Open Admissions at the City University of New York

To refurbish its reputation, a major university turns to standardized testing. Who gets left out when educational opportunity depends on test scores?

By William Crain

Lawrence Rushing, an African American psychologist, recalls that as a young man in the 1950s, he lived near the City College of New York in Harlem. "I can still vividly see, in my mind's eye, hundreds of students rushing to classes after exiting from the 145th Street subway stop. What I do not recollect—and didn't comprehend at the time—is seeing black faces in the crowd. I didn't know then that only an average of thirty-three African Americans graduated from City College each year."

Rushing's recollection, which appears in the spring 2003 issue of Encounter: Education for Meaning and Social Justice, highlights a historical inconsistency. African Americans were largely absent from a college that was emblematic of democratic opportunity. Established as the Free Academy in 1847, City College had given thousands of poor and working-class students and recent immigrants the chance for a college education they couldn't otherwise afford. But even during the 1960s, the student body of City College was largely white—as was the overarching City University of New York (CUNY), which City College and other campuses joined in 1961.

Open Admissions

Then, in 1969, riding the crest of the civil rights movement, a group of African American and Latino students shut down City College's South Campus. They demanded that the college reflect the racial and ethnic composition of Harlem. After numerous tense meetings, New York City's politicians agreed upon an open admissions policy that guaranteed every New York City high school graduate a place in CUNY. This often meant a place in one of CUNY's community (two-year) colleges. But as sociologists David Lavin and David Hyllegard point out in their 1996 book, Changing the Odds: Open Admissions and the Life Chances of the Disadvantaged, open admissions was groundbreaking because it was oriented toward the senior (four-year) colleges. The African American and Latino activists didn't want students of color to be restricted to two-year college degrees. So the activists sought and won an open admissions policy that permitted students who were in the top half of their high school class or who had a grade point average of eighty to enroll in the senior colleges.

The open admissions policy achieved its purpose: it opened CUNY's doors to people of color. In 1969, students of color accounted for less than a fifth of all CUNY undergraduates; today, they constitute a decided majority. (Today's student body is 32 percent African American, 25 percent Latino, 14 percent Asian, and 29 percent white.)

But open admissions also generated heated opposition. In 1971 U.S. vice president Spiro Agnew said that CUNY would give away "100,000 devalued diplomas." CUNY itself seemed ambivalent. During the early 1970s, the CUNY senior colleges began tightening admissions. In 1976, New York State and CUNY imposed tuition. It was a year of fiscal crisis in New York City, but it also was the year in which CUNY's student body became predominantly people of color. Before then—even during the Great Depression—CUNY had been tuition free.

Although the open admissions policy was weakened, its critics continued to assail it. Politicians and media pundits branded CUNY with labels such as "Remedial U." Open admissions, to be sure, required CUNY to offer much remedial work. Many students from the city's public schools weren't prepared for college. But CUNY faculty rose to the challenge; the university gained a national reputation for innovative remedial courses, and they were quite effective. For example, a 1998 CUNY Office of Institutional Research study indicated that three-quarters of the students completed their remedial work within a year at the senior colleges, and that those who completed remediation went on to graduate at nearly the same rate as those who hadn't needed it.

CUNY also continued to provide a high-quality education. Statistics on City College, where I teach, are telling. The National Research Council reported that between 1983 and 1992, in the heart of the open admissions era, 860 City College graduates went on to earn Ph.D.'s. This number is higher than that for our prestigious neighbor, Columbia College. To be sure, our student body is somewhat larger than Columbia's, but the comparison is illuminating. And set side by side with almost all other colleges in the New York City area, a higher percentage of City College graduates have earned Ph.D.'s. Another CUNY college, Hunter, has sometimes outperformed City College in this respect, and other CUNY colleges have made their own marks in business, the arts, and the helping professions.

Open admissions demonstrated that when people are given opportunities, they often achieve stunning success. But right-wing politicians, think tanks, and newspapers (especially the New York Post and New York Daily News) kept hammering away. Then, in the mid-1990s, George E. Pataki became New York State's governor, Rudolph W. Giuliani became New York City's mayor, and these two men mounted the biggest assault yet. They appointed people to the Board of Trustees who, in the name of "higher standards," began developing plans to restrict senior college admissions.

Social Justice

As the trustees considered new admissions plans, hundreds of faculty, students, and community members testified at public hearings about the importance of student access to CUNY. Many of us saw the issue as one of social justice. Through no choice of their own, children grow up in different social conditions. Some attend excellent schools and receive a fine preparation for college. Others, including many children of color, find themselves in much poorer schools. As long as such inequities exist, justice demands that colleges and universities try to make sure that all students have a chance to develop their minds and pursue their dreams.

To my surprise, some trustees were responsive to our arguments. But they also experienced intense pressure from the mayor and the governor and, in January 1999, voted to ban remediation in the senior colleges. The ban centers on standardized tests: if a student's test scores on the CUNY reading, writing, or math skills tests suggest any need for remediation, the student cannot enter a bachelor's degree program at a senior college. It doesn't matter if the student meets the standard admissions criteria-if she has solid high school grades and credentials. To gain admission, the student must pass the three skills tests.

The trustees agreed to certain modifications. Students can receive exemptions from the skills tests by specific scores on the SAT or New York State Regents exams. In addition, some English as a Second Language (ESL) students, as well as underprepared students who qualify for a special opportunity program titled SEEK, do not initially have to pass all three skills tests. These ESL and SEEK students, who usually account for about 20 percent of first-year students, must pass the tests before they proceed far in college, however.

Before the new policy was adopted, CUNY administrators presented data indicating that the testing requirement would disproportionately exclude students of color from the bachelor's degree programs. Specifically, the testing requirement would exclude larger percentages of African American, Latino, and Asian students than white students. This finding surprised no one—at least for the first two groups. Standardized tests have consistently yielded lower average scores among African American and Latino students.

People have often asked me if the new policy is really so bad. Can't students who fail a skills test simply begin at a community college? Yes, they can. But Lavin and Hyllegard reported in Changing the Odds that CUNY students who begin at a community college rather than a senior college have a 19 percent lower chance of ever earning a bachelor's degree. (This figure resulted from comparing students who entered community and senior colleges with the same academic credentials, aspirations, and income.)

For CUNY students, the new tests add to a long list of obstacles. Most are from poor families and must work long hours on jobs outside college. Many are the first in their families to go to college, so they are navigating uncharted waters. They have doubts and anxieties. New hurdles—such as the new testing requirement and the loss of remedial opportunities at senior colleges—can make the entire college enterprise overwhelming. Most of the nation's public four-year colleges place no such hurdles in the way of their applicants, who are predominantly white, so CUNY's new policy seemed unfair from the beginning.

These considerations should have been sufficient to halt the policy. In addition, one should never use any test—especially for a high-stakes purpose such as college admissions—before making sure the test is reliable and valid. Reliability refers to a test's consistency, which is important, but my focus for this article is on a test's validity. In college admissions, the most important kind of validity is predictive validity; if a test is used to determine whether a student can enter a program, the test should be a good predictor of success in the program. For at least a decade, CUNY researchers had been finding evidence that the skills tests lacked predictive validity. But the trustees paid little attention to this problem.

Thus, CUNY began phasing in a new admissions policy for its bachelor's degree programs despite evidence that the policy would disproportionately exclude students of color and do so on the basis of invalid tests.

A small group of us found this situation so outrageous that we soon filed a complaint with the U.S. Department of Education's Office for Civil Rights. The office agreed to gather data monitoring the new policy's effects on people of color. At the same time, the New York State Board of Regents, which holds ultimate authority over all educational institutions in the state, got involved. The board approved the new admissions policy, but it also directed CUNY to monitor its impact and scheduled a revote for December 2002.

What did the monitoring show? Sandra Del Valle, associate counsel of the Puerto Rican Legal Defense and Education Fund, and I examined the admissions data the Office for Civil Rights gathered for fall 2001, the first semester that the new policy was fully implemented. We specifically wanted to know how the policy affected the students who met the standard admissions requirements—the students whose grades, college-preparation units, and other credentials were sufficient for admission. These students would have been admitted to bachelor's degree programs in the past, but now faced additional testing hurdles. Among these students, 21 percent of the white applicants, 42 percent of the African American applicants, 42 percent of the Hispanic applicants, and 27 percent of the Asian applicants were turned away solely on the basis of the new testing requirement.

CUNY (that is, its central administration) analyzed the admissions data differently. It included the ESL and SEEK applicants who didn't have to pass all the tests to gain admission to the bachelor's programs, and it concluded that the policy had produced little ethnic or racial change in the admissions rates. CUNY also reported only small changes in the overall diversity of the entering first-year classes between 1999 and 2002, before and after the policy went into effect. The proportion of white and Asian first-year students increased two to three percentage points, while the proportion of African American and Latino first-year students decreased by two to three points. But even this relatively small change in diversity reverses a long-term trend that began in 1970. Is CUNY on a new course, becoming whiter and whiter?

Validity of Tests

Del Valle and I focused on the applicants who met the standard admissions requirements (such as grades and college-prep units). In combination, these requirements generally have respectable predictive validity; they predict success in first-year courses reasonably well. Can we say the same of the tests that now bar many students from the bachelor's degree programs?

Shortly after the trustees adopted the new admissions policy, the Rand Corporation, commissioned by a mayor's task force on CUNY, summarized its research findings on CUNY's tests. Rand found that the correlation coefficients between the writing, reading, and math skills tests and first-year grades in bachelor's degree programs were .14, .18, and .25, respectively. A correlation coefficient can range from 0 to 1, and although no single number unequivocally indicates adequate test validity, most authorities would have more confidence in the tests if the correlations coefficients were over .40.

Students are exempt from the skills tests if they receive specific scores on the SAT or the state Regents exams. Rand found that the SAT verbal and math scores correlated .19 and .24 with first-year grades—far lower than the SAT's general performance nationwide and, I would add, too low to instill much confidence in the SAT's usefulness at CUNY. The data with respect to the Regents tests were unknown.

To its credit, CUNY hired the ACT (American College Test) organization to develop better reading and writing tests. As the Board of Regents' December 2002 vote on the admissions policy approached, CUNY pulled together its data on the predictive validity of the tests. CUNY said that for technical reasons, it was inappropriate to calculate the correlation coefficients most researchers like to see. Instead, it reported data showing that students who meet cutoff scores earn a bit higher grades.

For example, CUNY uses a cutoff score of 480 on the SAT verbal test; students who earn a 480 do not need to pass the reading and writing skills tests in order to enter bachelor's degree programs. CUNY reported that students scoring between 480 and 520 earn a mean grade point average of 2.87 in first-year composition courses. In contrast, students scoring between 430 and 470 earn a grade point average of 2.81. In college courses that require a great deal of reading, the grade point difference is 2.61 versus 2.44. CUNY found slightly smaller grade point differences with respect to the Regents' English exam and the ACT reading test. I don't believe any of the differences are sufficiently large to justify using the cutoffs to exclude students from the bachelor's degree programs.

CUNY points out that students have several opportunities to meet the cutoffs. If they don't meet a cutoff on the SAT verbal or math test, they can do so on the Regents tests, and if they still fail, they can meet the cutoffs on the writing, reading, and math skills tests. They also can retake these skills tests, and CUNY provides workshops to help prepare them. For such students, the critical test is usually the ACT writing skills test. It gives most students the greatest difficulty. CUNY provided no data on the validity of this test's cutoff score prior to the Board of Regents' December 2002 vote. More recent CUNY data show only slight differences between the first-term grades of students who just meet the cutoff and students who fall a notch below it. What's more, those who fall a notch below the cutoff earn the slightly better grades.1 CUNY's testing program is indeed flawed.

Unwarranted Legitimacy

In the months leading up to the Board of Regents' December 2002 vote, many of us—faculty members and community activists—lobbied the regents to add flexibility and restore opportunity to CUNY admissions. We urged the regents to stop rejecting applicants solely on the basis of standardized tests, which so often plague low-income students and students of color. We pointed to the evidence that CUNY's tests even lack predictive validity. But the regents seemed uninterested in the legitimacy of the tests.2 They approved the new admissions policy, adding that CUNY should monitor it for an additional year.

In his 1999 book, Standardized Minds: The High Price of American's Testing Culture and What We Can Do to Change It, social critic Peter Sacks argues that our culture as a whole gives standardized tests unwarranted legitimacy and power. In part, we do so because tests generate numerical scores and seem "scientific." In addition, standardized tests serve a latent social function. By consistently yielding higher scores among economically advantaged white people, the tests help perpetuate the social class and racial structure of our society—an outcome that many may tacitly approve of.

California has been among the first states to limit the unwarranted power of standardized tests. The University of California's questioning of the SAT is well known. In addition, the Mexican-American Legal Defense and Education Fund persuaded California community colleges to use placement tests only if the tests clearly predict success in college courses. This regulation applies to placement—not admissions. At CUNY, where the issue is admissions, there is an even greater need to stop turning students back on the basis of invalid tests.

Notes

1. Students who fell below the writing test cutoff score but took bachelor's degree courses included ESL and SEEK students. Also, CUNY only examined data from the first administration of the test, so some students entered bachelor's degree programs after passing it at a later sitting. (Back to text)

2. The Board of Regents has also ignored problems with standardized tests when it comes to teacher certification. Education students must pass a state certification test to become teachers, and schools of education must achieve high passing rates to stay in operation. But there is no evidence that the test predicts good teaching. And school officials have told me that it's their clear impression that this test, too, disproportionately impedes students of color. (Back to text)

William Crain is professor of psychology at the City College of the City University of New York and author of the 2003 book Reclaiming Childhood: Letting Children Be Children in Our Achievement-Oriented Society.

Printer-Friendly Page