Cohort Default Rates or Graduation Performance Rank as College Quality Indicator? (Calling LA, MA, MO, NM, NC, WA residents especially!)

Parents Forum
21

I downloaded these spreadsheet (below) and then added a column that expressed balance owned after years as percentage of the original “principal”, then sorting by that.

At first I thought “bingo” (once I saw the extreme end of the list where people owe drastically MORE than what they started off with). But when I looked at a few universities I was familiar with, I realized that for some, their lower repayment rate could be related to the economic realities of their average student (e.g., whether more were forced to drop out at some point) than quality of education.

I also attempted facoring in the number of students - thinking about whether a lower repayment rate where only 30% of students needed loans, should be valued differently compared to a seemingly higher repayment rate at a place where 60% needed loans.

While the headline below is talking about graduate schools, it seemed to me as if some spreadsheets included bachelor degrees:

2 Likes
22

I’m curious about how the study obtained earnings information. The loan information for the school where I worked looks correct, and it should - this is available information. Earnings information is not collected by nor reported for the school, though. Collection and reporting are not required. If they use an “average earnings” for the degree, based on information not tied to the particular school, the information is not really good information. And because I am familiar with what our graduates do, I know that even if they had actual earnings information, the small size would really skew the averages.

1 Like
23

A confounding variable is likely to be the economic prospects within the student catchment area. This is especially true when talking about lower tier directional colleges who mostly serve local students who aren’t going to move away to more economically vibrant parts of the country. It is likely to explain a significant part of the differences between states and some of the geographic differences within states (eg coastal vs inland WA).

1 Like
24

As @FryingPanShoals mentioned, in NC the bottom six schools on your list are HBCUs and UNC-Pembroke, which is a historically Lumbee Indian university. I would guess a lot of students from lower income families may be the ones defaulting on these loans. I could see a situation where they have to choose between helping out other family members with living expenses or paying student loans. They also may be first gen students who got in over their heads on student loans without a parent who attended college to guide them, in contrast to students at UNC-CH who may be multi-generational legacies.

And you did miss UNC-Charlotte on your list as @bronzerfish mentioned. It’s one of the biggest schools in the state with over 30K students so a pretty big omission.

To answer your questions:

  • Do you see a relationship between the school’s default rate and the school’s reputation for quality in the state?

I guess I would say somewhat. I think everybody (except State fans) sees UNC- Chapel Hill as having the best rep in the state for a public. (Don’t get me started on the Dukies.) However I think NC Central and NC A&T have pretty good reps too as HBCUs.

  • In your experiences of dealing with the alumni from the various universities, do you see a relationship between how well you thought of them (work/thought quality) as compared to the default rates?

No not really at all. I am very slightly acquainted with an Elizabeth City State grad (she works in the public library in my town) and she seems nice and like she has her stuff together from our very casual acquaintance. I know lots of grads from the other schools and I have never thought, “oh, no wonder their work/thought quality is lacking, they only went to X school”

  • Basically, do the default rates surprise you? Confirm what you’ve seen in real life? Or do you think the default rates have no correlation to the quality of education received?

I don’t think the rates have that much correlation to the quality of education received. It does correlate closely to selectivity of the institution, but I don’t think selectivity always correlates to the quality of the education received – a student can learn a lot at a less selective university too. I think the default rate may more closely correlate to lower income of the students and their families and to first gen students. I would be interested to see how that compares to the default rate.

  • If you were to use a default rate as a filter on schools, where would you draw the line when considering schools for anyone you’d give guidance to…5%? 10%? 15%? Doesn’t matter?

I don’t think I would draw the line using the default rate. I would use the financial health of the institution to draw lines and that did factor into our decisions for our kids. If the default rate affects the overall financial health of the institution then that would factor into my advice. This is because financial health can impact the programs being offered – some back be cut due to budget shortfalls. If the institution is able to remain financially healthy with a higher default rate it would not factor. If a student I was counseling was using loans I would counsel to not default on them!

3 Likes
25

Not sure how I missed UNC-Charlotte, but it’s now been added (in bold) below, as I can’t edit my original post.

School State 2019 2018 2017
UNC-Chapel Hill NC 0.9% 1.4% 1.9%
North Carolina State NC 1.0% 2.0% 2.9%
Appalachian State NC 1.3% 3.6% 3.9%
UNC-Wilmington NC 1.6% 3.1% 3.9%
UNC School of the Arts NC 1.2% 3.4% 4.1%
East Carolina NC 1.7% 5.1% 5.1%
UNC-Greensboro NC 2.7% 5.3% 5.2%
UNC-Asheville NC 2.7% 5.8% 5.7%
UNC-Charlotte NC 1.7% 4.4% 5.9%
Western Carolina NC 2.3% 5.2% 6.5%
North Carolina Central NC 4.1% 8.2% 6.9%
North Carolina A&T NC 4.8% 9.1% 9.5%
UNC-Pembroke NC 5.4% 10.2% 9.5%
Fayetteville State NC 6.0% 11.6% 11.9%
Elizabeth City State NC 5.5% 12.9% 12.8%
Winston-Salem State NC 5.2% 11.2% 13.6%

I appreciate everyone’s thoughtful comments. I’m going ruminate more on people’s ideas and dig into some of the data that @digitaldad shared (though it’s unfortunate they don’t have the same info for undergrad programs). And I will say that I was surprised that only 14% of people have a graduate degree. Definitely shows how skewed our social circles can be (though being in the education field, I acknowledge that my environment unsurprisingly skews towards more education).

2 Likes
26

I live in one of the most highly educated places in the country and generally feel undereducated with only a BA.

Interesting to see the big difference between 2017 and 2019 for UNC-Charlotte. All getting a lot better across the board.

1 Like
27

I suspect the loan deferments that were put in place during Covid are the reason why there are fewer students/grads defaulting on their loans, rather than things universally getting better at all the schools.

3 Likes
28

Some of those programs in the linked article are truly predatory—and in some cases aren’t even really offered by the college whose name they’re using, but are actually offered by a third-party contractor (technically: online program managers, or OPMs), and nearly always a for-profit one.

So yeah, they really don’t care about outcomes as long as they make their money, especially since they aren’t even sullying their own name if students don’t succeed.

29

The WA numbers pretty much correlate straight down the line for all the reasons already mentioned. I live in Eastern WA which is rural with a much lower SES and know people who attend or did attend most of these schools. Many people never finish at the schools with the higher default numbers - esp if those who head there right after high school because they don’t know what else to do and/or are under family pressure to attend college. And yes, the numbers also correlate with things like professionalism and hustle in the workplace and community later on.

Also for UW-Bothell it’s in a very high SES area and has a reputation for having students who couldn’t get into Seattle (which is also super close by in location). So I wouldn’t be surprised if it’s numbers were very close to UW Seattle. Also, in state tuition is pretty reasonable and if you are commuting from home for either school it’s possible not as many loans are even taken?

2 Likes
30

Yup. If you’re living at home, and attending UW-T or UW-B as an in-state student, you only really have to pony up about $14k/yr (tuition & books). In particular Tacoma doesn’t have an “on-campus food service”, so even if you stay in their “dorm” (an apartment building exclusively populated by students), you don’t have to pay an exorbitant board fee, and the school has a “food pantry”, as they’ve found some significant portion of their students need help supplementing their food budgets. Students can “shop” (get food, not pay, cuz it’s a food pantry for indigent students) there once a week and reduce their overall costs. I don’t know if the same info holds true for Bothell (I have a kid attending Tacoma…)

Also, UW-Tacoma seems to have about 50% non-traditional students, who may be taking classes as they can afford to, cashflowing the whole thing, not taking loans, one term at a time until they’ve cobbled together a degree.

I do know that Edmonds in the WA list is a community college – so, as others have said, about the bottom half of the WA list is community colleges that have recently added some 4-yr degrees.

2 Likes
31

As mentioned above, and due to my own experience in public education, I know that certain factors are more of an indicator of the financial background of students and not so much what we think/hope it might be measuring. For instance, most analyses show that there is a very direct correlation between standardized test scores and family incomes. However, when looking at students’ growth on standardized tests (at least our state standardized tests), there is no correlation with family income. So a student can grow a lot or regress a lot, and there is no correlation to their income. Basically, it’s a way to assess the effectiveness of the instruction without the very real impact of family income coming into play.

As it appears that family income is likely a very big factor in the default rates (particularly since the default rates are not restricted to those who received their degree), I’ve been thinking about other possible ways to measure an institution’s effectiveness that (hopefully) is not as closely linked to income. The Wall Street Journal released its college rankings recently and one of the factors it included related to expected graduation rates (see thread). Although I am unable to read the WSJ’s details (not a subscriber), the Washington Monthly also produces college rankings AND it allows the data to be downloaded. Its 2023 guide & rankings are out, and the methodology section explains the graduation performance rank, which I partially use below.

Half of that score was determined by the reported graduation rate and the other half came from comparing the reported graduation rate to a predicted graduation rate based on the percentage of Pell recipients, the percentage of students receiving student loans, the admit rate, the racial/ethnic and gender makeup of the student body, the number of students (overall and full-time), and whether a college is primarily residential. We estimated this predicted graduation rate measure in a regression model separately for each classification using average data from the last three years, imputing for missing data when necessary. Colleges with graduation rates that are higher than the “average” college with similar stats score better than colleges that match or, worse, undershoot the mark.

Washington Monthly divided its lists by the categories National, Liberal Arts, Bachelor’s, and Master’s, so the rankings don’t necessarily correspond well to one another within a state when the school is in a different category. Thus, for the state comparisons, I’ve calculated the difference between the actual graduation rate and the predicted graduation rate and sorted them from biggest (positive) difference to lowest. The rank data for the category is still included.

As you look through the data (or look at the general data on WM’s website, or your state’s data), I’d love for people to think about the same questions as before, but substituting the the graduation rate performance gap metric for the default rate:

  • Do you see a relationship between the school’s graduation rate performance gap and the school’s reputation for quality in the state?
  • In your experiences of dealing with the alumni from the various universities, do you see a relationship between how well you thought of them (work/thought quality) as compared to the graduation rate performance gaps?
  • Basically, do the graduation rate performance gaps surprise you? Confirm what you’ve seen in real life? Or do you think the graduation rate performance gaps have no correlation to the quality of education received?
  • If you were to use a graduation rate performance gap as a filter on schools, where would you draw the line when considering schools for anyone you’d give guidance to…-2% -5%? -10%? -15%? Doesn’t matter?
  • NEW: When looking at graduation rate performance gaps, is there a percentage where you would say, “Wow, I really need to look into that school,” that you wouldn’t have considered before? (+5%, +10%, +15%, doesn’t matter)

Without further ado, here are the state-by-state tables for Louisiana, Massachusetts, Missouri, New Mexico, North Carolina, and Washington. The tables don’t align perfectly with the ones from #1 as some schools (probably the community colleges in WA were excluded by WM).

LOUISIANA

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
105 Grambling State LA 41% 34% Master’s 7%
155 Southern A&M LA 39% 37% National 2%
216 LSU-Baton Rouge LA 68% 68% National 0%
338 Southern U.-New Orleans LA 34% 35% Master’s -1%
368 Northwestern State U. of LA LA 41% 43% Master’s -2%
322 U. of Louisiana-Monroe LA 48% 51% National -3%
168 LSU-Alexandria LA 31% 36% Bachelor’s -5%
371 U. of Louisiana-Lafayette LA 52% 58% National -6%
472 Nicholls State LA 47% 53% Master’s -6%
466 McNeese State LA 47% 53% Master’s -6%
485 Southeastern Louisiana U. LA 42% 49% Master’s -7%
423 Louisiana Tech LA 52% 63% National -11%
420 U. of New Orleans LA 43% 54% National -11%
571 LSU-Shreveport LA 35% 48% Master’s -13%

MASSACHUSETTS

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
16 MA Maritime Academy MA 79% 62% Master’s 17%
95 Westfield State MA 65% 57% Master’s 8%
111 Bridgewater State MA 62% 55% Master’s 7%
26 Mass. College of Liberal Arts MA 60% 54% Liberal Arts 6%
154 Salem State MA 62% 57% Master’s 5%
181 Fitchburg State MA 59% 55% Master’s 4%
151 U. Mass-Amherst MA 79% 76% National 3%
240 Framingham State MA 60% 58% Master’s 2%
314 Worcester State MA 60% 61% Master’s -1%
270 U. Mass-Dartmouth MA 55% 57% National -2%
328 U. Mass-Boston MA 57% 61% National -4%
357 U. Mass-Lowell MA 61% 66% National -5%

MISSOURI

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
69 U. of Missouri-St. Louis MO 58% 51% National 7%
187 Truman State MO 72% 68% Master’s 4%
199 MO Univ. of Science & Tech. MO 70% 69% National 1%
180 Missouri State-Springfield MO 57% 56% National 1%
329 U. of Central Missouri MO 54% 54% Master’s 0%
257 U. of Missouri-Columbia MO 69% 71% National -2%
354 U. of Missouri-Kansas City MO 53% 57% National -4%
458 Missouri Southern State MO 40% 45% Master’s -5%
440 Southeast Missouri State MO 49% 54% Master’s -5%
471 Northwest Missouri State MO 52% 58% Master’s -6%
195 Lincoln MO 27% 35% Bachelor’s -8%
212 Harris-Stowe MO 18% 28% Bachelor’s -10%
577 Missouri Western State MO 34% 48% Master’s -14%

NEW MEXICO

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
284 New Mexico Highlands NM 41% 40% Master’s 1%
198 U. of New Mexico NM 52% 52% National 0%
450 Eastern New Mexico NM 45% 50% Master’s -5%
408 New Mexico State NM 53% 61% National -8%
529 Western New Mexico NM 34% 43% Master’s -9%
558 New Mexico Inst. of Mining & Tech NM 55% 66% Master’s -11%

NORTH CAROLINA

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
1 Winston-Salem State NC 63% 39% National 24%
17 East Carolina NC 67% 56% National 11%
56 UNC-Pembroke NC 46% 36% Master’s 10%
35 Elizabeth City State NC 44% 35% Bachelor’s 9%
73 Appalachian State NC 74% 65% Master’s 9%
87 North Carolina Central NC 49% 41% Master’s 8%
74 UNC-Chapel Hill NC 90% 84% National 6%
127 Fayetteville State NC 44% 38% Master’s 6%
111 Western Carolina NC 65% 60% National 5%
86 North Carolina A&T NC 51% 46% National 5%
102 UNC-Wilmington NC 72% 67% National 5%
122 UNC-Greensboro NC 58% 54% National 4%
146 North Carolina State NC 81% 78% National 3%
308 UNC-Charlotte NC 61% 63% National -2%
158 UNC-Asheville NC 63% 67% Liberal Arts -4%

WASHINGTON

Graduation Rate Performance Rank School State 8-year graduation rate Predicted Graduation Rate Category Difference Between Actual & Expected Graduation Rate
24 U. of Washington-Tacoma WA 75% 60% Master’s 15%
28 Evergreen State WA 66% 52% Master’s 14%
96 Western Washington WA 73% 66% Master’s 7%
99 Central Washington WA 63% 56% Master’s 7%
148 U. of Washington-Bothell WA 78% 73% Master’s 5%
121 U. of Washington-Seattle WA 84% 80% National 4%
175 Eastern Washington WA 58% 54% Master’s 4%
177 Washington State WA 65% 63% National 2%
1 Like
32

No use of incoming student academic stats like HS or prior college GPA to get predicted graduation rates?

2 Likes
33

Of the states that had previously racially segregated colleges, in LA and NC, the HBCUs (and the historically Native American university in NC) cluster near the top of the graduation rate performance rankings, while the HBCUs in MO are near the bottom.

1 Like
34

The HBCUs in NC and LA do appear to be doing quite well on this measure.

I found it interesting that state flagships were generally not the highest performers in their states, but those are generally (with some exceptions) the only publics that many people on CC will consider. Perhaps the non-R1 schools focus more on teaching and the undergraduate experience, getting their students further, or at least as far, as the students’ backgrounds might indicate?

Pinging @Shelby_Balik who might also have some thoughts on this topic.

2 Likes
35

Well, this was interesting, looking at Washington.

I thought it was nice that ALL the schools listed exceeded the predicted graduation rate, so yay Washington.

I was very surprised that Washington State was at the bottom statewide. It’s my impression that UW, Washington State, and Western Washington have the best local reputations (UW on top). Central Washington, Eastern Washington, and the UW Bothell and UW Tacoma schools are generally seen as a tier down.

It was interesting to see the variation among schools with reputations in the same category. It made me wonder how stable those numbers are and how much they change year to year.

I think Evergreen is an outlier because it doesn’t give traditional grades and has fewer barriers to graduation than a typical school, arguably not always for the best.

1 Like
36

Since the predicted graduation rates used here do not seem to account for incoming student academic stats, perhaps that means that states like WA and NC with relatively more selective state universities will find the predicted graduation rates easier to beat? In contrast, where the state universities are less selective in LA and NM, most of the state universities fall short of the predicted graduation rates.

1 Like
37

I dunno. Only UW is particularly selective. All the others have acceptance rates north of 80% and I think Western Washington and Eastern Washington admit 96 or 97%.

38

It has been a long time since my econometrics classes, but since one of the regression variables is admit rate, shouldn’t that account for selectivity? Unless you’re saying admit rate and selectivity are noticeably different variables? (or, I am missing something probably very obvious about the regression model. Like I said, it’s been along time since I was doing this stuff regularly!)

1 Like
39

Hmm. Other than spending a lot of time in colleges and universities, I don’t have a whole lot of background in the study of higher ed. But I have some ideas.

First of all, for universities with lower than expected graduation rates, it doesn’t mean students are dropping out – they might be doing that, or they might be transferring to other schools (less expensive schools, or maybe regional schools that specialize in a desired field, or are closer to home?) or taking a break and finishing later. So it’s hard to know why actual graduation rates are lower than expected. We would expect flagships to do better – students are wealthier, more academically prepared, probably come with stronger support systems that got them to these institutions in the first place. But it’s also easier to get lost in the shuffle, and professors at R1s may not prioritize teaching as much. That’s a gross generalization – many of them are wonderful teachers and dedicated mentors, and they do have lower teaching loads, which theoretically means more time for fewer students. But they also have grad students and high research expectations, and they might have TAs in large classes (and so don’t get to know the undergrads as well). Their attention is divided.

At more regional or local universities, which are more teaching-intensive, professors teach more classes but have to meet lower research expectations. Sometimes that translates into closer relationships with students, but not always, because teaching obligations can be overwhelming, there are probably more online offerings (with less student contact), and these universities also rely a lot on adjuncts who have a less stable presence because their jobs are more precarious and they might be teaching at multiple institutions in a given semester. The quality of advising at these universities is no better or worse on average than in larger flagships, but the most successful universities in this tier might have stronger and more deliberate student-retention programs in place, because their students need more support – that allows them to beat the odds with graduation rates. Also, at these lower tier universities, students tend to be more likely to be first-gen, they might be living closer to the margins financially, with less of a safety net and support system, and also less preparation for college. They might go in and out of school as life circumstances allow. So a school that exceeds low expectations is not necessarily doing better than a school that’s underperforming high expectations – it’s more about where the bar is placed to begin with.

The high performance of HBCUs doesn’t surprise me. I think that those school communities – families, students, staff, and faculty – share a sense of purpose and commitment that might not be as common at non-HCBUs, which would probably help contribute to higher retention and graduation rates.

3 Likes
40

After looking over the Washington Monthly rankings a little more, it appears the schools that rank highest are frequently ones that draw large numbers of commuter and/or nontraditional students — and not 4-year schools operating on a traditional model with residential requirement for first-years, etc. At least I think I see a pattern.

Despite studies that show living in dorms, etc., is the best way to integrate students and increase retention, I wonder if that is actually true for the lowest income students. Those students may find a greater degree of success in schools that allow them to be more in the driver’s seat regarding room and board costs, many evening and online classes they can fit around significant work and family commitments, etc. In other words, maybe it is less about the education provided and more about the kind of environment where different types of students flourish?