"By the Numbers"

Reconfiguring Admissions to Serve the Mission of Selective Public Higher Education
By Mindy L. Kornhaber

I. Introduction

On what bases should students be admitted to highly selective public colleges and universities? In Texas under Hopwood, and in California under Proposition 209, the answer is: "by the numbers."

With lawsuits pending in Michigan and a state initiative underway in Washington, the same response may soon be heard more frequently. Given the uneven circumstances from which students’ grades and class rank arise, the most potent numbers in admissions are often those from high-stakes standardized, norm-referenced tests (HSSNRTs), such as the SAT, ACT, and LSAT. Proponents of this approach argue that consideration of candidates' race or ethnicity violate laws against discrimination on the basis of race. They also assert that when HSSNRT scores are emphasized, merit is placed at the center of admissions decisions (See e.g., Thernstrom & Thernstrom, 1997; D'Souza, 1991). This score-ranked conception of merit is readily conveyed and grasped: the higher the score, the more an applicant deserves to be admitted. Nevertheless this conception is erroneously narrow: HSSNRT scores reveal little about who will succeed in higher education; they say far less about who will succeed thereafter.

Thus, this score-ranked conception does not clearly support selective higher education’s mission to achieve excellence in instruction, research, and service. Furthermore, when put into practice, this score-ranked conception yields devastating social consequences.

This paper first spells out several of these consequences. It then considers whether HSSNRTs are technically adequate to justify such consequences. Next, it offers a principled resolution to the debate between advocates of score-ranked admissions and proponents of diversity. This resolution is grounded less in the ideologies of individual merit and affirmative action than in the history and mission of American public higher education. Finally, it concludes by laying out various possible approaches to admission for policymakers and the wider public to consider. Each of these possibilities is considered in light of the mission of public higher education as well as its social consequences.

II. Consequences of Increased Emphasis on Testing in Admissions

Central to the debates that gave rise to Hopwood and Proposition 209 is the question of whether admission to a selective public institution should in any way be influenced by a candidate’s racial or ethnic affiliation (Hopwood v. Texas, 1994, 1995; see Brest & Oshige, 1995; Tierney, 1997). Proponents of admission by the numbers argue that individuals should be admitted, whatever their race, ethnicity, gender or socioeconomic status, as long as they have ‘made the grade.’ Yet, it is evident that individual applicants are not the ones to gain or lose under these policies (Brest & Oshige, 1995; Bowen & Bok, 1998). As highlighted below, an increased reliance on test scores also has consequences for particular groups, for selective public colleges and universities, and for the wider society:
Decreased admission for already underserved groups

The clearest and most immediate consequence of the increased emphasis on HSSNRTs is a sharp drop in the acceptance rates of African American and Hispanic students at selective institutions. Following Proposition 209, the number of African American students accepted to Boalt Hall, Berkeley's law school, plummeted from 75 to 14. All 14 then opted for other institutions, leaving just one African American student who had deferred enrolling the previous year (Chronicle of Higher Education, 1997; Kane, 1998a). The percentage of entering African American freshmen fell from 7.2 percent in 1997 to 3.3 percent in 1998 (Berkeley Office of Student Research, 1997a, 1998). This situation is mirrored in Texas. Following Hopwood, African Americans admitted into the University of Texas at Austin's Law School fell from 5.9 percent in 1996 to 1.1 percent in 1997. Admission of Mexican American dropped from 6.3 percent to 3.3 percent (Chapa & Lazaro, 1998). Of some 500 students who actually entered UT Austin’s Law School in the fall of 1997, there were just four African Americans and 26 Mexican Americans (Traub, 1998). In essence, these ‘new’ policies have restored patterns of segregation last seen prior to the passage of the Civil Rights Act of 1964 (see Karabel, 1998).

A demographic collision

The precipitous drop in Black and Hispanic students in these selective institutions collides with the increasing proportion of these minorities in the surrounding society. Between the mid-1990s and 2010, the proportion of Black and Hispanic residents in Texas will increase
from 39 to 43.5 percent1 (Texas Population Projections Programs, n.d.). In California, the proportion will rise from 35 percent to 43 percent, with the gain attributable to the growth of Hispanic population (Malson, 1998).

Given this growth, the new admissions policies will exclude an enormous portion of the population from selective public colleges and universities. At the same time, these institutions will concentrate their resources on society's decreasing proportion of white students and on certain segments of the Asian population. As discussed below, such enrollment patterns could well generate a backlash against selective public higher education.

Reducing public support for selective public institutions

Because tuition, endowment income, and grants are not sufficient to finance public higher education, state taxpayers subsidize these institutions. The subsidy occurs, even though many of the enrolled students are affluent. This is especially true among whites and Asians. For instance, in 1997, nearly 42 percent of white freshman at Berkeley had parental incomes over $100,000 a year, as did 27 percent of Asians. In contrast, 14 percent of African Americans and 10 percent of Chicanos had family incomes at that level (Berkeley Office of Student Research, 1997b).

Given the demographic trends described above, if the new admissions policies persist, the subsidy of an increasingly affluent student body will be borne by taxpayers who themselves have not benefitted from selective higher education and who are unlikely to see the benefit for their own children. Will taxpayers and their elected representatives continue to support selective higher education for cohorts of unrepresentative and affluent students? There is serious doubt, at least in Texas, according to Russell Weintraub, a University of Texas law professor: "If the majority of people in this state are going to be Mexican-American and African American, and they are going to assume many of the leadership roles in the state, then it's going to be big trouble if the law school doesn't admit many minority students -- it's going to be a bomb ready to explode" (In Traub, 1998, p. 20).

Some backlash is already evident. The Texas legislature, led by a black state assemblyman named Ron Wilson, has sought to have public universities apply the ten percent rule to all students, including athletes.2 This policy would demolish quality football teams and do away with millions of dollars in football-generated university revenue. Wilson has also considered efforts to transfer money away from selective state universities and to historically black colleges (Traub, 1998).

Thus, HSSNRT-driven admissions not only segregate campuses and exclude increasing proportions of the surrounding society. By alienating African American and Hispanic citizens, they threaten the financial resources that selective institutions need to sustain their work.

A cognitive collision

Any assessment system, from teacher made tests to those created by ETS, have systemic effects on teaching and learning (Frederickson & Collins, 1989): Teachers teach to the test. Students study to it (i.e., ‘Will this be on the test?’). Tests therefore influence the development of cognitive skills.

Unfortunately, because HSSNRTs are largely devoid of the kinds of problem solving skills now demanded by the wider society, they undermine the development of these skills. For example, unlike HSSNRTs, real world problems do not yield one right answer. Instead, many answers are possible and their feasibility must be considered in light of an often-shifting array of constraints. For another example, HSSNRTs must be completed by test takers working in splended isolation. In contrast, most real-world problems require people to identify and orchestrate a range of resources, including other people, computers, and books (See Gardner, 1991; Reich, 1991; Resnick, 1987; Zuboff, 1988, 1995). Given these qualities, real world problems, unlike HSSNRT questions, typically call for more than a minute or two of concentration.

http://osr4.berkley.edu/PUBLIC/STUDENT.DATA/PUBLICATIONS/UG/ugf97.html#table8

Berkeley Office

Copyright 2005, test scores, jean and bill bruce. Recommended books of Professor W.C. Bruce


_____