平权法案裁决之后：美国大学录取变了什么
一、2024 年 9 月，两组数字
2024 年 9 月，麻省理工学院公布裁决后第一届新生的族裔构成：非裔学生占 5%，亚裔学生占 47%。一年前，这两个数字分别是 15% 与 40%。同一天，耶鲁公布的数字走向相反：亚裔占比从 30% 降至 24%，非裔占比纹丝不动，仍为 14%。
这是 2023 年 6 月 29 日最高法院裁决后的第一届新生。裁决之后三年，各校的数据仍在变化，规则也在变化。
二、裁决：6 比 3，237 页
2023 年 6 月 29 日，美国最高法院就两宗合并审理的案件作出裁决：Students for Fair Admissions 诉哈佛大学（600 U.S. 181）与诉北卡罗来纳大学案。6 比 3，多数意见由首席大法官罗伯茨撰写。裁决认定，哈佛的录取政策违反《1964 年民权法》第六章，北卡的政策违反宪法第十四修正案平等保护条款。意见全文连同异议共 237 页。
多数意见给出三条理由：学校声称的录取目标不是"足够聚焦、可衡量"的目标；族裔被用作"负面因素"，且政策基于"同一族裔的学生想法相同"的假设—多数意见写道，"大学录取是零和博弈，给一些申请者好处必然以牺牲另一些申请者为代价"；学校声称的终点等于"赤裸裸的族裔平衡"，而族裔平衡早已被判违宪。多数意见总结："消除种族歧视意味着消除全部歧视。"
裁决没有正式推翻 1978 年 Bakke 案与 2003 年 Grutter 案，但以严格审查标准衡量，几乎无法通过。法院还驳回了学校"至少应允许用到 2028 年"的主张—Grutter 案曾说"25 年后将不再需要族裔偏好"，多数意见认为终点已经到了。裁决明确豁免军校，理由是军校"可能存在不同利益"。
多数意见结尾留了一个"开口"：裁决不应被理解为禁止大学考虑申请者在文书中讲述族裔如何影响了自己的人生，"无论是通过歧视、启发，还是其他方式"。索托马约尔大法官在异议开头写道："今天，本院挡住了数十年先例与重大进步的路……本院将肤浅的色盲规则奉为宪法原则。"
官司打了九年。原告 Students for Fair Admissions 由反平权活动人士 Edward Blum 创立，2014 年分别起诉哈佛与北卡，核心指控是哈佛在录取中故意歧视亚裔申请者，包括使用受族裔刻板印象影响的"个人评分"。2018 年 10 月的三周庭审中，地区法院的事实认定显示，亚裔申请者在学术与课外评分上高于白人申请者—60% 的亚裔申请者学术评分高于 3+（白人为 46%），28% 的亚裔申请者课外评分达到 2 或更高（白人为 25%）；但在"个人评分"上亚裔得分低于白人。2019 年 9 月 30 日，地区法院法官 Allison Burroughs 裁定哈佛胜诉；第一巡回上诉法院 2020 年维持原判。最高法院 2023 年的裁决没有直接处理"哈佛是否歧视亚裔"这一指控，而是以更宽泛的理由判定族裔考量本身违宪，下级法院"哈佛未非法歧视亚裔"的结论未被推翻。
三、三年数据
2028 届是裁决后第一届按新规则录取的新生。各校公布的数据均为入学新生占比，不是按族裔划分的录取率—没有一所顶尖大学公布过按族裔划分的录取率。
哈佛：非裔 18% 降至 14%，亚裔 37%（持平），拉丁裔 14% 升至 16%，8% 未披露族裔（上年 4%），国际生 16%。麻省理工：非裔 15% 降至 5%，亚裔 40% 升至 47%，拉丁裔 16% 降至 11%；非裔、拉丁裔与原住民合计从 25% 降至 16%。耶鲁：非裔 14% 持平，亚裔 30% 降至 24%，拉丁裔 19% 升 1 个百分点，白人 46% 升 4 个百分点。普林斯顿：非裔约 9% 微变，亚裔降约 2.2 个百分点（约 26% 降至 23.8%），拉丁裔约 9%。布朗：非裔 15% 降至 9%，亚裔 29% 升至 33%，拉丁裔 14% 降至 10%。宾夕法尼亚：历史上代表性不足群体从 25% 降至 23%。杜克：非裔 13% 持平，亚裔 35% 降至 29%，拉丁裔 13% 升至 14%。
口径警告：哈佛从 2028 届起改变了统计方法—族裔百分比的分母从全体新生改为"自报族裔的学生"，并回算了 2027 届数据。因此哈佛 2027 届亚裔占比有两个数字并存：旧口径 29.9%（占全体新生），新口径 37%（占自报族裔者）。跨年比较须注明口径；哈佛不公布白人占比。麻省理工招生主任 Stuart Schmill 明确表示，变化的原因就是最高法院的裁决。
变化并未在第一年定型。哈佛 2029 届（2025 年秋季入学，1,675 人）：非裔 11.5%（2028 届为 14%），拉丁裔 11%（从 16% 下降），亚裔 41%（从 37% 上升）；申请量 47,893 份，低于 2028 届的 54,008 份与 2027 届的 56,937 份；录取 2,003 人，录取率 4.18%（2028 届为 3.65%）。2029 届是疫情后首届强制提交 SAT/ACT 成绩的班级。普林斯顿 2029 届：非裔 5%，为 1968 年以来最低；亚裔从约 24% 升至约 27%。普林斯顿 2028 届几乎没有变化，变化滞后一年才出现。耶鲁 2029 届：非裔 12%（2028 届为 14%），亚裔占比上升；录取 2,227 人，申请 57,517 份，录取率 3.9%。耶鲁 2030 届（2026 年秋季入学，1,652 人）：白人 42%，亚裔 33%，拉丁裔 12%，非裔 10%，原住民 1%；国际生 10%。非裔占比为过去十届新生中最低。
四、亚裔之谜
2028 届数据公布后最引人注目的不是非裔占比的下降，而是亚裔占比在不同学校的相反走向：MIT 的亚裔占比从 40% 升至 47%，耶鲁却从 30% 降至 24%，杜克从 35% 降至 29%，普林斯顿从约 26% 降至 23.8%。同一裁决之下，MIT 的亚裔占比上升 7 个百分点，耶鲁与杜克却各降 6 个百分点。
2024 年 9 月 17 日，SFFA 致信耶鲁、普林斯顿、杜克三校的总法律顾问，称对三校"不遵守"裁决"深表关切"，以 MIT 为反例，要求三校解释其人口数据，并要求"保存所有可能相关的文件与通信"，警告若无解释，SFFA 将认定三校在"规避最高法院的裁决"，"准备通过诉讼执行裁决"。
2025 年 4 月 22 日，亚裔美国人教育联盟（AACE）向美国教育部民权办公室提交民权投诉，指控耶鲁在 2023–2024 录取周期中使用族裔代理变量歧视亚裔申请者。投诉援引教育部代理助理部长 Trainer 2025 年 2 月 14 日的"致同事信"，称该信是执行最高法院 SFFA 裁决的具体步骤。普林斯顿称审阅人员在班级最终确定前看不到族裔数据；多家学校强调一年数据存在自然波动。专家提醒，停用平权需要调整期，早期数据未必代表长期趋势。
五、规则连锁
裁决后，长期被批评为"富人平权"的校友子女优先成为下一个改革目标。约翰斯·霍普金斯大学自 2014 年入学的班级起逐步取消；阿默斯特学院 2021 年 10 月宣布取消；卫斯理大学 2023 年 7 月宣布"正式结束"（此前校友子女约占新生 7%）；西方学院 2023 年 7 月宣布取消（2014–2023 年间校友子女占新生 8%–11%）；弗吉尼亚理工大学 2023 年 7 月取消（校友子女曾占新生 20% 以上）；明尼苏达大学双城分校 2023 年 7 月取消。麻省理工与弗吉尼亚大学表示不考虑校友关系。哈佛 2023 年确认将继续在录取中考虑校友关系。
五州立法：科罗拉多州 2021 年（公立院校，首个）；弗吉尼亚州 2024 年 3 月 8 日签署、7 月 1 日生效（公立院校）；马里兰州 2024 年 4 月签署、7 月 1 日生效（公立与接受州经费的私立院校）；伊利诺伊州 2024 年 8 月（公立院校）；加利福尼亚州 2024 年 9 月 30 日签署 AB 1780、2025 年 9 月 1 日生效（与州助学金挂钩的私立非营利院校；加州大学早在 1998 年已取消校友子女优先）。加州是第五个立法禁止的州，也是继马里兰之后第二个将禁令覆盖到私立院校的州。据 Education Reform Now 2025 年 2 月简报，全美仍有 420 所四年制大学考虑校友子女身份；自 2015 年以来已有 452 所大学取消该做法，下降 52%；仅 2022–2023 年间就有 92 所大学停止。
标化考试也在回归。疫情期间超过 80% 的四年制大学实行标化可选政策；2022 年起风向逆转。麻省理工 2022 年 3 月宣布恢复，为首家恢复的学校。达特茅斯 2024 年 2 月 5 日宣布恢复，为首家常春藤；本校研究发现，考试成绩仍是校园学业成功的最强预测指标。耶鲁 2024 年 2 月恢复，初为"灵活"政策；2026 年 5 月收紧为仅限 SAT/ACT。布朗、哈佛、加州理工 2024 年恢复，康奈尔 2024 年宣布、适用 2026 年秋季，斯坦福 2025 年 6 月宣布，普林斯顿 2025 年 10 月宣布、为第七所恢复的常春藤。据 The College Investor 2026 年 10 月核对，2027 年秋季入学要求 SAT/ACT 的大学与大学系统共 72 所，包括八所常春藤中的六所、MIT、斯坦福与佛罗里达全部 12 所公立大学。恢复的连锁反应：哈佛申请量从 2027 届的 56,937 份降至 2029 届的 47,893 份；耶鲁单选早申申请量下降 14%；布朗早申减少 1,200 份；2024–25 申请季提交考试成绩的学生人数上升 11%，为疫情以来首次显著增长。
六、申请者与对照组
Common App 2024–25 申请季数据显示，独立申请人逾 149.7 万人（增长 5%），提交申请 10,193,579 份（增长 8%），人均申请 6.80 所（2013–14 年仅 4.63 所）。公立大学收到的申请量首次超过私立大学。早申申请量增长 17%，提前决定增长 4%。裁决后，拉丁裔申请人增长 15%，非裔申请人增长 12%；享受费用减免与低收入申请人增长近四倍。人口悬崖在前：美国高中毕业生人数在 2025 年达到约 390 万人的峰值，预计到 2041 年下降约 13%，东北与中西部受冲击最大。录取率：哈佛 2028 届 3.65%，2029 届 4.18%；耶鲁 2029 届 3.9%；普林斯顿 2029 届 4.6%。
加州是理解裁决后果的天然对照组—加州 209 号提案（1996 年）早在 28 年前就已禁止公立大学在录取中考虑族裔；加州大学自 2021 年起不再考虑 SAT/ACT 成绩。2024 年秋季录取：加州居民新生录取率 70%；历史上代表性不足群体的录取占比升至 45.4%；拉丁裔占加州新生录取的 38.6%，为最大群体；非裔录取人数增长近 10%。2025 年秋季录取：加州居民新生录取 102,031 人，创纪录；加州申请者的录取率为 78%；伯克利与洛杉矶分校录取率各约 10%。族裔构成同比几乎没有变化：拉丁裔仍为最大群体（38%）；低收入学生占 43%。长期趋势：加州居民新生录取从 2016 年至 2025 年增长 42.2%。这组数字同时被裁决的支持者与反对者引用。2025 年 3 月，美国司法部对伯克利、尔湾、洛杉矶三校展开调查；同年又指控圣地亚哥、洛杉矶、戴维斯三校医学院优待非裔与拉丁裔申请者、歧视白人与亚裔申请者。加州大学否认指控，称遵守加州已实施近三十年的平权禁令。
七、军校：豁免只维持了两年
裁决的军校豁免只维持了两年。裁决后不到三个月，SFFA 于 2023 年 9 月 19 日起诉西点军校，2023 年 10 月起诉海军学院，指控两校在录取中使用族裔基准违反宪法第五修正案。SFFA 称西点的做法"比哈佛本身更糟"，是"不受约束的种族歧视"。法院不予叫停：2024 年 1 月 3 日，Halpern 法官拒绝发布初步禁令；2024 年 2 月 2 日，索托马约尔大法官拒绝紧急禁令；2024 年 2 月，最高法院以未署名命令拒绝介入；2024 年 12 月 20 日，Bennett 法官拒绝叫停海军学院的族裔考量。
2025 年 1 月，特朗普总统签署第 14185 号行政命令，宣布"武装部队内任何个人或团体都不应因性别、种族、族裔、肤色或信仰而受到优待或不利对待"；国防部据此取消军校录取中的族裔考量。2025 年 6 月，海军学院同意停止在录取中考虑族裔；2025 年 8 月 12 日，西点军校与空军学院同意同样处理；2025 年 8 月 13 日，西点案经双方协议自愿撤诉，案件关闭。
八、收束：回到那两组数字
回到 2024 年 9 月的那两组数字：麻省理工非裔 5%、亚裔 47%；耶鲁亚裔 24%、非裔 14%。三年过去，第三届数据已经公布：耶鲁 2030 届非裔 10%，为过去十届新生中最低；亚裔 33%。哈佛 2029 届亚裔 41%，非裔 11.5%。各校的数字仍在动，规则仍在改—校友子女优先的立法、标化考试的回归、军校的终结，都是同一裁决之后的事。
After the Affirmative Action Ruling: What Changed in American College Admissions
1. September 2024: two sets of numbers
In September 2024, the Massachusetts Institute of Technology released the racial composition of its first freshman class admitted after the Supreme Court's ruling: Black students made up 5 percent, Asian students 47 percent. A year earlier, those figures were 15 percent and 40 percent. On the same day, Yale's numbers moved in the opposite direction: its Asian share fell from 30 percent to 24 percent, while its Black share did not move at all, remaining at 14 percent.
This was the first freshman class admitted under the new rules after the Supreme Court's June 29, 2023 decision. Three years on, the numbers are still moving, and so are the rules.
2. The ruling: 6–3, 237 pages
On June 29, 2023, the U.S. Supreme Court decided two consolidated cases: *Students for Fair Admissions, Inc. v. President & Fellows of Harvard College* (600 U.S. 181) and *Students for Fair Admissions, Inc. v. University of North Carolina*. By a 6–3 vote, with the majority opinion written by Chief Justice Roberts, the Court held that Harvard's admissions policy violated Title VI of the Civil Rights Act of 1964 and that the University of North Carolina's policy violated the Equal Protection Clause of the Fourteenth Amendment. The opinions, including the dissents, ran to 237 pages.
The majority gave three reasons: the schools' stated admissions goals were not "sufficiently focused and measurable" for a court to evaluate; race was used as a "negative" factor, and the policy rested on the assumption that students of the same race think alike — "College admissions are zero-sum. A benefit provided to some applicants but not to others necessarily advantages the former group at the expense of the latter"; and the schools' stated endpoints amounted to "outright racial balancing," which had long been held unconstitutional. The majority concluded: "Eliminating racial discrimination means eliminating all of it."
The decision did not formally overrule the 1978 *Bakke* case or the 2003 *Grutter* case, but under strict scrutiny almost no race-conscious policy could pass. The Court also rejected the schools' plea to be allowed to continue at least until 2028 — *Grutter* had said race preferences would no longer be needed in 25 years, and the majority held that the endpoint had arrived. The ruling expressly exempted the military academies, on the ground that they may have "potentially distinct interests."
The majority opinion ended with an opening: nothing in the opinion should be construed as prohibiting universities from considering an applicant's discussion of how race affected his or her life, "be it through discrimination, inspiration, or otherwise." Justice Sotomayor began her dissent: "Today, this Court stands in the way and rolls back decades of precedent and momentous progress... In so holding, the Court cements a superficial rule of colorblindness as a constitutional principle."
The lawsuit lasted nine years. The plaintiff, Students for Fair Admissions, was founded by anti-affirmative-action activist Edward Blum. In 2014 it sued Harvard in federal court in Massachusetts and the University of North Carolina in federal court in North Carolina, alleging principally that Harvard intentionally discriminated against Asian applicants, including through a "personal rating" shaped by racial stereotypes. At a three-week trial in October 2018, the district court's findings of fact showed Asian applicants scored higher than white applicants on academic and extracurricular measures — 60 percent of Asian applicants scored above 3+ on academics (compared with 46 percent of white applicants), and 28 percent scored 2 or higher on extracurriculars (compared with 25 percent) — but scored lower on the personal rating. On September 30, 2019, District Judge Allison Burroughs ruled for Harvard; the First Circuit affirmed in 2020. The Supreme Court's 2023 decision did not directly address whether Harvard had discriminated against Asian applicants, instead holding on broader grounds that race-conscious admissions were themselves unconstitutional; the lower courts' conclusion that Harvard had not unlawfully discriminated against Asian applicants was left undisturbed.
3. Three years of data
The Class of 2028 was the first freshman class admitted under the new rules. The figures schools released are enrolled-class shares, not admit rates by race — no top university has ever published admit rates broken down by race.
Harvard: Black students fell from 18 percent to 14 percent, Asian students held at 37 percent, Latino students rose from 14 percent to 16 percent, 8 percent did not disclose their race (up from 4 percent), and international students were 16 percent. MIT: Black students fell from 15 percent to 5 percent, Asian students rose from 40 percent to 47 percent, Latino students fell from 16 percent to 11 percent; Black, Latino, and Native American students combined fell from 25 percent to 16 percent. Yale: Black students held at 14 percent, Asian students fell from 30 percent to 24 percent, Latino students rose 1 point to 19 percent, white students rose 4 points to 46 percent. Princeton: Black students were little changed at about 9 percent, Asian students fell about 2.2 points (from about 26 percent to 23.8 percent), Latino students were about 9 percent. Brown: Black students fell from 15 percent to 9 percent, Asian students rose from 29 percent to 33 percent, Latino students fell from 14 percent to 10 percent. University of Pennsylvania: historically underrepresented students fell from 25 percent to 23 percent. Duke: Black students held at 13 percent, Asian students fell from 35 percent to 29 percent, Latino students rose from 13 percent to 14 percent.
A warning about denominators: starting with the Class of 2028, Harvard changed its methodology — the denominator for its race percentages changed from the entire freshman class to "students who self-reported their race," and it recalculated the Class of 2027 figures. As a result, Harvard's Class of 2027 Asian share has two numbers: 29.9 percent under the old denominator (share of the whole class) and 37 percent under the new one (share of those who reported a race). Year-over-year comparisons must state the denominator; Harvard does not publish a white share. MIT's dean of admissions, Stuart Schmill, stated plainly that the change was caused by the Supreme Court's decision.
The change did not settle in the first year. Harvard's Class of 2029 (entering fall 2025, 1,675 students): Black students 11.5 percent (14 percent in the Class of 2028), Latino students 11 percent (down from 16 percent), Asian students 41 percent (up from 37 percent); applications totaled 47,893, down from 54,008 for the Class of 2028 and 56,937 for the Class of 2027; 2,003 students were admitted, an admit rate of 4.18 percent (3.65 percent for the Class of 2028). The Class of 2029 was the first class since the pandemic required to submit SAT or ACT scores. Princeton's Class of 2029: Black students 5 percent, the lowest since 1968; Asian students rose from about 24 percent to about 27 percent. Princeton's Class of 2028 had shown almost no change; the change arrived a year late. Yale's Class of 2029: Black students 12 percent (14 percent in the Class of 2028); the Asian share rose; 2,227 students were admitted from 57,517 applications, an admit rate of 3.9 percent. Yale's Class of 2030 (entering fall 2026, 1,652 students): white students 42 percent, Asian students 33 percent, Latino students 12 percent, Black students 10 percent, Native American students 1 percent; international students 10 percent. The Black share was the lowest in ten freshman classes.
4. The Asian puzzle
What drew the most attention after the Class of 2028 data was published was not the fall in Black shares but the opposite movement of Asian shares at different schools: MIT's Asian share rose from 40 percent to 47 percent, while Yale's fell from 30 percent to 24 percent, Duke's from 35 percent to 29 percent, and Princeton's from about 26 percent to 23.8 percent. Under the same ruling, MIT's Asian share rose 7 points while Yale's and Duke's each fell 6 points.
On September 17, 2024, SFFA wrote to the general counsels of Yale, Princeton, and Duke, expressing "deep concern" that the three schools were "not complying" with the ruling, holding up MIT as the counterexample, demanding an explanation of their demographic data, and demanding that they "preserve all potentially relevant documents and communications," warning that without an explanation SFFA would conclude the schools were "evading the Supreme Court's decision" and were "prepared to enforce the decision through litigation."
On April 22, 2025, the Asian American Coalition for Education (AACE) filed a civil rights complaint with the U.S. Department of Education's Office for Civil Rights, alleging that Yale used race proxies to discriminate against Asian applicants in the 2023–24 admissions cycle. The complaint cited the February 14, 2025 "Dear Colleague Letter" from acting assistant secretary Trainer, describing it as a concrete step in enforcing the Supreme Court's SFFA decision. Princeton says its readers cannot see race data before the class is finalized; several schools emphasize that one year's data fluctuates naturally. Experts caution that ending affirmative action takes an adjustment period, and early data may not represent the long-term trend.
5. The rule changes that followed
After the ruling, legacy preference — long criticized as "affirmative action for the wealthy" — became the next target for reform. Johns Hopkins University began phasing it out with the class entering in 2014; Amherst College announced its end in October 2021; Wesleyan University announced in July 2023 that it was "formally ending" legacy preference (legacy students had been about 7 percent of freshmen); Occidental College announced its end in July 2023 (legacy students were 8–11 percent of freshmen from 2014–2023); Virginia Tech ended it in July 2023 (legacy students had been more than 20 percent of freshmen); the University of Minnesota Twin Cities ended it in July 2023. MIT and the University of Virginia say they do not consider legacy status. Harvard confirmed in 2023 that it will continue to consider it.
Five states have legislated: Colorado in 2021 (public institutions, the first); Virginia, signed March 8, 2024 and effective July 1, 2024 (public institutions); Maryland, signed April 2024 and effective July 1, 2024 (public institutions and private institutions receiving state funds); Illinois in August 2024 (public institutions); California, which signed AB 1780 on September 30, 2024, effective September 1, 2025 (private nonprofit institutions tied to state financial aid; the University of California had already ended legacy preference in 1998). California was the fifth state to legislate a ban, and the second after Maryland to extend it to private institutions. According to a February 2025 briefing from Education Reform Now, 420 four-year colleges still consider legacy status; 452 have dropped it since 2015, a decline of 52 percent; 92 stopped in 2022–2023 alone.
Standardized tests are also coming back. During the pandemic, more than 80 percent of four-year colleges adopted test-optional policies; from 2022 the direction reversed. MIT announced its return in March 2022, the first school to restore the requirement. Dartmouth announced on February 5, 2024, the first Ivy League school; its own faculty research found test scores the strongest predictor of academic success on campus. Yale returned in February 2024, initially with a "flexible" policy; in May 2026 it tightened the policy to SAT or ACT only. Brown, Harvard, and Caltech returned in 2024; Cornell announced in 2024, effective fall 2026; Stanford announced in June 2025; Princeton announced in October 2025, becoming the seventh Ivy League school to restore the requirement. According to The College Investor's October 2026 check of admissions websites, 72 colleges and university systems require SAT or ACT scores for fall 2027 entry, including six of the eight Ivy League schools, MIT, Stanford, and all 12 of Florida's public universities. The return had knock-on effects: Harvard's applications fell from 56,937 for the Class of 2027 to 47,893 for the Class of 2029; Yale's single-choice early action applications fell 14 percent; Brown's early decision applications fell by 1,200; the number of students submitting test scores rose 11 percent in the 2024–25 cycle, the first significant increase since the pandemic.
6. The applicants and the control group
Common App data for the 2024–25 cycle show more than 1.497 million unique applicants (up 5 percent), submitting 10,193,579 applications (up 8 percent), an average of 6.80 applications per applicant (only 4.63 in 2013–14). Public universities received more applications than private universities for the first time. Early action applications grew 17 percent, and early decision grew 4 percent. After the ruling, Latino applicants grew 15 percent and Black applicants 12 percent; applicants receiving fee waivers and low-income applicants grew nearly fourfold. A demographic cliff lies ahead: the number of American high school graduates peaks at about 3.9 million in 2025 and is projected to fall about 13 percent by 2041, with the Northeast and Midwest hit hardest. Admit rates: Harvard 3.65 percent for the Class of 2028 and 4.18 percent for the Class of 2029; Yale 3.9 percent for the Class of 2029; Princeton 4.6 percent for the Class of 2029.
California is a natural control group for understanding the ruling's consequences — Proposition 209 (1996) banned public universities from considering race in admissions 28 years ago, and the University of California has not considered SAT or ACT scores since 2021. In fall 2024 admissions, the admit rate for California-resident freshmen was 70 percent; the share of historically underrepresented students rose to 45.4 percent; Latino students were 38.6 percent of California freshman admits, the largest group; Black admits grew nearly 10 percent. In fall 2025 admissions, California-resident freshman admits totaled 102,031, a record; the admit rate for California applicants was 78 percent; Berkeley and UCLA each admitted about 10 percent. The racial composition was nearly unchanged year over year: Latino students remained the largest group (38 percent); low-income students were 43 percent. Over the long term, California-resident freshman admits grew 42.2 percent from 2016 to 2025. These figures are cited by both supporters and opponents of the ruling. In March 2025, the U.S. Department of Justice opened investigations into Berkeley, Irvine, and Los Angeles; later that year it accused the medical schools at San Diego, Los Angeles, and Davis of favoring Black and Latino applicants and discriminating against white and Asian applicants. The University of California denies the accusations, saying it complies with the state's nearly 30-year-old ban on affirmative action.
7. The academies: an exemption that lasted two years
The ruling's exemption for the military academies lasted only two years. Less than three months after the decision, SFFA sued West Point on September 19, 2023, and the Naval Academy in October 2023, alleging that the schools' use of racial benchmarks in admissions violated the Fifth Amendment. SFFA called West Point's practice "worse than Harvard itself" and "unconstrained racial discrimination." The courts declined to halt the practice: on January 3, 2024, Judge Halpern declined to issue a preliminary injunction; on February 2, 2024, Justice Sotomayor declined an emergency injunction; in February 2024, the Supreme Court declined to intervene by unsigned order; on December 20, 2024, Judge Bennett declined to halt the Naval Academy's consideration of race.
In January 2025, President Trump signed Executive Order 14185, declaring that "no individual or group within the Armed Forces should be favored or disadvantaged on the basis of sex, race, ethnicity, color, or creed"; the Department of Defense ended consideration of race in academy admissions. In June 2025, the Naval Academy agreed to stop considering race in admissions; on August 12, 2025, West Point and the Air Force Academy agreed to the same; on August 13, 2025, the West Point case was voluntarily dismissed by agreement, closing the case.
8. Close: back to those two sets of numbers
Back to those two sets of numbers from September 2024: MIT, Black 5 percent, Asian 47 percent; Yale, Asian 24 percent, Black 14 percent. Three years on, the third class's data is out: Yale's Class of 2030 has Black students at 10 percent, the lowest in ten freshman classes, and Asian students at 33 percent. Harvard's Class of 2029 has Asian students at 41 percent and Black students at 11.5 percent. The numbers at each school are still moving, and the rules are still changing — the legislation on legacy preference, the return of standardized tests, and the end of the academy exemption all followed the same ruling.
来源：美国最高法院判决与各大学公布数据
发布：GSM·美国（发布链接）
图片来源：图：黄昏时分的美国最高法院大楼。Wikimedia Commons，CC BY-SA 3.0。 / Photo: The U.S. Supreme Court building at dusk. Wikimedia Commons, CC BY-SA 3.0.
GSM Global · 本篇英文版 English edition of this story
