[Paleopsych] Espenshade and Chung: The Opportunity Cost of Admission Preferences at Elite Universities

Premise Checker checker at panix.com
Thu Jun 9 01:05:07 UTC 2005

The Opportunity Cost of Admission Preferences at Elite Universities*
Thomas J. Espenshade1 and Chang Y. Chung11Princeton University
Social Science Quarterly Volume 86 Issue 2 Pages 293-305 - June 2005


[First, the summary from CHE:

Dropping Affirmative Action Would Harm Black and Hispanic Applicants but
Help Asian Applicants, Study Finds
News bulletin from the Chronicle of Higher Education, 5.6.8

    Disregarding race in college admissions would cause sharp drops in the
    number of black and Hispanic students enrolled at elite institutions,
    according to a new study by two researchers at Princeton University.

    The study, described in an article published in the June issue of
    Social Science Quarterly, also found that eliminating affirmative
    action would significantly raise the number of Asian-American
    students, while having little effect on white students.

    If affirmative action were eliminated, the acceptance rates for black
    applicants would fall to 12.2 percent from 33.7 percent, while the
    acceptance rates for Hispanic applicants would drop to 12.9 percent
    from 26.8 percent, according to the study. Asian-American students
    would fill nearly 80 percent of the spaces not taken by black and
    Hispanic students, the researchers found, while the acceptance rate
    for white students would increase by less than 1 percent.

    The researchers who conducted the study -- Thomas J. Espenshade, a
    professor of sociology, and Chang Y. Chung, a statistical programmer
    at Princeton's Office of Population Research -- looked at the race,
    sex, SAT scores, and legacy status, among other characteristics, of
    more than 124,000 applicants to elite colleges and universities.

    "We're trying to put these admissions preferences in context so people
    understand that lots of students, including those with SAT scores
    above 1500, are getting a boost," Mr. Espenshade said in a written
    statement. "The most important conclusion is the negative impact on
    African-American and Hispanic students if affirmative-action practices
    were eliminated."

    Stephen H. Balch, president of the National Association of Scholars,
    which opposes racial preferences in admissions, said the study's
    findings revealed that affirmative-action policies are "about

    "That it's Asian students who bear the brunt of affirmative-action
    policies at elite institutions strikes me as an interesting finding in
    and of itself," Mr. Balch said on Tuesday. "One of the dirty little
    secrets in all of this is that one of the chief losers is a minority

    The article, "The Opportunity Cost of Admission Preferences at Elite
    Universities," is online for subscribers and can be purchased by
    nonsubscribers on the journal's [67]Web site. Social Science Quarterly
    is published by the [68]Southwestern Social Science Association

    Background articles from The Chronicle:
      * [69]U. of California Admitted Fewer Students With Low SAT Scores,
        a Recent Target of Critics, in 2004, Report Says (4/5/2005)
      * [70]Michigan: Who Really Won? (1/14/2005)
      * [71]Federal Court Declines to Set New Limits on Affirmative Action
      * [72]Affirmative Action Survives, and So Does the Debate (7/4/2003)
      * [73]For Asians, Affirmative Action Cuts Both Ways (6/6/2003)

      * [74]From 'Bastions of Privilege' to 'Engines of Opportunity'
      * [75]Putting the Michigan Rulings Into Practice (2/25/2005)
      * [76]In California, a Misguided Battle Over Race (5/21/2004)


   68. http://www.sssaonline.org/
   69. http://chronicle.com/daily/2005/04/2005040502n.htm
   70. http://chronicle.com/weekly/v51/i19/19a02101.htm
   71. http://chronicle.com/weekly/v51/i18/18a03401.htm
   72. http://chronicle.com/weekly/v49/i43/43s00101.htm
   73. http://chronicle.com/weekly/v49/i39/39a02401.htm
   74. http://chronicle.com/weekly/v51/i25/25b01801.htm
   75. http://chronicle.com/weekly/v51/i25/25b02801.htm
   76. http://chronicle.com/weekly/v50/i37/37b01601.htm

E-mail me if you have problems getting the referenced articles.


The Opportunity Cost of Admission Preferences at Elite Universities* Thomas J. 
Espenshade1 and Chang Y. Chung11Princeton University
Social Science Quarterly Volume 86 Issue 2 Page 293 - June 2005 

Objective. This study examines how preferences for different types of 
applicants for admission to elite universities influence the number and 
composition of admitted students.

Methods. Previous research with these NSCE data employed logistic regression 
analysis to link information on the admission decision for 124,374 applications 
to applicants' SAT scores, race, athletic ability, and legacy status, among 
other variables. Here we use micro simulations to illustrate what the effects 
might be if one were to withdraw preferences for different student groups.

Results. Eliminating affirmative action would substantially reduce the share of 
African Americans and Hispanics among admitted students. Preferences for 
athletes and legacies, however, only mildly displace members of minority 

Conclusions. Elite colleges and universities extend preferences to many types 
of students, yet affirmative action is the one most surrounded by controversy.

In an earlier article in this journal, Espenshade, Chung, and Walling (2004) 
examined the strength of admission preferences for underrepresented minority 
students, athletes, and alumni children at three highly selective private 
research universities in the United States. Using data from the National Study 
of College Experience on 124,374 applications for admission during the 1980s 
and the fall semesters of 1993 and 1997, they found that elite universities 
give extra weight in admissions to candidates whose SAT scores are above 1500, 
who are African American, and who are student athletes. A smaller, but 
nevertheless important, preference is extended to Hispanic and legacy 
applicants. African-American applicants receive the equivalent of 230 extra SAT 
points (on a 1600-point scale), and being Hispanic is worth an additional 185 
SAT points. Other things equal, recruited athletes gain an admission bonus 
worth 200 points, while the preference for legacy candidates is worth 160 
points. Asian-American applicants face a loss equivalent to 50 SAT points. The 
underrepresented minority advantage is greatest for African-American and 
Hispanic applicants whose SAT scores are in the 1200 [-] 1300 range, and not 
for applicants near the lower end of the SAT distribution as some have 
suggested (cf. Dugan et al., 1996 ). Finally, the advantage that athletes have 
over nonathletes in elite university admissions has been growing, whereas the 
strength of the minority student advantage, especially for Hispanic candidates, 
has been waning.

An important but unanswered question has to do with the opportunity cost of 
these admission preferences. Who are the beneficiaries and, by extension, who 
loses a seat at academically selective universities because some students are 
favored over others in the admission process?

The admission process at academically selective colleges and universities 
inevitably entails opportunity costs (Bowen and Levin, 2003; Shulman and Bowen, 
2001 ). A decision to admit one student involves a choice not to admit someone 
else. When preferences enter into the mix, applicants who are denied admission 
often feel that they would have been next in line to be accepted had 
preferences not played a part (Kane, 2003). In this article, using the same 
data, we extend the work of Espenshade, Chung, and Walling (2004) and ask two 
questions. First, what is the impact of affirmative action on the profile of 
students admitted to elite universities? In other words, who gains and who 
loses as a result of admission preferences for underrepresented minority 
students? And, second, to what extent do preferences for athletes and legacies, 
both of whom are disproportionately white, offset the effects of affirmative 

Answering these questions is inherently difficult. One reason is that the 
selection process at elite private institutions is typically more nuanced and 
subjective than the explicit point systems formerly relied on by undergraduate 
admission officers at the University of Michigan and other large public 
universities (University of Michigan, 2002; Zwick, 2002:39 ). With a more 
numerical approach, it would be relatively straightforward to see how 
applicants' comparative rankings would be reordered as points were removed for 
being a minority applicant, an athlete, or a legacy (Kane, 2003).

More importantly, many of the factors affecting the makeup of the first-year 
class are themselves endogenous to the choice of a particular preference 
regime.1 Eliminating racial and ethnic preferences, for example, could 
discourage applications from members of minority student groups (Bowen and Bok, 
1998; Conrad, 1999; Klitgaard, 1985).2 The proportion of admitted students who 
eventually enroll (the so-called yield rate) might also be adversely affected 
if minority students would be less likely to matriculate at campuses where 
there are relatively few members of their own group (Bowen and Bok, 1998; 
Conrad, 1999 ) or if financial aid is more restricted at less academically 
selective schools to which minority students might be more likely to apply in 
the absence of affirmative action (Dugan et al., 1996 ). Finally, institutions 
that are no longer able to consider an applicant's race or ethnicity may still 
try to meet representational goals by altering the weights assigned to other 
factors in the selection process. Fryer, Loury, and Yuret (2003) predict that 
schools will "flatten" the function that relates test scores and other measures 
of academic performance to the probability of admission and give greater 
emphasis to socioeconomic background and other personal factors. Indeed, in 
response to the Board of Regents' 1995 decision to end affirmative action at 
the University of California, the Berkeley law school faculty voted to reduce 
the importance of LSAT scores and other numerical indicators from "greatest" to 
"substantial" weight (Guerrero, 2002:91 [-] 92).

One way to gauge the effect of admission preferences on the composition of 
entering classes is to consult expert opinion. In 1976 [-] 1977 all U.S. law 
schools were asked how many minority students they had in their first-year 
classes and how many of these would have been admitted if it had been 
impossible to detect the racial background of applicants. Respondents believed 
the number of African-American students would have declined by 82 percent. Only 
27 percent as many Chicano students would have been accepted. Just 28 percent 
of all minority students, including Asians, would have been admitted under a 
race-blind procedure (Klitgaard, 1985:155).

A more satisfactory approach is to rely on a quantitative analysis of how 
individual applicants' probabilities of being admitted change depending on 
which preferences are in effect. In the remainder of this article we present 
the results of several micro-simulation exercises aimed at illustrating how the 
profile of students admitted to our three elite universities would differ 
depending on whether a candidate's racial background was considered in the 
admission decision and whether preferences were granted to athletes and to 
legacies. We combine athlete and legacy preferences because athletes and 
legacies comprise a relatively small proportion of the applicant pool and 
because both student groups are largely white. Our analysis is based on the 
1997 cohort of applicants to reflect recent conditions, and we assume that 
satisfactory answers to who loses and who gains under different preference 
structures can be obtained by turning selected preferences on and off and 
ignoring second-round effects.

More specifically, our simulations are based on the logistic regression model 
for the 1997 cohort in Table 7 in Espenshade, Chung, and Walling (2004) . This 
equation is used to predict a probability of admission (at the institution to 
which the application was sent) for each of the 45,549 applicants in the 1997 
cohort. Predictor variables include sex, citizenship status, SAT score, 
race/ethnicity, recruited athlete, and legacy status. Following a procedure 
suggested by Kohn, Manski, and Mundel (1976) , we also generated a random 
proportion on the uniform distribution between 0 and 1 for each applicant. An 
applicant was assumed to be accepted if the random proportion was less than or 
equal to the predicted probability of admission; otherwise they were put in the 
rejected category. The effect of removing race from consideration was captured 
by setting all regression coefficients on racial background to zero or, 
equivalently, by assuming that all applicants are white (the reference 
category). We eliminate preferences for athletes and legacies by setting the 
athlete and legacy coefficients to zero.3

Before examining the effects of withdrawing preferences for selected groups of 
students, we first want to ask how well our simulation methodology reproduces 
the actual distribution of students admitted in 1997. The results are shown in 
Table 1 . There is remarkably good agreement between the number and 
distribution of students actually admitted and those in the simulation. For 
example, 899 African-American candidates were accepted from the 2,671 who 
applied, in contrast to 910 who were expected to be admitted in the simulation. 
The overall acceptance rate for African-American applicants was simulated to be 
34.1 percent in contrast to an actual rate of 33.7 percent. This high degree of 
correspondence between the actual and expected profiles of admitted students 
adds credibility to the simulations we discuss next.

Table 2 shows the actual profile of admitted students in 1997 and the 
micro-simulation results of removing racial/ethnic admission preferences while 
keeping those for athletes and legacies (Simulation 1), retaining preferences 
for underrepresented minority students but eliminating them for athletes and 
legacies (Simulation 2), and removing preferences for both minority students 
and for athletes/legacies (Simulation 3).4 To understand the impact of 
affirmative action, we compare the actual distribution of students with 
Simulation 1, which ignores applicants' race or ethnicity. The result of 
eliminating admission bonuses for African-American and Hispanic applicants 
would be dramatic. Acceptance rates for African-American candidates would fall 
from 33.7 percent to 12.2 percent, a decline of almost two-thirds, and the 
proportion of African-American students in the admitted class would drop from 
9.0 to 3.3 percent. The acceptance rate for Hispanic applicants would be cut in 
half [-] from 26.8 percent to 12.9 percent, and Hispanics would comprise just 
3.8 of all admitted students versus an actual proportion of 7.9 percent. If 
admitting such small numbers of qualified African-American and Hispanic 
students reduced applications and the yield from minority candidates in 
subsequent years, the effect of eliminating affirmative action at elite 
universities on the racial and ethnic composition of enrolled students would be 
magnified beyond the results presented here.

White plaintiffs in Gratz v. Bollinger (2003) and Grutter v. Bollinger (2003) 
argued that they were unfairly denied admission while some less qualified 
minority students were accepted. Our results show that removing consideration 
of race would have a minimal effect on white applicants to elite universities. 
The number of accepted white students would increase by 2.4 percent, and the 
white acceptance rate would rise by just 0.5 percentage points [-] from 23.8 to 
24.3 percent. Many rejected white applicants may feel they would have been 
accepted had it not been for affirmative action, but such perceptions probably 
exaggerate the reality. It would be difficult to tell from the share of white 
students on campus whether or not the admission office was engaged in 
affirmative action.

Asian applicants are the biggest winners if race is no longer considered in 
admissions. Nearly four out of every five places in the admitted class not 
taken by African-American and Hispanic students would be filled by Asians. We 
noted earlier that Asian candidates are at a disadvantage in admission compared 
to their white, African-American, and Hispanic counterparts. Removing this 
disadvantage at the same time preferences for African Americans and Hispanics 
are eliminated results in a significant gain in the acceptance rate for Asian 
students [-] from 17.6 percent to 23.4 percent. Asians, who comprised 29.5 
percent of total applicants in 1997, would make up 31.5 percent of accepted 
students in the simulation, compared with an actual proportion of 23.7 percent. 
Other aspects of admitted students, including the distribution of SAT scores 
and, especially, the proportions of students who are athletes or legacies, are 
hardly affected by affirmative action.

The remaining question is the extent to which athlete and legacy preferences 
offset preferences for underrepresented minority applicants. White students 
comprise fewer than half of all applicants in 1997, yet they account for 
three-quarters of athletes (73.3 percent) and a similar proportion of legacies 
(75.6 percent). This fact alone suggests that preferences for athletes and 
legacies are likely to boost the proportion of whites among admitted students. 
We return to the simulation results to see the magnitude of these effects. 
Suppose we begin with a situation where admission officers give no extra 
consideration to minority applicants, athletes, or legacies (see Simulation 3). 
Now introduce race consciousness into the decision making (Simulation 2). The 
effect of affirmative action for African Americans and Hispanics and of what 
some might term "disaffirmative action" for Asians is a substantial increase in 
the African-American and Hispanic shares of admitted students and a sharp 
decline in the Asian proportion. The combined African-American and Hispanic 
proportion increases from just over 7 percent to 17.5 percent, while the Asian 
share falls from one-third to one-quarter. Acceptance rates for these groups 
move in the same direction.

Next, comparing Simulation 2 with the actual distribution of accepted students 
is equivalent to adding athlete-legacy bonuses on top of those for 
underrepresented minority applicants. With the inclusion of preferences for 
athletes and legacies, the proportion of admitted students who are white rises 
somewhat (from 49.5 to 51.4 percent) as does the acceptance rate for white 
applicants. Minority student effects go in the opposite direction, but they are 
not large. The African-American share among admitted students declines modestly 
from 9.2 to 9.0 percent, the Hispanic share falls from 8.3 to 7.9 percent, and 
Asians now account for 23.7 percent of all admitted students instead of 25.1 
percent. Acceptance rates for each minority student group also decline, but the 
changes here are mostly small as well. The impacts would be greater either if 
the athlete and legacy bonuses were larger or if athletes and legacies 
accounted for more than a small share of all applicants. If the time trends 
detected earlier in Espenshade, Chung, and Walling (2004) persist, there may 
come a time when the rising preference for athletes in combination with a 
relatively stable bonus for legacies is sufficient to fully offset the 
weakening preferences for underrepresented minority applicants. Not 
surprisingly, the proportions of athletes and legacies among admitted students 
increase when admission officers give these characteristics more weight in 
admission decisions.5

No other research of which we are aware has examined the potential for athlete 
and legacy preferences to counteract admission bonuses for underrepresented 
minority applicants. Our findings on the effects of affirmative action are 
consistent with results reported elsewhere. For example, Kane (1998:432) 
contends that: "The proportion of minority students at [elite colleges and 
universities] would be extremely low if admissions committees ignored the race 
or ethnicity of applicants."Bowen and Bok (1998:31) estimate the effect of 
"race-neutral" admissions policies in the 1989 entering student cohort by 
assuming that "black applicants, grouped by SAT ranges, would have the same 
probability of being admitted as white applicants in those same ranges." At the 
five academically selective schools for which they have admission data, 
acceptance rates for African-American applicants would fall from 42 to 13 
percent if the race of applicants were ignored, while the proportion of white 
applicants admitted would only increase from 25 to 26.5 percent (assuming that 
whites filled all the seats created by accepting fewer African-American 
applicants). The impact on African-American enrollment would be equally 
dramatic. The share of African-American students in the first-year class would 
be expected to fall from 7.1 to 2.1 percent. Using a nationwide sample from the 
National Education Longitudinal Study, Long (2004b) finds that eliminating 
affirmative action at all colleges and universities would reduce the 
underrepresented minority share of students accepted from 16.1 to 15.5 percent 
across all four-year institutions and from 10.6 to 7.8 percent at the highest 
quality 10 percent of schools.

Dugan et al. (1996) estimate the effect of eliminating affirmative action on 
graduate management education programs. Using data on a sample of all 
applicants in the early 1990s, they find that failing to consider a candidate's 
minority status in admission would reduce the probability of acceptance for 
African Americans from 70 percent (the actual figure) to 52 percent. The rate 
for Hispanics would decline from 78 to 60 percent. However, the acceptance rate 
for Asians, who experience a disadvantage in admission, would increase slightly 
from 53 to 57 percent. Similar results are obtained from an analysis of more 
than 90,000 applications to law school in the 1990 [-] 1991 application year. 
Wightman (1997:15 [-] 16) shows that of 3,435 African-American applicants who 
were accepted by at least one law school, just 687 or one-fifth as many would 
have been accepted if admission decisions were based solely on LSAT scores and 
undergraduate GPAs. If instead admission determinations were based exclusively 
on undergraduate GPAs, more than 60 percent of African-American candidates who 
were originally accepted by at least one law school would still be completely 
shut out. Wightman finds similar patterns for other racial and ethnic minority 
groups, but the impacts are most severe for African-American students.

A final test comes from a real-world "natural experiment." The Board of Regents 
for the University of California system voted in 1995 to eliminate affirmative 
action in higher education. This decision was reinforced in November 1996 by a 
statewide vote in favor of Proposition 209. Impacts on graduate programs took 
effect with the fall of 1997 entering classes. Effects on admission to 
undergraduate programs were delayed until the fall of 1998. The impacts are 
striking. Compared to the fall of 1996, the number of underrepresented minority 
students admitted to the University of California [-] Berkeley Boalt Hall Law 
School for the fall of 1997 dropped 66 percent from 162 to 55 (Guerrero, 2002 
). African-American applicants were particularly affected as their admission 
numbers declined by 81 percent from 75 to 14, but acceptances of Hispanics also 
fell by 50 percent. None of the 14 admitted African-American students chose to 
enroll. Of the 55 minority students admitted, only seven enrolled in the fall 
of 1997, a falloff that had the effect of reducing the underrepresented 
minority share in the first-year class to 5 percent in 1997 compared with 26 
percent in 1994 (Guerrero, 2002:159). Similar impacts were felt at law schools 
at UCLA and UC [-] Davis.

Numbers at the undergraduate level mirrored those in graduate programs. At UC 
[-] Berkeley, just 10 percent of all undergraduate students admitted for the 
fall of 1998 were underrepresented minority students compared with 23 percent 
admitted in the previous year (Guerrero, 2002:146 ). The largest declines 
occurred among African Americans, whose admission numbers fell by 66 percent 
between 1997 and 1998. Admission to the undergraduate College of Letters and 
Science at UCLA was similarly affected (Committee on Undergraduate Admissions 
and Relations with Schools, 1999 ). Acceptance rates for African Americans fell 
from 57 percent in 1997 to 31 percent in 1998. Those for Hispanics (including 
Latino Americans and Chicanos/Mexican Americans) declined from 51 to 30 
percent. These declines were offset by small increases in admission rates for 
Asian Americans. In general, our simulation results are in very good agreement 
with the California experience.6

Critics of affirmative action in American higher education often overlook the 
fact that elite universities give added weight in the admissions process to 
many different types of student characteristics. In this article, we use 
micro-simulation analysis to investigate the effect on the profile of admitted 
students of eliminating preferences for one or more categories of students. 
Data for the 1997 entering class indicate that eliminating affirmative action 
would reduce acceptance rates for African-American and Hispanic applicants by 
as much as one-half to two-thirds and have an equivalent impact on the 
proportion of underrepresented minority students in the admitted class. White 
applicants would benefit very little by removing racial and ethnic preferences; 
the white acceptance rate would increase by roughly 0.5 percentage points. 
Asian applicants would gain the most. They would occupy four out of every five 
seats created by accepting fewer African-American and Hispanic students. The 
acceptance rate for Asian applicants would rise by one-third from nearly 18 
percent to more than 23 percent. We also show that, even though athlete and 
legacy applicants are disproportionately white and despite the fact that 
athlete and alumni children admission bonuses are substantial, preferences for 
athletes and legacies do little to displace minority applicants, largely 
because athletes and legacies make up a small share of all applicants to highly 
selective universities.


•   Arenson, Karen W. 2003. "Change on Early Admission Produces Application 
Shifts."New York Times November 13.
•   Bowen, William G., and Derek Bok. 1998. The Shape of the River: Long-Term 
Consequences of Considering Race in College and University Admissions. 
Princeton, NJ: Princeton University Press.
•   Bowen, William G., and Sarah A. Levin. 2003. Reclaiming the Game: College 
Sports and Educational Values. Princeton, NJ: Princeton University Press.
•   Card, David, and Alan B. Krueger. 2004. Would the Elimination of 
Affirmative Action Affect Highly Qualified Minority Applicants? Evidence from 
California and Texas. NBER Working Paper 10366, March. Cambridge, MA: National 
Bureau of Economic Research.
•   Committee on Undergraduate Admissions and Relations with Schools (CUARS). 
1999. 1998 [-] 1999 Report to the Academic Senate at UCLA. Los Angeles, CA: 
•   Conrad, Cecilia A. 1999. "Affirmative Action and Admission to the 
University of California."Pp. 171 [-] 96 in Paul Ong, ed., Impacts of 
Affirmative Action: Policies and Consequences in California. Walnut Creek, CA: 
AltaMira Press.
•   Dugan, Mary Kay, Nazli Baydar, William R. Grady, and Terry R. Johnson. 
1996. "Affirmative Action: Does it Exist in Graduate Business 
Schools?"Selections Winter: 11 [-] 18.
•   Espenshade, Thomas J., Chang Y. Chung, and Joan L. Walling. 2004. 
"Admission Preferences for Minority Students, Athletes, and Legacies at Elite 
Universities."Social Science Quarterly 85(5): 1422 [-] 46.
[Synergy Abstract] [ISI Abstract]
•   Fryer Jr., Roland G., Glenn C. Loury, and Tolga Yuret. 2003. Color-Blind 
Affirmative Action. Unpublished manuscript, Harvard Society of Fellows and 
•   Gratz v. Bollinger (decided June 23, 2003). 123 S. Ct. 2411.
•   Grutterv. Bollinger (decided June 23, 2003). 123 S. Ct. 2325.
•   Guerrero, Andrea. 2002. Silence at Boalt Hall: The Dismantling of 
Affirmative Action. Berkeley, CA: University of California Press.
•   Kane, Thomas J. 1998. "Racial and Ethnic Preferences in College 
Admissions."Pp. 431 [-] 56 in Christopher Jencks and Meredith Phillips, eds., 
The Black-White Test Score Gap. Washington, DC: Brookings Institution Press.
•   Kane, Thomas J.. 2003. "The Long Road to Race-Blindness."Science 302: 571 
[-] 73.
[CrossRef Abstract] [ISI Abstract]
•   Klitgaard, Robert E. 1985. Choosing Elites. New York: Basic Books, Inc.
•   Kohn, Meir G., Charles F. Manski, and David S. Mundel. 1976. "An Empirical 
Investigation of Factors Which Influence College-Going Behavior."Annals of 
Economic and Social Measurement 5(4): 391 [-] 419.
•   Long, Mark C. 2004a. "College Applications and the Effect of Affirmative 
Action."Journal of Econometrics 121(1 [-] 2): 319 [-] 42.
[CrossRef Abstract] [ISI Abstract]
•   Long, Mark C.. 2004b. "Race and College Admissions: An Alternative to 
Affirmative Action?"Review of Economics and Statistics 86(4): 1020 [-] 33.
[CrossRef Abstract] [ISI Abstract]
•   Shulman, James L., and William G. Bowen. 2001. The Game of Life: College 
Sports and Educational Values. Princeton, NJ: Princeton University Press.
•   Sullivan, John. 2003. "At Rutgers, Weathering an Ordeal."New York Times 
November 30: Section 14, p. 1.
•   University of Michigan. 2002. Description of University of Michigan 
Undergraduate Admissions Policy. Ann Arbor, MI: University of Michigan.
•   Wightman, Linda F. 1997. "The Threat to Diversity in Legal Education: An 
Empirical Analysis of the Consequences of Abandoning Race as a Factor in Law 
School Admission Decisions."New York University Law Review 72(1): 1 [-] 53.
•   Zwick, Rebecca. 2002. Fair Game? The Use of Standardized Admissions Tests 
in Higher Education. New York: RoutledgeFalmer.


* Support for this research has been provided by grants from the Andrew W. 
Mellon Foundation and the National Institute of Child Health and Human 
Development Center Core Grant P30 HD32030. We are grateful to Elana Broch, 
James Snow, Kristen Turner, and Chengzhi Wang for bibliographic assistance. 
Kalena Cortes, Sara Curran, Bonnie Ghosh-Dastidar, Lauren Hale, Stephen 
LeMenager, Germán Rodríguez, Christopher Weiss, Charles Westoff, and, 
especially, Joyce Jacobsen and Mark Long contributed many useful suggestions.

1The interdependent nature of the decision-making process was observed by 
economist Robert Klitgaard (1985:78) nearly two decades ago: "The existence of 
incentive effects transforms the selection problem from a static to a dynamic 
framework. The classic selection problem is static [-] given an applicant pool 
with certain characteristics, choose those most likely to succeed along certain 
criteria of later performance. The dynamic problem is richer. The choice of 
this particular class of students must take account of the effects of the 
choice on applicant pools in the future [...] ." Further evidence that today's 
students respond quickly to altered incentives is provided by the effect of 
changes in admission policies at several elite universities. Yale and Stanford, 
both of which changed last year from a binding early decision admission program 
to nonbinding "single-choice" early action, saw applications for the 2004 
entering class increase by 42 and 62 percent, respectively. Early applications 
to Harvard fell 47 percent in response to a switch from nonbinding early 
action, where students could apply early to several institutions, to 
single-choice early action [-] a plan that prohibits students from applying 
early to any other institution. Princeton, which made no changes in its 
admission policies, saw a 23 percent decline in its early applications 
(Arenson, 2003).

2The magnitude of this effect has been estimated separately by Long (2004a) and 
Card and Krueger (2004) , with somewhat different results. Long finds that 
underrepresented minority students in California and Texas are predicted to 
send fewer SAT-score reports to top-tier in-state public colleges and 
universities after the elimination of affirmative action, while white and 
Asian-American students are predicted to send more. Card and Krueger find no 
change in the propensity of highly qualified African-American and Hispanic 
students to send their SAT scores to the most selective public institutions in 
either California or Texas. Eliminating affirmative action also left other 
features of the application process unaffected, including the number of schools 
to which scores were sent and the lower bound on the quality of such 

3Long (2004b) uses a comparable micro simulation to evaluate the effect of 
eliminating affirmative action.

4In the simulation reported in Table 1 , the average of the predicted admission 
probabilities for the 45,549 applicants was 0.219280, exactly the same as the 
actual proportion of applicants accepted (9,988/45,549). In the simulations 
described in Table 2, removing preferences for particular student groups has 
the effect of lowering the average predicted admission probability below 
0.219280. In these cases, the intercept of the logistic regression for the 1997 
cohort in Table 7 in Espenshade, Chung, and Walling (2004) was adjusted upward 
by enough in each simulation so that the average of the predicted admission 
probabilities equaled 0.219280.

5 We prepared an alternate simulation by ranking applicants on the basis of 
their SAT scores and admitting students having the top 9,988 scores (the actual 
number of students accepted). This is the closest that any of our simulations 
comes to choosing a class solely on the basis of academic merit. Applicants in 
this simulation average 1512 on their SATs. Compared to students who were 
actually admitted, the shares of most student groups decline in the simulation 
[-] from 51.4 percent to 47.7 for whites, from 9.0 to 0.9 for African 
Americans, from 7.9 to 2.2 for Hispanics, from 10.2 to 1.9 for athletes, and 
from 6.5 to 3.2 for legacies. Only the share of Asians increases when SAT 
scores dominate [-] from 23.7 to 38.7 percent. These results are qualitatively 
similar to effects reported by Klitgaard (1985:29) had Harvard's Class of 1975 
been chosen on the basis of SAT verbal scores alone. The percentage of admitted 
students who were alumni sons would have declined from 13.6 to 6.1, of athletes 
from 23.6 to 4.5, and of African Americans from 7.1 to 1.1. The proportion of 
scholarship students would have remained unchanged at 55 percent.

6The effects of rescinding affirmative action were not limited to California. 
Voters in the State of Washington passed a referendum forbidding affirmative 
action at the state university. In 1998 at the University of Washington, 1 in 
11 students in the first-year class was a member of a minority group. By the 
fall of 1999, when the new law had taken effect, the ratio fell to 1 out of 18 
students (Sullivan, 2003).

Social Science Quarterly
Volume 86 Issue 2 Page 293  - June 2005

1Princeton University

Direct correspondence to Thomas J. Espenshade, Office of Population Research, 
249 Wallace Hall, Princeton University, Princeton, NJ 
08544-2091tje at Princeton.Edu [>] .

* Support for this research has been provided by grants from the Andrew W. 
Mellon Foundation and the National Institute of Child Health and Human 
Development Center Core Grant P30 HD32030. We are grateful to Elana Broch, 
James Snow, Kristen Turner, and Chengzhi Wang for bibliographic assistance. 
Kalena Cortes, Sara Curran, Bonnie Ghosh-Dastidar, Lauren Hale, Stephen 
LeMenager, Germán Rodríguez, Christopher Weiss, Charles Westoff, and, 
especially, Joyce Jacobsen and Mark Long contributed many useful suggestions.

Image Previews

Number of Applicants in the 1997 Entering Cohort, Number Admitted, and 
Simulated Number Admit...

Number and Characteristics of Admitted Students Simulated Under Alternative 
Preference Scenar...

More information about the paleopsych mailing list