Alternative is back in style, so I said "What about Breakfast at Tiffany's?"

The editors of Nature deign to notice reality:

Alternative career paths should be celebrated, not seen as a compromise.

Do go on.

...but young scientists have more reason than most to be disillusioned when things do not go to plan. Almost all have completed a PhD. And almost all would have been told that the qualification — and the effort and dedication involved — was the first step on the ladder to a permanent academic position. 

Nature and others have long pointed out that this is a lie. There are simply too many PhD students and too few senior posts. Hence the purgatory of the postdocs: trapped in transition and trying to accrue the necessary credit to move on.

Well, I'm not sure how "long" they've pointed it out.  Their link is to 2011.  If they'd pointed this out in the 1970's I'd be more impressed.

Up next:  The editors of Nature suspect that these newfangled desktop computers might have an effect on how people do science.

PhD overproduction in the NYT

Every now and then the Education section of the New York Times publishes something sensible.  Today they have an article noting that there are too many STEM PhDs.  I'm pretty sure that my web browser must be broken, because nothing that sensible or that contrary to establishment narratives could possibly appear in that venue.

Metzger: Administrations and Bureaucracy

Now Metzger is discussing the growth of university bureaucracy and the backlash against it.  As universities expanded in the late 19th and early 20th centuries they inevitably became more bureaucratic.  Metzger has no problem accepting Thorstein Veblen's decision to blame boards of trustees (mostly businessmen) for some of it, but also notes that it helped to protect academic freedom to the extent that professors were subject to rules and procedures rather than the whim of the president of the university.  I definitely need to read something by Veblen at some point.

I also like how Metzger notes that expanded PhD production had an effect on the academic job market.  (Page 454)  This is something that some people still can't wrap their minds around, even though the National Science Foundation just released a report on how many new PhDs are taking jobs as postdocs rather than industry despite the alleged need for more PhDs in industry.

All of this has happened before and will happen again.

Stephan, final thoughts

I've finished the book.  Here are some parting thoughts.

First, I want to mention an important point that she made earlier in the book but I forgot to blog:  Research is a classic public good, in economics jargon.  Once the results of research are out there it's impossible to control who will benefit from the information, which means that the producers of the good cannot capture all of the benefit.  In most cases this means that we have an incentives problem, because if people can't capture the benefits they won't produce it and then society won't benefit.  Science solved this with the priority system:  The first to publish something gets credit that will benefit their career and funding.  This thus incentivizes people to put information out there as quickly as possible, for the benefit of themselves as well as everyone else.

Second, as far as her recommendations later in the book, I largely agree that we need a looser coupling between the production of research and the production of scientists.  Ideally, all of our scientists should be produced in environments that produce research, but we don't need to produce all of our research in environments that produce scientists.  To that extent, I support the idea of research institutions that don't have graduate programs.  However, I would caution against using national labs as the model, as Stephan does in her final chapter.  While the national labs produce great research, being federal institutions they have absurd overhead rates and procurement bureaucracies that rival Soviet central planning for inefficiency.  I have no easy prescription for how to create effective research institutions that are not educational institutions but aren't federal either (a quick look at federal contracting shows that there are plenty of ways to set up bloated and inefficient systems outside of the federal government), but I do think we should be wary of encouraging DOE to metastasize even more than it already has.

Maybe one intermediate step would be to make it easier for university researchers to support technicians and staff scientists on grants.  This is easier said than done, but having it as a target would keep the academic science community churning out just as much research (if not more) without the wasteful by-product of PhD over-production.

On the other hand, I am deeply skeptical of her proposal to give fellowships to students and then let graduate programs compete for them.  Partly there's the fact that the programs would have an incentive to structure projects in a way that gets results out quickly and gets students out quickly (without the formative "time in the wilderness" that students spending in deep frustration but also deep learning).  However, I recognize that that may reflect my biases as an old man.  A bigger problem is that it would be tantamount to putting the federal funding agencies in charge of graduate student admissions.  While this would certainly be great from a research standpoint (putting all sorts of information in one central place for subsequent analysis) I think there would be all sorts of political insanity in what would follow.

All in all, this book was a good, insightful read.  For those who aren't academic scientists, the first several chapters do a good job of describing how the system works.  For those who are academic scientists, the last few chapters (starting with the one on the job market) do a good job of raising issues that we have not grappled with as well as we could/should.

One more thing about Stephan and the economic return on a PhD

To her credit, Stephan's comparison of PhD earnings compared to BA/BS earnings is better than most I've seen.  For instance, she compares BA/BS holders who have been out of school for a few years to PhD holders who are new in the job market, to reflect opportunity cost. Not everyone does that. Also, she limits her analysis to PhDs and breaks them down into three categories, whereas I've seen analyses that lump together ALL doctorates.  As much as I'd love to think that my compensation is comparable to that of a dermatologist, the sad fact is that it isn't.

There is one weakness, though, albeit one that is common to all such analyses:  The average person with a PhD was probably more talented when they finished college than the average person who majored in the same subject but didn't get a PhD.  Note the "on average" part.  Of course there are dumb PhDs (just come to a faculty meeting...) and of course there are brilliant people who never went to grad school.  And of course there are people who never got a graduate degree but made outstanding use of people skills that aren't a direct part of a research degree (though they can of course contribute to research success) and there are people who got a PhD but can't work with others, and yadda yadda.

But with all of those necessary disclaimers out of the way (to head off a flood of anecdotes that people always feel the need to pour forth when you make a statement comparing two groups "on average"), it's likely that most (not all) of the people who get PhDs were in the upper tier of the talent pool before starting grad school, and they probably could have done quite well in the job market even without a PhD.  Her analysis doesn't account for that.  To be fair, nobody else does either, probably because it would be really, really hard to account for that.  Still, it's a caveat that needs to be applied when looking at those income premiums.  You don't know how much is a premium for the PhD itself and how much is a premium on traits that they possessed prior to the PhD and differentiate them from the average person with a BA/BS.

We have found a witch; may we burn her?

On page 163 Paula Stephan questioned the premise of the sweet, sweet "STEM pipeline" money:

The number of individuals receiving PhDs also depends on underlying demographics and college graduation patterns.  For example, the large increase  in the number of women receiving PhDs is due in large part to the increase in the number of women graduating from college, not to a change in the propensity of those going to college to get a PhD.41 The same is true for underrepresented minorities.  Indeed, the most effective way to increase the supply of underrepresented minorities receiving PhDs is to increase the number receiving bachelor's degrees.  This is not a trivial observation: a policy maker would achieve larger increases by building the base of students eligible to go to graduate school than by investing, as many institutions do, in changing the propensity of those who graduate to go to graduate school.

In other words, the best way to make progress at the graduate level is to make progress at the undergraduate level.  But the best way to make progress at the undergraduate level would be to make progress in k-12.  And progress in k-12 would be easier if we improved underlying social conditions, because there are plenty of things in a person's environment that matter more than their school.  And that takes us to problems that are hard for anyone to solve and are impossible for the scientific community to solve.

Also, the 41 in the quote is for a footnote, in which Stephan breaks it down more precisely and says:

Richard Freeman estimates that 70 percent of the increase in the ratio of women to men getting PhDs is due to growth in the ratio of women receiving bachelor's degrees relative to men receiving bachelor's degrees.  Likewise, 63 percent of the increase in the ratio of underrepresented minorities to non-minority PhDs is due to growth in the ratio of minority to non-minority bachelor's degree recipients.  Source: Freeman's tabulations from data obtained from the Survey of Earned Doctorates (National Science Foundation 2011c and the Appendix) and the U.S. Department of Health, Education, and Welfare.  See Stephan 2007b.

"Stephan 2007b" refers to a presentation at this meeting, but I cannot find a link to the presentation itself.

Stephan, Chapter 7

I'm several pages into the chapter on the job market for scientists.  Full credit to Stephan:  She tries to make her comparisons as apples-to-apples as possible.  For instance, on page 155 she doesn't show data comparing the salary of a new PhD with a new BA/BS recipient.  Rather, she compares PhDs 0-9 years out of grad school with BA/BS recipients ages 25-34, when they have a few years of work experience.

A few interesting things so far:
1) Although engineering PhDs generally make more than physical science PhDs, who generally make more than life science PhDs, the early-career income ratios (PhD salary in each category divided by BA/BS salary in each category) pretty much track each other over time.  All three types of PhDs can seek employment in multiple sectors, so they all seem to track the overall economy in tandem rather than the fortunes of any one particular sector.

UPDATED: 2) The PhD premium (relative to a BA/BS) is real but volatile and (apparently) declining.  She has data through 2006, showing that the PhD premium for an early-career life science PhD was barely 5%.  For physical science it was about 25% in 2006, and for engineering it was about 40% in 2006.

The volatility makes sense for a specialized credential attained by small numbers of people.  You'd expect it to be risky.

3) The later-career PhD premium is also real and somewhat more stable but still not what it used to be.  The ratios also converge, probably because people in every category branch out into more types of jobs (including management) as they progress in their careers.  An engineer working in management at a biomedical device firm probably doesn't make much more than a similarly skilled manager who works in the same firm but has a PhD in biology.  Either they know how to make their teams succeed or they don't.

4) You're still better off with an MBA.

5) For all the talk of scientists only caring about science, PhD production tracks the unemployment rate several years earlier (people are more likely to stay in school if jobs are hard to get) and the salary prospects for their field several years earlier.

It's almost as though scientists were human beings who responded to resource scarcity and incentives.

Stephan, first few chapters: Workforce issues

Paula Stephan's book is hard to blog because it is detailed and descriptive rather than argumentative.  It's an interesting read, painting a very wide and detailed picture of how science works, but it's hard to pick out points to ponder or grapple with because it's not about Big Points (which is fine).  The part about intellectual property was interesting:  According to page 50, data shows that universities get more patent revenue for themselves when they let their faculty keep a larger share of the revenue (because their faculty have more incentives to commercialize) but the universities that give a larger share of revenue to faculty don't actually produce more patents.  The difference, then, must be the type and quality of research that they generate, its commercial feasibility, and the aggressiveness of their faculty in going beyond the patent office and actually pushing their idea to market.

On page 68 I learned that one reason (though hardly the only reason) why US university labs rely more on grad students and postdocs while Europeans rely more on staff scientists and similar sorts of personnel is that in the 1980's NSF started pushing faculty to put grad students on their grants rather than technicians.  I suspect that part of the reason was economic, but only partly so:  While a grad student will just about always receive less in salary and benefits than a technician (especially an experienced technician), some universities charge massive tuition for grad students and that tuition then gets covered by the research grant supporting the student.  Consequently, while at some schools the economics will favor the use of students by a wide margin, at others the margin may not be quite as wide.  Moreover, depending on their level of training, level of experience, personal abilities, and opportunities for development, a technician may be more useful than a grad student--the grad students may ultimately go farther as scientists, but it takes a while to get them there, whereas a good technician can keep a project moving forward on a consistent basis.  I suspect that part of the reason for NSF's preference is that they have long had a mandate to expand the scientific workforce because of an alleged crisis in STEM, and supporting students will do that, whereas supporting technicians will simply keep the labs humming along and producing mere science.  And this focus on students apparently started in the 1980's, which is when "A Nation At Risk" came out.

(And let me assure you that I don't buy into a dichotomy:  Clearly the world needs good lab techs AND students training to be scientists, but the answer then is to fund projects that have people in whatever mix of roles makes sense for the success of the project, not a mandate to endlessly expand the STEM workforce.)

ADDENDUM: I decided to look for the original source for the assertion about NSF.  It's a 1998 interview with then-NSF Director Rita Colwell by Science magazine.  Here's the key quote:

On a recent report of a Ph.D. glut in the life sciences: 

In the 1980s, NSF asked investigators to put graduate students on their research budgets, saying that it preferred to fund graduate students rather than technicians. But as often happens, the pendulum swung too far. There needs to be a balance. Well-trained technicians are needed to run equipment and labs, and what better way to provide opportunities for them than to build it into research grants. 

There's great respectability for those who want to be technicians, but we don't give them the opportunity. We've made it a sign of failure if you don't get a Ph.D. Other countries don't have this [negative] attitude. … With all the emphasis on graduate students, I'm not so sure that the question we're facing is one of overproduction. I think we have to look more closely at how we're using our resources. We've created a situation in the career pipeline where there is a bulge at the end of a postdoc and no place to go.

Even in 1998, people were talking about PhD gluts.  All of this has happened before and will happen again.  Of course, that was around the time that NIH doubled its research spending, so for a while things were good for life sciences PhDs.  But exponential growth can only go on for so long.

de Tocqueville on questions of science, science and progress

Chapters 10 and 15 in Part 1 of Vol. 2 of Democracy in America have some points that I find particularly worthy of using as a lens for modern academia.

In Chapter 10 de Tocqueville argues that a more egalitarian society will be a more economically competitive one (on this point I completely agree with him) and thus people have less of the leisure time that is so necessary for intellectual contemplation, insight, and advancement.  He thus argues that democratic societies will make most of their advances in the applied sciences rather than the basic sciences.

I partly dissent, and not because the second half of the 20th century saw such great scientific advancement in the US in spite of being a time of comparative economic equality and rising racial equality.  The middle of the 20th century was something of an anomaly in US history, with great prosperity enabled by the fact that the US had better infrastructure than anyone else in the post-WWII era.  That fact alone means that de Tocqueville's analysis of economic competition in egalitarian societies does not really apply to the 20th century US; we were very far from the sort of equilibrium that he was envisioning.  In fact, it would seem to support his case, precisely because it was an exception to his economic assumptions and a counter-example to his assumptions about scientific progress.  Instead, I partly dissent because he gives short shrift to the relationship of the pure and applied sciences.  Advances in the applied sciences make advances in the pure sciences cheaper.  Imagine studying pure questions in turbulence without the insights enabled by modern computing power or high-speed cameras.  Imagine doing the purest work in cell biology without bioinformatics infrastructure built on our computing industry, the vast infrastructure of lab equipment suppliers who have made things cheaper and faster to compete for the budgets of labs and companies doing applied research, or the tools of modern microscopy and lasers.  How many people realize that the multi-photon microscope, a beautiful tool of neuroscience, would not be possible without the advances in pulsed laser technology enabled by the telecommunications industry?

On the other hand, it is definitely true that a leisured class, whether born into leisure or elevated to it through the academic system, is an invaluable element of a basic research apparatus.  In that regard, it is interesting to look at the structure of the modern academy.  It has become more and more unequal, with a class divide between high-status researchers (who do have a certain amount of time for contemplation) and low-status adjunct instructors, and a shrinking middle class that try to bridge teaching and research.  This certainly seems to mirror the aristocratic arrangement that de Tocqueville regarded as necessary for basic research.  However, even (especially?) the biggest of the bigshots seem to spend more time working on funding proposals than contemplating science.  The system has the superficial appearance of an aristocracy of philosopher-kings, but the reality is much less romantic.  OK, if there's one major theme of this blog it's that the past was never all that romantic compared to the present, but at the very least the current system demonstrates that pure research isn't really a product of leisure.

A more interesting place to look is at graduate study.  The best part of graduate school is that after a few hurdles it is largely unstructured.  I think every PhD looks back on graduate school and thinks "Man, if I'd known then what I know now I would have made better use of that unstructured time!"  The purpose of that unstructured time is to let the embryonic scientist enjoy an interlude of leisure (albeit impoverished leisure) so that they might focus on learning.  It is thus interesting to note that the National Science Foundation's Graduate Research Fellowship Program is putting more and more burdens on graduate students.  When I applied in 1998 they basically wanted to fund smart and motivated students, and the applications were almost trivial.  Nowadays they want research plans and outreach plans ("Broader Impact").  Honestly, since leisure is wasted on the young I sort of get why they want to give bright young minds more structure.  I still think it's a mistake, but I get why they do it.

Now, if we look at graduate research fellowships through the lens provided by de Tocqueville, we must note that there is no organization more openly devoted to a semblance of democratic spirit and social equality in science than the National Science Foundation.  Their Broader Impact Criterion is, for good or ill, all about that.  Making graduate students engage in Broader Impact projects is not only in keeping with the social goals of the NSF, but it also keeps graduate students in a state of hustle and multi-tasking.  They can't just focus on their subject, they must also show people their public impact.  You can like it or dislike it, and it's not the same as the commercial competition that de Tocqueville considered, but it's clearly different from a vision of graduate study as sheltered time.

I can't say that anybody consciously decided that graduate study should be a time of greater structure for the sake of democracy, but it is interesting to see this pattern noted by de Tocqueville in one context show up in another, and via the intervention of an organization whose mission and viewpoint match the democratic spirit that he described.

Lies, damn lies, and PhD job statistics

Today I went to the physics student club room (a place where physics majors hang out and study and socialize) to drop off a few science magazines for them to read when they come back for fall quarter.  On a table I found this poster from the American Physical Society:

(A bit of Googling did not turn up a digital version of the poster.  If I find one I will link it.)

This looks like a pretty nice set of numbers for PhD employment.  64% in the private sector, 22% in potentially permanent academic jobs (which can be a sweet gig), and those in the private sector start at $92k.  Sweet!  What on earth could I be complaining about?

READ THE FINE PRINT!

Those numbers exclude postdoc positions and temporary academic jobs like adjuncting. What percentage of physics PhDs start off as postdocs or adjuncts?  According to the American Institute of Physics, 65%. 56% go into postdoc positions, and another 9% go into other short-term positions.  Now, most of them will eventually jump off of the postdoc/adjunct/visiting/etc. treadmill and find a full-time, long-term job that is reasonably satisfying.  That's great.  If APS wants to put those statistics out there, go for it.  However, they instead chose to lie with statistics, and put a really rosy picture of PhD job outcomes in big print, while putting a partial caveat in fine print, and never mentioning at all that their data only applies to 1/3 of new PhD graduates.

Now, if APS were a commercial organization attempting to sell something, well, caveat emptor.  The world is full of people telling incomplete stories.  The fact that it's ubiquitous does not make it right, but it does make it rather unremarkable.  There aren't enough hours in the day to blog about every misleading sales pitch ever made.  APS, however, is not a commercial organization.  APS is a non-profit professional society that exists to serve the physics community at all levels, including students.  (Indeed, off-camera the poster includes a link to this student membership site.)  Professional societies best serve their members when they provide their members with accurate data on careers and education.  APS is not serving the interests of their student members when they provide misleading statistical information to people who are trying to decide what to do after college.

Now, to be fair, they did include some information about careers for people with bachelor's degrees in the second half of the poster:

(For some reason Blogger is making it possible to zoom in on the first image but not these two.  I'm working on it.  And yes, I'm aware of the irony of making the fine print visible in only one image while complaining about people who are sweeping things under the rug.  I'll try to find a fix.)

The bottom figure certainly sells graduate school as The Path, by lumping graduate school, k-12 teaching, government jobs, and non-profit jobs in the same category.  It is clearly a slanted poster.  There's no good reason for this.  It is left as an exercise for the reader to figure out why an organization that ostensibly exists to serve all of its members would target college students with misleading statistics that might influence their post-graduation decisions.