Friday, July 05, 2013

On the business side of doctoral education, where is it heading?

One has the feeling that we are descending backwards into a Darwinian approach.  When I started as a doctoral student, back in 1976, that class had 27 students.  I hadn't realized it at the time of entry, but I soon learned that people came to the program under different funding packages.  I was on fellowship.  Some of my classmates were not.  I had applied to eight programs in Economics.  Among the places where I was accepted, three had given me money.  Northwestern and Rochester each gave fellowships.  Minnesota bundled its money with some work requirement.  I had my heart set on Berkeley, but I didn't get funding from them.  Ditto for Princeton.  So, given the realistic alternatives, I ended up at Northwestern.  Robert Gordon had written a detailed "come on" letter to those students who were granted admission.  Much later I learned that letter was controversial among the rest of the faculty.  It probably had an impact on my choice, though from a geography point of view, I wanted to get away for New York.  That mattered too. 

I was ignorant of many things when I started that PhD program.  Principal among them was whether I was committed to following through all the way to the doctorate.  I had so little economics as an undergrad that I didn't have an informed basis to make such a commitment.  As it turned out those of my classmates who had been economics majors were rudely surprised by how far the doctoral program was from their expectations.  I told myself at the time that I would go all out for one quarter and in that time not ask at all whether I liked it or not.  Just do it for a while with full commitment.  Only then come to some decision about whether to continue or not. When it was time I realized I was completely hooked, though there were aspects of the program I didn't like.  It was brutal.   My first classmate to drop out of the program did so soon after the second quarter had begun.  By the end of that first year 14 students had left the program, including some who were pretty good students.  And by being so focused on the economics, the life was not intellectual in the way my last two years at Cornell had been.  I had no far ranging discussions with my classmates that first year.  Indeed, as an undergrad I could have those discussions because we were all studying different things.  In that first year in grad school, immersed in this common experience, our conversations were rather narrow.

I also hadn't realized that my fellowship wasn't guaranteed for four years.  It had to be renewed each year and there were implicit performance requirements that needed to be satisfied to justify renewal.  Somewhere in the middle of that first year we learned that fellowship money for the second year was going to be scarcer than it had been - the overall economy in the throes of "Stagflation" and doctoral programs not immune from the consequences.  Those who were not on fellowship were competing for this scarce funding as well.  It meant the competition among the students was rather cutthroat.  You'd think this would manifest via performance in class participation and on exams.  Mainly it did.  But there were other pernicious consequences.  During the first quarter most of our readings were in the Reserve Room, but in the second quarter we had to go to the stacks and find the bound periodicals where the readings were.  Some classmate(s) looking to get a leg up would get to the article first and then re-shelve the thing in an improper location, so it was harder to find.  On one or two occasions that I recall, the article had been ripped out of the bound periodical entirely.  This was bush league stuff and left a bad taste for me and the rest of my classmates.  Fairness is not part of survival of the fittest.

I really don't know whether this sort of attrition was part of the Northwestern model, to be planned for in advance, or if it was an aberration for our particular class.  I believe the next class was much smaller at the outset, but my memory of this is quite hazy.  I do know that the Economics Program at the University of Chicago had a reputation for using the first year to weed out the mediocre students, admitting as many as 70 students for the first year, with many fewer students moving on after taking the qualifying exams (at Northwestern administered at the start of the second year).  The idea of the first year of graduate education as a lottery and hence with some brutal aspects to it seems inhuman.  The tone is captured in the movie The Paper Chase.   One wonders whether it is necessary.  I know that experience had a profound impact on my teaching.  I vowed never to call on students by name and thereby embarrass some of them who couldn't come up with a good answer.  Fear motivates.  But it also contributes to the brutality of climate.  I did not want to do that.

After the first year the tone was much more collegial, which I preferred, but I know I stopped pushing myself as hard, particularly in courses that were required but where I didn't see myself heading.  I did put in quite a lot of energy into a Math Analysis class I took during the entire second year.  But in that first quarter of the second year, in Economic History and Econometrics, my effort was much less intense. 

Another thing I didn't anticipate at all was the change in mindset that would happen when the field courses were near conclusion and the dissertation stage would soon begin.  Students are enculturated to worry about their performance in each course and care a lot about their GPA.  But having reached the dissertation stage, all that matters not.  The only thing that matters is the quality of the paper(s) you produce.  There is a big deal question whether the training in required courses, on subject matter, is good preparation for writing the dissertation.  Perhaps it is not.  During the third quarter of my first year I took a course on topics in microeconomics, or something like that, where we read different papers on a variety of themes, many in working paper form and thus not yet published.  Students presented those and critiqued them.  I may have been the only first year student in that class.  From that I got into the habit of attending departmental seminars in economic theory.  Those activities were much more relevant for learning to write a research paper.  So I had some reasonable preparation for the dissertation stage, but I would have benefited from more and many of my classmates who were doing things other than economic theory may have had little to none of this. 

Let me turn to how the TA function was managed while I was a graduate student.  That happened during the second year.  During the first year the entire role was as graduate student.  In the sense that being a TA is a diversion from your studies, a time tax if you will, the idea was that up front you should give your full attention to your graduate studies.  I subscribe to that view even now.  The TA function was deemed part of your graduate education and therefore no money compensation was given for it beyond the fellowship.  When I started in the program the requirement was to TA four quarter courses (typically two recitation sections for two quarters).  Between my first and second year the requirement was raised to six quarter courses.  It is very difficult to see that change as stemming from a graduate education need.  (It is much more readily explained by the need to reconcile the staffing of recitation sections with the paucity of fellowship money at the time.)  There was a committee for graduate student issues in the department and I was the representative for the second year class on that committee.  I complained about the issue there, but to no avail.   After I had become a faculty member at Illinois I learned that this requirement about being a TA had come under investigation by the IRS and was found a form of tax avoidance that is inappropriate.  TAs should be paid.  I have no problem with that but I do believe in the TA function as mainly an apprentice role.  Much of that is learning by doing.  If your first job after getting the degree involves teaching,  particularly on the tenure track, you should have some prior experience.  Too much is riding on the performance of an assistant professor for it to be otherwise.

I did serve as a grader in an intermediate microeconomics class during my third year.  That was a unique situation as Leon Moses had just move back into the department from being director of the Transportation Center and he got a grader for his teaching as a perq associated with the move.  There were stipulations under the fellowship about not working elsewhere.  Somehow that was accommodated and I got paid for grading.  (Subsequently, Leon and I wrote a couple of papers on the economics of inventories, so this association was far more beneficial to me than just getting money.)  Some of my classmates taught during the third and fourth years in a nighttime program.  That was also for pay.  The idea that advanced graduate students teach independent courses has been with us for some time.  It makes some sense, though I'm not particularly enamored with the practice.  I think of the doctoral student near or at the dissertation stage as a monk-like existence, but then it should be funded well enough for that to be the person's universe.  Teaching to generate income is just as much a tax on the person's time as it is during the first year. 

The last part of the process is going on the job market.  There is no doubt that the preparation we had is for getting a research university level position.  But the market also includes teaching level positions, positions as government agencies such as the Fed, and private sector jobs in consulting firms or large financial institutions.  So there are issues of what makes for a good match between candidate and position.  There are also questions about when to go on the market and what factors determine that.  I was only 25 when I started at Illinois, so one might argue that I should have finished my thesis and tried to get some things published before going on the market.  But I was tired of that monk-like existence and wanted to earn a real salary.  And it did seem like decent jobs were available to me, given what progress I had made and how my dissertation committee saw where I was relative to others on the market, determining the quality of the letters of support they wrote. 

Economics has a two stage process where in the first stage you have interviews at the winter meetings of the ASSA.  These are usually about a half an hour in some hotel room.  That year the meetings were in Atlanta and I had in excess of twenty of these interviews.  Based on the interviews the schools decide which candidates should be invited out for a campus visit (or analog for the other types of jobs I mentioned above).  I made four such visits, then got two offers, which helped a tiny bit in the salary negotiation.  Subsequently, I got a couple of phone calls from places I hadn't interviewed with in Atlanta.  In economics jargon, market clearing takes quite a while.  The top of the market clears first, then on down the line.  More recently there has been some gaming of the market where before the interview stage candidates are invited out to give a paper (an unofficial campus visit).  Those departments further down the line want to get a jump on the market, if they can. 

* * * * *

The above is meant to show there are many different pieces of the puzzle needed to fit together to have a sensible business model for doctoral education.  So far, I've left out perhaps the biggest piece of all.   From where does the revenue come to cover the cost of this education? 

At a private institution like Northwestern there are three possible sources:  gifts from donors, grants from the government or outside foundations, and tuition dollars (from undergraduate students) reallocated to support graduate instruction.  These are to cover the faculty salaries, the salaries of support personnel, the funds to bring in outside speakers and to host conferences, any facilities cost and cost of common infrastructure, and other possible costs. 

At a public institution like Illinois there is a fourth possible source, general tax revenue given to the campus.  In relative terms if not absolute terms the contributions from that fourth source has been declining.  Further, health insurance costs for staff have been on the rise, eating away at much of the state contribution.  Moreover, in many fields, economics is an exemplar, salaries have been growing at a rate faster than the increase in inflation, pretty much for the entire time period since I started.  So the underlying question is whether how the puzzle gets solved should remain the same or if it should change fundamentally because of what is happening on the funding/cost side.

Before trying to get at this question, let me make a few other observations.  I like to think of each discipline as its own ecology.  Positions in aggregate are created by retirements that need to be filled or by growth of the entire discipline that generates fundamentally new positions.  Of course there is also churn - people leave one job and find another.  So the volume one sees in the market has both of these factors in play at the same time.  There is also washout of people from jobs well before the retirement phase.  One thinks, in particular, of folks go don't get tenure.  Thus the severance is involuntary from their point of view.  To the extent that they stay within the sector, but end up at less prestigious places, this is really no different for this observation than the people who get bid away to a job they prefer.  But some of these folks may wash out of the sector entirely pretty much in the same manner as my classmates in the first year of graduate school washed out.  If you ask what business model makes sense, the welfare of those who wash out should be taken into account.

That said, there is a question of whether the discipline does a reasonable job of equating supply and demand in the market.  Note that each graduate program decides on its own how large its doctoral program should be.  There is no central coordination to determine this.  In most markets we think of the "invisible hand" as guiding the market toward equilibrium.  Is something akin to that at work here as well? 

There is the potential for moral hazard to create persistent excess supply.  One source of this is the faculty preference to teach graduate students, given a fixed teaching load, and to have graduate students around to provide assistance with their research.  Note that this preference exists pretty much the same whether the institution is public or private. 

Another source of moral hazard is particularly at play at public institutions, in departments that have large general education offerings or that have many students in the major with large class early offerings there.  Then there is a need for a lot of TAs or some alternative way to staff these large undergraduate classes.  The undergraduate teaching needs then may conflict with the discipline-as-ecology needs.  I note that at Illinois, and I suspect at many public institutions, first year graduate students do TA, provided their English speaking skills are up to par.  Since I became aware of the situation, I have felt this is putting the cart before the horse. More on why follows in a bit.  But let's note first that the faculty preference and the need for TAs are mutually reinforcing.  So in aggregate public institutions are more likely to have larger doctoral programs than they should as compared to private institutions.

Let's also note that in some disciplines there may be chronic excess demand, for quite different reasons.  Potential graduate students have reasonable opportunities elsewhere, with more immediate payoffs and quite possibly higher lifetime earnings.  Fields where I'm aware this has been happening for some time include Computer Science and Accounting.  There may be quite a few other disciplines in this category as well.  The obvious solution in such areas is to hire clinical faculty in significant numbers and to encourage practitioners to embrace a clinical faculty position while in mid career.  But doing this is out of character for a research institution.  Clinical faculty are treated like second class citizens at an R1. 

Further, it is very hard for such an institution to make rules of behavior that differentiate across fields according to whether the field is in equilibrium, excess supply or excess demand, though from the viewpoint of this analysis, that would seem to be necessary.  The rules change only very slowly if at all.  The market conditions can change more rapidly, though as I've noted in some fields the situation persists out of equilibrium for some time. 

* * * * *

In this concluding section I want to take a pass a the quality issue by comparing my experience at Northwestern to that of Illinois.  The issue is how quality interplays with the moral hazards and what if anything should be done about it. 

In Gordon's letter to us back in spring 1976, I believe he had the Northwestern department rated eighth or possibly ninth.  There might have been a little self-promotion in that, but not too much.  Econ was one of the favored departments at Northwestern.  The president of the university at the time, Robert Strotz, was an economist. 

The rankings matter as to where a student applies for grad school.  While other factors might matter as well, presumably the applicant wants to attend the highest ranked program he or she can get into. 

The Illinois Econ department was somewhere in the mid 20's in ranking when I joined it.  But it was also very large at the time, with well over fifty faculty members.  Since some of the rankings are done by total number of pages in journals, weighted by journal quality, the rankings were biased toward large departments.  On an average quality level per faculty member, Illinois was lower than that.  In an event, one might infer simply from the rankings that the quality level of Illinois graduate students in Economics was lower than that of Northwestern.  Of course, there is always the possibility of a diamond in the rough. I have a friend who was an Illinois graduate student when I was a graduate student at Northwestern.  (Actually, he was one year ahead of me.)  He now has an endowed chair at Yale.  But he is the exception that proved the rule. 

My impression on relative graduate student quality was confirmed when in the mid 1980s I started to teach students in the microeconomics core courses.  It was my distinct impression that these students were okay, but not as good as my classmates at Northwestern.  Admittedly, this is a subjective evaluation and it is only a look at these students from the point of view of microeconomics.  Nonetheless, I'm going to take this quality differential across students as a fact. 

So Northwestern has (had) better students and they only TA during the second year.  Illinois has many of its students teaching the entire time in grad school.  Further, while Illinois had a prelim like Northwestern had when I started, it got rid of it around 1990 because it was controversial.  After that there was no process for weeding out students or students self-selecting out like what I experienced my first year in graduate school.  So Illinois was easily turning out more PhDs in Economics than Northwestern was - greater volume and average lower quality.

The only possible rationale for this, in my view, is from considering undergraduate enrollments.  The principles classes are very high enrollment and the economics major tends to be large, quite independent of department ranking.  That teaching must be staffed.  If graduate students are not so relied on then some other method must be found for staffing these courses.   But this is not a pretty picture. 

There is yet another factor to consider here - the role of international students, particularly if those students intend to return to their home country after they get their degree.  Above I tried to consider a discipline as an ecosystem.  If that is a reasonable metaphor one should ask whether the ecosystem is global in scope, continental in scope, or national in scope.  If the ecosystem is national in scope, the international students returning to their home country is akin to students dropping out of the program and pursuing a career in a different field.  Knowing this in advance, why then should such students get an assistantship when they are doctoral students? 

Here is one possible business reason.  The Economics department has a high tuition masters program in policy economics aimed at professional students who get their companies or their governments to pay their tuition.  Most of these students are content with this program as is but some are really there as a gateway to get into the doctoral program.  So that possibility helps in recruiting for the high tuition program.  As such, the business model must consider both programs acting in concert. 

Let me make one more point and then close.  The quality dimension cuts across departments (ecosystems) and at a university like Illinois there will be some departments that raise average quality while others lower it.   One might want the rules to vary by department ranking because of this, but as with the issue of excess supply or excess demand it is very difficult to have rules do this. Further, it raises philosophical questions.   For example, what should be done about a highly ranked department but in a niche field as compared to a moderately rated department but in a core field?

It has been my view for some time, for example see this post on the business and ethical dilemmas of undergraduate education at public R1s, that with inflation-adjusted undergraduate tuition on the rise for in-state students, now on a par to what Northwestern tuition was like when I attended there, that the undergraduate students should command more attention from the research faculty.  I don't see that happening, not yet anyway.  If that did occur then perhaps some of the moral hazard would be mitigated and disciplines producing excess supply of students would come closer into balance.  In most markets resources are directed to be responsive to the revenue source.

Higher Education may be unlike most markets in that it is very traditional in its approach.  The jury is out on whether there will ultimately be a sensible response of if the system will fracture instead because no sensible response is found.  I'm keeping my fingers crossed, but I'm not holding my breath.

No comments: