From the important NYT report here. Read it all.
Yes, the universities that educate students from modest backgrounds face big challenges, particularly state budget cuts. But many of them are performing much better than their new stereotype suggests. They remain deeply impressive institutions that continue to push many Americans into the middle class and beyond — many more, in fact, than elite colleges that receive far more attention.
Where does this optimistic conclusion come from? The most comprehensive study of college graduates yet conducted, based on millions of anonymous tax filings and financial-aid records. Published Wednesday, the study tracked students from nearly every college in the country (including those who failed to graduate), measuring their earnings years after they left campus. The paper is the latest in a burst of economic research made possible by the availability of huge data sets and powerful computers.
… Obviously, colleges don’t deserve all the credit for their graduates’ success. But they do deserve a healthy portion of it. Other research that has tried to tease out the actual effects of higher education finds them to be large. And they’re not limited to money: Graduates are also happier and healthier. No wonder that virtually all affluent children go to college, and nearly all graduate.
The question is how to enable more working-class students to do so. “It’s really the way democracy regenerates itself,” said Ted Mitchell, Obama’s under secretary of education. The new research shows that plenty of successful models exist, yet many of them are struggling to maintain the status quo, let alone grow. It’s true in red states as well as in many blue and purple states, and it’s a grave mistake.
This figure shows student incomes at age 34 as a function of family income. Our students average about $45K at that age. But they come in from wealthy families. All the dots to the left of UO are schools that were able to take lower income students than UO and produce the same outcomes for their graduates as we do for our rich kids.
What can we do to fix this?
Simplest fix: don’t take half our students from out of state, paying out of dtate tuition jacked up as high as the market will bear, from well to do families.
Of course, this would wreck UO finances.
while there are likely many ways to interpret the widely scattered data on this graph but clearly, the UO is significantly underrepresented, for whatever reason, in students with median family income < 100 K.
Note also that the boxes (grid lines) are 50K x 20K
What this plot shows me is simply that public schools
produce people who at age 34 have an income of 35-55 K – no surprise here.
I think the UO would like to be one of those few yellow dot schools in the box 50-100 x 60 -80
Given the density of the dots below UO, I read that scatter graph as indicating that at $45K UO is on the upper third of graduate salaries for public schools. That seems cause for celebration, not alarm.
I also read that graph as indicating that the median family income for UO graduates is around $125,000. While $45K may seem low compared to $125K, its an apples to oranges comparisons. Can’t really conclude our graduates are doing worse than their parents when household income likely includes two earners (as compared to the single figure for the graduate), and there is no information on locales (Californians earn more than Oregonians and Washingtonians, and I suspect UO moves a fair number of people who grew up in California to the NW).
My eyeball linear regression says that UO is way below the line for all schools and somewhat less below the line for all publics. UO is doing worse than average given what it has to work with.
Not sure what line you are drawing to get to the conclusion that UO is below average. There is a HUGE concentration of dots below the $45K mark where UO sits. So big a concentration they become a mass. You have to look at the concentration of dots, not the range. UO is just above that mass, where the dots can be seen individually. Seems like you are looking at the median (mid-point of the data) not the mean (average). (And, Yes, I cop to the fact I used the word “median” above in discussing family income when it should have been “mean” or “average,” so this is not a spelling, er, statistics flame.)
The scatterplot shows that there is a very strong and almost linear relationship between median family income and median graduate income, something we already know from about a million other sources. If you draw a best-approximation line through the mass, you get a sort of average prediction for what you would expect a college-educated student to be earning given their family income. Schools that improve on this lie above this line, and schools that fall short lie below it.
At $43K or $44K ($45K seems too high to me), UO falls short. The slope of the best-approximation line through the publics is a little shallower than the line through all schools, but UO grads should still be earning a median of $52K or $53K. This is more than the 15% difference in the average cost of living between California and Oregon. So even in the most generous scenario, UO isn’t making the grade.
Here’s another way to think of it: UO grads might be earning the the top third (your horizontal line), but their families are earning in the top tenth or higher (a vertical line). Lousy ROI if you ask me.
Except, anecdotally, I know a number of UO’s California students come from the SF Bay Area (around here UO is sometimes referred to as “U.C. Eugene”), and in the SF Bay Area salaries (and housing costs) are inflated over Oregon norms by well over 15%. Try more like a 150%+. So the scatterplot is meaningless because it fails to correct for the locales of the student’s “family income.” I would like to see an apples to apples comparison before I can declare the conclusions being drawn here as valid.
I suspect the very first poster on this thread got it right.
can’t see to find the data for that graph
would like to know where UW, PSU, and UC Berkely
are on this graph
I wonder if we could expand beyond a mere correlation and consider other factors as well, specifically the role of level of public support. For example, the UO, as we all know, gets among the smallest level of state support of any R1 public school, expressed as a percentage of the total budget. So compared to many peers on that graph we are much more reliant on tuition to fund everything, which means we must rely on those able to pay lots in tuition, which means quite a few wealthier students from California.
If Salem and the people of the state stepped up to properly support our school, in the way that Washington supports UW, for example, we could charge less and admit more modest and lower income folks.
It is not necessary to always run down our school. I suppose concerned tenured professors could give back some of their salary or Tier 1 PERS to help poor students attend the UO.
Sure–I’m just calling balls and strikes. The numbers say UO is underperforming by this one measure. There are lots of contributing factors, not all of which are the fault of the way people do their job at UO.
I don’t buy the explanation that high Bay Area incomes are a major reason for pushing UO to the right edge of the distribution. Cost of living is much higher, but incomes for the same kinds of jobs are not. Even if every single CA student at UO came from the Bay Area and then got a job in Portland, it still wouldn’t account for where UO is on the graph.
who cares what the reasons are for pushing the UO to the right edge of the distribution – the point is, that is where they are. The UO seems to admit students with a significantly higher proportion of high family incomes compared to most other state schools. That should be the ONLY message seen in this diagram.
There are two axes in the diagram. The message isn’t just that UO students are wealthier, but that they’re rather downwardly mobile relative to similarly wealthy students pretty much anywhere else. Why don’t you think that’s important?
I think any importance of the second axis what you get from just looking at the dots on the graph without any context.
Here is some context:
1. Party school graduates may not be interested in a career.
2. Coming from more wealthy families such students may not need to get a “high paying job”.
3. Most UO degreess (e.g. no engineering) do not allow students to get entry level jobs at high salaries.
4. If these students want to say in Oregon, salaries are generally depressed there.
So, in sum, I do not think the position of the UO on the Y (compressed) axis, is meaningful. I also think
the position of the UO (44K) relative to the eyeball regression line (55K) is mostly lost in the noise.
Focused — I’m sure some tenured professors/Tier 1 retirees do give money to UO to help less well-off people. But I think it would be a bit much to expect them to “give back” in this way. Why should they? True, you didn’t quite say that. But the professors are hardly responsible for whatever problems are being caused by PERS — nor is PERS the sole reason why UO is not so affordable to less well-off families.
I was just being flip on that bit Uncle. I am sure many profs do help and they and they pensions are not really the reason for the shortfall that must be made up for in tuition. Still when older tenured profs drawing upwards of 100k a year and in T1 PERS complain about how the univeristy doesn’t have money to help poor students it is a little awkward. Not really their fault but they sure could be more sensitive to the optics for students and those of us toiling in a new school compensation environment.
I see two big contributing factors.
1. We have a ton of students come from rich families out of state, but they are also were not academically strong enough (usually) to get into one of the best UC’s. So we are selecting students based solely on ability who will on average be lowerly mobile.
2. The best departments at UO are in sciences. The sciences, generally do not pay well. This includes undergrads, masters, and Ph.D.s. The professional schools have not be investing in enough (because the salaries look to crazy to be competitive in recruiting), and we lack any major presence in engineering.
We could partially overcome this by focusing on computer sciences, data science and machine learning. That’s the most exciting think happening in the skilled labor market right now. But it would cost of a lot of money.
The big news from the study is how universities serve the rich instead of almost everyone else. The UO is a particularly bad offender and getting worse. It ranks in the bottom 10 percent on almost every measure.
https://www.nytimes.com/interactive/projects/college-mobility/university-of-oregon
There is, as is noted above, a pretty decent line that describes this apparent correlation between a university’s median family income and its’ graduates income. It’s interesting to think, however, about the many factors that can drive this. One that pops to mind is majors available. At UO, for example, there are no engineering majors; fields which dominate the lists of highest-earning degrees (nine of the top ten, from what I recall). As examples, from the fun graphs at http://www.hamiltonproject.org/charts/career_earnings_by_college_major/ , electrical engineering, mechanical engineering, business administration, and psychology majors on average earn $85, 83, 56, and 45k, respectively, 12 years after the “start of their career,” which I’d guess corresponds roughly to age 34. At a school like Berkeley, approximately 10% of students are in the college of engineering. If we replaced 10% of our student body with ones earning $30k higher, it would shift us noticeably on the graph. (Not quite to the regression line, I’d guess, but there are of course other demographic factors as well.)
None of this actually matters, though, even to the point of the article. The more interesting graphs would be, for each school, income vs. family income, and then a plot of the regression parameters for each school.
Another exercise just for kicks: of the majors available at UO, which are the highest earning? Playing around with the above site briefly, my guess would be Physics, but I might be biased…
It possible that most “party schools” let rich kids in that don’t give a shit about a career. That profile would place schools in the part of the diagram occupied by the UO.
It’s possible. I don’t see many Econ majors that fall into that category.