A recent CTViewpoints opinion — Connecticut’s four year public state university graduation rates fall short — correctly observed that Connecticut’s state universities “have a responsibility to help students graduate.”  Their success would “provide the state with more educated individuals equipped to enter the workforce and ultimately, enable them to become more productive citizens.”

The good news is that the CSCU universities are in fact successful in achieving that objective. But that was not the conclusion of the author of the op-ed, who argued that six-year graduation rates of the CSCU universities were unacceptably low.

The false impression caused by a flawed methodology should be corrected.

Unfortunately, the author of the op-ed relied on a metric that is egregiously flawed – although it is uncritically accepted as valid in too many higher education circles.

The problem is based on a definition of “graduation rate” which was established in 1990. The defective “graduation rate” methodology does not reflect the actual performance of a university in producing credentialed graduates who receive bachelors’ degrees.

Instead, the widely-accepted definition of “graduation rate” only measures the number of graduates who entered the university a specified number of years earlier as first-time (i.e. freshmen), full-time, fall-term (“FTFTFT”) students.  [Using this metric, Donald Trump would not have been counted in the graduation rate at Fordham, where he first enrolled, because he did not graduate from Fordham, nor would he have been counted in the graduation rate at UPenn, since he did not originally enroll there as a freshman, but subsequently transferred in.

This measure may have made sense at some past time for traditional institutions such as liberal arts colleges and research universities that primarily admitted students as freshmen in the fall term, who were on campus full time, who did not stop out or go part-time periodically to earn money to pay tuition and then come back to school to continue their education. But across the nation, today’s student generation is “markedly different,” as the organization Student Achievement Measure (SAM) notes: One-third of today’s college students are over 25 years old, 20 per cent are employed full-time when enrolled, and one in four have children.

CSCU-type universities serve students like this: where they are, not where they might ideally be.  They certainly admit full-time freshmen in the fall term, but they also admit full-time transfer students then, and part-time students then and in subsequent terms.  They admit full-time students in the spring term.  They admit working adults. They recognize that some students drop out, or drop back to part-time status, in order to earn money for tuition and fees – and living expenses – in subsequent semesters.  And, most notably, they admit substantial numbers of transfer students from other institutions, especially community colleges, throughout these students’ college careers, a phenomenon called “student swirl.”

[This fact is partially recognized by using a six-year “graduation rate” metric instead of a four-year “graduation rate”: because of the difficulty of “student swirl” students to complete a bachelor’s program in what formerly was regarded as the “normal” or linear four year “time to completion.”  The more appropriate measure now used is graduation within 150 percent (six years) of “normal time” to program completion.]

The number of bachelors’ degrees granted at CSCU universities exceeds the number of freshman students who entered six years earlier.

The substantial inflow of students who do not count in the officially-measured “graduation rate” more than balances the outflow of “FTFTFT” students.  The consequence is that – at CSCU – the number of graduates six years after its official cohort class enters the institution is SUBSTANTIALLY GREATER than the number of first-time, full-time, fall-term students who enter in the cohort.

The following table provides data for the entering cohort in Fall Term 2010 (Column B), compared with (1) the number of graduates from that FTFTFT cohort in AY 2015-16 (Column D), and (2) the total number of bachelors’ degree graduates in AY 2015-16 (Column E).  As is evident, CSCU System universities graduate far MORE graduates each year than were admitted as FTFTFT students six years earlier.

Column D in the table shows that, on average, only about 51 percent of the graduates in any one year are part of the FTFTFT cohort that began college work six years before (2,232 out of 4,358).  The rest (3,450 out of 5,682) – as presented in Column G in the table – are “bonus,” providing a return on investment not reflected by calculating ROI on the basis of those who count in the accepted “graduation rate.”

In short, CSCU, like other universities of its kind, is providing substantial benefits to the economic competitiveness of the state – credentialing significant numbers of future workforce members, who are not measured or measurable by the traditional “graduation rate” analysis.

ABCDEF=(E – B)G=(E – D)
UniversityFall 2010 FTFTFT cohortListed “Grad Rate” (%)Number of 2010 Cohort who Graduate in 6 years (B x C)Actual Bachelors Degrees AY 2016Excess of Actual Degrees over “FTFTFT” Entering CohortExcess of Actual Degrees over “FTFTFT” Cohort Graduates
WCSU *84544%3721,084239712
All CSCU4,35851%2,2325,6821,3243,450

*  This data set is based on the 2010 cohort.  Data for WCSU in this table are for the Fall 2011 FTFTFT Cohort, the “Graduation Rate” in AY 2017, and Actual Bachelors Degrees Granted in AY 2017.

A Better Measure:  Count All Students

Two other data sets provide additional light on the contributions of the CSCU System universities to the future workforce

First, data are now being collected that provide more detailed information about the educational careers of the FTFTFT cohort.  It turns out that many students from this cohort (1) transferred to another institution and received a bachelors’ degree from there (see the example of Donald Trump), (2) continued to be enrolled at the home university, and/or (3) transferred to another institution and continued enrollment there. Assuming that all of these are indicators of educational accomplishment, the cumulative known “success rate” is far higher than the flawed “graduation rate” measure.   (Please note that these data are, with the exception of SCSU, for the Fall 2011 cohort.)

UniversityFTFTFT Cohort in Fall 2011Listed “Grad Rate” (%)Transferred to another institution and graduated w/ Bachelors’ Degree (%)Continued enrollment at original CSU univ.(%)Transferred to another 4-year  institution and continued enrollment there (%)Success Rate



SCSU *1,24851%12%5%5%73%

* This data set is based on the 2011 cohort.  Data for SCSU are for Fall 2010 cohort, and AY 2016 status.

Second, it has become evident that a considerable number of students (ranging as high as 75 percent of the number of FTFTFT entering students) entering in the fall term are full-time transfers from another institution.  Many of these transferring students commenced work at a community college: it may have been close to home, or it may have been far less expensive than a four-year institution for the first year or two of college.  In light of the increasing demand for low- or no-tuition charges for community college students, the numbers of these students – already very substantial – are likely to increase. The data set below, analogous to the data set just above, also demonstrates a significant “success rate.”

University Full-Time Transfer Student Cohort in Fall 2011Listed “Grad Rate” (%)Transferred to another institution and graduated w/ Bachelors’ Degree (%)Continued enrollment at original CSCU univ.(%)Transferred to another 4-year  institution and continued enrollment there (%)“Success Rate” of Full-Time Transfer Student Cohort
SCSU *94753%9%3%5%70%

* This data set is based on the 2011 cohort. Data for SCSU are for Fall 2010 cohort, and AY 2016 status. Comparison with peer institutions nationally.

Even if one does rely on the outmoded “graduation rate” methodology, the CSCU System universities generally compare favorably with institutions like them in other states.  These peer institutions are, like CSCU universities, NOT private liberal arts colleges or private or public research universities – institutions which may not serve many students who are first-in-their-family to attend college, or are highly selective in their admissions, or whose self-selected applicants are able to afford higher tuition and fees.

UniversityCSU System University “Graduation Rate” of FTFTFT Cohort entering in Fall Term 2010 (%)Peer Institution “Graduation Rate” of FTFTFT Cohort Graduating in AY 2016 (%)

CSCU System graduates constitute a significant share of bachelors’ degree-holders in Connecticut.

Although the number of bachelors’ degrees granted by the state’s flagship research university (UConn) is important to the overall economic competitiveness of Connecticut, so too is the number of bachelors’ degrees granted each year by the four CSCU System universities.

In fact, the number of CSCU graduates is GREATER than the number of UConn bachelors’ degrees. In AY 2016, UConn produced 4,989 bachelors’ degree graduates, while the four CSCU System universities produced 5,682.

These CSCU degree holders are a major reason why, together with the degrees granted by UConn and private colleges and universities in the state, the number of adults with bachelors’ degrees in Connecticut ranks fourth in the nation – making it one of the most “highly educated” states, with a workforce primed to take on future economic challenges.


CSCU System Universities provide substantial benefits to the economic competitiveness of the state by graduating students fully prepared to become productive workforce members.

Post Script – Additional Analysis Needed

However, it must be admitted that there are still substantial gaps in our knowledge. What is truly lamentable is not a low “graduation rate,” but the absence of good data – and the consequent inability to assess critical outcomes and analyze factors which might contribute to the success or failure of students to be able to contribute to the workforce.

The astute reader will observe from the data reported above that many students fall off the radar at some time during their higher education experience. Even the use of the “success rate” metric misses students whose status is unknown after six years, even after others have been identified as graduates from somewhere, or continue to be enrolled in some higher education institution. As many as 30 percent of full-time students initially enrolling in a fall term just disappear from the charts (See the second and third data sets above). And what about students admitted at other times, or admitted as part-time students?

Also, there is a paucity of data about the successful integration of college graduates into the workforce.  We know that they receive degrees in substantial numbers.  But we don’t know if they find employment, either in-state or out-of-state, and whether their jobs are substantial and productive, or if they are “under-employed.”  Measuring real outcomes would provide a substantial advancement of knowledge, and potentially enable the development of public policy to address and correct any shortcomings of the educational process.

A substantial research effort is needed to collect and analyze data to address these two issues.

Bill Cibes is Chancellor  Emeritus of the Connecticut State University System.

CTViewpoints welcomes rebuttal or opposing views to this and all its commentaries. Read our guidelines and submit your commentary here.

Leave a comment