2/23/2012: I’ve had requests for Michael Raymer’s slides from the union forum. He sent them along, 4 pages, here, with this explanation:
My Conclusions from the Data:
There is an undeniable inverse correlation between tenure-track faculty (TTF) unionization and average school quality. (Presence of union correlates to lower ranking.)
Of course, causality is not indicated. Does low quality lead to unions, or do unions lead to low quality, or neither?
The question we should ask ourselves:
If UO wants to strive to increase its quality, should it associate itself with the top schools, the vast majority of which are not TTF unionized? Or with the lower-tier schools, where nearly all of the unions exist?
I’m happy to post more figures/links/data if people send them to me. Slide 1:
My Conclusions from the Data:
There is an undeniable inverse correlation between tenure-track faculty (TTF) unionization and average school quality. (Presence of union correlates to lower ranking.)
Of course, causality is not indicated. Does low quality lead to unions, or do unions lead to low quality, or neither?
The question we should ask ourselves:
If UO wants to strive to increase its quality, should it associate itself with the top schools, the vast majority of which are not TTF unionized? Or with the lower-tier schools, where nearly all of the unions exist?
I’m pretty skeptical of unionization, but in the name of intellectual rigor I have to say I don’t find these particular data very convincing. There are lots of 3rd variables to consider, some of which are already pretty well known. For example, private universities are blocked from having faculty unions (NLRB v. Yeshiva), and privates have a number of advantages that allow them to do better on rankings. I realize private vs. public doesn’t account for everything on this figure, but it’s a pretty potent example of how such correlations might be undiagnostic. (And by similar logic, we want to “associate with” top-tier schools, shouldn’t we abandon our public mission and go private?)
Note that every point on the horizontal axis is one school, whether called out by a name box or not. So this is a large data set, that is, every school (318) ranked or listed unranked by US News and World Report. You can find their complete rank list easily online (although it is not searchable).
Dog says
I am not sure this plot does anything more than illustrate what we already know;
The UO does not place well on the X-axis by this agency. I seriously doubt
that if Harvard got unionized, they would migrate towards larger numbers on
the X-axis. The only thing I would conclude that if X > 50, the higher the probability that a union forms but the existence of a union, I don’t believe,
determines your position on the X-axis.
If the UO wants to increase its quality then it needs to
a) get more graduate students
b) build better classrooms
c) have smaller classes
d) start moving toward more interdisciplinary undergrad degree programs
e) greatly improve its faculty-to-student ratio (no matter how you count it)
if the union can facilitate these 5 points, that would be great.
This comparison also fails to take into account the fact that tenure-related faculty at private colleges and universities cannot unionize, whether they want to or not, thanks to the Supreme Court’s decision in NLRB v. Yeshiva (1980). So let’s compare institutions that are comparable, and exclude Harvard, Northwestern, and Duke from the listing.
I’ll agree with the importance of the points mentioned, although I think “d” is underway both inside established departments and in new programs. (Some of those are really vehicles for pulling tuition money away from established faculty units, but I digress.) A union could present faculty demands on the other issues but their remit really covers faculty pay and working conditions so you’d have to couch it in those terms. Student needs, even as we identify them, get sacrificed early and often whether or not there’s a union.
Regarding grad students: we don’t need more of them, we need better funding (higher levels and more fellowship support) so we can get the best grads to actually come here. That will build a better research profile, but that requires investment, and there we have not only administrative resistance but additional problems concerning the need to bring the GTFF in on the discussions. Right now there’s plenty of push to get more grad students, but the intent is (speaking from English here) to use them as cheap creators of tuition dollars (the WR course sequence) and work them like adjuncts to the very limit of the GTFF contract. At the same time so many grads come in, without any concomitant increase in quality via University investment in faculty and resources, that the value of the degree on the job market rapidly falls.
dog says
given that our grad student participation is down at 13%, way below
the Carnegie I research university threshold of 20% and well below
the average of 30% for top flight research universities (e.g. UW,
Michigan, etc) we do actually need more and having more fellowhip
support would help.
Disqualification of Carnegie I status actually precludes the UO from
participating in certain grant programs.
Point taken–I’d been thinking more in the local backyard and not across the UO as I ought. But dog is as so often the case right–we need the funds to do either. If we had that, we could have more grads AND effectively better support. But it will be harder to expand numbers before quality I suspect, and both may turn on institutional support PERIOD.
The idea that UO should “associate itself with the top schools” by *refraining* from some action — in this case, establishing a union — makes no sense in the absence of any credible proposals to improve our quality. Now that the New Partnership is dead, there are no such proposals, nothing to get Dog’s 5 points done, or anything else for that matter.
Wow – finally some hard data. I’ll pass on the union.
Really hard data? I’m kind of amazed that intellectuals accept data that claims no causality as hard proof. Very scientific!
I think this correlation analysis is “stupid statistics.” Yes, there MIGHT be a causation. But there was no evidence presented of causation, just a correlation. And the data set is tiny. This is like graphing which excellent Universities have more than 50% out of state students. There can be a dozen reasons for why a university has more or fewer students from out of state. Just off the top of my head there is marketing, location, state policies on enrollment, AND excellent academic programs that weigh into whether a non-local student attends a University.
If one track down when the schools unionized, and data from the chronicle, I will construct credible counterfactual estimates of the effect of unionization on salary profiles.
This would be good to see. Three dates I quickly found:
University of Vermont: 2001, http://www.unitedacademics.org/inthenews.php?id=6
University of Delaware: 1972, http://www.udel.edu/aaup/newnews/1002.html
Rutgers: 1970: http://www.rutgersaaup.org/about.htm
Only Vermont is recent enough to be worthwhile to look at. I have no idea how to easily find ranking information for a range of years, but perhaps someone here does.
First off, I’m suspicious of anything published by USN&WR. Usually simplified for their lower-end
readers. More data is needed, like faculty/student ratios, funding per student, graduate programs
and enrollment, sponsored research funding, faculty salaries, faculty support, dept rankings, entrance
standards, facilities, etc. I suspect that several of the schools on the far right are lacking in these
key metrics. In other words, poorly supported. And years of poor support led to a faculty union. Sort
of a last gasp by those faculty that didn’t bail out after all else failed to improve the situation over
the years. And thus we can see UO’s future in that graph, unless something major happens to reverse
course. Look at it this way: If funding and governance are good, faculty and staff treated right,
including above average pay and benefits, then who would want to unionize? I sure wouldn’t. But at UO
we’ve heard the same old line for over 30-years about the administration improving salaries. Instead,
we’ve seen the retirement benefits head south and the pay remains below par while the JH people
reward themselves with monthly beamer payments, UO paid family vacations to away football games,
retirement golden parachutes, extraordinary raises, etc. Not to mention all the shenanigans related
to Athletics.
After granting that the data shows nothing at all about causation, Professor Raymer suggests that UO can “improve its quality” by “associat[ing] itself with the top schools” by not unionizing. This smuggles an assumption about causation back in with no evidence whatsoever. The tacit assumption that United Academics doesn’t care about quality is entirely unjustified.
The data in the graph are questionable as well. The U.S. News rankings are pretty dubious, and it’s worth noting that since the Yeshiva University Supreme Court decision of 1980, most major private university full-time faculty don’t have the legal right to organize and bargain collectively. Moreover, many states severely restrict collective bargaining for public employees, so many state university faculties don’t have the liberty to choose collective bargaining.
I hope our faculty recognizes the advantages of collective bargaining, an enhanced voice in Salem and an added stakeholder in determining the direction of the University of Oregon and in enhancing its quality.
Rereading the initial post here, it appears that “anonymous,” not Michael Raymer, is the author of the comments about “associating ourselves” with higher-ranked schools by rejecting collective bargaining. As the author of the above post, my apologies to Professor Raymer for associating him with the questionable logic in the remarks under his slide.
Actually, I think you were correct the first time. I did post the comment both above and below the slide, referring as “My Conclusions from the Data.” I stated it as a rhetorical question, “If UO wants to strive to increase its quality, should it associate itself with the top schools, the vast majority of which are not TTF unionized? Or with the lower-tier schools, where nearly all of the unions exist?” I wanted to leave it open to more careful analysis, as posters here have done admirably. But, knowing where I stand on the issue (opposed), one can guess how I answer the question.
Note that if you click on the link, you get all four slides, which contain much more information than the cropped slide above.
As someone who has first-hand experience of Rutgers’ AAUP-AFT chapter, the union may be one of the few reasons why the university has remained as highly ranked as it is. The union served as a bulwark against administration attempts to sidestep faculty governance and faculty voice against the move to ever greater numbers of adjunct faculty, assured that grad students (the only unit where they are part of the same union as faculty) got raises to a living wage, and generally fought against administrative fiat. It wasn’t perfect–administrators still found ways around it, and have done incredible damage to the research mission of the university by attacking whatever funding wasn’t shielded from them–but it could have been so very, very much worse. It may yet go south, if Chris Christie gets his way, but union membership did pretty well by us.
I agree that “USN&WR Rankings” are suspect as a measure. Even if one accepts them as useful, you have to define which ranking you’re talking about. It varies from program to program. As for the public vs. private issue, that’s also a valid one to consider; they’re not under the same strictures regarding unionization. What is it about a union that people think will somehow bar faculty from doing good research, from getting grants, and from defending standards for teaching and research? The only argument with any water that I’ve heard is that it will somehow “poison our working relationship with the administration and the OUS.” But we’re already under rule by fiat and open mismanagement from that angle, without any meaningful voice to oppose it, so the alternative is apparently to ignore the poisoning already going on. Unionization won’t magically fix everything (it still takes work, after all) but if it gives just a little pause to the Pernsteiners of the world, I’ll pay those dues and happily so.
Yikes, so, this is what happens when you have a research university without a statistics department.
Statistical Dog says
The RaymerGram is a qualitative visual representation of how schools
separate in the USN&WR Rankings – that’s it. There is nothing there you can
do statistics on. I do plan to make some diagrams that you can do statistics on.
Isn’t the point that we’re looking for association between predictors and school quality? The whole sad joke of the Raymer/Keyes diagram is picking a variable at random to display correlation rather than building a model from an honest analysis over a rich data set.
Once you start associating with statisticians, your whole world view changes.
(and god forbid you ever use the word random incorrectly!)