The PhD problem

http://arstechnica.com/science/news/2011/04/the-phd-problem-what-do-you-do-with-too-many-doctorates.ars

by brianwillis • 15 years ago

36 27 15 years ago

melling• 15 years ago

I think for the hard sciences and engineering we should be creating more research jobs in the private industry for PhD's. We have no shortage of problems that need to be solved and discoveries to be made.

Private industry needs to be incentivised to do more long-term R&D.

brianobush• 15 years ago

Agreed. In my company, a fortune 500 company, we are concerned more with revenue than building technology to solve big problems. This leads to quick hacks that solve short term issues only to leave technical debt and little invested in truly solving a problem.

zheng2• 15 years ago

I'd be willing to guess that you are more concerned with revenue because you have shareholders to answer to. R&D is something that doesn't show up on the bottom line with regularity. Sure, you might develop something huge and make tons, but probably not.

What would be interesting is to see some heavy quantitative studies done on the ROI of research in a corporate setting. If we could prove that R&D provides a decent ROI, I'd be willing to bet that the business types would be more willing to fund it.

URSpider94• 15 years ago

What you are suggesting was the norm 20-30 years ago. Major corporate research centers like Bell Labs and IBM Watson were the crown jewels of their respective corporate owners. Today, these programs are mere shells of their former selves.

Corporate America just doesn't have the stomach to fund fundamental research with a payback varying from years in the future (optimistically) to never (pessimistically). In fact, we have exactly the opposite trend -- university researchers have taken on all kinds of applied research projects that would have been considered far too commercial in the past. In turn, the universities are looking to license out their IP portfolios and take stakes in companies founded by their students and faculty to raise more cash.

pbsurf3• 15 years ago

The list of inventions that came out of Bell Labs is breathtaking: solar cells, information theory, the first transistor, the first MOSFET, the laser, CCDs, UNIX, and many more. Silicon Valley was essentially founded by William Shockley, one of the inventors of the transistor at Bell Labs.

At Bell Labs the absolute best scientists could come together and just work, without having to deal with writing grant applications, teaching courses, or mentoring grad students (the three activities which now consume the vast majority of an academic researcher's time).

I am convinced that if Bell Labs was as strong today as it was 50 years ago, we'd have kilo-core microprocessors (and know how to program them), all our genomes would be sequenced already, and we'd all be driving electric cars recharged by ultra-efficient solar power stations. Or rather they'd be driving themselves. Maybe Google can help get us back on track.

michaelochurch3• 15 years ago

Corporate America just doesn't have the stomach to fund fundamental research with a payback varying from years in the future (optimistically) to never (pessimistically).

Most corporations are completely beholden to Wall Street these days, and stock traders can't evaluate long-term research initiatives.

jcampbell1• 15 years ago

The universities also have a major cost advantage. They dangle the possibility of a tenured job in exchange for getting PhDs to work for $45k as post doc researchers. Even if primary research does have a payoff in the future, it makes more sense for universities to do it at 1/3 the cost and guaranteed government backing.

Maybe the government and universities have squeezed the corporate sector out of research?

teyc2• 15 years ago

The same problem was true for engineering/CS graduates in developing countries. The problem in that case was that the skills available did not meet the economic profile of the countries they were working in. For instance, labour may be so cheap it is not cost efficient to automate processes.

In many cases, we need to properly analyze the chain of value creation, from basic sciences to applied engineering, and iron out the inefficiencies. Once this occurs, PhDs will again enjoy their proper place in the economy.

jgamman• 15 years ago

>we need to properly analyze the chain of value creation

Central planning is ver rarely the answer. Besides, who is the 'we' in this proposition?

teyc2• 15 years ago

> Central planning is very rarely the answer

I disagree. That's dogma. Dogma is rarely the answer. The a lot of university's research already relies on public funds which is disbursed by a central committee. There is little private money in this space because the timelines and risks are too high.

As for solutions, I heard of a few entrepreneurs who regularly visit university's technology board looking for interesting solutions that they can take to market. This is a kind of gap filling move that can be done broadly, provided the frameworks are realistic. I remembered a tale about IdeaLabs where they were spinning off a startup and decided to give the entrepreneurs 80% share of the company instead of the regular 20%. If a university has unrealistic expectations, then this can also curb the introduction of a technology/technique.

Then there are some schools who do very good basic research but little in the way of translational research (e.g. medical trials, FDA approvals etc). The reason these may be occurring is because the way research funding works is that these activities are outside of their scope and ability.

dvse• 15 years ago

One of the problems with CS PhD graduates at least is the lack of broad familiarity with their and related fields.

Outside of a handful of top US schools the resources (i.e. high level graduate courses taught by people who actually understand what they are talking about) are simply not available. Indeed even there one needs to make an effort to take classes from other departments.

Somewhere like UK with no course work requirements at all, it is not as easy to find a PhD student with an understanding of the proverbial bigger picture (of course there are exceptions etc).

It is the broad understanding of the relations between methods employed in different fields that makes a PhD useful in an industry setting and not the narrow topic of their research (for example someone familiar with ideas used in numerical computing can do research level work in many fields).

pbsurf3• 15 years ago

I would suggest that graduate programs begin requiring applicants work for a year or two before applying for or starting a PhD (not unlike MBA programs). This would help bridge the gap between the academic and working worlds and provide students with some much needed savings to carry them through graduate school as well as a potential placement after graduating. And, having seen what they can earn in the working world, only the truly dedicated would accept a graduate student stipend for five years, solving the PhD overproduction problem.

papaf• 15 years ago

Until recently I was living this problem. I finished my PhD in Computational Biology and then joined the ranks of the long term unemployed. I got 1 job interview in 1 year. Things have improved and I now have a position in academia (the first academic position that I applied for).

I personally don't see that doing a PhD should be a issue for recruiters in industry. I worked as a programmer for 10+ years before doing a PhD and can confidently say that doing research was one of the most rewarding times of my life. I learnt an incredible amount - especially how to independently research and tackle hard problems. Does industry not have any hard problems?

Perhaps this is just a period of adjustment. The people recruiting don't have PhDs so they don't know what research involves or what someone with a PhD can bring to a company.

I have seen both sides of the fence and if I was ever in a position to recruit I would certainly consider someone with research experience on an equal footing with a new graduate. I often joked that I would have had the same employment chances if I'd have spent 4 years sitting on a beach instead of doing a PhD. In a complex modern world such as ours this is truly twisted and bizarre.

apl• 15 years ago

  > I got 1 job interview in 1 year. Things have
  > improved and I now have a position in academia
  > (the first academic position that I applied for).

  > I worked as a programmer for 10+ years before
  > doing a PhD

What kind of jobs were you applying for? It seems somewhat unusual to do 10 years in the industry, then get a somewhat but not really related PhD, then get back into the industry. It'd be great if you elaborated a bit.

varjag3• 15 years ago

One rarely mentioned thing is that most PhD degrees are worthless.

This IMO is a key factor to understanding many problems plaguing postgraduate world. Of maybe dozen PhDs (spanning two different countries and fields of mathematics, physics and CS) I've been acquainted with, only one had uncompromising, endurance runner attitude that serious research takes. He was the only one who landed at research career after the defense.

I fail to fully realize the reasons behind the motivation of the less successful ones. Whatever it was, it was certainly not a thirst for knowledge. Their contribution to advancing the state of art was zero. Those in the fields were I have a glimpse of understanding, should not have defended. They were net waste of grants/taxpayer money, and their own life.

As it is now, postgraduate research is like uranium enrichment: a costly process with majority of output being waste. Of course the thing wasted in this case is people's effort and time which could (maybe) better spent elsewhere.

michaelochurch3• 15 years ago

This is sad but true. A lot of people go into PhDs at 22 thinking they're intellectually invincible. They've gotten A's their entire life, never had any kind of serious setback, and fail to process the fact, when they hear PhD programs' abysmal job placement numbers, that those numbers will apply to them. This is even worse in the humanities, where the numbers are closer to 1-in-5 than 1-in-2. People who have been in the top 10% for their whole life fail to apprehend that all their classmates have been in the top 10% forever as well, and that they're counting on luck more than anything else.

Five years in, the more successful PhDs have moved on or are at least in the process of finishing up, but about half of the students are long past burned out but will still be around for 2-5 years. Most careers encourage the less successful people to leave or fire them (which seems harsh, but it's less harsh than wasting peoples' time) but academia doesn't, or to whatever extent that it does, people don't get the hint. The result is that you have a lot of people who reach 29-31 when they finish their PhDs, and because they took too long they can't get great post-docs, and it's downhill from there.

I think graduate school is a great way to learn a lot for people who have the ambition and ability, but the problem with it, from a career perspective, is that it's regressive rather than progressive for a lot of people. What I mean is that I'd rather be a first-year graduate student than a bottom-half investment banker, for sure, but the people in the "real world", even if they take some lumps early on, have already moved on, learned from a mix of success and failure (in most cases) and had a chance to establish themselves by 29-30. The less successful half of the graduate students haven't had this opportunity; they start out as graduate students at elite departments and, 7 years later, they're "wraiths" with nothing to show for their time. It's a really broken and unfair system. They should be given the hint much earlier (like after 1-2 years) so they can move on and do something where they have a better shot at success.

varjag3• 15 years ago

I agree that there ought to be better exit routes for failure, and, as you say, the system just doesn't have an incentive for that. Luckily for me I was simultaneously both fulltime employee in the industry and a postgrad student, and by the 2nd year was forced to choose. In retrospect, it spared the world from another worthless dissertation.

You obviously thought about it before; do you have any ideas how the situation could be improved? Sometimes I feel the whole mentor/advisor thing is flawed, and postgrad students would do better if it was abolished, but you'd think if it was that simple, it'd already be sorted out..

hjkdsfhjkasdf• 15 years ago

> Their contribution to advancing the state of art was zero. Those in the fields were I have a glimpse of understanding, should not have defended.

The research committees that decided to accept these dissertations were composed of multiple experts in their respective fields of research. Every expert took months to evaluate each thesis and decide whether the contribution was sufficient to grant the degree. And yet your glimpse of knowledge is apparently enough to disqualify their judgement.

varjag3• 15 years ago

I don't see why you have to post this from a throwaway account, as this is a legitimate question and you unlikely to suffer for it at HN.

To answer though, the committees are much more inbred and affiliated than they should be in theory. It is a complicated mix of academic politics, budgetary allocations and feudal bonds that advisor-student relationships are. There is a lot of leeway in interpreting things like "novelty" and "importance", and amazingly often it is treated in the broadest way possible.

As to me, yes, I'm qualified enough in certain fields to make a judgement, and have enough understanding of some others to cut through bullshit. The issue here is not the knowledge (or lack of) the judges, but rather their complacency with mediocrity due to life realities. I didn't imply they were green-lighting falsified results or outright crackpotry. Just that the amount of meh that shouldn't belong to PhD domain at all but gets through is incredible. If you throw all those dissertations out, the state of respective field wouldn't degrade a iota.

numeromancer• 15 years ago

tl;dr: I'm on board the SS PhD now, so let's pull up the boats before it gets too crowded up here.

sebkomianos• 15 years ago

The problem with not only PhDs but even the first levels of higher education is only one and it is universal: We see it from an economic perspective.

A bachelors degree is good if it helps you do a masters degree.

A masters degree is good if it improves your chances at finding a job.

And then a PhD is good if it increases your expected salary.

Well, I am sorry, but we shouldn't be complaining about "the PhD problem" then. If you do a PhD, do it to learn. Don't do it to find a job that will pay more.

The PhD problem

Related Stories