Estimating the Size and Impact of Affirmative Action in Undergraduate Admissions at the University of Cape Town

• Author: Patrizio Piraino,Andrew Kerr,Vimal Ranchhod
• Published date: 01 December 2017
• DOI: http://doi.org/10.1111/saje.12174
• Date: 01 December 2017

ESTIMATING THE SIZE AND IMPACT OF AFFIRMATIVE

ACTION IN UNDERGRADUATE ADMISSIONS AT THE

UNIVERSITY OF CAPE TOWN

ANDREW KERR*, PATRIZIO PIRAINO

†

AND VIMAL RANCHHOD

‡

Abstract

In this paper, we estimate the extent and targeting of affirmative action at the University of Cape

Town (UCT), a large public university in South Africa. To do this we use admissions data from

the UCT, as well as South African population census data and administrative enrolment and

graduation data from the South African Department of Higher Education. We find that affirma-

tive action does have a significant effect on the racial distribution of who is made an offer by the

university. We also find that affirmative action is well targeted, with those who we estimate to be

beneficiaries being of much lower socioeconomic status than those who we estimate are displaced

by affirmative action. Beneficiaries of affirmative action have low graduation rates on average,

with those beneficiaries who attend UCT being less likely to graduate than those beneficiaries

who enrol at other public universities.

JEL Classification: I20, I23, I28

Keywords: Affirmative Action, Education, University of Cape Town, South Africa.

1. INTRODUCTION

Affirmative action in higher education is a policy designed to lower inequalities across

different racial groups. South Africa has large and persistent inequalities between race

groups in education and labour market outcomes, amongst many other measures of well-

being. These inequalities stem from a long history of racial discrimination, most recently

in the Apartheid period, and despite some progress, life chances are still strongly corre-

lated with an individual’s racial group.

* Corresponding author: Senior Research Officer, DataFirst, School of Economics Building,

Middle Campus, University of Cape Town, Rondebosch, 7701, South Africa. Tel: 127 21

6504952. E-mail: andrew.kerr@uct.ac.za

†

School of Economics, University of Cape Town

‡

SALDRU, University of Cape Town.

The authors thank Tim Brophy for his help with merging the applicant address data with the

2011 Census small area data. They thank Jean Skeane and Norman Nkwana from the South

African Department of Higher Education and Training for assistance in matching the HEMIS

data to the applicant data. Carl Herman in the UCT admissions office, Hugh Amoore the

UCT registrar and Judy Favish and Jane Hendry in the UCT Institutional Planning Depart-

ment provided very helpful background information and helped to source the applicant data.

The two anonymous referees and participants in two SALDRU seminars at UCT, the MASA

2014 conference in Durban, as well as a UNU-WIDER 2014 conference in Helsinki all pro-

vided helpful comments.

V

C2017 Economic Society of South Africa. doi: 10.1111/saje.12174

515

South African Journal of Economics Vol. 85:4 December 2017

South African Journal

of Economics

The Higher Education Act that regulates higher education in South Africa notes that

it is desirable that there is redress for past discrimination and ensuring representivity and

equal access. The University of Cape Town (UCT) is a public higher education institu-

tion that uses its admissions policy to implement affirmative action in selecting which

students to make offers of places at the university to. Affirmative action is targeted

towards black African, coloured, Indian and Chinese South Africans, as a result of the

history of discrimination against these groups by the white minority group.

In this paper, we explore the impact of an affirmative action admissions policy at the

UCT. The policy that we study was applied until 2015. For applicants entering in 2016

a new admissions policy was implemented which took account of both race and various

measures of socioeconomic status (SES) in the decision to make an offer of admission.

We do not investigate the effects of this new policy in this paper.

We study the size of the effect on the racial composition of the distribution of offers

made to potential undergraduate students, the extent to which the policy targeted low-

income students and how the policy impacted both the beneficiaries of the policy and

those who were displaced. To do this we combine university applicant data, population

census data and government administrative higher education records of applicants.

In particular, we estimate the number of offers of admission that can be attributed to

affirmative action for two different years at UCT. We then calculate to what extent this

affirmative action policy changes the racial distribution of offers of admission. The next

question that we answer is whether relatively rich Black African students are benefiting

from affirmative action and displacing relatively poorer White students. We utilise a geo-

matching algorithm to match the 2013 application data to t he 2011 Census. This allows

us to obtain a measure of SES of the applicant and to evaluate the extent to which the

policy targeted low-income students. We next combine data from the national Depart-

ment of Higher Education and Training and estimate the fraction of displaced white stu-

dents who subsequently enrolled, either at UCT or at some other university in South

Africa. Finally, we compare subsequent graduation rates of both beneficiaries and dis-

placed students who were affected by affirmative action in admissions.

We find that affirmative action has a substantial impact on the racial distribution of

offers of admission. The most important finding is that there is no evidence to support

the concern that a substantial amount of displacement is being driven by high SES Black

Africans at the expense of low SES Whites. In addition, most displaced Whites do subse-

quently enrol at some university, with a non-trivial proportion enrolling for some other

degree at UCT. On the other hand, when we consider eventual graduation rates, we do

find some evidence consistent with the “mismatch hypothesis,” in the sense that the ben-

eficiaries of affirmative action who enrolled at UCT are substantially less likely to gradu-

ate, relative to similar beneficiaries who were made an offer of admission at UCT but

subsequently enrolled at some other university.

2. RESEARCH ON AFFIRMATIVE ACTION

Affirmative action in college admissions has attracted significant levels of controversy and

debate. As succinctly summarised by Bertrand et al. (2010), the theoretical debates can

be grouped into two broad issues of contention. First, there are concerns about how

well-targeted affirmative action is. In particular, critics point out that such policies may

privilege the most advantaged members of disadvantaged groups, at the cost of relatively

516 South African Journal of Economics Vol. 85:4 December 2017

V

C2017 Economic Society of South Africa.