August 2020 Paper (working paper, 2020).  Also, a wide variety of summary tables on admissions, enrollment, and graduation patterns for both undergraduates and graduates in the UC system can be found online at UC Infocenter, though some work is required to ferret out the relevant data.
• 14Sources are census data and UC data.
• 15The methodology and sources for this analysis are elaborated in much more detail in a memorandum to the UC Regents that I prepared in September 2020.  That memorandum is on file with the Law Review.
• 16Roland G. Fryer & Steven D. Levitt, Understanding the Black-White Test Score Gap in the First Two Years of School, 86 Rev. Econ. & Statistics 447 (2004).