choosing transfer institutions: examining the decisions of ......2015; velez and javalgi 1987;...
TRANSCRIPT
FPO
Research Supported by:
The opinions expressed inthis report are those of the
author(s) and do not necessarily reflect the views of
Greater Texas Foundation.
Choosing Transfer Institutions: Examining the Decisions ofTexas Community College Students Transferring to
Four-Year Institutions
Huriya Jabbar and Wesley Edwards
The University of Texas at Austin
2
1. Introduction
Community colleges in the U.S. have multiple institutional goals, including workforce education and
academic programs, and serve a broad range of non-traditional students (Grubb 1991). Yet a key, ongoing
aim of community colleges is to facilitate transfer to four-year institutions. Community colleges play an
important “democratizing” role, providing open access to postsecondary education for historically
disadvantaged students (Gonzalez and Hilmer 2006; Leigh and Gill 2003). Today, one-third of all current
college enrollees in the U.S. attend community colleges (Snyder, de Brey, and Dillow 2016). However,
community colleges have a complex role in fostering student completion and success (Bowen, Chingos,
and McPherson 2009). Some researchers argue that they also “divert” students from higher education
(Brint and Karabel 1989; Long and Kurlaender 2009), in part due to their high attrition rates and their
complex structures, which make it difficult for students to navigate course, degree, and transfer
requirements (Scott-Clayton 2015). National reports have found that while 80% of students entering
community college intend to earn a bachelor’s degree, only 25% of students actually transfer to a four-
year university within 5 years, and only 17% earn a bachelor’s degree within 6 years of transferring
(Jenkins and Fink 2015). However, we know little about the reasons for these low rates of transfer and
completion.
One key factor may relate to students’ choice of transfer destination. A number of studies over
the past two decades have examined transfers from two- to four-year institutions, focusing on either the
factors that predict student transfer to a four-year college (Crisp and Nuñez 2014; Deil-Amen and
Rosenbaum 2003; Gandara, Alvarado, Driscoll, and Orfield 2012; Lee and Frank 1990; Scott-Clayton
2015; Velez and Javalgi 1987; Wassmer, Moore, and Shulock 2004), or the effects of attending
community college on four-year college completion rates and outcomes (Belfield and Bailey 2011; Castro
and Cortez 2017; Dougherty 1992; Doyle 2009; Gonzalez and Hilmer 2006; Grubb 1991; Hilmer 1997;
Lee and Frank 1990; Leigh and Gill 2003; Lockwood Reynolds 2012; Long and Kurlaender 2009;
Melguizo and Dowd 2009; Surette 2001). However, despite the large number of studies examining high
3
school students’ initial choices of postsecondary institutions (Beattie 2002; Black, Cortes, and Lincove
2015; Grodsky and Jones 2007; Long 2007; Manski 1993; Radford 2013; Tierney 1983; Turley 2009),
and students’ decisions about whether to attend community colleges in the first place (Bers and Galowich
2002; Somers et al. 2006), there has been almost no research examining how community college students
choose among four-year institutions (for a recent exception, see Backes and Velez 2015, which examined
community college students’ choices of transfer institutions in Florida).
Given that community college students represent a different population than average high school
students—less academically prepared, more likely to come from historically disadvantaged groups, and
more financially and geographically constrained (Backes and Velez 2015)—it is important to understand
the factors that matter in prospective transfer students’ educational decisions, and their choice of four-
year institution. Understanding the characteristics of schools selected by students (e.g., geographical
location, financial support, institutional quality) may help to explain the mechanisms by which
community college students do—or do not—transfer to four-year institutions, as well as later outcomes,
and may inform programs and policies that help low-income, first-generation students in particular to
successfully apply and transfer to high-quality four-year institutions.
This study draws on theories and models of college choice to examine how transfer-intending
community college students decide between four-year institutions. To do so, we draw on a statewide data
set that captures the destinations of all Texas students who began at a community college and made a
transfer to a four-year university between 2011 and 2015. We observe the schools that community college
students ultimately transfer to, and the institutional and geographic factors associated with those choices.
We use conditional logit regressions (Backes and Velez 2015; Long 2004; McFadden 1973) to model
choice behavior and examine what institutional features (e.g., distance, cost, selectivity) predict students’
decisions by observing the institutions they ultimately selected in comparison to their other options.
4
We find that an overwhelming majority of community college transfer students select public and
mostly research universities, despite the fact that nearly half of the institutional options were private
schools. We also confirm prior work showing students’ likelihood of choosing a school increases for
transfer options that are geographically closer to the community college attended. Further, we note
interesting preferential differences between specific subgroups of students (e.g., economically
disadvantaged vs. non-economically disadvantaged, traditional vs non-traditional) related to measures of
institutional support and quality present amongst the options in student choice sets. Our analysis comes
from recent data in a demographically diverse state with a steadily growing population. Therefore, our
findings have implications for polices that lessen existing barriers to successful transfer and for
institutions seeking to increase student enrollments through increased community college transfer rates.
2. Background
Researchers have proposed three stages to the higher education choice process (Hossler and Gallagher
1987). In the first stage, the “predisposition stage,” students decide whether they will attend college, or, in
our case, whether they intend to transfer (Hossler and Gallagher 1987). A key part of the decision-making
process is the formation of a choice set, which typically occurs next in the process. The choice set
includes the particular set of choices considered by an individual when making a decision. The
alternatives considered are important to understand because examining how a student’s ultimate choice
compares to the potential set of options they considered can provide valuable insight into their
preferences and decisions. In this stage of the process, students search for schools, obtaining information
about institutions of higher education and developing criteria for judging schools before they actually
decide on a college or university to attend, which constitutes the third stage, and the termination of a
sequential choice process (Bell 2009; Castleman, Schwartz, and Baum 2015; Hossler and Gallagher 1987;
Tierney 1983).
5
Empirical examinations of college choice among high school students have explored both the
decision to enroll in college as well as the choice of which institution among many options. These studies
reveal several important factors driving students’ decisions, including cost, school quality, geography, and
student demographics. Studies have also found that college choices vary by student background, race,
gender, income, and wealth (Kinsler and Pavan 2011; Long 2007; Lovenheim and Reynolds 2013; Pallais
and Turner 2006). One important factor is attending a university that is close to the student’s home
(Hillman and Weichman 2016; Long 2004; Tierney 1983; Tinto 1985). This is especially important for
low-income students and students of color (Turley 2009), groups that are over-represented among
community college students. In Texas, for example, Latino students are most sensitive to distance,
although across all race and ethnicity groups, students are less sensitive to distance as income level
increases (Black, Cortes, and Lincove 2015). And some studies have found that, in general, students are
becoming increasingly sensitive to distance over the past several years as researchers have been
monitoring this trend (Skinner 2016), while others find that distance is becoming relatively less important
than factors such as quality and cost (Long 2004). These studies all focus on high school students
selecting colleges, but community college students, who also tend to be older and perhaps more
constrained due to work or family responsibilities, may be even more sensitive to distance (Fishman
2015; Jepsen and Montgomery 2009).
The cost of higher education, including tuition, fees, room and board, and financial aid, is an
important factor in students’ decisions about which institution to attend (Avery and Hoxby 2004; Hurwitz
2012; Fishman 2015; Long 2004). Long (2004) in a national study of college decisions among over 2,000
universities, found that price played an important role in the deciding between colleges. Similarly,
Fishman (2015), in a national survey of students who planned to enroll in higher education within the
next 12 months, found that, for all but the wealthiest students, cost outweighed other factors, including
location. This information, however, may not always be transparent. For example, there can be
misunderstandings amongst students about the cost of higher education, particularly in the case of
6
university “sticker prices,” which may overestimate the costs for low-income students in particular.
Indeed, research suggests that low-income students of color, and their families, often overestimate the
costs of attending college, compared to middle-class or White parents (Grodsky and Jones 2007), and
low-income students may be more sensitive to small price changes (Mundel 2008). However, one recent
study found that the importance of cost in college choice is declining over time (Skinner 2016).
Community college transfer students have already attended a low-tuition college for at least a part of their
higher education. They may be more sensitive to costs in general, or they might be more willing to pay a
higher tuition for fewer years of study, given that they may have saved on tuition while at the community
college.
The quality or ranking of the institution may also factor into students’ decisions. This, too, is
shaped by students’ backgrounds, as students vary, based on their social resources, in the extent to which
they are aware of or actively use rankings in their decisions (Ball et al. 2002). Understanding how
students’ decisions are influenced by institutional quality is important, particularly due to the growing
body of research on academic “undermatching” in higher education, i.e., when high school students apply
to schools that are less selective than the schools they could have been admitted to, based on their
academic records (Bastedo and Jaquette 2011; Hoxby and Avery 2012; Page and Scott-Clayton 2016;
Smith, Pender, and Howell 2013). For example, high-achieving, low-income students tend to have choice
sets that are more similar to their socioeconomic peers than their academic peers (Page and Scott-Clayton
2016). And some recent research suggests that attending a higher-quality four-year institution, one that is
better funded, with higher skilled peers, regardless of the student’s own performance, is beneficial for all
students, (Goodman, Hurwitz, and Smith 2015), since schools with better conditions and supports may
help improve persistence and completion (Castleman et al. 2015; Page and Scott-Clayton 2016). Among
high school students, institutional quality appears to be increasingly important over time (Long 2004;
Skinner 2016). Therefore, it is important to understand how students choose among universities of
varying quality.
7
School demographics may also play a role in students’ decisions about higher education
institutions. One study in Texas found that students of color preferred colleges with greater proportions of
same-race students, or those where students from their high schools have been successful in the past
(Black et al. 2015). Black students are also more likely to select Historically Black Colleges and
Universities, if those options are available (Ehrenberg, Rothstein, and Olsen 1999). Students of color may
perceive that such schools will be better equipped to support them or believe that these institutions will
foster a greater sense of belonging.
While there is a robust body of research examining the choices of high school students and the
importance of various factors (e.g., geography, tuition, selectivity, family) in high school students’
decisions, there is almost no research on the decision-making processes of community college transfer
students. Indeed, researchers have argued that existing models for choice are “less effective in predicting
nontraditional or delayed-entry students’ search and choice processes than they are of traditional-aged
students” (Hurtado et. al 1997, 45). Without understanding how students select among these four-year
alternatives, our understanding of choice behavior is incomplete (Tierney 1983), particularly since so
many students begin their higher education experiences in community college institutions. The choice of
higher education institution, namely its institutional quality, has important implications for transfer
student outcomes, including post-college earnings (Hilmer 2002). In particular, while some research has
found an earnings penalty for students who start at a community college prior to transferring to a four-
year university (Gill and Leigh 2003), this disparity shrinks when considering the quality, or selectivity,
of the four-year institution attended (Hilmer 2002). Therefore, it is important to understand students’
choices of transfer destinations in order to inform the “diversion” or “democratization” debate regarding
the role of community colleges.
Studies have examined the decision to attend a community college (Grubb 1988; Hilmer 1998;
Lovenheim and Reynolds 2013; Rouse 1994), but we found only one study that examined college choice
among transfer institutions using an administrative data set. Backes and Velez (2015) examined the
8
choices of community college transfer students in Florida by observing transfers to public universities in
the state. They used a statewide administrative data set including nearly 80,000 high school graduates
who entered postsecondary education through a community college. Employing a conditional logit model
(similar to Long 2004), they found that transfer students were sensitive to distance, meaning they were
less likely to choose a four-year institution located further away, relative to recent high school graduates.
They included 10 public four-year institutions in Florida to represent students’ choice set, along with
institution-specific characteristics as measures that varied based on unique student-institution match.
Backes and Velez were able to disaggregate their findings by student subgroups. They note that students
in their sample eligible for free- or reduced-price lunch were the most sensitive to larger distances and
students with stronger academic backgrounds were the least sensitive to larger distances between the
community college and four-year transfer destination. Finally, they found that instructional expenses to be
the most salient predictor of transfer amongst a set of various institutional characteristics.
In an attempt to expand on what Backes and Velez found in Florida, in this study we examine
transfer choice in a new state context, Texas. Specifically, we examine the institutional characteristics that
are associated with Texas community college transfer students’ decisions about where to enroll. We draw
on five years of data from more recent years and a broader set of institutions (including all private
universities in the state), and we include more characteristics of the transfer destinations. Specifically, we
include a measure of the percentage of students attending the institution that are of the same race as the
chooser, based on recent research which suggests this is an important factor, particularly for students of
color (Black et al. 2015), as well as a measure of tuition or cost that is more tailored to the student’s
economic background. We are also able to capture both traditional and non-traditional college students
who transfer during the period observed. Understanding transfer students’ choice behaviors and pathways
may help to guide and target future policies and programs focused on improving community college
student transfer and completion rates.
Texas Transfer Context
9
In Texas, 75% of all students completing bachelor’s degrees were at some point enrolled in a community
college (National Student Clearinghouse Research Center 2017). Just over 20% of community college
students successfully transfer to a university within six years, however 56% of the students who do
transfer graduate within six years of starting their postsecondary career (THECB 2018). The transfer rate
in Texas is disparate for traditionally underrepresented groups. African American and Hispanic students
have the lowest transfer rates amongst all other racial/ethnic groups in the state. While relatively few
students are able to transfer from a community college to a four-year school, of those who do, most go on
to earn a bachelor’s degree.
The low transfer rate in Texas is likely due to a decentralized system of transfer policies where
there are few incentives for two-year and four-year institutions to collaborate towards improving transfer
pathways (Schudde, Bradley, and Absher 2018). The responsibility to create supportive transfer pathways
falls to individual institution-to-institution agreements. Currently, community colleges and four-year
schools across the state create bilateral articulation agreements where students can follow a particular
course pathway and avoid taking unnecessary credit hours prior to transferring. However, according to
survey research by the THECB, not all Texas public universities have active faculty participation in the
creation of articulation agreements, and there is large variation in the quality of available information
about these agreements (THECB 2014).
As previously noted, the most recent analysis of student transfer decision-making was conducted
in Florida where in comparison to Texas there exists a much more supportive system of community
college to four-year transfer policies (Backes and Velez 2015). For example, a state policy in Florida
created a transfer student Bill of Rights, providing students who graduate from a community college with
an Associate’s degree admission into one of the 11 state universities. Further, as Backes and Velez (2015)
note, Florida’s “well developed articulation system” gives community college students in the state an
advantage compared with community college students in other states (7). Thus, our study provides a look
10
at transfer patterns in a less centralized state context, potentially providing for an opportunity to compare
student outcomes and inform future state-level transfer policy decision making.
3. Data and Method
In this study we ask two main research questions. First, for students who start at a community college,
what institutional characteristics or features (e.g., distance, selectivity) predict where community college
students ultimately transfer? Secondly, how does the importance of these characteristics vary across
different types of students?
To answer these questions, we draw on administrative data from Texas, which includes
longitudinal student-level data from the Texas Higher Education Coordinating Board and the Texas
Education Agency. All files are housed in the Education Research Center (ERC) at the University of
Texas at Austin. The ERC provides access to longitudinal data from multiple sources, allowing
researchers to follow individual Texas students through their K-12, postsecondary, and workforce
trajectories. The Texas data includes student demographic measures (e.g., gender, year of birth, ethnicity,
financial aid status, socioeconomic status), and institutions of enrollment. We created a measure of
“vertical transfer” when a student’s primary institution (defined as the institution at which they attempted
the most credits, in the case of co-enrolled students) changed from one long semester (i.e., Fall or Spring)
to the next, from a two-year to a four-year institution.
For our main set of independent variables, or predictors, we used publicly available data from the
Integrated Postsecondary Data System (IPEDS). This allowed us to capture various institutional
characteristics including: instructional expenses per full time equivalent student (FTE), student to faculty
ratio, number of full-time, first-time degree-seeking students receiving financial aid, six-year graduation
rate, institutional control (e.g., public vs. private), status as a flagship university, net price cost of
attendance, full tuition sticker-price, and a measure representing the 75th percentile SAT math scores for
first-time students. We matched this information from IPEDS to the year prior to a student’s first year
11
attending their choice school since these characteristics would be observed when a student was deciding
where to transfer.
Other predictor variables were calculated to provide more precise estimates of factors influencing
student choice. For example, a unique estimate of student cost was created by matching a student’s
socioeconomic status to a specific breakdown of average net price that each institution provides by family
income quintile. We were also able to capture whether a student qualified for a tuition exemption due to
prior military service, indicating that for any public institution they would have the complete cost of
tuition covered. This factor went into our unique calculation of student cost. We also created a measure of
the distance from the community college a student first attended to each four-year institution in their
choice set. To do this, we utilized latitude and longitude coordinates for each institution from the IPEDS
data, which allowed us to calculate a measure of total distance using the ‘geodist’ command in Stata. This
command calculates the length of the shortest curve between two points along the surface of a
mathematical model of the earth and reports the figure in kilometers (Picard 2012).
To examine the schools that community college students ultimately transfer to, and the
institutional and geographic factors associated with those choices, we restricted the sample to students
who made a vertical transfer (i.e., from a community college to a four-year university) between the years
2011 and 2016. Therefore, we focus not on the decision to transfer, but on the choice of transfer
institution. The sample consisted of 94,710 students who made a vertical transfer during the time period
observed.
Our data set is not without limitations. First, we can only capture a single state, and we are unable
to observe transfers out of state. Furthermore, we were unable to match a sufficient number of students to
their high-school records; thus we were unable to identify distance of the institution from the student’s
high school (as a proxy for their hometown), or use student test scores/performance to split the sample or
identify institutional “match” based on academics.
12
4. Analytic Approach
We used conditional logit regressions to predict students’ transfer destinations and generate policy-
oriented behavioral models (McFadden 1973). This is an approach used in previous literature on student
college choice to model choice behavior and examine what institutional features (e.g., distance, cost,
selectivity) predict students’ decisions by observing the institutions they ultimately selected in
comparison to their other options (Backes and Velez 2015; Long 2004). In particular, we examined the
role of various explanatory factors, drawing on previous literature and our qualitative analyses (including
interviews with over 150 community college students in Texas, see Jabbar, Epstein, Sanchez, and
Edwards, forthcoming), such as distance and institutional characteristics in predicting college choice.
As with prior college-choice studies, we assume that students maximize utility in their selection
of four-year institutions, with individual i choosing between j four-year institutions, characterized by
various features of each institution (Backes and Velez 2015; Long 2004). The utility associated with
attending a particular four-year institution for an individual is a function of institutional characteristics
(e.g., price, quality), as well as characteristics that vary by the individual (e.g., distance, percentage that
are the same race). The individual selects a four-year university alternative (the choice of transfer
institution) if the utility of that option is greater than the alternatives, subject, of course, to the student’s
budget constraint. In our case, the alternatives students may select from include all public and private
four-year institutions in the state of Texas.
Our dataset thus included pairwise combinations of each student i with each school j in the state,
with observations stratified by individual into groups of j. Each individual appears 77 times in the data,1
by each potential match of four-year institution in that given year. We excluded any institution that did
not offer a bachelor’s degree, as well as specialty schools, such as medical schools, health professions
schools, schools of business or management, and theological seminaries. Our rationale for excluding these
institutions included the fact that most were classified as majority graduate or professional degree
13
granting institutions and therefore would not be relevant as choices for our sample. Additionally, some
institutions that were excluded did not admit first-time full-time degree-seeking students.2 Each
observation also contains different values for the four-year institution-specific variables, such as
institutional type, selectivity, and quality measures, as well as variables that vary by individual, such as
distance from community college attended to the four-year institution, percentage of students who are the
same race/ethnicity, or estimated tuition cost. Finally, there is an indicator of the transfer destination or
choice, equal to one if the alternative is chosen, zero otherwise.
We then split the sample to compare students based on race, socioeconomic status, and age to see
if there were differences in the choices of these groups, by comparing the coefficients of the regressions.3
5. Results
A total of 94,710 students were included in our final analytic sample. Table 1 includes descriptive
statistics of these students. Approximately 43% of our sample identify as white, over one third (36%)
identify as Hispanic/Latino, 9% as Black/African American, and 7% as Asian. Close to 46% of students
were economically disadvantaged. On average, the students in our sample were around 22 years old at the
time of transfer, and student choice sets included approximately 77 schools.
14
Table 1
Descriptive Statistics of Students
Mean Standard Deviation
Female 0.547 0.498
Black 0.094 0.292
White 0.428 0.495
Hispanic 0.356 0.479
Asian 0.069 0.253
Pacific Islander 0.001 0.033
Native American 0.002 0.043
International 0.018 0.133
Transfer age 21.859 5.733
Economically disadvantaged 0.457 0.498
Academic disadvantage 0.314 0.464
Disability 0.021 0.144
Limited English Proficiency 0.036 0.185
Choice-set size 77.279 0.960
N 94710
Note. - A small proportion of students in the sample had no data for the Hispanic
race/ethnicity variable category (376 students). Academic disadvantage is defined by the
THECB as any student based on a Texas Success Initiative (TSI) approved test that does not
have college entry-level skills in reading, writing, or math.
Table 2 provides institutional characteristics of only the universities to which students transferred.
These summary statistics show that transfer destinations are in most cases public institutions, representing
nearly 98% of all transfers observed. In addition, close to 74% of the transfers were to research
institutions. Students transferred to institutions with an average acceptance rate of close to 67% of
applicants and six-year graduation rates slightly greater than 50%. Across all chosen institutions,
expenses per each full-time-equivalent student were around $8,300 and student-to-faculty ratios were
15
approximately 21:1. The annual mean estimated cost of attendance (“Tuition cost”) for chosen institutions
was $11,881, with a maximum observed cost of attendance across all observations at $40,724. In terms of
financial aid, an average of more than 80% of students within the group of chosen institutions receive
some form of financial aid. Finally, our descriptive results indicate that greater than three-fourths of all
transfers were to institutions located in urban contexts, with 13% choosing schools in rural areas, and just
under 10% selecting a school in a suburban area.
Table 2
Descriptive Statistics of Chosen Institutions
Mean Standard Deviation Min Max
Distance 119.714 154.611 0 1190.488
Public 0.979 0.142 0 1
Flagship 0.147 0.354 0 1
HBCU 0.015 0.121 0 1
Research 0.739 0.439 0 1
Percent admitted 66.865 12.818 15 100
Expenses 8333.385 3147.148 2683 40056
Enrollment 22825.740 14847.510 135 55471
Grad rate 52.272 15.355 0 100
Student to faculty ratio 21.039 2.577 6 27
Tuition cost 11881.710 5634.072 0 40724
Sticker-price tuition rate 7678.331 3112.767 1392 41750
Financial aid 82.164 7.948 51 100
SAT 75th percentile score 596.433 57.559 405 800
Urbanicity
Urban 76.86
Suburban 9.72
Rural 13.42
N 94,710
Note. – Each variable represents institutional characteristics the year prior to each student’s transfer. Distance is
the total kilometers between a community college and four-year institution. Expenses are average institutional
16
expenditures per full time equivalent undergraduate student. Enrollment is a full-time enrollment count for
undergraduates. Tuition cost is a measure of net price which varies depending on student background
characteristics and institutional cost. Financial aid is the percent of first-time full-time undergraduates receiving
any form of financial aid. SAT 75th percentile score is the SAT match 75th percentile scores for first-time full-time
freshman at each institution.
Figure 1 shows the most frequently chosen transfer destinations. In this figure we display
frequency of transfer for the top 20 most-chosen institutions. As previously noted, most of the transfers
we observed were to large, public, research institutions. The set of institutions in Figure 1 represents only
one-fourth of the 78 schools we observed transfers to across all years. Notably, the number of transfers to
these 20 institutions represents greater than 86% of total transfers observed across all years. This could
suggest that while there is a diverse set of four-year institutions in Texas to choose from, students are
actually transferring to only a subset of these options. This descriptive finding could also point to existing
relationships between four-year institutions with high transfer rates and community colleges which ease
the process of transfer.
17
Figure 1. Total recorded transfers between 2011 and 2015 to the top twenty most-chosen four-year
institutions.
Table 3 shows the characteristics of the institutions in students’ choice sets (all eligible transfer
destinations, as well as separate descriptive statistics for public and private universities). On average,
students’ transfer options were 370 kilometers (230 miles) away. However, this figure is the mean for
their entire choice set for which there is considerable variation within the measure of distance. Just over
half of the institutions in students’ theoretical choice sets were public and only one-third were research
institutions. This descriptive finding is interesting considering the proportion of students who actually
transferred to a public or research university (98 and 74% respectively). The average six-year graduation
rate for all institutions in student choice sets was just over 44%, nearly ten percentage points lower than
the same figure for actual transfer institutions. This could indicate that given a wide selection of schools,
students in our data are prioritizing schools with more success in terms of student completion rates. In
18
Table 3 we also compare average characteristics for public vs. private institutions within the choice sets.
On average, public institutions within student choice sets had student enrollments more than four times
larger than those of private institutions. Our comparison also indicates that average annual tuition cost for
private institutions included was more than three times greater than the same measure for public
institutions ($21,870 and $6,222 respectively). The large difference in average tuition cost between
private and public institutions could potentially explain what we observe as a strong preference towards
transferring to a public institution amongst students in our sample. Our descriptive results show public
institutions enroll a greater proportion of Latino students (39 vs. 23%) and a lower proportion of African
American students (14 vs. 21%) compared with private schools.
Table 3
Descriptive Statistics of
Student Choice Sets
All Institutions (n=78)
Public Institutions
(n=41)
Private Non-Profit Institutions
(n=37)
Mean SD Mean SD Mean SD
Distance 369.689 122.616 406.3483 145.1565
Public 0.526 0.503
Flagship 0.026 0.159 0.0487805 0.2180848
HBCU 0.103 0.305 0.0487805 0.2180848
Research 0.333 0.474 0.4878049 0.5060608
Percent admitted 68.230 19.701 71.71657 17.02833
Expenses 8444.831 4614.272 7338.053 2357.634 9671.261 6035.157
Enrollment 7294.470 10026.790 11525.390 12103.400 2606.157 3047.888
Grad rate 44.145 19.057 41.73124 16.51509
Student to faculty ratio 16.681 4.289 19.70845 2.783671
Tuition cost 13644.460 9572.314 6221.833 1709.096 21869.530 7775.636
Financial aid 89.402 10.485 84.27192 10.3768
SAT 75th percentile 562.131 68.375 556.467 53.219 568.408 82.315
% enrollment Black or African American 17.425 22.926 13.983 17.904 21.239 27.196
% enrollment Hispanic 31.424 23.582 39.114 26.874 22.903 15.654
% enrollment White 39.376 22.651 36.152 21.181 42.948 23.957
19
Urbanicity
Urban 70.51
Suburban 6.41
Rural 23.08
N 94,710 92,766 1,944
Note. - Each variable represents institutional characteristics the year prior to each student’s transfer. Distance is the total kilometers between a
community college and four-year institution. Expenses are average institutional expenditures per full time equivalent undergraduate student.
Enrollment is a full-time enrollment count for undergraduates. Tuition cost is a measure of net price which varies depending on student
background characteristics and institutional cost. Financial aid is the percent of first-time full-time undergraduates receiving any form of
financial aid. SAT 75th percentile score is the SAT match 75th percentile scores for first-time full-time freshmen at each institution.
Our conditional logistic regression results are shown in Table 4. We find noteworthy patterns in
terms of student choice and transfer when running our full predictive model.4 Next, we briefly discuss
trends emerging from our predictive model including all students, then report results for specific
subgroups of students when comparing key variables in the paragraphs that follow.
On the whole, our results confirm prior findings reporting sensitivity to distance (Backes and
Velez, 2015). The coefficient predicting the influence of distance from community college to four-year
institution indicates that an increase in distance lowers the predicted probability of transfer. Texas covers
a large geographic area and this result could indicate that attending community colleges near popular
transfer institutions increases the likelihood of eventual transfer.
20
Table 4
Conditional Logit Predicting Transfer Choice
All Students Economically
Disadvantaged
Non-
Economically
Disadvantaged
Under-
represented
Minority
Non Under-
represented
Minority
Non-
traditional Traditional
B SE B SE B SE B SE B SE B SE B SE
Distance -0.010*** 0.000 -0.010*** 0.000 -0.009*** 0.000 -0.009*** 0.000 -0.010*** 0.000 -0.013*** 0.000 -0.009*** 0.000
Expenses -0.000*** 0.000 -0.000 0.000 -0.000*** 0.000 0.000 0.000 -0.000*** 0.000 -0.000*** 0.000 -0.000*** 0.000
Student to
faculty ratio 0.003 0.002 -0.021*** 0.004 0.028*** 0.003 -0.003 0.003 0.011** 0.003 -0.046*** 0.006 0.015 0.003
Financial aid -0.002** 0.001 0.000 0.001 -0.008*** 0.001 -0.001 0.001 -0.007*** 0.001 -0.004* 0.001 0.000 0.001
Grad rate 0.017*** 0.000 0.014*** 0.001 0.020*** 0.001 0.015*** 0.001 0.016*** 0.001 0.014*** 0.001 0.017*** 0.000
Enrollment 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000
Public 3.328*** 0.032 4.454*** 0.062 2.439*** 0.042 3.409*** 0.048 3.062*** 0.045 5.056*** 0.116 3.118*** 0.034
Research 0.076*** 0.012 0.081*** 0.018 0.072*** 0.017 0.139*** 0.017 0.044** 0.018 0.163*** 0.034 0.077*** 0.013
Flagship -2.465*** 0.034 -3.179*** 0.054 -2.041*** 0.044 -2.649*** 0.050 -2.211*** 0.047 -2.485*** 0.112 -2.511*** 0.035
Percent
admitted -0.001* 0.000 -0.007*** 0.001 0.005*** 0.000 -0.006*** 0.000 0.001* 0.001 -0.002** 0.001 -0.001*** 0.000
Racial match 0.029*** 0.000 0.029*** 0.000 0.027*** 0.000 0.019*** 0.000 0.042*** 0.001 0.024*** 0.001 0.030*** 0.000
Tuition cost 0.000*** 0.000 0.000*** 0.000 0.027*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** 0.000
21
SAT 75th
percentile -0.003*** 0.000 -0.003*** 0.000 -0.003*** 0.000 -0.006*** 0.000 -0.001** 0.000 -0.001** 0.000 -0.004*** 0.000
HBCU -1.648*** 0.035 -1.646*** 0.047 -1.897*** 0.064 -1.446*** 0.043 -1.503*** 0.093 -1.540*** 0.091 -1.684*** 0.039
Urbanicity
Suburban -0.411*** 0.017 -0.410*** 0.027 -0.408*** 0.022 -0.347*** 0.024 -0.583*** 0.024 -0.555*** 0.049 -0.369*** 0.018
Rural 0.220*** 0.013 0.227*** 0.018 0.277*** 0.018 0.223*** 0.018 0.043* 0.019 0.186*** 0.034 0.241*** 0.014
Pseudo R-
squared 0.471
0.496
0.456
0.469
0.478
0.508
0.472
Number of
Observations 7303738a
3334354b
3969384c
3565403d
3738335e
886891f
6416847g
Number of
Students 84052
Note. *p < .05. **p < .01. ***p < .001. The columns display results for our model estimation first for all students then reported by specific subgroups.
Creation of the economically disadvantaged subgroup follows the THECB definition as any student with income below the annual federal poverty line,
eligible for family aid or assistance programs, or eligible for a Pell Grant or comparable state program of need-based financial aid. The subgroup
underrepresented minority is defined as students who identify as Black/African American, Pacific Islander, Native American, or Hispanic. Traditional students
are defined as students who were 24 years old or younger at the time of transfer whereas nontraditional students were defined as any student 25 years or older
at the time of transfer. The variable distance is the total kilometers between a community college and four-year institution. For the variable Urbanicity the
reference group is a transfer to an urban institution. The number of observations in each column represents all unique student-institutional combinations.
a 821,687 observations dropped because of all positive or all negative outcomes.
b 438,573 observations dropped because of all positive or all negative outcomes.
c 383,114 observations dropped because of all positive or all negative outcomes.
d 465,382 observations dropped because of all positive or all negative outcomes.
e 356,305 observations dropped because of all positive or all negative outcomes.
f 179,620 observations dropped because of all positive or all negative outcomes.
g 642,067 observations dropped because of all positive or all negative outcomes.
22
In line with our descriptive findings, our regression results demonstrate that students prefer public
research institutions. Further, these results indicate that students prefer schools with higher six-year
graduation rates and a greater proportion of the student body that matches their own racial/ethnic identity.
As an indication of aversion to more selective institutions, we see a small but negative coefficient for
SAT math 75th percentile score (-.003), meaning that as this measure increases the probability of transfer
decreases. Finally, in terms of geographic locale, our results suggest an increased probability of transfer to
a rural school (.22) and lower likelihood of transferring to a suburban school (-.41) when urban is the
reference category.
Geographic Variables
In all of our models the variable distance is a measure of kilometers between a community college
attended and the institutions in each students’ choice set. Our results that measure students’ sensitivity to
distance when transferring from a community college to four-year institution largely confirm prior
findings from Backes and Velez (2015). Across all subgroups of students, the farther away a potential
transfer destination is from the community college attended, the less likely a student is to transfer there.
Specifically, for every one-unit increase in distance (one km) we expect a -.01 decrease in the log-odds of
an institution being chosen, holding all other independent variables constant. This result was consistent
for each of our subgroup model estimates, and non-traditional students appear to be the most sensitive to
distance (-0.013), although the difference in coefficient size is marginal. This finding has important
implications for transfer patterns in a state as large as Texas. Students at community colleges that are far
from the top transfer destinations in the state could be at a disadvantage when it comes to transferring due
to what is likely a high cost of commuting or relocating.
When compared to institutions located in urban contexts, students across all subgroups express a
preference for four-year schools in rural and urban vs suburban settings. Students in our full sample
model had a lower likelihood of transferring to a suburban school compared to a school located in an
23
urban context (-0.41) and a greater likelihood of transferring to a rural school compared to an urban
institution (0.22). Coefficients representing the log-odds for each of or subgroup analyses, or likelihood of
transfer dependent on locale type, were similar to those for the full sample.
Figure 2.- Coefficients for the continuous predictor variables representing log odds of transfer by student
economic status subgroups. Key: Triangles=Economically disadvantaged students, Circles=Non-economically
disadvantaged students. Note.- We used the institutional coding of student economic disadvantage which follows the
THECB definition as any student with income below the annual federal poverty line, eligible for family aid or
assistance programs, or eligible for a Pell Grant or comparable state program of need-based financial aid.
24
Figure 3.- Coefficients for the categorical predictor variables representing log odds of transfer by student economic
status subgroups. Key: Triangles=Economically disadvantaged students, Circles=Non-economically disadvantaged
students. Note.- We used the institutional coding of student economic disadvantage which follows the THECB
definition as any student with income below the annual federal poverty line, eligible for family aid or assistance
programs, or eligible for a Pell Grant or comparable state program of need-based financial aid.
Figure 4.- Coefficients for the continuous predictor variables representing log odds of transfer by students’
underrepresented status. Key: Triangles=Underrepresented students, Circles=Non-underrepresented students. Note.-
Student observations were coded as underrepresented if their racial/ethnic identity was Black/African American,
Pacific Islander, Native American, or Hispanic.
Figure 5.- Coefficients for the categorical predictor variables representing log odds of transfer by students’
underrepresented status. Key: Triangles=Underrepresented students, Circles=Non-underrepresented students. Note.-
25
Student observations were coded as underrepresented if their racial/ethnic identity was Black/African American,
Pacific Islander, Native American, or Hispanic.
Figure 6.- Coefficients for the continuous predictor variables representing log odds of transfer for traditional vs.
non-traditional students. Key: Triangles=Non-traditional students, Circles=Traditional students. Note.- Student
observations were coded as non-traditional if their age was 25 years or older in the year prior to transferring.
Figure 7.- Coefficients for the categorical predictor variables representing log odds of transfer for traditional vs.
non-traditional students. Key: Triangles=Non-traditional students, Circles=Traditional students. Student
observations were coded as non-traditional if their age was 25 years or older in the year prior to transferring.
Cost and Financial Variables
Cost of attendance, which estimated students’ tuition costs based on a number of individual
characteristics, had little relationship to institutional choice when looking at all students. However, for
26
students who were not economically disadvantaged, there was a small positive association. For these
students, an increase in expected tuition costs resulted in slightly higher odds of transferring to that
institution. We also sought to examine whether the percent of students receiving any form of financial aid
at an institution predicted students’ choices. Overall, a higher share of students receiving financial aid
lowered the likelihood of transferring to that destination. However, this was not statistically significant for
economically disadvantaged students or under-represented minorities. Non-traditional students, those who
were not from under-represented minority groups, and those who were not economically disadvantaged
had lower odds of transferring to institutions as the percentage of aid-receiving students increased.
Institution Quality
Included in each of our models were a collection of variables representing institutional quality (e.g.,
student to faculty ratio, percent admitted, instructional expenses, etc.). Overall, the relationship between
instructional expenses and choice was relatively weak. This result held across all subgroups, potentially
because this information is not easily accessible to students or because other factors matter more to
community college transfer students. However, other results indicate students’ sensitivity to certain sets
of quality measures. For example, in our models economically disadvantaged students were more
transfer-averse to institutions with larger student to faculty ratios (-0.021) compared with non-
economically disadvantaged students who were slightly more likely to transfer to schools with larger
student to faculty ratios (0.028). This trend was consistent for our other subgroup analyses, especially for
non-traditional students who appear to be the most choice-averse to institutions with larger class sizes (-
.046) compared with all other subgroups. These coefficients suggest that traditionally disadvantaged
groups of community college students might prefer four-year institutional contexts where class sizes are
smaller and as result students receive more support during their post-secondary trajectory.
In terms of average six-year graduation rates, students across all models appear more likely to
transfer to institutions with higher graduation rates. The coefficients representing log-odds of transfer in
27
each model we ran were positive and statistically significant. In the model where all students are included,
a one-percent increase in the six-year graduation rate of an institution increases the log-odds of transfer by
0.017, holding all other predictors in the model constant. While small, these coefficients could indicate
meaningful differences in preferences when comparing institutions with sizable gaps in six-year
graduation rates. The descriptive statistics shown in table 4 demonstrate a relatively low mean six-year
graduation rate across all potential four-year options (44.15%) with substantial variation between
institutions (SD 19.1%). Our results show a weaker preferential relationship for a separate relative
measure of institutional quality, the percent of first-time full-time freshman admitted amongst all eligible
annual applicants.5 Finally, we added the SAT math 75th percentile score as another measure of four-year
institution quality. Across all models, our results show small, negative, but statistically significant
coefficients. In other words, for every one percentage point increase in the SAT math percentile score of
incoming freshmen at the institutional level, we find that the probability a student in our sample transfers
slightly less likely.
Institutional Type and Characteristics
We also examined the relationship between choice and several variables representing aspects of
institutional type (institutional size and whether the institution was public, a flagship university, research
university, or HBCU). School size, measured by full-time enrollment, did not seem to influence students’
choices, both overall and for subgroups. Students did have a strong preference for public institutions
(3.328), which held across subgroups. Our estimates suggest that compared to all other subgroups of
students, non-traditional students are the most likely to choose a public institution within their choice set.
When an institution is public, we expect a 5.056 increase in the log-odds of an institution being chosen by
non-traditional students, holding all other independent variables constant.
When a university was designated a “research” university, based on Carnegie classifications,
there was a higher likelihood of it being selected (0.076). This result held for all subgroups, but the
28
relationship was stronger for under-represented minorities and non-traditional students. There was a
negative association between choice and an institution being designated a flagship (e.g., University of
Texas at Austin and Texas A&M College Station). This result also held across all groups, although the
relationship was stronger for economically disadvantaged students. These institutions may be less
accessible to community college students across the board.
Given that students of color are over-represented among community college students, we also
examined whether a designation of being a historically black college or university (HBCU) was related to
students’ choices. Across the board, there was a negative relationship between choice and HBCU
designation. This held for under-represented minorities as well. However, the racial demographics of an
institution, particularly the match between the student’s racial/ethnic background and the college’s, did
drive student choices. Specifically, for every one percentage increase in the university’s student
population that was from the same background as the student’s, there was a greater likelihood that
students would select the institution (0.029). This general relationship held across all subgroups, but was
weaker for students from under-represented groups.
6. Discussion
Our study examined the institutional characteristics associated with Texas community college transfer
students’ decisions about where to enroll. We extend the literature in this area though our analysis of a
larger and more recent sample of students compared to related work (Backes and Velez 2015).
Importantly, this study draws on a broad set of transfer destination options, where each option contains a
robust set of institutional characteristics, and thus more nuanced student choice sets compared to prior
analyses of community college transfer. Our findings contribute to what is known about the predictors of
student choice when transferring in a number of ways. At the descriptive level of analysis, we find that
students across all groups have a nearly exclusive preference for public universities even though close to
half of the transfer options were private. In fact, 98% of the transfers were to public universities, with
29
74% to research institutions. For private institutions seeking to increase transfer enrollments, our results
point to the need for policies supporting partnerships with nearby community colleges. Private four-year
schools might consider financial supports to offset their generally higher cost of attendance.
Our analysis related to students’ sensitivity to distance aligns with prior research on transfer
destination choice; however, we are able to parse this pattern of sensitivity by student subgroups. We find
that all students in our sample were less likely to enroll in an institution the farther it was from the
community college attended, but that this preference was the strongest for non-traditional students. This
result is intuitive but nonetheless important considering that community colleges are a likely pathway to
four-year degree attainment for older students who might juggle school with full-time employment as
well as family obligations (Fishman 2015). In fact, prior qualitative work confirms the relative constraint
felt by non-traditional community college students when deciding where to transfer in terms of whether a
four-year option is feasible given work or family responsibilities (Jabbar, Serrata, and Epstein 2017).
Considering our full sample results showing a consistent aversion to distance, we can conclude that
students who are geographically isolated are likely to be at a disadvantage when it comes to transfer. This
has important policy implications for community colleges located in more rural areas and nearby
institutions with the resources to expand access to four-year programs in less urban areas of Texas.
The institutional characteristics associated with students’ choice of transfer destination suggest
that all students were less likely to choose a flagship university. This was particularly true for
economically disadvantaged students, which could indicate access issues such as strict admissions
requirements, limited space, or even cost constraints. We also find that higher graduation rates increase
the probability of transfer across all subgroups. Given the increased time to degree for most students
starting at community colleges and transferring, this result demonstrates that students are more likely to
consider schools with strong degree outcomes. A higher graduation rate could also signal a more
supportive institutional environment overall, and future research in this area should probe which specific
supports systematically predict higher rates of degree completion for this population. Instructional
30
expenses per each full-time enrolled student marginally increased likelihood of transfer, potentially
because information on instructional expenses is less transparent to students. And a smaller student to
faculty ratio as a measure of both support and quality predicted transfer, especially for non-traditional and
economically disadvantaged students. This finding could point to the benefit of smaller class sizes and
more available advising time valued by traditionally disadvantaged groups as they enter their pathway to
four-year degree attainment.
It could be that cost of attendance and a lack of financial support present significant barriers to
transfer and degree completion for community college students. However, in our analysis we found very
small coefficients associated with our measure for cost of attendance and percent of students receiving
financial aid, thus no evidence in our full sample that transfer students weigh cost of attendance factors
when deciding where to transfer. Our subgroup analysis looking at differences between students based on
economic disadvantage provided some additional results in this area. Affluent students are more likely to
transfer to institutions with higher tuition costs and less likely to transfer to institutions with a larger share
of students receiving financial aid. This finding could underline a choice pattern where more advantaged
students seek institutions with students for whom they share a similar socioeconomic background. Or
these students might assume that higher tuition signals a better-quality institution.
Finally, we find interesting results related to our measures of student body demographics and
their association with transfer choice. Overall, we observe a preference for schools offering more of a
racial/ethnic student match. This preference was somewhat stronger for Asian and White students
compared to students historically underrepresented in higher education (e.g., African American, Latino,
Native American). Though this finding warrants a more detailed investigation, it could be that on many
campuses in our sample the racial breakdown is a plurality, but that when Whites and Asians are grouped
together, they represent both a majority of our sample and of enrollments in top-chosen schools. Given
this apparent trend, it would be interesting to know whether a larger proportion of underrepresented
31
students at top-chosen schools in Texas would then influence underrepresented groups of community
college transfer students considering a suite of institutional options.
Taken together, this analysis begins to explain the mechanisms by which community college
students do—or do not—transfer to four-year institutions with the potential to inform programs and
policies that help low-income, first-generation students in particular to successfully apply and transfer to
high-quality four-year institutions.
7. Conclusion
This study contributes to the empirical literature on community college transfer and college choice. First,
ours is one of the few studies of community college choice. In particular, we expand on the most robust
prior analysis of community college choice in Florida (Backes and Velez 2015) with a model that takes
into consideration a more exhaustive set of institutional characteristics. Importantly, our study also creates
a much larger theoretical choice set of institutions with more meaningful variation in institutional
characteristics and geographic locational range.
At the same time, our study also faces several limitations. First, we were unable to link students
to their high school data, resulting in a great deal of missing information. Therefore, we were unable to
explore differences in students’ decisions based on prior achievement levels. Second, studies differ in
how college choice sets are specified. Our study included every postsecondary institution (see, e.g., Long
[2003, 2004]), while other studies simulate a set of institutions that would likely grant the student
admission (see, e.g., Montgomery [2002]). However, it may be unrealistic to assume that students are able
to consider all of the possible alternatives (Niu and Tienda 2007). Our approach thus aligns more with a
human capital model of decision making, rather than a sociological model (Perna 2006), although a
sociological approach to college transfer decisions would add nuance to the literature on the choices of
transfer students.
32
Community colleges are seen both as a way to democratize and potentially divert access to higher
education, yet we know little about the actual choices and decision-making that transfer students
encounter as they pursue a four-year degree. By elaborating the decisions of transfer students and the
characteristics of the schools they select (e.g., geographical location, financial support, and institutional
quality), our study may help to inform programs and policies that help low-income, first-generation
students to successfully apply and transfer to high-quality four-year institutions. This study thus builds on
an important and understudied area, with the aim to provide insight to leaders and policy makers who
work in expanding access to a post-secondary education.
Acknowledgements
The authors would like to thank Uri Treisman and Lauren Schudde for their helpful comments. We also
want to acknowledge Mark Olofsen, Meghan Shea, and Ibrahim Bicak for their research assistance. The
research presented here utilizes confidential data from the State of Texas supplied by the Texas Education
Research Center (ERC) at The University of Texas at Austin. The authors gratefully acknowledge the use
of these data. The views expressed are those of the authors and should not be attributed the funders or
supporting organizations mentioned herein, including The University of Texas and the ERC. Research
was generously supported by the Greater Texas Foundation Faculty Fellows Program.
Notes
1. Some individuals only appear 76 times due to the opening or closing of four-year options during the
time frame of our study.
2. In order to identify additional information used to exclude institutions based on degree offerings or a
specialty programmatic focus we referenced the College Navigator database of NCES
(https://nces.ed.gov/collegenavigator/) and we used the Carnegie Classification system variable in
IPEDS (https://surveys.nces.ed.gov/ipeds/VisGlossaryAll.aspx)
3. In order to formally compare results from each of our model results we ran postestimation Wald tests
of linear hypothesis, assessing whether there was a significant difference between each pair of
coefficients (Korn and Graubard 1990) indicating that the majority of our coefficients are
significantly different from each other. In our discussion, we focus on differences that were
statistically significant.
4. As a check for robustness, we ran our model with a sample of students that restricted to Texas
residents. We also ran a model with a sample excluding students who were tuition exempt. Results
from both of these models show no significant differences when compared to those with our full
student sample. The tables with coefficients from robustness checks are not included in this paper but
are available upon request.
5. On the whole, the percent admitted predictor variable was statistically significant, yet in all models
the values were close to zero and in most cases slightly negative.
33
References
Avery, Christopher, and Caroline M. Hoxby. “Do and Should Financial Aid Packages Affect Students’
College Choices?” In College Choices: The Economics of Where to Go, When to Go, and How to
Pay for It, edited by Caroline M. Hoxby, 239–301. University of Chicago Press, 2004.
Backes, Ben, and Erin Dunlop Velez. 2015. “Who Transfers and Where Do They Go? Community
College Students in Florida.” Washington, D.C.: National Center for Analysis of Longitudinal
Data in Education Research (CALDER).
https://caldercenter.org/sites/default/files/WP%20126.pdf.
Ball, Stephen J., Jackie Davies, Miriam David, and Diane Reay. 2002. “‘Classification’ and ‘Judgement’:
Social Class and the ‘Cognitive Structures’ of Choice of Higher Education.” British Journal of
Sociology of Education 23 (1): 51–72. doi:10.1080/01425690120102854.
Bastedo, Michael N., and Ozan Jaquette. 2011. “Running in Place: Low-Income Students and the
Dynamics of Higher Education Stratification.” Educational Evaluation and Policy Analysis 33
(3): 318–39. Beattie, I. R. (2002). Are All “Adolescent Econometricians” Created Equal? Racial,
Class, and Gender Differences in College Enrollment. Sociology of Education, 75(1), 19–43.
doi:10.3102/0162373711406718.
Belfield, Clive R., and Thomas Bailey. 2011. “The Benefits of Attending Community College: A Review
of the Evidence.” Community College Review 39 (1): 46–68. doi:10.1177/0091552110395575.
Bell, Courtney A. 2009. “All Choices Created Equal? The Role of Choice Sets in the Selection of
Schools.” Peabody Journal of Education 84 (2): 191–208. doi:10.1080/01619560902810146.
Bers, Trudy H., and Pamela M. Galowich. 2002. “Using Survey and Focus Group Research to Learn
About Parents’ Roles in the Community College Choice Process.” Community College Review 29
(4): 67–82. doi:10.1177/009155210202900404.
Black, Sandra E, Kalena E Cortes, and Jane Arnold Lincove. 2015. “Apply Yourself: Racial and Ethnic
Differences in College Application.” Working Paper 21368. National Bureau of Economic
Research. doi:10.3386/w21368.
Bowen, William G., Matthew M. Chingos, and Michael S. McPherson. 2009. Crossing the Finish Line:
Completing College at America’s Public Universities. Princeton University Press.
34
Brint, Steven, and Jerome Karabel. 1991. The Diverted Dream: Community Colleges and the Promise of
Educational Opportunity in America, 1900-1985. Oxford, New York: Oxford University Press.
Castleman, Benjamin L., Sandy Baum, and Saul Schwartz. 2015. “Behavioral Economics and
Postsecondary Access: A Primer.” In Decision Making for Student Success: Behavioral Insights
to Improve College Access and Persistence, 1–19. New York, NY: Routledge.
Castro, Erin L., and Edén Cortez. 2017. “Exploring the Lived Experiences and Intersectionalities of
Mexican Community College Transfer Students: Qualitative Insights Toward Expanding a
Transfer Receptive Culture.” Community College Journal of Research and Practice 41 (2): 77–
92. doi:10.1080/10668926.2016.1158672.
Crisp, Gloria, and Anne-Marie Nuñez. 2014. “Understanding the Racial Transfer Gap: Modeling
Underrepresented Minority and Nonminority Students’ Pathways from Two-to Four-Year
Institutions.” The Review of Higher Education 37 (3): 291–320. doi:10.1353/rhe.2014.0017.
Deil-Amen, Regina, and James E. Rosenbaum. 2003. “The Social Prerequisites of Success: Can College
Structure Reduce the Need for Social Know-How?” The ANNALS of the American Academy of
Political and Social Science 586 (1): 120–43. doi:10.1177/0002716202250216.
Dougherty, Kevin J. 1992. “Community Colleges and Baccalaureate Attainment.” The Journal of Higher
Education 63 (2): 188–214. doi:10.1080/00221546.1992.11778349.
Doyle, William R. 2009. “The Effect of Community College Enrollment on Bachelor’s Degree
Completion.” Economics of Education Review 28 (2): 199–206.
doi:10.1016/j.econedurev.2008.01.006.
Ehrenberg, Ronald G., Donna S. Rothstein, and Robert B. Olsen. “Do Historically Black Colleges and
Universities Enhance the College Attendance of African American Youths?” In A Nation
Divided: Diversity, Inequality, and Community in American Society, edited by Phyllis Moen,
Donna Dempster-McClain, and Henry A. Walker, 171–88. Ithica: Cornell University Press, 1999.
https://vtechworks.lib.vt.edu/handle/10919/83084.
Fishman, Rachel. 2015. “2015 College Decisions Survey: Part I: Deciding to Go to College.”
Washington, D.C.: New America Foundation. https://www.newamerica.org/education-
policy/policy-papers/deciding-to-go-to-college/.
35
Gandara, Patricia, Elizabeth Alvarado, Anne Driscoll, and Gary Orfield. 2012. “Building Pathways to
Transfer: Community Colleges That Break the Chain of Failure for Students of Color.” Civil
Rights Project. https://civilrightsproject.ucla.edu/research/college-access/diversity/building-
pathways-to-transfer-community-colleges-that-break-the-chain-of-failure-for-students-of-
color/Fullpaper-Building-Pathways-final-2-1-12b.pdf.
Gill, Andrew M., and Duane E. Leigh. 2003. “Do the Returns to Community Colleges Differ between
Academic and Vocational Programs?” Journal of Human Resources 38 (1): 134–55.
doi:10.3368/jhr.XXXVIII.1.134.
Gonzalez, Arturo., and Michael Hilmer. 2006. “The Role of 2-year Colleges in the Improving Situation of
Hispanic Postsecondary Education.” Economics of Education Review 25 (3): 249–257.
doi:10.1016/j.econedurev.2004.12.002.
Goodman, Joshua, Michael Hurwitz, and Jonathan Smith. 2015. "Access to Four-Year Public Colleges
and Degree Completion." NBER Working Paper 20996, National Bureau of Economic Research,
Cambridge, MA. doi:10.3386/w20996.
Grodsky, Eric, and Melanie T. Jones. 2007. “Real and Imagined Barriers to College Entry: Perceptions of
Cost.” Social Science Research 36 (2): 745–66. doi:10.1016/j.ssresearch.2006.05.001.
Grubb, W. Norton. 1988. “Vocationalizing Higher Education: The Causes of Enrollment and Completion
in Public Two-Year Colleges, 1970–1980.” Economics of Education Review 7 (3): 301–19.
doi:10.1016/0272-7757(88)90003-9.
Grubb, W. Norton. 1991.” The Decline of Community College Transfer Rates: Evidence from National
Longitudinal Surveys.” The Journal of Higher Education 62(2): 194–222.
doi:10.1080/00221546.1991.11774115.
Hillman, Nicholas, and Taylor Weichman. 2016. “Education Deserts: The Continued Significance of
‘Place’ in the Twenty-First Century.” Viewpoints: Voices from the Field. Washington, D.C.:
American Council on Education. https://www.acenet.edu/news-room/Documents/Education-
Deserts-The-Continued-Significance-of-Place-in-the-Twenty-First-Century.pdf.
36
Hilmer, Michael J. 1997. “Does Community College Attendance Provide a Strategic Path to a Higher
Quality Education?” Economics of Education Review 16 (1): 59–68. doi:10.1016/S0272-
7757(96)00018-0.
Hilmer, Michael J.1998. “Post-Secondary Fees and the Decision to Attend a University or a Community
College.” Journal of Public Economics 67 (3): 329–48. doi:10.1016/S0047-2727(97)00075-3.
Hilmer, Michael J. 2002. “Human Capital Attainment, University Quality, and Entry-Level Wages for
College Transfer Students.” Southern Economic Journal 69 (2): 457–69. doi:10.2307/1061683.
Hossler, Don, and Karen S. Gallagher. 1987. “Studying Student College Choice: A Three-Phase Model
and the Implications for Policymakers.” College and University 62 (3): 207–21. https://scholarworks.iu.edu/journals/index.php/jiuspa/article/download/4595/4218/0.
Hurtado, Sylvia, Karen Kurotsuchi Inkelas, Charlotte Briggs, and Byung-Shik Rhee. 1997. “Differences
in College Access and Choice Among Racial/Ethnic Groups: Identifying Continuing Barriers.”
Research in Higher Education 38 (1): 43–75. doi:10.1023/A:1024948728792.
Hurwitz, Michael. 2012. “The Impact of Institutional Grant Aid on College Choice.” Educational
Evaluation and Policy Analysis 34 (3): 344–63. doi:10.3102/0162373712448957.
Jabbar, Huriya, Eliza Epstein, Joanna Sánchez, and Wes Edwards. Forthcoming. Diving into the Pool: An
Analysis of Texas Community College Students’ Transfer Institution Choice Sets. Teachers
College Record.
Jabbar, Huriya, Carmen Serrata, and Eliza Epstein. 2017. “Échale Ganas”: Family and Community
Support of Latino/a Community College Students’ Transfer to Four-Year Universities. Journal of
Latinos and Education, 1–20.
Jenkins, Paul Davis, and John Fink. 2015. “What We Know About Transfer.” Community College
Research Center: Teachers College, Columbia University. doi:10.7916/D8ZG6R55.
Jepsen, Christopher, and Mark Montgomery. 2009. “Miles to Go before I Learn: The Effect of Travel
Distance on the Mature Person’s Choice of a Community College.” Journal of Urban Economics
65 (1): 64–73. doi:10.1016/j.jue.2008.08.004.
37
Kinsler, Josh, and Ronni Pavan. 2011. “Family Income and Higher Education Choices: The Importance
of Accounting for College Quality.” Journal of Human Capital 5 (4): 453–77.
doi:10.1086/663649.
Korn, Edward L., and Barry I. Graubard. 1990. “Simultaneous Testing of Regression Coefficients with
Complex Survey Data: Use of Bonferroni T Statistics.” The American Statistician 44 (4): 270–76.
doi:10.1080/00031305.1990.10475737.
Lee, Valerie E., and Kenneth A. Frank. 1990. “Students’ Characteristics That Facilitate the Transfer from
Two-Year to Four-Year Colleges.” Sociology of Education 63 (3): 178–93. doi:10.2307/2112836.
Leigh, Duane, and Andrew Gill. 2003. “Do Community Colleges Really Divert Students from Earning
Bachelor’s Degrees?” Economics of Education Review 22(1): 23–30. doi:10.1016/S0272-
7757(01)00057-7.
Lockwood Reynolds, C. 2012. “Where to Attend? Estimating the Effects of Beginning College at a Two-
Year Institution.” Economics of Education Review 31 (4): 345–62.
doi:10.1016/j.econedurev.2011.12.001.
Long, Bridget Terry. 2004. “How Have College Decisions Changed Over Time? An Application of the
Conditional Logistic Choice Model.” Journal of Econometrics 121: 271-296.
doi:10.1016/j.jeconom.2003.10.004.
Long, Bridget Terry. 2007. “The Contributions of Economics to the Study of College Access and
Success.” Teachers College Record 109 (10): 2367–2443.
Long, Bridget Terry, and Michal Kurlaender. 2009. “Do Community Colleges Provide a Viable Pathway
to a Baccalaureate Degree?” Educational Evaluation and Policy Analysis 31 (1): 30–53.
doi:10.3102/0162373708327756.
Lovenheim, Michael F., and C. Lockwood Reynolds. 2013. “The Effect of Housing Wealth on College
Choice: Evidence from the Housing Boom.” Journal of Human Resources 48 (1): 1–35.
doi:10.3368/jhr.48.1.1.
38
Manski, Charles. 1993. “Adolescent Econometricians: How Do Youth Infer the Returns to Schooling?” in
Studies of Supply and Demand in Higher Education. C. Clotfelter and M. Rothschild, eds.
Chicago: University of Chicago Press.
McFadden, Daniel. 1973. "Conditional Logit Analysis of Qualitative Choice Behavior" in Frontiers of
Econometrics, ed. by P. Zarembka, 105–142. New York: Academic Press.
Melguizo, Tatiana, and Alicia C. Dowd. 2009. “Baccalaureate Success of Transfers and Rising 4-Year
College Juniors.” Teachers College Record 111 (1): 55–89.
Mundel, David. “What Do We Now About the Impact of Grants to College Students?” In The
Effectiveness of Student Aid Policies: What the Research Tells Us, edited by Sandy Baum,
Michael McPherson, and Patricia Steele, 9–38. New York, NY: The College Board. 2008.
National Student Clearinghouse Center. 2017. Two-Year Contributions to Four-Year Completions.
https://nscresearchcenter.org/snapshotreport-twoyearcontributionfouryearcompletions26/
Page, Lindsay C., and Judith Scott-Clayton. 2016. “Improving College Access in the United States:
Barriers and Policy Responses.” Economics of Education Review, Access to Higher Education, 51
(April): 4–22. doi:10.1016/j.econedurev.2016.02.009.
Pallais, Amanda, and Sarah Turner. 2006. “Opportunities for Low-Income Students at Top Colleges and
Universities: Policy Initiatives and the Distribution of Students.” National Tax Journal 59 (2):
357–86. doi:10.17310/ntj.2006.2.08.
Picard, Robert. 2012. “GEODIST: Stata Module to Compute Geodetic Distances.” EconPapers.
https://EconPapers.repec.org/RePEc:boc:bocode:s457147.
Radford, Alexandria Walton. 2013. Top Student, Top School?: How Social Class Shapes Where
Valedictorians Go to College. University of Chicago Press.
Rouse, Cecilia Elena. “What to Do after High School: The Two-Year Versus Four-Year College
Enrollment Decision.” In Choices and Consequences: Contemporary Policy Issues in Education,
edited by Ronald G. Ehrenberg. Cornell University Press, 1994.
39
Schudde, Lauren, Dwuana Bradley, and Caitlin Absher. 2018. “Ease of Access and Usefulness of
Transfer Information on Community College Websites in Texas.” Community College Research
Center CCRC Working Paper No. 102. https://ccrc.tc.columbia.edu/media/k2/attachments/ease-
access-usefulness-transfer-information-community-college-websites-texas.pdf.
Scott-Clayton, Judith. “The shapeless river: Does a lack of structure inhibit students’ progress at
community colleges?” In Decision making for student success: Behavioral insights to improve
college access and persistence, edited by Benjamin L. Castleman, Saul Schwartz, and Sandy
Baum, 102–123. New York: Routledge, 2015.
Skinner, Benjamin. “Modeling College Enrollment via Conditional Logit: How has College Choice
Changed? Paper presented at Association for Education Financial and Policy Annual Meeting,
Denver, CO, March 2016.
Smith, Jonathan, Matea Pender, and Jessica Howell. 2013. “The Full Extent of Student-College Academic
Undermatch.” Economics of Education Review 32 (February): 247–61.
doi:10.1016/j.econedurev.2012.11.001.
Snyder, Thomas, Cristobal de Brey, and Sally Dillow. 2016. “Digest of Education Statistics.” NCES
2016-006. Washington, D.C.: National Center for Education Statistics, Institute of Education
Sciences, U.S. Department of Education. https://nces.ed.gov/pubs2017/2017094.pdf
Somers, Patricia, Kevin Haines, Barbara Keene, Jon Bauer, Marcia Pfeiffer, Jennifer McCluskey, Jim
Settle, and Brad Sparks. 2006. “Towards a Theory of Choice for Community College Students.”
Community College Journal of Research and Practice 30 (1): 53–67.
doi:10.1080/10668920500248886.
Surette, Brian J. 2001. “Transfer from Two-Year to Four-Year College: An Analysis of Gender
Differences.” Economics of Education Review 20 (2): 151–63. doi:10.1016/S0272-
7757(00)00013-3.
Texas Higher Education Coordinating Board. 2014. “Texas General Academic Institutions: Increasing
Successful Community College Transfer.” Austin, Texas: Texas Higher Education Coordinating
Board. http://www.thecb.state.tx.us/reports/pdf/6073.pdf.
Texas Higher Education Coordinating Board. 2018. “Texas Public Higher Education Almanac: A Profile
of State and Institutional Performance and Characteristics.” Austin, Texas: Texas Higher
40
Education Coordinating Board. http://www.thecb.state.tx.us/reports/PDF/10900.PDF?CFID=93113602&CFTOKEN=46841789.
Tierney, Michael L. 1983. “Student College Choice Sets: Toward an Empirical Characterization.”
Research in Higher Education 18 (3): 271–84. doi:10.1007/BF00979600.
Tinto, Vincent. "Dropping Out and Other Forms of Withdrawal from College." In Improving Student
Retention, edited by Noel Levitz and Diana Salvri. San Francisco: Jossey-Bass Inc, 1985.
Turley, Ruth N. López. 2009. “College Proximity: Mapping Access to Opportunity.” Sociology of
Education 82 (2): 126–46. doi:10.1177/003804070908200202.
Velez, William, and Rajshekhar G. Javalgi. 1987. “Two-Year College to Four-Year College: The
Likelihood of Transfer.” American Journal of Education 96 (1): 81–94. doi:10.1086/443882.
Wassmer, Robert, Colleen Moore, and Nancy Shulock. 2004. “Effect of Racial/Ethnic Composition on
Transfer Rates in Community Colleges: Implications for Policy and Practice.” Research in
Higher Education 45 (6): 651–72. doi: 10.1023/B:RIHE.0000040267.68949.d1.