Why Affirmative Action is Critical to an Equal Society

Both sides of the affirmative action debate are arguing the wrong question.

Opponents look at LSAT and MCAT averages, see a gap, and call admissions a violation of merit. Proponents argue for affirmative action as reparations, as fairness, as a corrective for historical and ongoing discrimination. The remediation case is real and was the original purpose. But by itself it loses the public argument, because it keeps the debate stuck on whether the country owes a debt instead of what the country loses when it doesn’t pay it.

What it loses comes down to two things. Everything else is downstream.

The Two Reasons That Actually Matter

1. The professions need to serve the country, and patients and clients get measurably worse outcomes when their doctor or lawyer doesn’t share their background.

2. The classrooms training those professionals need to be diverse, because future doctors and lawyers can’t learn to serve a diverse country if their entire training cohort comes from one demographic.

That’s the whole case. Everything else in this post is the evidence behind those two claims, and a breakdown of why the public argument keeps stalling out.

Reason #1: Why Diversity in the Profession Matters

Same-race providers save lives. The data isn’t subtle.

In 2018, Marcella Alsan and colleagues ran a randomized controlled trial in Oakland. 1,300 Black men were randomly assigned to either Black or non-Black doctors at a free screening clinic.

What happened:

• The men assigned to Black doctors took up substantially more preventive care
• Especially the invasive stuff: flu shots, diabetes screening, cholesterol screening
• The men with the least trust in the medical system benefited the most
• The researchers calculated that Black doctors could close 19% of the Black-white cardiovascular mortality gap

For newborns, a 2020 PNAS study of 1.8 million Florida births found that under white physicians, Black newborns died at 430 more per 100,000 than white newborns. Under Black physicians, the gap was 173 per 100,000.

⚠️ A 2024 reanalysis controlling for very low birth weight weakened that finding substantially. So take the strong version of the infant mortality result with caution. The weaker version (that some of the gap closes with race concordance) is consistent with the broader literature.

The mechanism is trust and communication. A patient who doesn’t disclose key symptoms doesn’t get the workup. A patient who doesn’t fill the prescription doesn’t get the benefit. Concordance moves both.

And the bias runs the other way.

Cross-race providers don’t just fail to help. Sometimes they actively hurt.

The Hoffman et al. 2016 PNAS study surveyed 222 white medical students and residents. About half endorsed false beliefs about biological differences between Black and white people. Things like:

• Black people’s skin is thicker than white people’s
• Black people’s blood coagulates faster
• Black people feel less pain

Those who endorsed more of those beliefs rated a Black patient’s pain as lower than the same pain in a white patient. And they made less accurate treatment recommendations.

This isn’t a one-study fluke. Black Americans are systematically undertreated for pain across multiple clinical contexts. In one study of cancer patients cited in Hoffman’s research, just 35% of racial-minority patients with metastatic or recurrent cancer received pain prescriptions appropriate under WHO guidelines, versus 50% of non-minority patients.

The obstetrics data is even uglier:

• Black mothers die from pregnancy-related complications at 3-4x the rate of white mothers
• That disparity exists regardless of education or income
• Per the CDC, Black women with a college degree or higher die from pregnancy-related causes at 1.6x the rate of white women with less than a high school diploma
• Serena Williams nearly died postpartum because her doctors initially dismissed her concerns about a pulmonary embolism

If income and education explained the gap, Serena Williams would have been fine. She wasn’t. Something else is doing the work, and the evidence converges on bias plus communication failure plus race-correction algorithms baked into clinical tools (eGFR for kidney function, VBAC calculators, pulmonary function tests, and others have historically used race adjustments that can delay or alter care).

Minority providers serve the communities that need them most.

A 1996 NEJM study of California found:

• Communities with high proportions of Black and Hispanic residents were 4x more likely than others to have a physician shortage, regardless of income
• Black physicians practiced in areas where the Black population was nearly 5x higher on average than where other physicians practiced
• Black and Hispanic physicians disproportionately served poor and uninsured patients

A 2012 California study (Walker et al.) of 48,388 physicians across specialties confirmed the same pattern, even among specialists, not just primary care.

So when you reduce the pipeline of Black and Hispanic medical students, you don’t just hurt those individuals. You drain the supply of doctors who would have served the communities with the worst existing shortages. Those communities lose providers twice. Once through the admissions decision. Again through the practice-location decision minority physicians disproportionately make.

The legal field has the same problem.

The empirical literature on attorney-client concordance is thinner than the medical literature. But the structure is identical:

• Black and Hispanic communities have the most contact with the criminal legal system, family court, immigration court, and the eviction docket
• They face a profession that is 78% non-Hispanic white
• People going through deportation, custody disputes, criminal charges, and eviction are routinely represented by lawyers who don’t share their cultural context, dialect, or experience of police and institutions

The downstream effect is the same. Less trust. Worse communication. Worse outcomes.

Reason #2: Why Diversity in the Classroom Matters

This is the piece almost nobody in the debate talks about. The case for diversity in medical and legal education isn’t just about who graduates. It’s about what they know when they graduate.

Research central to the Grutter ruling found that diversity in higher education classrooms produces:

• Better active thinking
• Higher intellectual engagement
• Stronger preparation across viewpoints
• Better consensus-building

Apply that to professional training and the implication is direct.

✅ A medical student who never sat next to a Black classmate, never heard a Latino classmate explain a family’s experience with diabetes management, never had an Asian classmate push back on a cultural assumption baked into a case study?

That student is a less-trained doctor by graduation.

✅ A law student who never argued constitutional questions across racial lines in a small section, never workshopped a brief with a clinical partner whose family had been through deportation proceedings?

Less-trained lawyer.

This is the case people dismiss as feel-good. It isn’t feel-good. It’s pedagogy.

Homogeneous classrooms produce graduates who haven’t been challenged on assumptions their future patients and clients will live by. Those graduates become the doctors who code Black chest pain as drug-seeking. The lawyers who misread silence in a Latino client.

The diversity isn’t the point. It’s the training.

The Admissions Data People Argue About

OK, now the data. This is the supporting evidence, not the main argument.

Average score gaps by race

LSAC and AAMC publish means by race; medians by race aren’t publicly available. For tests with a fairly symmetric distribution, mean and median are close.

MD program acceptance rates, AAMC 2024-25

What the “merit” crowd seizes on:

Within any given score band, Black applicants were more likely to be admitted than higher-scoring Asian or white applicants. For applicants in the 2013-2016 window (old MCAT scale) with average GPAs (3.40-3.59) and average MCATs (27-29):

• Black applicants were ~4x as likely as Asian applicants to be admitted
• Black applicants were ~3x as likely as white applicants to be admitted

📌 Both things are true at the same time:

• Aggregate acceptance rates are lower for Black applicants because the Black applicant pool scores lower on average
• Within score bands, Black applicants were strongly preferred

The right-tail nuance worth knowing: the gap between racial groups is widest at the top of the score distribution. With Black and white mean LSAT scores roughly 10-12 points apart for 25+ years running, the pool of Black applicants scoring at top-14 medians (170+) is small. That’s the arithmetic reason race-conscious admissions had their biggest proportional effect at elite schools.

Population vs. profession representation

So the population that has the most medical need, the most legal exposure, and the longest history of being mistreated by both systems is served by a workforce that doesn’t resemble it. That’s the policy problem affirmative action was responding to.

It’s also the gap that gets worse when the pipeline shrinks.

This Isn’t a Snapshot. It’s a 25-Year Standstill.

Here’s the part that gets lost when people argue about a single year of admissions data.

These gaps aren’t new. They aren’t the residue of a recent policy change. They aren’t something that’s in the process of self-correcting if we just leave it alone.

They’ve been frozen, across multiple measures, for decades.

The LSAT score gap has been stuck for 25+ years

Sources: Journal of Blacks in Higher Education (1998, 2004); Brookings analysis of 2013-14 LSAT data; AccessLex/LSAC reporting 2017-2023.

A quarter-century. The gap hasn’t moved.

That’s the central fact people on both sides of this debate keep dancing around. The “merit” argument acts like the gap is about to close on its own if we’re patient. The progressive argument sometimes acts like the gap is recent or accelerating. Neither is true. The gap is stable. Stability across this many years is its own data point.

Black share of physicians: 120 years of glacial movement

Source: Ly (JGIM 2021), using Decennial Census and ACS data; AAMC 2024 active physician workforce.

Over 120+ years, the Black share of physicians moved from 1.3% to 5.3%. Most of the gain came between 1940 and 2018. From 2018 to 2024 the share declined. And at no point has it come close to the Black share of the U.S. population.

Now zoom in on Black male physicians: 2.7% in 1940. 2.6% in 2018.

Negative progress over 78 years.

For most of a century, the line has been almost flat — and in the most recent window, it bent the wrong way.

Black share of lawyers: completely flat at 5% while everyone else grew

Source: ABA Profile of the Legal Profession, multiple years. Asian American lawyer counts pre-2023 are undercounted because California did not report attorney race/ethnicity until 2023.

In the same 14-year window:

• Lawyers of color nearly doubled from 12% to 23% of the profession
• Asian American lawyers grew sharply (partly from California finally reporting data)
• Hispanic lawyers grew by half (4% → 6%)
• Black lawyers stayed at 5%

That’s not a system slowly converging on representation. That’s a system where every other underrepresented group is moving and one isn’t.

What the trend tells you

If the score gap has been frozen at 10-11 points for a quarter century, the right-tail gap can’t mathematically be closing.

If the Black physician share has barely moved in 120 years, the pipeline isn’t about to self-correct in the next decade.

If Black lawyers have been stuck at 5% while every other group expanded, then something specific to that pipeline is sticking, not “the system is gradually working it out.”

This is what stable inequality looks like. The political framing keeps shifting (Bakke, Grutter, Fisher, SFFA, post-SFFA, the Trump executive orders). The underlying numbers don’t.

You can argue about why. You can argue about what to do about it. But you can’t argue that the underlying problem is shrinking. It isn’t.

What the Numbers Don’t Measure

Set the racial-loading question aside for a moment.

Even if the LSAT and MCAT were perfectly fair tests, they still wouldn’t measure who becomes a great lawyer or doctor.

They measure a slice of academic readiness: test-taking under time pressure, reading speed, pattern recognition, content recall, reasoning on artificial problems. That slice matters. It is not the job.

What the LSAT doesn’t measure

• Whether a client will trust you enough to tell you the fact that changes the case
• Whether you’ll spot the issue the senior partner missed because you grew up around the industry it involves
• Judgment under uncertainty
• Listening, interviewing, factual investigation
• Moral courage and strategic restraint
• The ability to translate a human life into institutional language
• Whether you’ll still want to be a lawyer at year five

What the MCAT doesn’t measure

• Whether a patient will trust you enough to disclose the symptom they’re embarrassed about
• Whether they’ll take the medication
• Whether they’ll come back for the follow-up
• Whether you’ll catch the diagnosis that keeps getting missed in patient populations the system has failed
• Bedside manner, diagnostic humility, cultural fluency
• Teamwork under pressure
• The ability to keep someone alive at 3 AM

These are the actual work.

Even U.S. News knows this

The single most numbers-obsessed actor in legal education is U.S. News & World Report. That ranking is what every law school dean optimizes for. And even they don’t weight admissions credentials anywhere close to where the “merit” obsessives think they should be.

Here’s the actual 2025-26 U.S. News law school rankings methodology:

Source: U.S. News & World Report 2025-26 Best Law Schools methodology.

Run the math:

• Outputs (employment + bar passage): 58% of the score
• Inputs (LSAT + GPA + acceptance rate): 10% of the score

The single most influential ranking system in legal education weights what graduates do nearly six times more heavily than the credentials they brought in with.

That’s not a moral statement. It’s not profound. It’s an admission of the obvious: the point of a law school is to produce lawyers who pass the bar and find legal jobs. Not to assemble the highest-scoring class possible.

NALP reported that 93.4% of the Class of 2024 secured jobs within 10 months of graduation, with 84.3% in bar-admission-required positions, both all-time highs. The profession is judged on output, not input.

And here’s the kicker for this whole debate: the racial disparities continue after law school, when the LSAT is no longer in the equation. Among the Class of 2024, Black graduates secured bar-admission-required jobs at 74.3% versus 86.5% for white graduates. Graduates of color made up 33.7% of the class but obtained only 27.1% of judicial clerkships. The gap doesn’t disappear when you take the test out, which says something about where the gap lives.

Why this matters for affirmative action

The hidden premise of the anti-affirmative-action argument is:

Higher score = more deserving = better future lawyer or doctor.

That chain breaks in multiple places.

Higher score may mean stronger test performance. It may mean slightly better odds of first-year grades. It may mean marginally better bar-passage odds, controlling for the school attended. It does not automatically mean better lawyer, better doctor, better advocate, better surgeon, better client counselor, or better public servant.

And if a slightly lower-scoring applicant is more likely to serve underserved patients, speak the language of a client population, build trust across the lines the profession has historically failed across, or do the work the system needs done, then that applicant is more professionally valuable for the institution’s mission.

Not as charity. Not as decoration. As function.

Scores matter to the extent they help identify people who can do the work. They do not define the work.

Why the “Merit” Frame Is Broken

Here’s the part the anti-affirmative-action argument refuses to engage with.

The LSAT and MCAT don’t measure innate ability. They measure the accumulated effect of 18+ years of opportunity, in a country where opportunity is racially stratified by history and ongoing policy.

A few brutal facts:

• 📊 Per Chetty et al. (Opportunity Insights), nearly a third of students from the top 1% of households score 1300 or higher on the SAT, compared to just 2.5% from low-income families
• 📊 By the first day of kindergarten, measurable differences in vocabulary, readiness, and preparation already exist between low-income and high-income kids
• 📊 Wealthy students attend better K-12 schools, hire test prep, and access enrichment that lower-income students cannot

Stratification by neighborhood, school funding, family wealth, and chronic stress all feed into the score. Admissions committees then treat that score as a clean thermometer reading of the applicant’s mind.

It’s closer to a thermometer reading of the applicant’s zip code, with race-of-zip-code baked in by housing policy going back generations.

So the “merit” position has a hidden premise: that the LSAT and MCAT are race-neutral. They aren’t. They’re noisy measures of upstream inputs that are themselves racially distributed, plus a real but incomplete signal of individual ability layered on top.

If the inputs that produce the score are racially loaded, admitting purely on the score is racially loaded by construction.

Calling that arrangement “merit” is a refusal to look at how the score is built.

One Opposing Argument Worth Naming and Dismissing

Richard Sander’s “mismatch theory” claims Black students at elite law schools pass the bar at lower rates than they would have at less-selective schools. The argument has been refuted multiple times. Ayres and Brooks showed it doesn’t reduce the total number of Black lawyers. Alon and Tienda showed minority graduation rates rise with institutional selectivity. Thaxton showed Black and Latinx students pay a “race tax” from the cumulative cost of navigating constant racial stereotypes in white institutional spaces, which mismatch theory can’t capture. It’s the strongest opposing argument and it’s still weak.

Post-SFFA: The Natural Experiment Is Running Right Now

The 2023 Students for Fair Admissions ruling ended race-conscious admissions in higher education. The first post-ruling data is in.

Medical school enrollment, first post-SFFA cycle:

Law school:

• Harvard Law’s Black 1L enrollment hit a 60-year low in the first post-SFFA class
• National 1L diversity held roughly steady (7.71% Black in 2024 vs. 7.84% in 2023)
• The decline is concentrated at the most selective schools

The professions need graduates trained at every tier of medical and law school. Lawyers and doctors come from every tier. Cutting the pipeline anywhere shrinks the pool of fully-trained Black and Hispanic professionals available to serve the patients and clients they’re best positioned to reach.

Final Word

Affirmative action wasn’t about lowering standards. It wasn’t about quotas. It wasn’t about racial preferences for their own sake.

It was about two functional problems:

1. The country needs lawyers and doctors who can serve the country, not just the parts of it that look like the current workforce.
2. The schools training those lawyers and doctors need to produce graduates who can competently practice in a multiracial country, not just graduates who passed a test.

Strip away the noise and that’s the whole case.

Applying to law school in the post-SFFA landscape? Reach out to work with me directly.

Related Reading

→ What Holistic Law School Admissions Really Means
→ How to Get Into Law School Below Both Medians
→ Your Softs Are Probably Better Than You Think
→ Diversity Statement Guide
→ Sample Diversity Statements
→ Blog Directory