How many people go to graduate school and a few other questions

Here are a few snapshots of graduate school enrollment data I’ve been playing. All of this comes from 12 month (unique head count) enrollment data (IPEDs) and US Census population estimates by age.

First, when it comes to grad school, the real college towns are cities and their suburbs.  No surprise there, because that’s where 80% of people in the US live and work. Still, although millennials in general may or may not be drifting away from urban areas, the millennials who seek advanced degrees are certainly not.

In 2015 just under 4.5% of US millennials were enrolled in graduate school someplace. That’s been a stable fraction of 20-39 year-old people enrolled in grad school since 2010.

Prior to that, the fraction of people in that age range who are in grad school has been growing steadily over the past several decades (because of the increasing popularity of the master’s degree). More and more people were going to grad school. But not anymore.

For a lot of reasons, I think that 4.5% is the key number to watch, rather than absolute enrollment numbers. Why has it plateaued? Is graduate school too expensive? Too much student debt? Has the Great Recession fundamentally changed this in some way? Will graduate schools be able to increase further the percentage of baccalaureates who go on to get an advanced degree? Or is this 4.5% a “hard” cap?

I have a sneaky suspicion that most universities are built on an enrollment growth model…which is to say that attracting a growing percentage of college grads has been key to keeping tuition rates lower than they’d be otherwise. Yikes!

Individual graduate schools will struggle to hit their enrollment targets if that 4.5% fraction remains stable. Because that means competition for students will stiffen and pressures to raise tuition revenue will become stronger.

The last, if not most, remarkable story is what I like to call the privatization issue. Graduate school enrollment at private universities has caught–and clearly surpassed–that in public institutions. This turnabout is kinda stunning. The graph shows that the growth in the percentage of people enrolled in grad school over the past 15 years can be attributed mostly to private universities.

I think there are two explanations for this trend.

First, public institutions don’t spend nearly the same on student recruitment as do private institutions. This is certainly the case at the undergrad level where privates outspend the publics at a rate of about 4 to 1 in terms of recruiting one student. So there is that institutional ethos in putting some effort into marketing that differs.

But at the graduate level, recruitment is so Balkanized finding “typical” recruitment costs is difficult. The spend per student recruited is almost certainly much lower than for undergrad recruitment at both public and private institutions. That’s mostly because nobody is really in charge of it in graduate school. Graduate school deans expect programs to fend for themselves whereas programs assume the deans are helping them out with marketing and recruitment.

When on the phone with recruiters at a public university on several occasions I’ve heard something along the lines of, “And of course I’m at a public university so I have no recruiting budget.” Of course.

That’s just not something we hear a lot when chatting with private university recruiters.

A second reason for privatization is that as the cost of attending a public university has grown, the price advantage they once had over private schools has diminished, if not evaporated completely.


Graduate School Enrollment is Higher in Private than in Public Universities and Colleges

public v private

Just a quick observation. Graduate school enrollment in public universities has been dropping since the early 90’s, while enrollment in private (not-profit) universities has been steadily rising.

It is hard not to imaging a steal effect going on there, where the privates get students who might have otherwise enrolled in the publics. But I’m certain the causes and the effects are much more complicated than simple competition between schools.

Some responsible forces off the top of my head range from reduced state support of the publics, narrowing tuition rates between the two, to endowment growth at the privates to relative successes in growing research infrastructure and the graduate enrollment that research will drive.

Of course, ultimately these numbers reflect where students choose to attend. This just relfects the evolving marketplace for the Master’s degree. If the students hold the cards–in the end, nobody can force a student to attend someplace they don’t wish to attend–what the graduate students seem to be saying is that the graduate school value proposition at private universities and colleges exceeds that of the public institutions.

If you add the private for-profit graduate school enrollment to that of the private non-profits, you get the inset result: Graduate school enrollment is higher now in private than in public colleges and universities. The inflection point actually occurred in 2008 and it is really difficult to see how this trend will ever be reversed.


Are We At Peak Master’s Degree? Part 2

While the enrollment trouble at for-profit universities has been making all the news the past few years, graduate school enrollment has been on the decline at traditional non-profit universities, too.  The numbers of Master’s degrees, but not doctorates, awarded in the past three years has been flat. And seventy-five percent of all graduate degrees awarded each year are the Master’s. Therefore, it stands to reason that declining graduate school enrollment can be attributed largely to a reduction in Master’s program enrollment.

This begs the question; is this a transient or have we seen peak Master’s degree? Here’s a bit more evidence from the government’s data vaults that indicated Master’s program enrollment is declining, and that it probably reflects a reduction in demand for the degree. And so, yes, peak Master’s may be something in the past.

Unfortunately, although government data reports the number of heads enrolled in graduate school any given year, it doesn’t report program-level enrollment. Strictly, when we see a dip in overall graduate school enrollment, we don’t really know if that dip is due to Master’s students, to certificate students or to doctoral students.

However, given a few simple assumptions on attrition rates and time-to-degree, it is possible to infer enrollment in different degree program categories by using the number of graduate awards conferred at these levels by an institution each year.

Here, I estimate graduate enrollment as function of the ratio of certificates plus masters awards, divided by doctorates, the CM/D ratio in the figure above. In these calculations I assume a average of 2 years to degree for a Master’s student and 5 years for a doctorate, and 10% annual attrition. These assumptions work as a good first approximation for institutions that have a complicated mixture of programs, times to degree and attrition rates.

Thus, the higher the CM/D ratio, the higher the relative enrollment of Master’s to doctoral students. For example, this model calculates Master’s to doctoral student enrollment is over 4 to 1 at institutions with a CM/D award rate ratios of about 10. In other words, it is hard to have their reported award ratios without having these calculated enrollment ratios.

If reductions in enrollment are afflicting Master’s program more than others, we’d expect to see institutions with the highest percentages of Master’s students most adversely affected. Indeed, by sorting institutions by their Carnegie classification, we see the enrollment dips are more pronounced in those groups of institutions that award higher percentages of Master’s degree relative to doctorates.

For example, the most precipitous enrollment drop happens in the institutional category of large program Master’s Colleges and Universities. This large group of 377 institutions, on average, award 47.4 Master’s plus graduate certificates for every one doctorate. I calculate their ratio of enrolled Master’s to Doctoral students is almost 20 to 1, which is very high. These are institutions that must on average be taking in about 30 new Master’s matriculates for every new doctoral matriculate.

Collectively, enrollment in the large program Master’s Colleges and Universities institutional category is down a whopping 6.5% from their 2010 peak. That’s just a remarkably high shift over a short period, as the graph dramatically illustrates.

In contrast, although graduate school enrollment at Research Universities (very high research activity; aka, Research 1 institutions) is not down, it is nonetheless flat. These institutions award 3.1 Master’s and certificate degrees for every doctoral degree.   At these elite institutions, that means only a little more than half of all enrolled graduate students are in Master’s programs. Since they are such large institutions, as a group they award high absolute numbers of Master’s degrees, but the high percentage of doctoral students enrolled in their student bodies may be serving as a buffer to lessen any dramatic impact of reduced Master’s demand on the overall graduate enrollment.

Reduced graduate school enrollment is evident at institutions that award intermediate CM/D ratios. In particular, enrollment is down markedly at the Research Universities (high research activity) and Doctoral/Research Universities, where the ratio of Master’s and certificates to doctoral awards is higher than at Research 1 institutions. Enrollment at Research Universities had been relatively stable since before the year 2000, but is now down a gob smacking 7.5% from a 2011 peak!

Meanwhile, enrollment is rising in the Medical School and Medical Center category of institutions, which unlike all of these other categories, generally award more doctorates than Master’s.

Taken together, these observations along with what I’ve pointed out previously argue strongly that a systemic reduction in graduate school enrollment is on-going throughout the entire landscape of US higher education. This afflicts Master’s programs because the demand for the Master’s degrees is falling. Call it a burst Master’s bubble, call it a Master’s virus, or call it the peak Master’s degree. The appetite for the Master’s seems to have diminished markedly compared to what it was even just a few years ago.

To believe a reduction in demand doesn’t explain flat and dropping graduate school enrollments you’d need to argue that all of these hundreds of institutions simultaneously raised admissions standards.

The data I’ve analyzed here comprises that for about 1800 non-profit institutions in total. Yet, most graduate education happens in only about 500 of these institutions, which are responsible for awarding over 90% of all doctorates and 70% of all Master’s. Collectively, the graduate enrollment in these 500 institutions represents about a third of their overall student enrollment. That’s a lot.

This impact must not be easy for them to absorb, particularly if they continue to set tuition rates and treat graduate student recruitment as business as usual. In fact, on the phone recently I was told by someone working at a fairly severely impacted institution they are cutting their graduate student recruiting budget. Good grief. That’s precisely the opposite of what they should be doing given the problem they have specifically, and what is happening more generally.


Hey, Non-Profit Graduate School Enrollment is Declining Too

I’ve been munging around a bit with graduate school enrollment and award data for US universities to get a sense of the overall trends and, at least in my mind, the picture is coming into focus. We’re in an unstable period right now and I think most of the people in graduate admissions and recruitment offices that we speak with on a daily basis are sensing that things are different than before. Our sense is that those at the front lines, the program directors we speak with, seem to feel things more sharply. If I had to summarize those conversations simply, we’re hearing stories of excess seat capacity and that fewer good applicants are applying.

The data visualization that seems to tell the full story for the non-profit universities only is right here in this graph, showing declines in both total and population-adjusted enrollment trends. What this shows is graduate school enrollment has declined sharply from a peak in 2011. The decline is a lot worse when the data are adjusted for US population levels. That’s almost 4% below the peak year of 2011 in a system used to growing about 0.6% year-on-year.

Separately, since starting the Gradschoolmatch project I’ve pretty much arrived at the conclusion that graduate enrollment is demand-driven. We look at our acceptance rates and like to think we’re highly selective, but that’s not exactly true. For example, as one data point to illustrate this conclusion, decent applicants (the ones who actually get into graduate school–the ones programs all want) don’t apply to many programs and are accepted by most to which they apply. They just hold the cards.

Another reason to conclude this is a demand-driven market is that for the most part, as a group, graduate schools are really poor at creating more demand. We either don’t adequately support a marketing effort or consider the marketing of our programs as an unnecessary expense. We therefore leave it to chance whether prospects find us. We don’t calculate very well the return on what a modest marketing investment can bring. We fail to see in our cost/benefit metrics how strongly even a single matriculate can drive tuition revenue or perhaps even research dollars from their discoveries.

When we recruit prospects we do so reactively, rather than proactively. For example, for the programs that do recruit, for most it only happens after an application comes in over that transom, not before it. This process of waiting and hoping is not one that creates demand.

Let’s face it, we academics are so crappy at communicating the value proposition of our universities and degree programs to prospects that we’d also probably struggle to sell a life jacket on the deck of the Titanic. When we have tried to be more proactive, we’ve bought into disappointing high cost/low precision/low yield services, when the very nature of our product, the single highly specialized graduate program, requires low cost/high precision/high yield marketing services.

Taken in this light, these enrollment numbers are just another data point I see suggesting quite strongly that the overall demand for graduate degrees is falling. Although this may just be regression to the mean, I also can’t help but wonder if we’re in a period where this is, in fact, a new normal. What concerns me is that the toxic combination of changing economic demographics, the continued spiraling costs of attending universities and excessive undergraduate student loan burden has fundamentally disrupted the appetites of people for advanced degrees. Despite the fact that the economy pays them so well and can’t seen to employ enough of them.

Declining demand is almost certainly exacerbated by the inability of specific degree programs to effectively communicate their value proposition to prospective students.

For example, if you read the twitter and blog internets of the many unhappy students out there regularly sharing their feelings, you’d be surprised to know that the vast majority of current graduate students are actually quite satisfied with their enrollment decisions. They should be, since they put a ton of effort into making their decisions and often without a lot of help from those of us in programs guiding them to us.

The graph above has enrollment numbers at for-profit universities stripped out of the data. The enrollment at for-profits is low relative to that at non-profit institutions, but it is not insignificant. For-profits are actually experiencing a more dramatic bust cycle than are the non-profits. I’ve excluded those data even though that cycle probably shares some of the drive affecting non-profit enrollment, because for-profits are also complicated by a very different dynamic than what problems might be afflicting the traditional universities.


US Graduate School Enrollment Has Been Flat Since 2010

Peak Graduate Enrollment (1)This is pretty interesting. It may be the proverbial canary in the coal mine… or it may just be regression to the mean. US graduate school enrollment has been flat since 2010, at just below 3 million heads per year. We haven’t seen anything like this over the past 30 years, a period for which we’ve enjoyed small but steady year-on-year increases. What factors might be responsible for this effect? What does it mean for graduate schools and prospective students?

First, the data. They are fall enrollment head counts from the IPEDS data center, a US Department of Education enterprise. Due to their validation process, there is a bit of a lag, such that the most recent available data is for 2103. Fall enrollment is used because the time series is longer. Another enrollment variable that IPEDS keeps, 12 month enrollment, shows that flat enrollment extends to 2014.

I’ve omitted data for the year 1999, which seem to have a wildly spurious under count relative to its neighboring years. I’ve included enrollment data for all non-profit (both public and private) and for-profit universities that award a masters or higher degree. Of the 2176 institutions in this group, enrollment data was missing for only about 50. This is a comprehensive picture because these are, in essence, anywhere that anybody goes to graduate school in the US for whatever field of specialization.

What factors might be responsible for this plateau in graduate enrollment?

  1. Economic prosperity-Probably not. Graduate school is widely thought to be anti-cyclical. The idea is to time your graduate education during a recession, so that jobs are more plentiful when you emerge. Inspecting the graph here sort of disputes that view. First, there are never enough recessions for this idea to explain the overall enrollment growth that has occurred up until recently. Furthermore, it is hard to discern enrollment boosts in recessions and pauses during the inter-recession periods with more than wishful thinking. Perhaps the best argument against this idea, however, is that although we are now a few years out of the recession, the US economy has hardly been bicycling on spinners over this period. The Great Recession has had long-lived negative effects that may have become baked-in to thwart economic progress.
  2. Declining enrollment at for-profit universities. No effect. The graph above looks the same if for-profit enrollment is removed from the data. Although it is true that enrollment is dropping at for-profits due to their notorious problems, graduate enrollment in the for-profit university sector is dropping at a lesser rate than undergraduate enrollment. Furthermore, the graduate enrollment at for-profits is only a small fraction of the overall graduate enrollment in the US. The simple fact is that enrollment at non-profit public and private universities is flat all on its own.
  3. Demographics-Quite possibly. The rate of growth has also slowed markedly for undergraduate enrollment in college. Since 2011, undergraduate enrollment has also been flat, holding pretty steady at 9.1 million heads (as a general rule of thumb, graduate enrollment stays around 1/3rd of undergraduate enrollment). Since about half of all baccalaureates earn graduate degrees, flat undergraduate enrollment growth predicts flat baccalaureate award rates, suggesting that the flatness in graduate enrollment will persist for some time to come.
  4. Tuition cost/Student loan burden-Possibly. College students have been graduating with increasingly higher student loan debt as they’ve been asked to bear a higher share of their cost of education. More people may be avoiding graduate school because the tuition cost is too high and/or they are unwilling to take on additional debt, despite the clear economic benefits of an advanced degree.
  5. Flat research spending-Quite Possibly. No matter the cost to students, producing people with advanced degrees is expensive for universities. There are a couple of reasons to believe that university research activity is, in fact, the strongest driver of overall graduate student enrollment, both of which I intend to write more about shortly. As research funding is constrained, the demand for and resources to train graduate student researchers is also constrained. The flat graduate enrollment we see above mirrors the flat to declining research spending over the same period. Not coincidentally, US universities also haven’t experienced year-to-year research spending this flat over the same period for over 30 years. Austerity is coming to roost at the university and it is likely impacting the vitality of graduate programs.

Flat graduate student enrollment seems is probably due to a combination of these latter three factors, in my opinion. The demographics and loan burden suggest to me that competition between programs for outstanding students, which is always high, will prove even higher. Fewer such people are considering graduate school and the value proposition for those who do attend will need to be stronger than it appears today. We hear data points here and there from our clients in the enrollment office trenches that suggest universities have already entered a period of evaluating their programmatic needs and preparing to mothball graduate programs that cannot attract enough talent. This may be the new normal.

The flat research spending means that fewer slots are available for students. Given the extraordinarily inefficient way people go about finding their graduate school options, this could negatively impact those students seeking research experiences as either doctoral or masters students, due to increasing competition between applicants for the available resources. Such students may need to widen their application strategies to ensure they find a program that not only fits, but is likely to have the resources to carry them through.

Finally, persistent flat enrollments is probably not a good thing in the overall picture of what a thriving knowledge economy needs. There is no reason at all to believe that the US economy suffers from too many highly educated workers, given the simple observation that unemployment rates are the lowest and earnings the highest for the most highly educated. A failure to produce more workers with advanced degrees could stymie the growth of companies and industries that depend upon such highly educated talent.