Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Compare: A Journal of Comparative and International

Education

ISSN: 0305-7925 (Print) 1469-3623 (Online) Journal homepage: http://www.tandfonline.com/loi/ccom20

The stratification of higher education in the USA


and Taiwan: a comparative analysis of students’
college-choice outcomes

Patricia Yu

To cite this article: Patricia Yu (2018): The stratification of higher education in the USA and
Taiwan: a comparative analysis of students’ college-choice outcomes, Compare: A Journal of
Comparative and International Education, DOI: 10.1080/03057925.2018.1448971

To link to this article: https://doi.org/10.1080/03057925.2018.1448971

Published online: 03 Apr 2018.

Submit your article to this journal

Article views: 5

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=ccom20
Compare, 2018
https://doi.org/10.1080/03057925.2018.1448971

The stratification of higher education in the USA and Taiwan:


a comparative analysis of students’ college-choice outcomes
Patricia Yu‡
Department of Education, College of Education, National Taiwan Normal University, Taipei City, Taiwan

ABSTRACT KEYWORDS
This comparative study modifies Cabrera and La Nasa’s model to Social stratification; college
examine social and class stratification in students’ college-choice choice; financing higher
outcomes in the USA and Taiwan, and compares the mechanisms education; comparative
that perpetuate this stratification between the two systems. Students’ analysis
college-choice outcomes are defined by both institutional selectivity
and control, in order to test not only stratification in the quality of
institution students attend but also their financial burden of college
costs. The findings show that students from lower-socioeconomic
backgrounds are less likely to attend high-selective institutions
and institutions that embrace better resources and provide more
subsidies to students. Current financing structures embedded in
both systems increase stratification, but through different ways. In
addition to enhancing academic preparation for socioeconomically
disadvantaged students, fundamentally rethinking the financing
structures – especially the distribution of public funds – is imperative
to increase equality for both countries in the cost-sharing era.

Introduction
Higher education is viewed as one of the main channels for upward social mobility (Brennan
and Naidoo 2008). There has seen an increasing need for college education at the societal
and individual levels worldwide. After higher education expansion, however, as Lucas (2001)
argues, ‘for levels of education that are universal, competition will occur around the type
of education attained’ (1653). Because higher education has become universal in both the
USA and Taiwan,1 This study seeks to understand more about social and class stratification
in students’ college-choice outcomes – which will influence their college education, degree
completion, and labour-market outcomes (Brewer, Eide, and Ehrenberg 1999; Cabrera,
Burkum, and La Nasa 2005; Melguizo 2008; Strayer 2002) – within these two systems. In
addition, this study compares the mechanisms that contribute to this stratification between
these two different systems.

CONTACT  Patricia Yu  pyu@ntnu.edu.tw



902 Education Building, 162 Heping East Road, Department of Education, College of Education, National Taiwan Normal
University, Taipei City 10610, Taiwan.
© 2018 British Association for International and Comparative Education
2   P. YU

Differences between the USA and Taiwan higher education systems


Although both the USA and Taiwan are high-income, developed countries (World Bank
2004) and charge up-front college tuition fees (or tuition fees that are paid upon matricu-
lation) (Johnstone and Marcucci 2010), the primary distinction between their higher edu-
cation systems falls into the logic of financing. In the USA, higher education is more often
viewed as an individual investment. The federal and state governments, higher education
institutions, and philanthropic organisations all provide direct student aid to support indi-
vidual decisions on both public and private higher education investment (Cohen and Kisker
2010). While the US higher education system employs a marketised model, higher education
in Taiwan is more centralised and considered as a national investment to educate students
with academic talents for strengthening the country’s workforce. Instead of offering direct
student aid, public funding has been largely appropriated to national universities (public
institutions) to lower their tuition fees. As national universities set a higher admission
standard, the Taiwan government indirectly subsidises a small group of academically able
students from taxpayers (Luoh 2004).
In addition to higher education finance, the US and Taiwan higher education systems are
also different in their size, arrangement, and culture. However, with these differences, there
has been similar growing stratification in higher education opportunities beyond college
access, among socioeconomic groups (see, for instance, Roksa et al. 2007; Tsai and Shavit
2007). Prior comparative analyses focus on comparing the effect of family background on
educational opportunity between countries but they seldom explore the mechanism behind
this effect (Schnabel et al. 2002). In order to add to the comparative higher education liter-
ature, this study seeks not only to better understand factors that influence student college
choice in the USA and Taiwan but to compare how these two vastly different systems con-
tribute to similar social and class stratification in college choice. Therefore, the following
two specific research questions are addressed in this study: (1) What are the roles of family
background, academic and financial factors in shaping students’ college-choice outcomes
in the USA and Taiwan? and (2) In what ways do the two different systems contribute to
social and class stratification in students’ college-choice outcomes?

Review of literature and conceptual model


Different disciplines, such as economics, sociology, and psychology, provide different lenses
to understand student college choice. Economists view education as an investment and
thus frame college choice as a rational cost-benefit analysis (e.g. Becker 1993; Coleman
and Fararo 1992). Under this framework, when making a college-choice decision, students
would maximise the returns by considering college costs and expected future earnings and
comparing higher education investment with other investments (Leslie and Brinkman 1987;
Paulsen 1998). Sociologists emphasise the effect of family background on student college
choice and the role of school contexts in reinforcing this effect (e.g. McDonough 1997;
Perna 2000). Improving social and cultural capitals would compensate for disadvantaged
background and enhance students’ opportunities and outcomes of college choice (Bourdieu
1973, 1986; Coleman 1988). Psychologists argue that student college choice is a sequence of
conflicts between alternatives, and each decision made is shaped by both internal intentions
and external environment (e.g. Deci and Ryan 2012; Hansen 1976). Therefore, information
COMPARE   3

search and student-institution fit play important roles in the college-choice process before
and after college enrollment (Hossler 1984; Jacoby 2000; Wiese 1994).
However, each of these three perspectives has its own limitations in explaining student
college choice. In order to have a comprehensive understanding of factors that influence
students’ decisions, this study used Cabrera and La Nasa’s (2000) college-choice model to
guide the analysis because their model integrates economic, sociological, and psychological
perspectives. This model not only illustrates three important stages of student college choice,
but also explains how each factor in these stages influences the others in a complex way
to eventually mould students’ final choices, following the temporal dimension they build
(see Cabrera and La Nasa 2000). Although their model is US based, it is also appropriate
to be used as the framework to guide the Taiwan analysis. Research in Taiwan has studied
college choices from the sociological perspective (e.g. Tsai and Shavit 2007) or the economic
perspective (e.g. Luoh 2002, 2004) and has examined the effects of student backgrounds, aca-
demic performance, locations, and school contexts. However, the Taiwan literature seldom
integrates different perspectives and has not yet empirically explored the role of financial
factors in studying student college choice. Therefore, it is vital to ground the Taiwan analysis
in the Cabrera and La Nasa model that provides a comprehensive framework.
In addition, student college choice has not yet been investigated in cross-cultural con-
texts. Quantitative, comparative analyses have been used in studying the stratification of
educational outcomes such as academic achievement at the K–12 level (see, for instance,
Chudgar and Luschei 2009; Holloway et al. 1990; Schnabel et al. 2002) or postsecondary
enrollment (see, for instance, Belley, Frenette, and Lochner 2014). However, prior compar-
ative studies did not focus on college choices and lacked a unified theoretical grounding
for analysis. Although they have consistently found social and class stratification in various
educational outcomes around the globe, still little is known about the mechanism embedded
in different educational systems that contributes to this universal phenomenon. This study
seeks to fill these holes in the literature and uses a simplified version of the Cabrera and
La Nasa (2000) model to understand the factors that perpetuate, or even reinforce, the effect
of family background on students’ college-choice outcomes between the USA and Taiwan.
This study simplifies the Cabrera and La Nasa (2000) model in the following ways, not
only because of data availability (which will be discussed later), but also for other more
important reasons. First, while they propose a three-stage process of college choice, this
study primarily focuses on the third phase – the choice stage – because this is the most
important stage that determines students’ college destinations and because of its high rel-
evance to policy. Given that the predisposition (first) and search (second) stages in the
Cabrera and La Nasa model are deeply influenced by student background and ability, they
are hard to change through policy intervention. Therefore, this study chooses a smaller
focus by highlighting the importance of the choice (final) stage, which is shaped not only
by student background but also by academic and financial factors. Second, this study mod-
ifies the Cabrera and La Nasa model such that the variable of interest is not student college
choice, but rather the outcome of college choice (namely the type of institution that students
actually attend) because college choice is a two-way, interactive process between students
and institutions, and using the type of institution attended as the dependent variable more
closely reflects both decisions made by students and by institutions. Student background
and factors that directly influence students’ decisions at the final stage of college choice (see
Cabrera and La Nasa 2000) are included in the conceptual model for this study. Constructs
included in the model and their corresponding measures are listed in Table 1.
4   P. YU

Table 1. Constructs, variable list, and coding.


Construct Variable Coding
Outcome of college choice Institutional control 1 = Private
0 = Public
Institutional selectivity 1 = High-selective
0 = Non-high-selective
Ascriptive characteristics Gender 1 = Female
0 = Male
Race 1 = Disadvantaged
0 = Advantaged
Socioeconomic backgrounds Annual family income Low: 1st–20th percentile (1, 0)
Middle: 21st–80th percentile (1, 0)
High: 81st–100th percentile (1, 0) (reference
group)
Father's education High school or less (1, 0)
Some post-secondary school (1, 0)
College (1, 0)
Graduate school (1, 0) (reference group)
Parental encouragement Parental encouragement 1 = Parental encouragement rated important in
college choice
0 = Parental encouragement rated not important
in college choice
Academic preparation High school grade point average C or below C (1, 0)
B (1, 0)
A (1, 0) (reference group)
Math ability Below average (1, 0)
Average (1, 0)
Above average (1, 0) (reference group)
Aspirations Educational aspiration 1 = Aspiring to obtain a graduate or professional
degree
0 = Not aspiring to obtain a graduate or profes-
sional degree
Occupational aspiration 1 = Aspiring to seek a specific career or occupa-
tion
0 = Not aspiring to seek a specific career or
occupation
Financial considerations Financial considerations 1 = Cost of attendance, financial aid, or living
costs rated important in college choice
0 = Cost of attendance, financial aid, or living
costs rated not important in college choice
Financial resources Family resources 1 = Using family resources to finance college
attendance
0 = Not using family resources to finance college
attendance
Student resources 1 = Using student resources to finance college
attendance
0 = Not using student resources to finance
college attendance
Grants 1 = Using grants to finance college attendance
0 = Not using grants to finance college attend-
ance
Loans 1 = Using loans to finance college attendance
0 = Not using loans to finance college attendance
Note: For variables that are listed with multiple categories (mutually exclusive), including annual family income, father’s
education, high school GPA, and math ability, a set of dummies are created for the categories embedded in each variable
when using the logit models for analysis. In other words, within a categorical variable, each category was coded as a
dichotomous variable, and one of the categories was selected as the reference group, in the analytical model. Because
students can use more than one type of financial resources to support college attendance, the four categories of financial
resources are not mutually exclusive. Therefore, unlike mutually exclusive multiple categories that were bounded under a
variable, these four financial resources categories were treated as four dichotomous variables separately.
COMPARE   5

Data and methods


Using a simplified version of the Cabrera and La Nasa (2000) model as the conceptual model,
this study analysed data from two sources: (1) the American Freshman2 Survey of 2005,
which was collected by the Higher Education Research Institute (HERI) at the University
of California at Los Angeles, and (2) the Taiwan Freshman Survey of 2005, which was
administered by the Taiwan Integrated Postsecondary Education Database (TIPED) at the
National Taiwan Normal University. These two surveys were used not only because their
measures properly describe the constructs of Cabrera and La Nasa’s (2000) model, but also
because they include the best comparative quantitative data that are currently available for
examining stratification in student college choice between the two countries. Both surveys
include nationally representative student-level data of the cohort who started college in
fall 2005.
Given that the longitudinal comparative data for the USA and Taiwan higher education
systems are currently unavailable, this comparative analysis is limited to the single year of
2005. This study does not intend to make causal claims due to the cross-sectional nature of
the data, which is a common limitation of many comparative studies. Comparative analyses
‘are limited in their ability to draw causal inferences, primarily because such studies are con-
strained to work with cross-sectional rather than longitudinal data’ (Chudgar and Luschei
2009, 649). The HERI and TIPED freshman surveys used in this study were administered
upon students’ arrival to their college destinations and did not follow these students from
the beginning of the college-choice process. However, by asking students to recall their
educational experiences before college enrollment, these surveys did provide most of the
important information needed to examine direct relationships of factors defined in the
three stages of Cabrera and La Nasa’s (2000) model to the final college-choice outcome.
Because the community college system and pathways to this system are fundamentally
different and incomparable between the USA and Taiwan, only students who enrolled in a
four-year institution or programme were included in this study, making the samples whose
college choice that the conceptual model intends to capture. Taking advantage of the large
sample sizes, after deleting individual cases for which data on any of the variables used in
this study were missing, the USA and Taiwan samples included 163,165 and 41,989 first-year
college students, respectively. In order to conduct a comparative analysis, this study recoded
variables for consistent measurement and meaningful simplification (as discussed below).
Although the data-set only includes student-level data and does not have contextual factors
that describe the differences between the USA and Taiwan systems, this study discusses
the results by situating them in each country’s unique context – which helps compare the
mechanisms behind social and class stratification between the two higher education systems
in a meaningful way.

Measuring college-choice outcomes


The variable of interest in this analysis was students’ college-choice outcomes – the type of
institutions in which students first enrolled after high school graduation. This study consid-
ered institutional control (public or private) and selectivity (high-selective or others) when
defining students’ college-choice outcomes, because these two dimensions show both the
quality and cost of institution attended. The HERI freshman survey already includes two
6   P. YU

variables that describe the selectivity and control of the first college a student attended.
However, the TIPED survey only has the institutional control variable. Because national
universities (public and academic-oriented) – which enjoy a longer history and better
resources – have been publicly recognised as the most selective institutional type in Taiwan,
this study used the intersection of institutional control and orientation to construct a var-
iable describing selectivity for the Taiwan analysis. In other words, control and orientation
were combined into four institutional types in Taiwan: (1) public, academic-oriented (i.e.
national universities); (2) private, academic-oriented; (3) public, technical-oriented; and
(4) private, technical-oriented.3 Although academic-oriented institutions have a higher
status than those that are technical-oriented, and within each type of orientation, public
institutions generally enjoy a higher quality of college education relative to private ones,
national universities are the only high-selective type among the four types of institutions in
Taiwan. Measuring institutional selectivity as a dichotomous variable (high-selective = 1,
others = 0), this study defined national universities as high-selective institutions, and the
other three institutional types combined were classified into the others category.

Students’ characteristics, academic and financial factors


The set of control variables were selected based on the conceptual model, as illustrated in
Table 1. Three main components of these controls included students’ demographic char-
acteristics, academic preparation, and financial circumstances. Students’ ascriptive char-
acteristics were measured in terms of gender and race. Given that racial groups in the USA
and Taiwan are not comparable, this study recoded racial groups into two: advantaged and
disadvantaged. For the USA, the racial advantaged group included Whites and Asians, while
the disadvantaged group had Blacks, Hispanics, Native Americans and other races. Asians
were combined with Whites into the advantaged group because of their higher levels of
college participation and success in general (Osajima 2005). Thus, it is reasonable to separate
Asians from other minority groups as this study is interested in the type of college attended.
For Taiwan, aborigines represented the only racial minority group who were disadvantaged
in terms of population size and economics, while the racial advantaged group included the
Hokkien, Hakka, and Mainlanders.
Students’ demographic characteristics also include their socioeconomic backgrounds,
which present a particularly important construct in this study because the core of interest is
to understand social and class stratification in students’ college-choice outcomes. A student’s
socioeconomic backgrounds were measured by annual family income (low, middle, and
high) and father’s education (high school or less, some postsecondary school, college, and
graduate school). Parental occupation, which is in the Cabrera and La Nasa (2000) model,
was not included in the analysis because of the difficulty in quantifying occupation and
incompatibility between the two countries. In order to compare between the two countries
with different income levels, this study measured students’ family wealth relative to their
counterparts’ in each country, rather than using absolute dollar amounts to define family
income. Measuring by the quintile thresholds of each country’s national income distribu-
tion, three categories of students’ family income were defined as follow: (1) low (1st–20th
percentile); (2) middle (21st–80th percentile); and (3) high (81st–100th percentile). Father’s
education, rather than mother’s education, was used because more data were available. For
COMPARE   7

observations that the data on father’s education were missing, this study inserted the values
of mother’s education for replacement.
This study is not only interested in how students from various family backgrounds enrol
in different types of institutions, but also in the roles that academic and financial fac-
tors play as opportunities or constraints in shaping different socioeconomic status groups’
college-choice outcomes. Academic factors include parental encouragement, academic
preparation, and aspirations. In this study, parental encouragement was measured by
whether a student rated parental encouragement as important for choosing to attend a
specific institution. Academic preparation was indicated by high school grade point average
(GPA) and math ability. High school GPA was on a four-point scale, which was coded in
the following way: A (4.0), B (3.0), C (2.0), D (1.0), and F (0.0). Because a relatively small
number of students whose high school GPAs were D or F in the samples, this variable was
collapsed into three categories (A, B, and C, or below C). This study defined math ability as
students’ self-rated ability (above-average, average, and below-average) compared to their
counterparts’. As this study included students who already attended four-year institutions, it
was reasonable to hypothesise that most of these students were with an aspiration to a bach-
elor’s degree. Hence, in order to better reflect differences in educational aspiration among
students in the samples, this study coded the educational aspiration measure as whether
or not a student aspired to an advanced degree after college graduation. Occupational
aspiration was measured by a student’s report of aspiring to a specific career or occupation
upon arriving at college. This measure distinguished between students who knew which
occupations they wanted to pursue and those who did not, at the time of college enrollment.
Two constructs – financial considerations and financial resources – were included in
the model to describe students’ financial circumstances. This study measured a student’s
financial considerations by using a dichotomous variable indicating whether or not col-
lege costs (cost of attendance, financial aid, or living costs) were considered as important
when making a college-choice decision. In terms of financial resources, four dichotomous
measures were employed to capture whether or not a student used family resources, student
resources, grants, or loans to finance college attendance, respectively.
This study acknowledges that, due to survey data limitations, the construct – saliency
of potential institutions – which has a direct relationship to students’ final college-choice
outcomes in the Cabrera and La Nasa (2000) model was not included in the analysis. With
more data collected on the order of perceived institutional attributes nested under each
student, future research is able to model the hierarchy of student college choice. Despite
some limitations, the data used for the analysis were guided by a comprehensive conceptual
model and all coded as dichotomous variables consistent across the two countries – which
allowed to directly compare the effects of controls on students’ college-choice outcomes
between the USA and Taiwan model. In other words, this study used the best quantitative
data that are currently available for comparing the mechanisms behind social and class strat-
ification in student college choice that are embedded in the two higher education systems.

Statistical model for analysis


This study used a nonlinear probability model called the logit model for analysing the
data. The logit model is used when the variables of interest are binary. In this study, the
outcome of student college choice was first defined by institutional control and second by
8   P. YU

institutional selectivity. Estimation of binary choice models is usually based on the method
of maximum likelihood, and each observation is treated as a single draw from a Bernoulli
distribution – binomial with one draw (Greene 2011). The logit model can be derived by
constructing a model in which the predicted probability, Pr(y = 1|𝐱), is forced to be within
the range 0 to 1 (Long and Freese 2006). In the linear probability model,
Pr(y = 1 | 𝐱) = 𝐱𝜷 + 𝜀
the predicted probabilities can be greater than 1 and less than 0. To constrain the predictions
to the range 0 to 1, the probability is first transformed into the odds,

Pr (y = 1 | 𝐱) Pr (y = 1 | 𝐱)
Ω(𝐱) = =
Pr (y = 0 | 𝐱) 1 − Pr (y = 1 | 𝐱)

for example, in this study, which indicate: (1) how likely a student attends a private institu-
tion (y = 1) relative to how likely that student attends a public institution (y = 0), or (2) the
relative probability of attending high-selective institutions compared to non high-selective
institutions. The odds range from ‘0 when Pr(y = 1|𝐱) = 0“ to “∞ when Pr(y = 1|𝐱) = 1
.’ The log of the odds, or logit, ranges from −∞to ∞. This range suggests a model that is
linear in the logit:
ln Ω(𝐱) = 𝐱𝜷
In order to compare the mechanisms behind the stratification of student college choice
between the USA and Taiwan, this study used the logit model to understand how students’
characteristics, academic, and financial factors affect the odds ratio of attendance at pri-
vate (versus public) institutions and high-selective (versus non high-selective) institutions
defined as the outcome (cost and quality) of student college choice.

Results
Table 2 presents the average characteristics of the USA and Taiwan samples, which were
comprised of first-year college students who attended four-year institutions in 2005 in
the USA and Taiwan, respectively. There were more females than males, and more stu-
dents from racial advantaged groups than those from a racial disadvantaged background
in both college-going samples. Women, who were historically underrepresented in higher
education, exceeded men in the number of college attendees; however, racial disparities in
college participation still existed, which were likely related to socioeconomic differences
among racial groups (Shavit et al. 2007). It is notable that the very poor were underrepre-
sented but the very rich were overrepresented in the USA college-going sample: only 7.6%
were from the bottom fifth of the national income distribution, whereas 34% were from
the top fifth.4 Nearly three-quarters (74.3%) of students in this sample had fathers who
had received at least some postsecondary education. Unlike the USA, the socioeconomic
composition indicates that more students in the Taiwan sample came from a lower socio-
economic background – 45.5% were from the lowest-fifth income group, and more than
two-thirds (67.6%) had fathers not receiving any postsecondary education. This may be
because higher education expansion in Taiwan started later, not until 1985, and thus college
Table 2. Descriptive means of the USA and Taiwan samples, by college-choice outcomes.
USA Sample Taiwan Sample
College-Choice Outcomes College-Choice Outcomes
Institutional Control Institutional Selectivity Institutional Control Institutional Selectivity
Non High High Non High High
 Variables All Public Private Selective Selective All Public Private Selective Selective
Ascriptive
Characteristics
Gender
Male 44.3% 45.5% 43.2% 42.0% 46.2% 45.6% 46.5% 45.2% 45.8% 45.1%
Female 55.7% 54.5% 56.8% 58.0% 53.8% 54.4% 53.6% 54.8% 54.2% 54.9%
Race
Disadvantaged 19.3% 20.5% 18.3% 24.0% 15.4% 2.0% 2.0% 1.9% 2.0% 2.0%
Advantaged 80.7% 79.5% 81.7% 76.0% 84.6% 98.0% 98.0% 98.1% 98.0% 98.0%
Socioeconomic
Backgrounds
Annual Family
Income
1st–20th 7.6% 8.8% 6.6% 9.8% 5.8% 45.5% 41.4% 47.7% 48.4% 36.4%
percentile (low)
21st–80th per- 58.4% 61.5% 55.8% 64.3% 53.5% 48.8% 52.7% 46.8% 46.4% 56.5%
centile (middle)
81st–100th per- 34.0% 29.8% 37.6% 25.9% 40.8% 5.7% 5.9% 5.5% 5.2% 7.2%
centile (high)
Father's Education
High school or 25.8% 29.4% 22.8% 33.7% 19.2% 67.6% 61.6% 70.8% 71.7% 54.7%
less
Some postsec- 3.9% 4.1% 3.7% 4.5% 3.4% 15.4% 17.1% 14.5% 14.5% 18.4%
ondary school
College 43.2% 44.5% 42.1% 44.0% 42.5% 13.0% 15.7% 11.5% 10.9% 19.3%
Graduate school 27.2% 22.1% 31.5% 17.9% 34.9% 4.1% 5.7% 3.2% 2.9% 7.5%
Parental
Encouragement

(Continued)
COMPARE 
 9
10 

Table 2. (Continued).
USA Sample Taiwan Sample
College-Choice Outcomes College-Choice Outcomes
 P. YU

Institutional Control Institutional Selectivity Institutional Control Institutional Selectivity


Non High High Non High High
 Variables All Public Private Selective Selective All Public Private Selective Selective
High School GPA
C or below C 3.2% 3.8% 2.7% 5.7% 1.1% 68.8% 57.1% 75.0% 71.4% 60.5%
B 39.5% 43.4% 36.3% 51.2% 29.7% 29.1% 40.1% 23.2% 26.5% 37.4%
A 57.3% 52.9% 61.1% 43.2% 69.2% 2.1% 2.8% 1.8% 2.1% 2.1%
Math Ability
Below average 16.9% 16.8% 17.0% 21.1% 13.3% 45.9% 37.8% 50.3% 48.7% 37.1%
Average 32.7% 33.1% 32.4% 35.8% 30.2% 35.6% 38.0% 34.3% 35.0% 37.5%
Above average 50.4% 50.1% 50.6% 43.1% 56.5% 18.5% 24.2% 15.4% 16.3% 25.4%
Aspirations
Educational Aspi-
ration
Not aspiring 23.6% 26.9% 20.9% 30.2% 18.1% 30.4% 20.9% 35.5% 34.9% 16.6%
to obtain a
graduate or
professional
degree
Aspiring to obtain 76.4% 73.1% 79.2% 69.8% 81.9% 69.6% 79.1% 64.5% 65.2% 83.5%
a graduate or
professional
degree
Occupational
Aspiration
Not aspiring to 13.5% 12.8% 14.1% 10.9% 15.6% 54.4% 55.0% 54.1% 54.7% 53.6%
seek a specific
career or occu-
pation
Aspiring to seek a 86.5% 87.3% 85.9% 89.1% 84.4% 45.6% 45.0% 45.9% 45.3% 46.4%
specific career
or occupation
Financial
Considerations
Cost of attend- 11.6% 9.8% 13.2% 7.9% 14.8% 29.6% 28.3% 30.3% 28.7% 32.3%
ance, financial
aid, or living
costs rated not
important in
college choice
Cost of attend- 88.4% 90.2% 86.8% 92.1% 85.3% 70.4% 71.8% 69.7% 71.3% 67.7%
ance, financial
aid, or living
costs rated
important in
college choice
Financial
Resources
Family Resources
Not using family 11.1% 13.2% 9.3% 13.8% 8.8% 21.0% 14.8% 24.3% 23.6% 12.8%
resources to
finance college
attendance
Using family 88.9% 86.8% 90.7% 86.2% 91.2% 79.0% 85.2% 75.7% 76.4% 87.2%
resources to
finance college
attendance
Student Resources
Not using student 24.6% 25.3% 24.0% 24.6% 24.6% 91.5% 91.9% 91.3% 90.9% 93.5%
resources to
finance college
attendance
Using student 75.4% 74.7% 76.0% 75.4% 75.4% 8.5% 8.1% 8.7% 9.1% 6.5%
resources to
finance college
attendance
(Continued)
COMPARE 
 11
12 

Table 2. (Continued).
USA Sample Taiwan Sample
College-Choice Outcomes College-Choice Outcomes
 P. YU

Institutional Control Institutional Selectivity Institutional Control Institutional Selectivity


Non High High Non High High
 Variables All Public Private Selective Selective All Public Private Selective Selective
Grants
Not using grants 19.3% 25.0% 14.5% 19.1% 19.4% 97.8% 97.7% 97.9% 97.8% 97.9%
to finance col-
lege attendance
Using grants to 80.7% 75.1% 85.5% 80.9% 80.6% 2.2% 2.3% 2.1% 2.2% 2.1%
finance college
attendance
Loans
Not using loans to 36.6% 43.2% 31.0% 34.4% 38.4% 78.6% 86.1% 74.6% 75.6% 87.7%
finance college
attendance
Using loans to 63.4% 56.8% 69.0% 65.6% 61.6% 21.4% 14.0% 25.5% 24.4% 12.3%
finance college
attendance
Total 100.0% 46.0% 54.0% 45.7% 54.3% 100.0% 34.9% 65.1% 75.8% 24.2%
Number of 1,63,165 75,008 88,157 74,571 88,594 41,989 14,642 27,347 31,814 10,175
Observations
COMPARE   13

access opportunity for parents of the 2005 college-going sample was limited – which would
result in a lower socioeconomic background of this sample.
In the USA sample, parental encouragement was rated not important in making a
college-choice decision by about 60% of students. More than half of students in this sample
were academically well-prepared: 57.3% earned A as an average in high school, and 50.4%
had above-average math ability. Educational and occupational aspirations were also prev-
alent in this sample. The majority of students aspired to obtain a graduate or professional
degree (76.4%) and to seek a specific career or occupation (86.5%). Most students (88.4%)
considered cost of attendance, financial aid, or living costs as important when choosing
among institutions. Students in this sample relied on a combination of family resources,
student resources, grants, and loans to pay for college – the majority of them were supported
by family resources and grants in particular. The Taiwan sample also shows high academic
aspiration. However, in contrast to the USA, it demonstrates strong parental encouragement
in students’ college choice, low academic preparation, and a lack of occupational aspiration
at the time of enrollment. Because data on standardised test scores were not available, this
study used high school GPA to measure students’ academic preparation. However, aca-
demic standards vary across high school in Taiwan. Those ranked top usually adopt rigour
standards and thus result in lower GPAs for their students who are academically able. In
addition, although most students (70.4%) in the Taiwan sample also reported a financial
concern when making a college-choice decision, the majority of students only relied on
family resources to pay for college. These descriptive statistics support Marginson’s (2011)
contention that higher education in East Asia, including Taiwan, has been immersed in the
Confucian model – which results in two primary phenomena: (1) parents are committed
to helping their children to achieve educational success, and they provide support through
a variety of ways, such as parental participation in students’ choice and contribution to
financing higher education; and (2) public funds are invested to promote academic research
and world-class universities but not to assist individual students with need. Table 2 also
describes the samples according to the control and selectivity of institution they attended.

Institutional control: attending private institutions


The empirical analysis begins by examining how demographic characteristics, academic,
and financial factors influence a student’s odds ratio of attending private (versus public)
institutions in the USA and Taiwan (see Models 1 and 2 in Table 3). Institutional control
is first tested because this study seeks to know which students are most likely to enrol in
privates, the more expensive college options in both countries.
In Model 1 of Table 3, socioeconomic backgrounds yield significant effects on the odds of
attending privates (versus publics) in the USA. Compared with students from high-income
families, low- and middle-income students have 40 and 36% lower odds of enrolling in a private
institution, respectively (p < 0.01). Having a lower level of father’s education also significantly
reduces a student’s odds of attending privates in the USA – the odds are 34–43% lower for stu-
dents whose fathers did not go to graduate school than for those with graduate school-educated
fathers (p < 0.01). These findings indicate a positive relationship between social class and private
college enrollment within the context of US higher education. In terms of other demographic
effects, women experience higher odds than do men of attending private institutions, while racial
disadvantaged students are less likely to do so compared to their more advantaged counterparts
14   P. YU

Table 3. Logit model, college-choice outcomes (attendance at private and high-selective institutions),
in the USA and Taiwan.
Private High selective
USA Taiwan USA Taiwan
Variables (1) (2) (3) (4)
Ascriptive characteristics
 Female 1.06*** 1.08*** 0.81*** 1.08***
(0.01) (0.02) (0.01) (0.03)
 Racial disadvantaged 0.97** 0.70*** 0.79*** 1.42***
(0.01) (0.05) (0.01) (0.12)
Socioeconomic backgrounds
Annual family income
  (Reference: 81st–100th percentile; high)
  1st–20th percentile (low) 0.60*** 0.89** 0.72*** 0.89**
(0.01) (0.04) (0.02) (0.05)
  21st–80th percentile (middle) 0.64*** 0.77*** 0.71*** 1.21***
(0.01) (0.04) (0.01) (0.06)
Father's education
  (Reference: Graduate school)
  High school or less 0.57*** 1.64*** 0.44*** 0.40***
(0.01) (0.09) (0.01) (0.02)
  Some postsecondary school 0.62*** 1.34*** 0.51*** 0.58***
(0.02) (0.08) (0.01) (0.03)
 College 0.66*** 1.23*** 0.61*** 0.73***
(0.01) (0.07) (0.01) (0.04)
Parental encouragement
Parental encouragement rated important 1.02* 1.03 0.97**
in college choice 1.02
(0.01) (0.03) (0.01) (0.03)
Academic preparation
High School GPA
  (Reference: A)
  C or below C 0.75*** 2.06*** 0.14*** 0.94
(0.02) (0.15) (0.01) (0.08)
 B 0.79*** 0.89 0.40*** 1.54***
(0.01) (0.06) (0.00) (0.13)
Math ability
  (Reference: above average)
  Below average 1.17*** 1.90*** 0.77*** 0.54***
(0.02) (0.06) (0.01) (0.02)
 Average 1.06*** 1.36*** 0.85*** 0.72***
(0.01) (0.04) (0.01) (0.02)
Aspirations
 Educational aspiration
 Aspiring to obtain a graduate or 1.24*** 0.53*** 1.60***
professional degree 2.31***
(0.02) (0.01) (0.02) (0.07)
Occupational aspiration
 Aspiring to seek a specific career or 0.89*** 1.15*** 0.63***
occupation 0.92***
(0.01) (0.03) (0.01) (0.02)
Financial considerations
  Cost of attendance, financial aid, or 0.57*** 0.82*** 0.56***
living costs rated important in college
choice 0.98
(0.01) (0.02) (0.01) (0.03)
Financial resources
  Family resources 1.32*** 1.04 1.29*** 0.98
(0.02) (0.05) (0.02) (0.06)
  Student resources 0.93*** 1.11*** 1.09*** 0.74***
(0.01) (0.05) (0.01) (0.04)
 Grants 2.33*** 1.04 1.15*** 0.98
(0.03) (0.08) (0.02) (0.08)
(Continued)
COMPARE   15

Table 3. (Continued).
Private High selective
USA Taiwan USA Taiwan
Variables (1) (2) (3) (4)
 Loans 1.97*** 2.04*** 1.15*** 0.51***
(0.02) (0.10) (0.01) (0.03)

Log likelihood –106305.71 –25174.39 −100238.47 −21436.98


χ² Statistic 12522.52 3955.14 24511.10 3628.48
Prob. > χ² 0.00 0.00 0.00 0.00
Observations 163,165 41,989 163,165 41,989
Note: Odds ratios are reported. Standard errors are in parentheses. This Table uses logistic regression techniques to compare
full college-choice models, which examine attendance at private (versus public) institutions and high-selective (versus
non-high-selective) institutions, between the USA and Taiwan.
***p < 0.01; **p < 0.05; *p < 0.1.

(p < 0.05 or smaller). Model 1 in Table 3 further reveals the relationship between academic
factors and private college enrollment in the USA. Students with parental encouragement in
the college-choice process, the highest GPAs in high school, and advanced degree aspiration
have greater odds of attending privates (p < 0.1 or smaller). However, having above average
math ability and aspiration to seek a specific career or occupation significantly reduces these
odds (p < 0.01). Considering the impact of financial factors, this model shows that students
who consider cost of attendance, financial aid, or living costs as important when making a
college-choice decision are less likely to attend privates in the USA (p < 0.01). Among four
financial resources types, family resources, grants, and loans have positive effects, while
student resources yield a negative effect, on private college enrollment (p < 0.01). Specifically,
students who receive grant aid have over twice (2.33) the odds of attending privates, relative
to students without this type of financial support.
In Taiwan, students from lower-income backgrounds are also less likely than their
high-income counterparts to attend private colleges and universities (p < 0.05 or smaller).
However, it is especially notable that, among four-year college attendees in Taiwan, there
appears to be the reverse of the USA in the stratification in private college enrollment
based on father’s education level. As shown in Model 2 of Table 3, compared to students
with graduate school-educated fathers, those whose fathers received formal education at
college, some postsecondary school, or high school or less do not have lower odds, but
instead have 23, 34, or 64% higher odds of attending private institutions in Taiwan, respec-
tively (p < 0.01). With respect to gender and racial effects, being a female increases the
odds of enrolling in a private institution, whereas being a racially disadvantaged student
decreases these odds (p < 0.01). It appears that students with poorer academic preparation
are more likely to enrol in private institutions in Taiwan. Specifically, students whose high
school GPAs were ‘C or below C’ and students whose math ability was below average,
have about twice (2.06 and 1.9, respectively) the odds of attending private institutions
than do ‘A’ average students, and those with above average math ability (p < 0.01). Even
those with average math ability have 36% higher odds of attending privates compared to
those with math ability above average (p < 0.01). Aspiring to a graduate or professional
degree upon arriving at college reduces the odds of going to the private sector; however,
a specific occupational aspiration is positively associated with private college enrollment
(p < 0.01). Model 2 in Table 3 also shows that students who regard cost of attendance,
16   P. YU

financial aid, or living costs as important in choosing among colleges have lower odds
of attending privates (p < 0.01). There appears to be a positive relationship of using
student resources or taking on loans to private college enrollment (p < 0.01). In other
words, if attending privates in Taiwan, students tend to depend on their own or loan aid,
especially the latter, to cover a higher tuition fee. Specifically, students who take on loans
have double (2.04) the odds of enrolling in privates.

Institutional selectivity: attending high-selective institutions


This section presents results for the second measure of college-choice outcomes –
attending high-selective institutions in the USA and Taiwan. Models 3 and 4 of Table 3
examine the effect of social class on how likely a student enrols in a high-selective insti-
tution, taking into account gender and race, as well as academic and financial factors.
As seen in Model 3 of Table 3, in the US, students from less advantaged socioeconomic
backgrounds have lower odds of attending a high-selective institution. Compared with
high-income students, the odds for low- and middle-income students are decreased by
28 and 29%, respectively (p < 0.01). Father’s education even has a greater magnitude of
effects. Having father’s education at high school or less, some postsecondary school,
and college (rather than at the graduate school level) reduces these odds by 56, 49, and
39%, respectively (p < 0.01). The results of other demographic factors demonstrate
that being a female or coming from a racial disadvantaged group hinders students
from attending a high-selective institution in the USA (p < 0.01). In addition, this
full college-choice model (Model 3 of Table 3) shows significant and positive effects
of better academic preparation and higher educational aspiration. High school GPAs
lower than ‘A’ and math ability at/below average result in a 60–86% and 15–23% lower
odd ratio of enrolling in high-selective institutions (p < 0.01). Having advanced degree
aspiration increases this odd ratio by 60% (p < 0.01). However, early occupational aspi-
ration (aspiring to a specific career upon arriving at college) yields a negative effect on
high-selective college attendance (p < 0.01). It is surprising to find a negative effect of
parental encouragement, but the magnitude of this effect is quite small (p < 0.05). With
respect to financial measures in this model, considering cost of attendance, financial
aid, or living costs when making a college-choice decision reduces a student’s odds of
attending a US high-selective institution by 44% (p < 0.01). All the four types of finan-
cial resources yield positive effects, and family resources have the largest magnitude of
effect among the four types – being financially supported by parents has a 29% higher
odd ratio of attending high-selective institutions in the USA (p < 0.01).
The results in Model 4 of Table 3 also show that social-class background plays a signifi-
cant role in the propensity of students to attend high-selective institutions, namely national
universities (relative to other institutional types), in the Taiwan context. Students from
low-income families have 11% lower odds of attending national universities than do students
from high-income backgrounds (p < 0.05). Surprisingly, middle-income students, however,
have 21% greater odds than do their high-income counterparts (p < 0.01). Compared with
family income, father’s education yields a more consistent and a larger magnitude of effect.
Students whose fathers did not receive education at the graduate school level are less likely
to enrol in a national university than students with graduate school-educated fathers – their
odds are 27–60% lower (p < 0.01). In addition, there appear to be associations of gender
COMPARE   17

and race with high-selective college attendance in Taiwan – women and racial disad-
vantaged students experience higher odds of going to national universities than do
men and those from a more advantaged racial group (p < 0.01). A 42% greater odds for
racial disadvantaged groups reflects that the affirmative action admissions policy may
work in Taiwan. This model (Model 4 in Table 3) further reveals that better academic
preparation (measured by math ability) and educational aspiration have significant and
positive effects on student enrollment in high-selective institutions in Taiwan (p < 0.01).
The odds that a student with below-average math ability will attend a national univer-
sity are only 54% of the odds for students with above-average math ability (p < 0.01).
In other words, a student with above-average math ability has nearly twice the odds
of entering a high-selective institution in Taiwan as an otherwise comparable student
whose math ability is below average. Students who aspire to an advanced degree have
over twice (2.31) the odds of going to national universities, relative to students without
this aspiration (p < 0.01). It is important to note, however, that high school GPA shows
a different direction of effect – students with a ‘B’ average are more likely to attend
national universities than students who earn an ‘A’ average (p < 0.01). This surprising
finding may be due to early stratification in the high school stage. In Taiwan, top-ranked
high schools, which would more likely enrol academically-able students, implement
rigorous curriculum and assessment for better preparing them for college. Academic
rigour might lower these students’ high school GPAs, but, with rigorous training and
the original preparation, their propensity to attend a high-selective institution would
be higher. Occupational aspiration yields a negative effect on high-selective college
attendance (p < 0.01). Student resources and loans are the only two financial measures
that significantly predict whether a student attends a national university in Taiwan.
Using either student resources or loans to pay college tuition fees lowers students’ odds
of attending national universities by 26 and 49%, respectively (p < 0.01).

Comparisons and policy implications


Comparing students’ college-choice outcomes between the USA and Taiwan provides a win-
dow into different mechanisms that shape social and class stratification in higher education.
The interaction of higher education financing philosophy with institutional opportunities
available to disadvantaged students is involved in these mechanisms. By testing the control
(publics versus privates) and selectivity (high-selective versus others) of institution attended,
this study finds stratification not only in the quality of college education that students receive
but also in college affordability.
In this study, college-choice outcomes consist of two dimensions: institutional selectivity
and control. When considering institutional selectivity, this study finds that coming from
lower socioeconomic backgrounds decreases a student’s odds of attending a high-selective
institution in both the USA and Taiwan, although the greater propensity of middle-income
students (compared to their high-income counterparts) to attend national universities in
Taiwan represents one exception. However, similar stratification in the selectivity of institu-
tion attended across the two contexts has different implications. For the USA, this means that
the opportunity to obtain a higher quality college education and postsecondary experience
is unequal towards low-income students and those with lower levels of father’s education.
However, for Taiwan, the meaning extends beyond this layer. National universities (which are
18   P. YU

all publics) represent the only high-selective institutional type in Taiwan and often use govern-
ment funds to provide students with an indirect subsidy by charging them lower tuition fees.
Therefore, a lower propensity of socioeconomically disadvantaged students to access national
universities not only indicates their lower likelihood of attending a high-selective institution but
their higher probability of paying more for college.
The other dimension of college-choice outcomes is institutional control. In the analysis of
whether students attend privates (in comparison to publics), this study finds different scenarios
between the USA and Taiwan. In the USA, students from less advantaged socioeconomic back-
grounds (relative to high-income students or students with graduate school-educated fathers)
experience lower odds of attending privates. While lower-income students in Taiwan are less
likely to attend a private institution as well, their counterparts with lower levels of father’s edu-
cation are more likely to do so. Private institutions in the USA often use their endowments and
private donations to provide subsidies to students and constitute a higher proportion (63.3%)
of high-selective institutions in the USA sample than do public institutions. Therefore, a lower
propensity of socioeconomically disadvantaged students to attend privates in the USA reflects
social and class stratification in the affordability to invest in a college education that would gen-
erate higher returns during college and/or after graduation. However, in Taiwan, none of private
institutions are classified as high selective, but they charge a higher price because of limited
endowments and other donative wealth. Thus, a greater propensity of lower-socioeconomic sta-
tus students (measured by father’s education) to enrol in privates again demonstrates a regressive
higher education system that increases stratification – students from families that have lower
levels of father’s education likely pay more but obtain a lower quality college education. These
findings support McCowan’s (2016) argument that inequalities have increased in the contexts
of higher education expansion around the globe – which are mainly evidenced by increasing
horisontal stratification of institutional type. He finds that socioeconomically disadvantaged
students disproportionally enrol in lower quality and/or higher price institutions (e.g. vocational
universities in England and private institutions in Brazil). Competitive admissions exams and
college tuition fees are barriers that further disadvantage those from disadvantaged backgrounds.
Although the scenarios of social and class stratification in students’ college-choice out-
comes are different between the USA and Taiwan, this study finds that better academic
preparation and advanced degree aspiration are factors that promote high-selective college
attendance across the two contexts. From a policy perspective, enhancing academic prepa-
ration and aspiration for socioeconomically disadvantaged students, and avoiding early
stratification in the high school stage, can be a substantive and viable avenue for reducing
stratification in the selectivity of college attended in the USA and Taiwan. Particularly
in Taiwan, promoting high-selective college attendance, namely enrollment in national
universities, for socioeconomically disadvantaged students also helps ease their financial
burden of paying college tuition and fees.
It is important to recognise that financial resources available to students also matter for
college-choice outcomes, but the influence of finances is different between the USA and
Taiwan. In the USA, all the four types of financial resources, especially those from parents
or families, are positively related to attendance at high-selective institutions. While family
resources, grants, and loans also benefit private college enrollment, using students’ own
resources to pay for college hinders them from attending privates in the USA These findings
reveal that the two dimensions of college-choice outcomes may be intertwined, and that
college becomes unaffordable for most in the USA. Especially when attending privates,
COMPARE   19

grants and loans play a remarkably important role in financing college enrollment. Although
providing direct financial aid and subsidies to increase student enrollment and choice is the
financing philosophy in US higher education, recent trends of the privatisation of public
higher education and unequal distribution of student financial aid (Cohen and Kisker 2010;
Heller 2006) all work to disadvantage already-disadvantaged students who seek a better
opportunity in today’s US higher education marketplace.
By contrast, college is much more affordable in Taiwan. However, the philosophy of
financing higher education in Taiwan contributes to a regressive system that increases strat-
ification and causes problems of college affordability. In Taiwan, large government funds are
appropriated to national universities to keep tuition fees low, rather than directed to individ-
ual students with need. Given that this study shows a lower propensity of socioeconomically
disadvantaged students to attend national universities in Taiwan, public funding (from
taxpayer allocations) is thus used to indirectly subsidise more advantaged students through
government appropriations to national universities. Analysing data across 31 countries in
South Asia and sub-Saharan Africa, Ilie and Rose (2017) find that the rich benefit more
from public spending on higher education because they are more likely to access college.
Similarly, this study finds that students from a more advantaged background – who tend
to succeed in college admissions and enrol in national universities – are more likely than
their disadvantaged counterparts to benefit from public spending. This study also finds that
the relationship of grants to private college enrollment is not significant, whereas student
resources and loans yield positive effects. Without grant-aid as a support, socioeconomically
disadvantaged students, who are more likely to attend privates in Taiwan, heavily rely on
their own financial resources (most from working part-time) and loans to pay for college.
There appear to be different philosophies of financing higher education between the USA
and Taiwan; however, they both contribute to increasing stratification. As higher education
is widely viewed as a major means for social mobility, it is important for the two nations with
the intent to improve equality to craft financial policies for increasing lower-SES students’
opportunity to attend high-selective institutions and for reducing their financial barriers.
In addition to continued calls for increasing student grant aid, it needs to carefully examine
the distribution of grants in the USA. For Taiwan, government funds need to be directly
distributed to individual students in the form of need-based financial aid rather than dis-
proportionately appropriated to national universities to lower their tuition.
In the global trend of cost sharing, tuition increases become more and more promi-
nent and merit the attention of scholars and policymakers who seek to assist disadvan-
taged students. Particularly in the USA, the distribution of public funding has increasingly
favoured students from an advantaged background since the late-1970s, which has been evi-
denced by an expansion of the federal Pell Grant programme to include support of middle-
income students (Savoca 1990), a change in federal student aid from grants to loans
(St. John and Noell 1989), and a shift in state and institutional aid from need- to merit-based
grant programmes (see, for instance, Doyle 2010; Heller 2002; Ness and Tucker 2008). As
this study shows a positive relationship of family resources to both attendance at high-
selective institutions and at privates in the USA, developing programmes that subsidise
middle-income families in long-term, systematically setting aside the resources to support
their children’s college education is a promising approach (for parental savings, see Burke
and Sheffield 2006; Dynarski 2004). Encouraging middle-income families’ contributions to
and investments in their children’s college education would not only increase their children’s
20   P. YU

educational aspiration, academic preparation, and college involvement, but also reduce
their competition with lower-income students for financial aid and allow more equitable
distribution of public funds.
However, in contemporary higher education, loans inevitably play a more and
more important role in financing postsecondary education. From a policy perspective,
income-contingent loans, which collect the debt (or loan repayments) based on students’
future levels of income, can be considered to use for avoiding potential future stratification
(for more discussions, see Chapman 2006; Johnstone 1972; Nerlove 1975). Other tuition
policies such as deferred tuition fees (Johnstone and Marcucci 2010) and Pay It Forward
(PIF) models (Delaney and Dharmapala 2017) – which shift college payment from the front
to the back end, and thus eaken the influence of family background – are found in the USA
literature to increase low-income students’ college access and the equilibrium level of higher
education subsidies. Whether these models are appropriate to use in Taiwan deserves further
investigation. Beyond the discussions of who – the government(s), taxpayers, parents, or
students – most generally should pay for college, future research ought to focus more on
‘how they pay’, in other words, resource distribution and financing patterns among different
stakeholders, and between different contexts.

Conclusions and future research directions


The US and Taiwan higher education systems have expanded and created many choices;
however, the two systems are increasingly stratified and the mechanisms behind are differ-
ent. Thus, more studies on social and class stratification beyond college access are needed. In
this comparative study between the USA and Taiwan, stratification is found not only in the
quality of higher education institutions that students attend, but also in college affordability
and finance. Family backgrounds, academic features, and financial factors are all important
in shaping students’ college-choice outcomes. Future research is needed to know more about
stratification in college choices – such as choice of major, single or dual degree programmes,
and continuously or interruptedly enrollment, and so on – in contemporary higher educa-
tion marketplaces. Enhancing academic preparation for socioeconomically disadvantaged
students would be an active way of reducing stratification, and providing them with adequate
financial support would at least ease their financial barriers or burdens. As the distribution
of public funding seems to be unequal towards socioeconomically disadvantaged groups in
both countries, more attention should be paid to the role that finances and related policies
play in reducing or perpetuating social and class stratification. When more cross-national,
longitudinal data become available, including contextual factors of different higher educa-
tion systems in the analysis would allow future research to better explore and compare the
mechanism behind social and class stratification in college choices among various nations,
and also have important implications for policy within each unique context.

Notes
1. 
The net college enrollment rate of 18–21-year-olds surpassed 50% by 2004 in Taiwan and by
2012 in the USA (National Centre for Education Statistics 2014; Taiwan Ministry of Education
2008). Both countries have achieved the universal level of college enrollment according to
Trow’s (1972, 1976, 2000) classification of higher education growth.
2. 
Here, the word, ‘freshman,’ is referred to as first-year college student.
COMPARE   21

3. 
These institutions, either academic- or technical-oriented, are four-year institutions or include
four-year baccalaureate programmes. This study only includes first-year college students
whose first postsecondary institutions or programmes lead to a bachelor’s degree.
4. 
By using the 20th and 80th percentile thresholds from the national income distribution to
create a three-category income structure for the analysis, there should be 20% in the lowest
income category, 60% in the middle category, and 20% at the top, for each income group to
be proportionally represented in higher education.

Acknowledgements
The author would like to thank Jennifer A. Delaney for her valuable comments and suggestions to
improve this work.

Disclosure statement
No potential conflict of interest was reported by the author.

References
Becker, G. S. 1993. “Investment in Human Capital: Rates of Return.” In Human Capital: A Theoretical
and Empirical Analysis, with Special Reference to Education, edited by G. S. Becker, 59–160. Chicago,
IL: University of Chicago Press.
Belley, P., M. Frenette, and L. Lochner. 2014. “Post-Secondary Attendance by Parental Income in
the U.S. and Canada: Do Financial Aid Policies Explain the Differences?” Canadian Journal of
Economics 47 (2): 664–696.
Bourdieu, P. 1973. “Cultural Reproduction and Social Reproduction.” In Knowledge, Education, and
Cultural Change, edited by R. Brown, 71–112. London: Tavistock Publications.
Bourdieu, P. 1986. “The Forms of Capital.” In Handbook of Theory and Research for the Sociology of
Education, edited by J. G. Richardson, 241–258. New York: Greenwood Press.
Brennan, J., and R. Naidoo. 2008. “Higher Education and the Achievement (and/or Prevention) of
Equity and Social Justice.” Higher Education 56 (3): 287–302.
Brewer, D. J., E. R. Eide, and R. G. Ehrenberg. 1999. “Does It Pay to Attend an Elite Private College?
Cross-Cohort Evidence on the Effects of College Type on Earnings.” Journal of Human Resources
34 (1): 104–123.
Burke, L. M., and R. Sheffield. 2006. “Continuing the School Choice March: Policies to Promote
Family K-12 Education Investment.” Backgrounder 2683: 1–5.
Cabrera, A. F., K. R. Burkum, and S. M. La Nasa. 2005. “Pathways to a Four-Year Degree: Determinants
of Transfer and Degree Completion.” In College Student Retention: Formula for Student Success,
edited by A. Seidman, 155–214. Westport, CT: American Council on Education and Praeger
Publishers.
Cabrera, A. F., and S. M. La Nasa. 2000. “Understanding the College-Choice Process.” In Understanding
the College Choice of Disadvantaged Students: New Directions for Institutional Research, edited by
A. F. Cabrera and S. M. La Nasa (No. 107), 5-22. San Francisco, CA: Jossey-Bass.
Chapman, B. 2006. “Income Contingent Loans for Higher Education: International Reforms.” In
Handbook of the Economics of Education. vol. 2, edited by E. A. Hanushek and F. Welch, 1437–1503.
Philadelphia, PA: Elsevier.
Chudgar, A., and T. F. Luschei. 2009. “National Income, Income Inequality, and the Importance of
Schools: A Hierarchical Cross-National Comparison.” American Educational Research Journal 46
(3): 626–658.
Cohen, A. M., and C. B. Kisker. 2010. The Shaping of American Higher Education: Emergence and
Growth of the Contemporary System. 2nd ed. San Francisco, CA: Jossey-Bass.
22   P. YU

Coleman, J. S. 1988. “Social Capital in the Creation of Human Capital.” The American Journal of
Sociology 94: S95–S120.
Coleman, J. S., and T. J. Fararo. 1992. Rational Choice Theory: Advocacy and Critique. London: Sage
Publications.
Deci, E. L., and R. M. Ryan. 2012. “Self-Determination Theory.” In The Handbook of Theories of Social
Psychology. vol. 1, edited by P. A. M. Van Lange, A. W. Kruglanski, and E. T. Higgins, 416–437.
London: Sage Publications.
Delaney, J. A., and D. Dharmapala. 2017. “‘Pay It Forward’ and Higher Education Subsidies: A Median
Voter Model.” Contemporary Economic Policy 35 (4): 615–629.
Doyle, W. R. 2010. “Changes in Institutional Aid, 1992–2003: The Evolving Role of Merit Aid.”
Research in Higher Education 51 (8): 789–810.
Dynarski, S. 2004. “Who Benefits from the Education Saving Incentives? Income, Educational
Expectations and the Value of the 529 and Coverdell.” National Tax Journal 57 (2, Part 2): 359–383.
Greene, W. H. 2011. Econometric Analysis. 7th ed. Upper Saddle River, NJ: Pearson Education.
Hansen, F. 1976. “Psychological Theories of Consumer Choice.” Journal of Consumer Research 3 (3):
117–142.
Heller, D. E. 2002. “The Policy Shift in State Financial Aid Programs.” In Higher Education: Handbook
of Theory and Research. vol. XVII, edited by J. C. Smart, 221–261. New York: Agathon.
Heller, D. E. 2006, March. “Merit Aid and College Access.” Paper presented at the symposium on
the consequences of merit-based student aid for the Wisconsin Center for the Advancement of
Postsecondary Education, Madison, WI.
Holloway, S. D., B. Fuller, R. D. Hess, H. Azuma, K. Kashiwagi, and K. Gorman. 1990. “The Family's
Influence on Achievement in Japan and the United States.” Comparative Education Review 34 (2):
196–208.
Hossler, D. 1984. Enrollment Management: An Integrated Approach (Report No. ISBN-0-87447-1931).
New York: College Board. (ERIC Document Reproduction Service No. ED249906).
Ilie, S., and P. Rose. 2017. “Who Benefits from Public Spending on Higher Education in South Asia
and Sub-Saharan Africa?” Compare: A Journal of Comparative and International Education.
doi:10.1080/03057925.2017.1347870.
Jacoby, J. 2000. “Is It Rational to Assume Consumer Rationality? Some Consumer Psychological
Perspectives on Rational Choice Theory.” Roger Williams University Law Review 6 (1): 83–161.
Johnstone, D. B. 1972. New Patterns for College Lending: Income Contingent Loans. New York:
Columbia University Press.
Johnstone, D. B., and P. N. Marcucci. 2010. Financing Higher Education Worldwide: Who Pays? Who
Should Pay? Baltimore, MD: Johns Hopkins University Press.
Leslie, L. L., and P. T. Brinkman. 1987. “Student Price Response in Higher Education: The Student
Demand Studies.” Journal of Higher Education 58 (2): 181–204.
Long, J. S., and J. Freese. 2006. Regression Models for Categorical Dependent Variables Using Stata.
2nd ed. College Station, TX: StataCorp.
Lucas, S. R. 2001. “Effectively Maintained Inequality: Education Transitions, Track Mobility, and
Social Background Effects.” American Journal of Sociology 106 (6): 1642–1690.
Luoh, M. C. 2002. “誰是台大學生?–性別、省籍與城鄉差異 [Who are NTU Students? – Differences
across Ethnic and Gender Groups and Urban/Rural Discrepancy].” 經濟論文叢刊 [Taiwan
Economic Review] 30(1): 113–147.
Luoh, M. C. 2004. “升學機會與家庭背景 [Educational Opportunities and Family Background in
Taiwan].” 經濟論文叢刊 [Taiwan Economic Review] 32(4): 417–445.
Marginson, S. 2011. “Higher Education in East Asia and Singapore: Rise of the Confucian Model.”
Higher Education 61 (5): 587–611.
McCowan, T. 2016. “Three Dimensions of Equity of Access to Higher Education.” Compare: A Journal
of Comparative and International Education 46 (4): 645–665.
McDonough, P. M. 1997. Choosing Colleges: How Social Class and Schools Structure Opportunity.
Albany, NY: State University of New York Press.
Melguizo, T. 2008. “Quality Matters: Assessing the Impact of Attending More Selective Institutions
on College Completion Rates of Minorities.” Research in Higher Education 49 (3): 214–236.
COMPARE   23

National Center for Education Statistics. 2014. Digest of Education Statistics, 2013. Washington, DC:
U.S. Department of Education.
Nerlove, M. 1975. “Some Problems in the Use of Income-Contingent Loans for the Finance of Higher
Education.” Journal of Political Economy 83 (1): 157–183.
Ness, E. C., and R. Tucker. 2008. “Eligibility Effects on College Access: Under-Represented Student
Perceptions of Tennessee’s Merit Aid Program.” Research in Higher Education 49 (7): 569–588.
Osajima, K. 2005. “Asian Americans as the Model Minority: An Analysis of the Popular Press Image
in the 1960s and 1980s.” In A Companion to Asian American Studies, edited by K. A. Ono, 215–225.
Malden, MA: Blackwell Publishing.
Paulsen, M. B. 1998. “Recent Research on the Economics of Attending College: Returns on Investment
and Responsiveness to Price.” Research in Higher Education 39 (4): 471–489.
Perna, L. W. 2000. “Differences in the Decision to Attend College among African Americans,
Hispanics, and Whites.” Journal of Higher Education 71 (2): 117–141.
Roksa, J., E. Grodsky, R. Arum, and A. Gamoran. 2007. “United States: Changes in Higher
Education and Social Stratification.” In Stratification in Higher Education: A Comparative
Study, edited by Y. Shavit, R. Arum, A. Gamoran, and G. Menahem, 165–191. Stanford, CA:
Stanford University Press.
Savoca, E. 1990. “Another Look at the Demand for Higher Education: Measuring the Price Sensitivity
of the Decision to Apply to College.” Economics of Education Review 9 (2): 123–134.
Schnabel, K. U., C. Alfeld, J. S. Eccles, O. Köller, and J. Baumert. 2002. “Parental Influence on Students'
Educational Choices in the United States and Germany: Different Ramifications – Same Effect?”
Journal of Vocational Behavior 60 (2): 178–198.
Shavit, Y., R. Arum, A. Gamoran, and G. Menahem, eds. 2007. Stratification in Higher Education: A
Comparative Study. Stanford, CA: Stanford University Press.
St. John, E. P., and J. Noell. 1989. “The Effects of Student Financial Aid on Access to Higher Education:
An Analysis of Progress with Special Consideration of Minority Enrollment.” Research in Higher
Education 30 (6): 563–581.
Strayer, W. 2002. “The Returns to School Quality: College Choice and Earnings.” Journal of Labor
Economics 20 (3): 475–503.
Taiwan Ministry of Education. 2008. Education Statistics in Taiwan, 2008 (Government Publication
No. 2004600003). Taipei, Taiwan: Author.
Trow, M. 1972. “The Expansion and Transformation of Higher Education.” International Review of
Education 18 (1): 61–84.
Trow, M. 1976. “Elite Higher Education: An Endangered Species?” Minerva 14 (3): 355–376.
Trow, M. 2000. “From Mass Higher Education to Universal Access: The American Advantage.”
Minerva 37 (4): 303–328.
Tsai, S. L., and Y. Shavit. 2007. “Taiwan: Higher Education – Expansion and Equality of
Educational Opportunity.” In Stratification in Higher Education: A Comparative Study,
edited by Y. Shavit, R. Arum, A. Gamoran and G. Menahem, 140–164. Stanford, CA:
Stanford University Press.
Wiese, M. D. 1994. “College Choice Cognitive Dissonance: Managing Student/Institution Fit.” Journal
of Marketing for Higher Education 5 (1): 35–48.
World Bank. 2004. Beyond Economic Growth: An Introduction to Sustainable Development. 2nd ed.
Washington, DC: Author.

You might also like