Enriching Diversity Beyond Affirmative Action

By considering both race and class, Colorado’s flagship university is increasing minority enrollment.

For many students, including protesters outside the Supreme Court  last week, affirmative action remains important to them.
National Journal
Sophie Quinton
Oct. 21, 2013, 2 a.m.

Michigan’s 2006 ban on af­firm­at­ive ac­tion means de-facto se­greg­a­tion, a co­ali­tion fight­ing the ban ar­gues in a brief filed to the Su­preme Court. The chal­lenge to the law, which the Court is now con­sid­er­ing, starts with the as­sump­tion that the num­ber of minor­ity stu­dents ad­mit­ted to Michigan’s top col­leges would shrink un­der the ban.

A new ap­proach at the Uni­versity of Col­or­ado (Boulder) sug­gests that some col­leges could main­tain ra­cial di­versity while de-em­phas­iz­ing race in the ad­mis­sions pro­cess. Since 2011, the state flag­ship has giv­en ex­tra con­sid­er­a­tion to dis­ad­vant­aged ap­plic­ants and to those whose grades and test scores are above av­er­age for their eco­nom­ic cir­cum­stances.

Even if the state were to ban race-based af­firm­at­ive ac­tion, the class-based frame­work should in­crease low-in­come ad­mis­sions, main­tain minor­ity rep­res­ent­a­tion, and even bet­ter, pre­dict suc­cess in col­lege. It’s a mod­el that could help oth­er uni­versit­ies be­come more di­verse.

“What we hope is that this res­ult will ease fears — at least a little bit — that an end to race-based af­firm­at­ive ac­tion will be ab­so­lutely dev­ast­at­ing for ra­cial di­versity on col­lege cam­puses,” says Mat­thew Gaert­ner, a re­search sci­ent­ist in the cen­ter for col­lege and ca­reer suc­cess at Pear­son. Gaert­ner helped the ad­mis­sions of­fice de­vise its new frame­work while a gradu­ate stu­dent at Boulder.

CU-Boulder is a mod­er­ately se­lect­ive in­sti­tu­tion that primar­ily serves Col­or­ado and the West. Only about 20 per­cent of the uni­versity’s over 29,000 stu­dents are Afric­an-Amer­ic­an, Asi­an-Amer­ic­an, His­pan­ic, or Nat­ive Amer­ic­an. The uni­versity has been work­ing to in­crease un­der­rep­res­en­ted-minor­ity en­roll­ment for the past six years.

Like oth­er se­lect­ive col­leges, CU-Boulder has a set of cri­ter­ia it uses to weigh mar­gin­al ap­plic­ants — those whose grades and test scores aren’t high enough to guar­an­tee ad­mis­sion, but who would still add value to the fresh­man class. Factors in­clude ex­tra­cur­ricular activ­it­ies, strength of the seni­or sched­ule, and race.

The uni­versity had to re­think its ap­proach in 2008, when a state bal­lot ini­ti­at­ive threatened to ban af­firm­at­ive ac­tion. “I was very con­cerned about what could po­ten­tially hap­pen to our cam­pus,” says Kev­in MacLen­nan, CU-Boulder’s dir­ect­or of ad­mis­sions. He vis­ited flag­ship in­sti­tu­tions in some of the eight states that have banned af­firm­at­ive ac­tion, and he found that they had ex­per­i­enced a sig­ni­fic­ant drop in eth­nic and ra­cial di­versity.

Even in Texas, where the ma­jor­ity of pub­lic-school stu­dents are Latino and where the top 10 per­cent of every high school class gain auto­mat­ic ad­mis­sion to state uni­versit­ies, the state flag­ship has found it ne­ces­sary to con­sider race in or­der to in­crease di­versity. Earli­er this year, the Su­preme Court sent a chal­lenge to the af­firm­at­ive-ac­tion plan at the Uni­versity of Texas (Aus­tin) back to a lower court.

So­cial and eco­nom­ic forces are mak­ing elite schools both more white and more af­flu­ent, a trend doc­u­mented by An­thony Carne­vale and Jeff Strohl of Geor­getown Uni­versity’s Cen­ter on Edu­ca­tion and the Work­force. In­equal­ity is a big part of the prob­lem. Low-in­come Afric­an-Amer­ic­ans and Lati­nos tend to grow up in more-con­cen­trated areas of poverty than their white peers. Oth­er factors may also be at work: Afric­an-Amer­ic­an and Latino stu­dents are less likely to en­roll in se­lect­ive col­leges even when they have sol­id grades and test scores.

Gaert­ner set out to cre­ate a class-based af­firm­at­ive-ac­tion frame­work for CU-Boulder that would take in­to ac­count re­sources avail­able to a child at home and in high school. He came up with two meas­ures. The “dis­ad­vant­age in­dex” meas­ures the like­li­hood that the ap­plic­ant will en­roll in col­lege at all, giv­en his so­cioeco­nom­ic status. The “over­achieve­ment in­dex” meas­ures wheth­er an ap­plic­ant’s grades and test scores ex­ceed the scores usu­ally achieved by stu­dents of his so­cioeco­nom­ic status.

“When stu­dents ap­ply and they demon­strate severe so­cioeco­nom­ic dis­ad­vant­age, or ex­traordin­ary over­achieve­ment re­l­at­ive to that dis­ad­vant­age, they’re giv­en a sub­stan­tial boost in the ad­mis­sions pro­cess,” Gaert­ner says. Over­achiev­ers from all back­grounds get ex­tra con­sid­er­a­tion, and so do severely dis­ad­vant­aged stu­dents, even those who aren’t achiev­ing at high levels.

Each in­dex is a stat­ist­ic­al mod­el that weighs many factors, in­clud­ing the teach­er-stu­dent ra­tio at the ap­plic­ant’s high school, the num­ber of de­pend­ents the ap­plic­ant’s par­ents are sup­port­ing, and wheth­er the stu­dent’s first lan­guage was Eng­lish. Gaert­ner cre­ated a com­puter pro­gram that runs the math and pro­duces a num­ber that ad­mis­sions of­ficers use to clas­si­fy stu­dents as mod­er­ately dis­ad­vant­aged, severely dis­ad­vant­aged, or not dis­ad­vant­aged; and as ex­hib­it­ing high over­achieve­ment, ex­traordin­ary over­achieve­ment, or none at all.

Two ex­per­i­ments prove that the in­di­cies are work­ing. In 2009, Gaert­ner had ad­mis­sions of­ficers re­view 478 ap­plic­a­tions, first un­der CU-Boulder’s race-based policy and then un­der the new class-based policy, with all ra­cial iden­ti­fi­ers re­moved. Of­ficers ended up ad­mit­ting 9 per­cent more un­der­rep­res­en­ted minor­ity stu­dents un­der the race-blind policy than and 20 per­cent more stu­dents of very low so­cioeco­nom­ic status.

In 2010, CU-Boulder ran an­oth­er ex­per­i­ment, this time on 2,000 ap­plic­a­tions deemed bor­der­line for ad­mis­sion. Half were eval­u­ated us­ing the new class-plus-race ap­proach, and half us­ing the old ap­proach that used race alone. The hy­brid ap­proach res­ul­ted in a 13 per­cent in­crease in ac­cept­ance rates for the poorest stu­dents, a 17 per­cent in­crease for un­der­rep­res­en­ted minor­ity stu­dents, and a 32 per­cent in­crease in the low­est-in­come, minor­ity stu­dents. The res­ults of both ex­per­i­ments were re­cently pub­lished in Har­vard Law and Policy Re­view.

Col­or­ado’s bal­lot ini­ti­at­ive didn’t pass, so today CU-Boulder uses both race and the new in­di­cies as factors in ad­mis­sions de­cisions. In gen­er­al, dis­ad­vant­aged stu­dents ad­mit­ted through this pro­cess don’t per­form as well at CU-Boulder as the typ­ic­al stu­dent, Gaert­ner says. “Their grades and test scores don’t in­dic­ate they’re auto­mat­ic ad­mits, so you’d sort of ex­pect them to have a little bit lower out­comes than typ­ic­al un­der­gradu­ate,” he says.

But stu­dents iden­ti­fied as over­achiev­ers ex­cel. “Their col­lege out­comes — that’s grades, cred­it hours earned, and gradu­ation at four years and six-year gradu­ation rate — is ac­tu­ally high­er than typ­ic­al un­der­gradu­ates,” Gaert­ner says of over­achiev­ers.

Con­struct­ing the new ad­mis­sions factors has had an ad­di­tion­al be­ne­fit. “The more we got in­to this, it ac­tu­ally ex­pan­ded our view of di­versity,” MacLen­nan says. The ad­mis­sions team now pays more at­ten­tion to rur­al stu­dents and stu­dents com­ing from un­der-re­sourced high schools.

While the in­di­cies are a good fit for CU-Boulder, it’s not clear wheth­er they’d work just as well at very se­lect­ive schools, Gaert­ner says. And it’s im­port­ant to re­mem­ber that im­prov­ing on-cam­pus di­versity in­volves a spec­trum of ef­forts.

“The in­di­cies are won­der­ful, but they can’t just stand by them­selves — or even part of the ad­mis­sions pro­cess,” MacLen­nan says. Out­reach and re­cruit­ment comes first. Then, once stu­dents are ad­mit­ted, the uni­versity needs to help every ad­mit­ted stu­dent make it to gradu­ation.

What We're Following See More »
Trump Not on Ballot in Minnesota
1 days ago
Trump on Immigration: ‘I Don’t Know, You Tell Me’
1 days ago

Perhaps Donald Trump can take a plebiscite to solve this whole messy immigration thing. At a Fox News town hall with Sean Hannity last night, Trump essentially admitted he's "stumped," turning to the audience and asking: “Can we go through a process or do you think they have to get out? Tell me, I mean, I don’t know, you tell me.”

Trump Enriching His Businesses with Donor Money
3 days ago

Donald Trump "nearly quintupled the monthly rent his presidential campaign pays for its headquarters at Trump Tower to $169,758 in July, when he was raising funds from donors, compared with March, when he was self-funding his campaign." A campaign spokesman "said the increased office space was needed to accommodate an anticipated increase in employees," but the campaign's paid staff has actually dipped by about 25 since March. The campaign has also paid his golf courses and restaurants about $260,000 since mid-May.

Trump Cancels Rallies
3 days ago

Donald Trump probably isn't taking seriously John Oliver's suggestion that he quit the race. But he has canceled or rescheduled rallies amid questions over his stance on immigration. Trump rescheduled a speech on the topic that he was set to give later this week. Plus, he's also nixed planned rallies in Oregon and Las Vegas this month.

Sean Hannity Is Also Advising Trump
4 days ago

Donald Trump's Fox News brain trust keeps growing. After it was revealed that former Fox chief Roger Ailes is informally advising Trump on debate preparation, host Sean Hannity admitted over the weekend that he's also advising Trump on "strategy and messaging." He told the New York Times: “I’m not hiding the fact that I want Donald Trump to be the next president of the United States. I never claimed to be a journalist.”