By Clifford Adelman
The title of this piece, a play on that of Wallace Stevens' poem, "Thirteen
Ways of Looking at a Blackbird," reflects the nearly infinite variety of nuances
in the basic question about attainment in American higher education.
When newspaper reporters or state legislative aides telephone and ask, "What
proportion of college students earn a bachelor's degree?" or "What percentage
of community college students earn credentials?" the answers are not clear until
the questioner defines who is in the denominator (or what we call the "universe"):
all students who graduated from high school? all students who ever said they aspired
to a bachelor's degree? all students who entered college? all students who attended
a four-year college at any time? all students who entered a four-year college directly
from high school?
After the denominator, the numerator is comparatively easy, but still must be
specified. For the bachelor's degree, the numerator must include a time "censor":
within five years of high school graduation? within five years of entering college?
within six years? by age 35? ever?
Of course we can complicate the numerator with conditions other than those of
time, but with each additional degree of complexity, we no longer are asking the
In July of this year, I was asked the basic question by staff at the Congressional
Budget Office. The context for the question was a proposed amendment to the reauthorization
of the Higher Education Act, and required a long-term time censor-something beyond
five or six years.
The purpose of this presentation is not to discuss the proposed amendment, but
rather to share with you what I told my colleagues at the Congressional Budget Office.
The source for the data is the postsecondary transcript file of the High School
& Beyond/Sophomore Cohort longitudinal study. This is the second national longitudinal
study conducted by the National Center for Education Statistics. The postsecondary
transcripts in those studies enable us to tell very accurate student histories, histories
that cross state lines and involve many institutions - for the same student.
At present, this is the only data source in the nation that can answer the basic
question about long term degree completion rates in recent years. In these data,
we are looking at the history of the scheduled high school graduating class of 1982,
from the time of their graduation up to 1993, when most of the cohort was 29 or 30
The Four-Year College Story
The basic position I take in Table 1 ("Postsecondary Fate to Age 30") is
that one is not in the denominator for the calculation of bachelor's degree attainment
rates unless one has gone to the trouble of actually enrolling in a bachelor's degree-granting
institution. With that simple gesture, one says far more than repeating 100 times
"I want to get a bachelor's degree" or "I am working toward a bachelor's
degree." It is neither accurate nor fair to judge attainment among those who
did not make a minimum attempt by age 30.
Table 1 first lays out the destinations of those who actually attended four-year
colleges at any time, no matter how many credits they earned, then indicates what
happened to those on-time high school graduates who went directly to college ("no
delayed entry"), then repeats both sets of "fates" for those whose
true first institution of attendance was a four-year school. Let us call these the
"benchmark BA" numbers.
The histories of approximately 1.3 million students are being described here.
Table 1 then ratchets up the threshold of earned credits in the histories of these
students three times: first to ten credits, then to 30 credits and 30 credits from
four-year colleges, and then to 60 credits and 60 credits from four-year colleges.
The point is obvious: The more credits one earns, the more likely one is to complete
At each threshold level, entering college directly from high school and first
entering a four-year college results in a higher degree completion rate, though this
phenomenon holds statistical significance only at the lower thresholds. Earning one's
threshold credits from four-year colleges - as opposed to any combination of college
types - also has a positive impact on degree completion rates, and these effects
are significant at every credit threshold.
This is all common sense: and long-term national degree completion rates are very
high, no matter how many schools a student attends (54 percent of students in this
sample - and 58 percent of the bachelor's degree recipients - attended more than
If one enters a four-year college directly from high school, and gets by the 60th
credit, the odds are about 7 in 8 of completing a bachelor's degree by age 30. That's
pretty good! In an age of multi-institutional attendance, such system graduation
rates make far more sense than "institutional graduation rates." Institutions
may "retain," but it is students who persist. And the last time I looked,
federal higher education policy was directed at students, not institutions.
Since all these data came from the college transcript records, I could add the
following note: At age 30, and among those who had earned more than 60 credits, a
relatively low proportion (nine to 12 percent) of the low proportion (13 to 21 percent)
who had not earned a bachelor's degree were still in school.
In other words, the vast majority of non-completers had drifted away from higher
education at age 30. While we always are confident that some will return, it is eminently
apparent that we are children of time, and that other demands and possibilities of
life come to supersede those of formal education after we have passed through our
The Community College Story
The community college story presented in Table 2 is both very different and very
exciting. For years, we have been beating up on community colleges because of what
we perceive to be low degree completion rates.
But students use community colleges for a variety of reasons, not all of which
are connected to credentials. Of the entire universe of students who ever enter community
colleges, nearly one out of six never earns even a semester's worth of credits. These
"incidental" students are excluded from the analysis in "Community
College Fate at 30" because they are just that - incidental - and it is not
fair, or accurate, to include them in a universe with which we judge institutional
Table 2 thus sets a minimum threshold: students who earned more than ten credits
from community colleges as undergraduates. (It excludes a small number of students
who attended community colleges after earning a bachelor's degree as well as those
four-year college students who took a course or two at community colleges.)
Table 2 does something else that is very important to judging community college
performance in terms of labor market preparation. It takes students who did not complete
any credential, looks at their transcript records, and asks whether one can describe
a dominant "tone" of study, something analogous to a college major or a
balanced general education program.
For example, a student may have accumulated 36 credits (and no credential), of
which half are in finite mathematics, electronics, computer programming, computer
organization and architecture. There is no doubt of a dominant "tone" to
this record. One can say that this student is prepared to enter the labor market
in the general area of computer technologies.
This does not mean the student is an automatic JAVA-whiz, nor does it mean that
we have witnessed the end of the individual's education and training. What it does
mean is that the student has derived something from the community college experience
that anyone - including employers - can describe. And I don't need to remind readers
that a majority of community college students attends in order to establish specific
trajectories into the labor market.
At age 30, then, those who have attended community colleges in non-incidental
ways have "separated" from the system in a satisfactory manner if they
have accomplished one of four ends: 1) transferred to a four-year college and received
a bachelor's degree; 2) earned a terminal associate's degree; 3) earned a certificate
indicating a coherent course of study that is nonetheless not a full degree program;
or 4) taken a sufficient amount of coursework that can be described as a partial
major or complete lower-division general education program.
Table 2 ratchets its thresholds only twice, producing three levels of community
college engagement. At the third level, which I call "community college dominant,"
we have students who earned 30 or more credits from a community college and fewer
than 11 credits from four-year colleges. In this group, the de facto "completion
rate" (associate's degree + certificate + a classifiable cluster of coursework)
is an astounding 89 percent - equivalent to the bachelor's degree completion rate
of four-year college students who entered directly from high school and earned more
than 60 credits.
For skeptics who say that this "community college dominant" group is
small, I beg to differ: We are looking at about 325,000 people from a single high
school graduating class.
For the record (and because someone is bound to ask), at the "community college
dominant" level, the distribution of classifiable course clusters of those who
left without a credential was: complete general education program (23 percent); business
and marketing (15 percent); business support occupations (ten percent); arts and
applied arts (8 percent); computer science and technology (7 percent); other science
and tech (7 percent); trades and crafts (7 percent); and other fields (23 percent).
Not all of these are statistically significant percentages, but they give one a decent
idea of the distribution.
So What's the Point?
Our system may appear sloppy to some, but our results are better than the popular
myths, most of which use the institution, and not the student, as the unit of analysis.
We have to be prepared to provide solid answers to those who hold us accountable
in terms that the public understands. Our judges have many ways of asking the question,
and any appearance of uncertainty on our part in answering will be taken as a sign
Our focus must on the student. We have to be clear about the terms of our answers:
The terms must be those of common sense, and we must be able to combine them quickly
Clifford Adelman is a senior research analyst with the U.S. Department of Education.
The analysis and opinions offered in this article are his own, are intended to stimulate
discussion, and do not necessarily reflect departmental positions or policy.