In praise of credentialism

“The stress on formal credentialism – the specific requirement for an educational qualification to be a member of a defined profession – is a phenomenon whose time has passed.”

Not sure about this. cf which I think made a bit of a ripple.

I agree very strongly with the second part on education not being a panacea for inequality.

I don’t think credentialism has to imply ‘too many young people are getting too much formal education, at too high a level’. I see it as pathology of managerialism in hiring practices which sets irrational requirements for candidates for jobs. It doesn’t have to mean formal credentials and is perhaps more typically years of experience in a specific role. (To be a barman you two years experience of bartebding etc). Unpaid internships and gap year/CV-boosting stuff maybe also qualify.

Imo credentialism is a real problem and it’s also a problem that the expansion of higher education (which I agree is a good thing) has gone hand in hand with the tightening grip of a brutally instrumental view of the purposes of education, as a process that socialises nascent wage-labourers for a life of wage labour. So I don’t think it’s too jaded to see the vaunted expansion of ‘educational opportunities’ in the last couple of decades as little more than an arms race for access to an ever-dwindling number of marginally privileged positions within an increasingly exploitative system of production funded by a burgeoning debt burden on workers.

It doesn’t have to be this way! But I fear that equating credentials (i.e. formal or informal qualifications explicitly demanded by employers for a specific economic role) with education (study, reading, learning, the life of the mind, …) may not be conducive to progress here.

I was brought up short by the discussion around “It Takes a B.A. to Find a Job as a File Clerk”.

“Someone seeking a job as a file clerk, for instance, would be well advised to acquire a knowledge of computer programs such as Microsoft Office, and an understanding of database management. This is likely to be done more efficiently in a classroom setting than by osmosis in a busy office.”

It is no doubt true that a recent university graduate is guaranteed to have acquired “a knowledge of computer programs such as Microsoft Office” (not, in general, of database management). Some of them – hopefully not many! – may even have waited until university to acquire this knowledge, through some regrettable failure of their high schools. But spending years at university, and thousands of dollars in tuition fees, to learn this is massive overkill. You could learn to use MS Office in a dedicated course in a week or two – or on your own even faster, depending on your personality. The other 95% or so of the time and money you’ve spent on your university training is going to be irrelevant to your job as file clerk. Even if we grant that it’s OK for the employer to pass the financial burden of training entirely on to job-seekers – which is kind of the crux of the problem here! – how is it reasonable for an employer to discriminate against someone for taking the obvious shortcut and learning all the skills that are going to be relevant to the job without passing through university?

More broadly, there’s no automatic contradiction between jobs now requiring greater skill and employers demanding unnecessary or excessive qualifications. Suppose a job that used to need high school levels of achievement now needs extra skills equivalent to a year’s worth of university-level study. There’s no such thing as a 1-year university degree (rightly), so the easiest solution for employers is to demand a university degree for the post – 2 years of which would be superfluous to their requirements. That requires would-be employees to spend thousands of dollars extra of their own money.

I am quite sure conservatives hate the idea of a more educated population, since it is harder for them to sell their snake oil. On the other hand, there really are a lot of jobs now that expect, if not actually require, applicants to have some college, but where that is not actually needed for the job.

It seems to me that the usual complaint about credentialism is not that education is a bad or superfluous thing, but that the credentials presently in use don’t have a lot to do with the work they are supposed to reflect, and are mostly an artefact of the education industry and the class structure of the surrounding society. One or more horror-lite anecdotes upon request.

Compared to people accumulating mountains of student loan debt and struggling to find decent jobs, this is a really petty complaint, but: credentialism also makes higher education worse for students who are interested in more than getting that degree to unlock the game progression to “a good job.”

I attended a SLAC around the turn of the millennium. I picked that school instead of a cheaper state school because I bought into the “life of the mind” campus tour messaging. My professors and fellow students would be personally engaged with interesting questions across many disciplines! It was true of the professors, actually. It was also true in smaller elective courses that nobody attended simply to fill requirements. But in mandatory courses or “easy-looking” electives there were a lot of seatwarmers who’d complain outside of class that history or philosophy or whatever was never going to be useful for a job, so we shouldn’t have to take these boring useless courses. I thought: yes, I too would be happier if you all weren’t attending these classes that just bore and annoy you.

It wasn’t until graduate school that I felt like my fellow students were there primarily due to a thirst for knowledge. It reminds me of the aphorism that a master’s degree is the new bachelor’s degree.

Maybe the purpose of a ceredential is not that it’s owner knows what is needed for the job you are offering, but is evidence of a certain level of displine and intellectual capability. So the odds of hiring someone for whom on-the-job training doesn’t stick, -or who lacks basin self-discipline are greatly reduced by requiring the credential.

One more example for the vicious cycle of educational requirements. The organization my wife spent a fair bit of her career working for had a mixture of graduates and non-graduates doing the same roles. The graduates were simply younger – hired when HR mandated (because they could due to the rise in university education) a degree. While the job was pretty skilled, she never noticed any difference in ability.

Another example would be that of clerks, where computerization has reduced the skill level required. File keeping, tracking orders and book keeping without computers is bloody hard.

Third, although I’d be the first to admit that general higher education skills are very useful in the workplace, I’m very skeptical about any classroom teaching of specific job-oriented skills.

The Engineering profession will not allow you to become a full engineer until you have completed a certain number of years in the workplace.

I sometimes think that professional class people, because they lack experience of the day to day of pink collar work and “business” education, are a little bit vulnerable to talk of new technologies, etc.

The example of this that drives me crazy is computer programmers. The reality is that most computer programming jobs are not difficult – many of them could be carried out (and sometimes are) by a smart high schooler. Generally when graduates do these jobs – they use nothing that they learnt at college to do them. They’re carried out by business, or English, or (sometimes) science/maths graduates.

Now there are definitely exceptions to this. There are computer programming jobs that are very hard, and for which you do need a Computer Science, or Maths degree, and a lot of technical skill. But these are the exception. The ‘App economy’, and the corporate IT world, largely consist of mediocre programmers – who are astonishingly ignorant of basic Computer Science concepts, but muddle through creating mediocre software.

From the employer side, though, the choice is often between five candidates who have the BA, and twenty who don’t. Since it’s more likely that the BA candidates will be capable employees, it makes some sense to eliminate the less credentialled immediately (because interviewing people isn’t free any more than on the job training is). I have been part of a hiring process where we got hundreds of applications every time we advertised, and had to grind through them looking for that one unicorn-like candidate who was worth hiring. Immediately binning anyone who didn’t follow the instructions or possess the requisite qualifications was simply essential to save time. As it was we spent probably $20,000 of staff time filtering ~600 applications down to 15 interviewees, then interviewing the 10 who turned up.

Related is the Paper Qualifications Syndrome, which was of such great concern to some in the 1980s that the ILO spent money studying it. The PQS was the trend by employers to favour recruits with formal qualifications over those without, even when those without may have had relevant work experience.

One reason employers may favour graduates for low level office jobs is that employers know that technology is no longer static. It is not merely that file clerks need today to know how to run databases, but also that such employees will need an entirely different set of technical office skills in 20 years time. Graduates, having learnt how to learn, are generally better able to cope in this environment than people whose formal education ended at 18.

I’m all for more education, of all sorts. The problem we have now with some types of education is that it’s thin, delivered in the same way all the way from maybe Grade 3 to post-secondary, and is starting to NOT signal anything to potential employers. This is especially true for several post-secondary majors like Communications, Marketing, Family Science, etc. And there are other majors that graduate far more young people than the economy can absorb — Graphic Arts, Education, Hospitality, etc. Yes, it signals that the student had enough self-discipline (and money) to persist, but is this really enough?

John, what figures are you using? Surely there are diminishing returns. I would strongly suspect that the step to near universal high school education shows huge gains. The step to say 1/3 2 year college is probably 2/3 of that gain. The step to 4 year college is probably 2/3 of that again. Past that I would be pretty surprised to find that the step from Masters degrees to PhDs are correlated with increased output per worker nearly as much as the step from high school to college. (Or even college to Masters degrees, which is probably the hardest step to measure).

It doesn’t take 4 years of college to be computer proficient enough for MS Word, or some simple database that a file clerk might use. (Programming the database would be a different issue, but that wouldn’t be the file clerk’s job.)

@Frowner, as another lurker and accounting prof, I’d be interested in hearing more. My link will lead you to contact info.

Its probably a statistical decision from the employers. Hire a grad and the odds he’s a keeper is ninety percent, hire the gal with the 2 year degree and they drop to 80%, and that smart seeming HS grad, maybe its 70%. The hassle and cost of terminating someone and starting the hiring process all over again is such that it only makes sense to minimize your chance of a hiring mistake.

I read once about an integer overflow bug in PHP that had a deployed patch that worked by seeing if the value stored in the variable was larger than the variable could hold.

I don’t think it was deployed for very long, but even so it’s a terrifying thing to contemplate. “Any high-school kid can program” is true only because the standards we expect of programmers are grotesquely low.

I’m not sure you’re getting at the serious complaints about credentialism.

Back when I had an HR/employment law focus, I would constantly see people who had been doing a job for years or decades, had great reviews, but were trapped at their current jobs because they were not unhire-able without an advanced degree. I’d see it on the HR side (someone, usually an engineering lead or scientist would want to hire someone for a mid-level technical job–say lab animal care or supply sourcing–and the HR dept wanted to stick to “the spec” which always included unnecessary degrees. I’d see it on the employee side with an often black or Hispanic male in his late 40s to early 50s asking if he was being discriminated against because he’d done the job for 20 years but wasn’t even getting call backs. (Answer, probably yes, but good luck proving it when the company can just say “we require a degree”).

I’m also really resistant to the idea that 4 years of college is necessary/helpful to file clerking type jobs. Yes computer skills are necessary that weren’t before, but that is testable, and nowadays usually picked up in high school or earlier. A lot of this seems traceable to the rise of the HR department. They are always in full cover your ass mode, and this is one of the ways.

The “get a degree and get a pay bump” phenomenon is huge in Education (in the US at least). There is even research documenting NO additional effectiveness for holders of MEd. degrees. In this instance, I think the MEd degree is simply an easy way for employers to limit the number of teachers who are eligible for certain raises.

Moz of Yarramulla:
With the “modern careers” like programming, there’s usually a huge gap between people who have the credentials and those who don’t.

As someone who has worked in the industry for 20 years, and has the credentials, I see little evidence for this. Some of the best programmers I’ve encountered have been self-educated. Some of the worst programmers had PhDs in Computer Science. I have found that the average programmers has little interest in their subject, never read a book (or blog post) on the subject unless they have to, and are not even aware of how bad they are at their craft.

Somebody else mentioned how hard it is to design databases effectively. I’m not sure I’d agree with that, but certainly most Computer Science graduates I come into contact with cannot do it. They make really bad, really basic, errors. Normalization? What’s that. The ones who do know what they’re doing, know because they were taught on the job (the way I was). The guy who taught me how to design databases didn’t even have a degree (if you have a logical mind, it’s really not that hard – I imagine a lot of clerks from yester-year would be good at it).

I’m not devaluing Computer Science education – I think it’s very difficult to be a highly skilled computer programmers without one. But the vast majority of computer programmers are not highly skilled, and don’t really need to be. Your average programmer is closer to a mechanic, than an engineer.

One final point on programming – most of the skill in programming is practical. It’s knowing how to look for and find bugs. It’s the experience to know that if you do X bad things (or sometimes good things) will result. These are skills that you learn doing the job, and are almost impossible to teach in college.

I’m with 35. Vetting is what we want. From the wiki:
investigate (someone) thoroughly, especially in order to ensure that they are suitable for a job requiring secrecy, loyalty, or trustworthiness.
“each applicant will be vetted by police”
Our system failed with our “So Called”. Time to change our political vetting system.

Speaking as a hirer – I agree with a mix of the comments made both in the paper and in comments.

My company hires dozens of interns every year, all from top colleges and all with top credentials. In my experience, the sorting factor normally has to do more with the individuals willingness to work hard, take direction and be conscientious in all that they do whether it is getting sandwiches from the deli downstairs or contributing to a major project.

The credentials themselves usually contribute more to them having a vocabulary and an understanding of what we are trying to accomplish than any first-hand experience with the task at hand.

Someone up-thread mentioned an accounting background – I’d agree that little of what one learns in an accounting class applies in day-to-day application. But if I hand over a stack of annual reports and say calculate the recievables/payables ratio for all of these companies over all years presented and put them in an excel table sorted best to worst – based on rolling 3 year averages I expect the typical accounting major would know what I’m talking about and could probably get it done with about 10 clarifying questions and on average a history major would need far more clarifying questions to get to the same point. Without question, I could teach the history major or even I high schooler what they need to do if they were any sort of student at all….but since there will be 100s of things to train either of them on every one that can be saved the better.

The fact is hiring is never an “all else equal” proposition between any two or even 10 candidates – every hire is a trade-off. The credentials are just one sort in the process that hopefully helps. Someone else above mentioned batting average or percentage of success – that’s certainly part of it. Sorting, improving the odds….then still its some amount of intuition, learned (or unlearned) biases and luck whether the hire winds up being everything hoped for.

Indeed, my experience tracks with SJ’s. University is, in most cases, a poor place for vocational training, and the increased stress placed on degrees (first Bachelors, then MBA/etc) means that the University’s ability to do what it’s occasionally good at – liberal educations and teaching critical thinking skills – is increasingly eroded, all while student debt becomes a bigger and bigger problem. The Modern University benefits a lot of people, but students are at the very bottom of that list.

“In reality, though, the newer professional occupations associated with the information economy, such as programmers, systems analysts and social media experts, have no such formal structure.”

Okay, I take exception to a number of things you’re saying here. I’ll take the side against what you’re calling “credentialism”.

I work as a computer programmer. I have multiple technical degrees, but none specificially in programming. I’ve had some formal training in programming (maybe 3-4 classes), but not that much. So, to a large extent, I’m a self-taught and/or learned-on-the-job programmer: a lot of us are.

There’s an academic discipline called “Computer Science” which likes to think it owns this field– they’ve decided that certain things are Fundamental, like say searching and sorting algorithms. They behave like this is the F=ma of the field, something that everyone must know to be competent, but why that’s so is difficult to grasp. CS researchers did indeed do some solid work in this area, but the things that they learned are embodied in standard libraries, and someone working in the field pretty much just uses the libraries. Most of us have only a vague knowledge of the issues involved, and are happy to leave it that way: we feel like we could learn it if we needed it (and probably we could), but it’s an odd specialty, not something Fundamental.

“Computer Science” itself seems like a very odd field to me: it’s dominated by mathematicians who don’t do any science (e.g. no one conducts research to see if a new language has any actual benefit out in the field), and they don’t do all that much with computers. Philip Greenspun once complained that the killer-app of Computer Science was apparently Microsoft Word.

Google was founded by two guys from Stanford’s Computer Science Dept, and since Google is riding high, there’s now a fetish for asking what I would call Computer Science trivia questions on job interviews (note that Microsoft and Apple were founded by hobbyist dropouts). There’s now a small sub-industry similar to the SAT-prep racket that help you cram this material in order to make it through job interviews. There’s an argument that these interviewing processes are essentially affinity tests in disguise: the question they’re designed to answer is “Are you like me?”.

The point that I’m making here is that CS-geeks like to believe that they’re in touch with the one true doctrine of programming, but there’s no evidence to that effect: you would have to get some social scientists to study the productivity of programmers trained in different ways, and the tradition in the field is to rely on argument from authority and wave your hands a lot.

So I’ve got two points here:

(1) Yes, the field I’m in has been getting by quite well without much in the way of credentialism– but it could easily be that that’s changing.

(2) If you standing up in defense of requiring credentials, my question is *which* credentials, and who gets to decide?

(Ah, I see I’m largely in agreement with Cian@45)

I found a lot of the comments above very interesting and helpful in further clarifying this issue in my mind. Following up on that, let me propose a definition of the apparently oxymoronic notion of “too much education”. A person is getting too much education when they take a course such that:

a) they don’t have any independent desire to learn what’s being taught
b) they don’t need to learn what’s being taught in order to accomplish their goals (for their careers or otherwise)

A person in such a situation is going to be understandably resentful, suffering from both the cost and the opportunity cost of the course, and is usually also going to make life worse for any fellow students who actually do want to be there (as noted by Matt above). Why would anyone be in such a situation? Well, clearly because:

c) in order to make any progress in their career, they DO need a piece of paper certifying that they’ve taken the course and gotten a satisfactory grade in it.

The comments above of Moz of Yarramulla, Sebastian H, EB, Sonny Jim, Dr. Hilarius, and probably a few I’ve missed all seem to confirm an impression I already had: that quite a few people are unfortunate enough to find themselves needing to get “too much education”, in this sense.

For educational institutions, “too much education” is a poisoned chalice. In the short term, many of us benefit from the fees of people in such situations, which (hopefully) serve to cross-subsidise people who actually are benefiting from being there. In the long term, such situations risk creating ill-will towards education in general, and helping erode public support for it.

LFC@49
If your friend went to Harvard, he was particularly thick or disinterested. Until a somewhat recent change in rules, 91% of Harvard College students graduated with honors. That “scandal” caused Harvard to cap honors at 60% of the student body. An A is the most common grade. I’ve linked to one of the many articles on this subject.
https://www.bostonglobe.com/metro/2013/12/05/with-its-most-common-grade-harvard-earns-disapproval-but-has-company/kCeheDYfuDjSRcM1sVljfK/story.html
The push back of corporate clients against using first year lawyers at top firms is also well known and has been going on for some time. It’s considered paying for training despite the Ivy law degree.
https://dealbook.nytimes.com/2014/10/10/corporations-drive-drop-in-law-firms-use-of-starting-lawyers-study-finds/

My company hires dozens of interns every year, all from top colleges and all with top credentials. In my experience, the sorting factor normally has to do more with the individuals willingness to work hard, take direction and be conscientious in all that they do whether it is getting sandwiches from the deli downstairs or contributing to a major project.

Nonsense. The sorting factor has to do with class. Mostly you’re selecting people from the upper classes, because that’s who mostly goes to top colleges. And that works for your company because they fit in to the culture.

Someone up-thread mentioned an accounting background – I’d agree that little of what one learns in an accounting class applies in day-to-day application. But if I hand over a stack of annual reports and say calculate the recievables/payables ratio for all of these companies over all years presented and put them in an excel table sorted best to worst – based on rolling 3 year averages I expect the typical accounting major would know what I’m talking about and could probably get it done with about 10 clarifying questions and on average a history major would need far more clarifying questions to get to the same point.

1) None of these things take 4 years to learn.
2) Implicit in this is that your company has no interest in training employees, even though from a social perspective it would be far more efficient for them to learn these skills while working. But companies have little interest in doing this because employees are seen as disposable assets.

Being a file clerk requires database management skills? I have been employed as a file clerk twice, in the 1990’s. It involved shoving papers into file folders and finding said folders for people. Whatever job you are describing isn’t a file clerk. As far as written communication skills are concerned, shouldn’t a high school graduate be capable of that? As it happens, I do have database management skills, SQL and Access. I am a data analyst, although my current job is accounting. And I learned none of it in college. I have a BA in History. Classics, mostly. I find the article at the link completely unpersuasive.

I was just looking through some of my old stuff on credentialism, and I’m reminded of the case of Freeman Dyson: he’s a physicist who did work in a number of areas, for example on the foundations of Quantum Electrodynamics back in 1949, and he did this all without earn8ing a PhD in the field. He’s commented since then that he doesn’t think anyone could do what he did back then: credentialism has a tighter grip now– at least in the field of physics.

I suspect that there’s no society-wide trend one way of the other, but rather it’s a matter of the maturity of a field– in the early days, things are more wide-open, later on professional societies get a grip on things and set themselves up as gate-keepers.

Anyway, if anyone cares, here’s my (excessively long) piece on credentialism, though the emphasis is more on “who do you believe?” rather than “who do we hire?”:

http://www.dailykos.com/stories/2015/11/6/1446086/-Anti-Nuclear-Amory-vs-Atomicrod-and-Other-Fun-with-Credentials

“Credentialism is at best a very quick-and-dirty method of evaluating a speaker, and elsewhere I’ve argued that things like this are useful and absolutely necessary, and yet they should also ‘leak’, we should be on the lookout for cases where exceptions deserve to be made.”

I think Cian@56 is kind of all-over-the-place with his response, and I would guess there are reasons for this: the “Computer Science” intellectual hegemony is pretty strong, to the point were it’s hard for people to think about a challenge to it…

“So Joseph is wrong. Computer Science is essential to the field – take it away and there’s nothing (I am comfortable defending this position). But he’s also right, you can get by just fine without them. Well sort of. ”

The thing is, I’m not claiming that the CompSci folks have never done anything useful, what I am saying is that the idea that they’re the sole gate-keepers of what’s useful is very dubious. (Analogy: “Where would your surgeons be without us mechanical engineers? Clearly you should come to *us* for medical advice!”)

Exploring an alternate reality scenario where computers were developed without much guiding academic expertise (steam punk?) is interesting, but probably beyond the scope of blog commentary. It strikes me as possible, but it could be I’m missing something.

“Because the flip side of this is that most jobbing programmers are terrible. They’re the equivalent of the guy who comes to your house and builds a bathroom that leaks. Typically they don’t know what they don’t know, which is dangerous. I regularly encounter programmers (and plenty of them have Computer Science degrees) who don’t know how to design a database, or write code that is dangerous/unstable.”

Sure, definitely. Fundamentally none of us know what we’re doing, because we don’t have any solid way of knowing what we’re doing, because the “Scientists” in this field aren’t interested in doing any science to find out, they just assert that they know already, and you’re supposed to trust them.

We try stuff, we fall in love with what we know, we sneer at what we don’t know, we wave our hands a lot and shout at each other, sometimes we trust experts, sometimes we dismiss them as unrealistic, sometimes we chase after marketing hype, sometimes we sneer at it with knowing cynicism.

Yes, PHP is a botched job– and yet, wikipedia. So what do we really know?

And against PHP, I give you Pascal, which was once the Latest Thing that was going to save us all– but it was an attempt at creating an idealized, isolated world of perfection, where even any form of I/O required a non-standard extension.

LCF@49 & Joseph@550 & JQ

LCF–My response may have been a bit strong. However, certain fields and professions tend to be more susceptible to signaling through credentialism than others. Law certainly is one. Just take a look at the SC. Or top ranked law schools for that matter. The judges and faculty are typically from 3 or 4 schools.

Joseph — Your point about CS interview-prep courses is interesting and well taken. A not small share of the credentialism is socialization. You gotta act like you belong.

JQ-What I find troubling about many of these comments is that practitioners in technical fields find credentialism run amuck. Those are the fields where I thought credentialism would be most justified. Thoughts?

One thought – much or most of the selection process has now been outsourced to specialist firms, or to in-house HR areas. These people typically have little or no knowledge of the actual attitudes or practical expertise involved, but they can screen CVs for what look like relevant credentials. So that’s what they do.

JQ – “every country in the developed world”? Not Germany or Austria, nor (I think) Japan.

I tried to act on the implications of this belief, with considerable success. HR put a stop to it and went to credentials, with less success. Last I heard, HR still ran the process. Money can be left on the table if it’s the wrong kind.

In Computer programming there is an expectation that everyone goes to college for four years. Which is equally ridiculous. Most programmers, the vast majority, need very little theory. This doesn’t mean everyone can be a programmer, anymore than everyone can be a bricklayer. It just means that the skill required to be a jobbing programmer is not particularly academic. And most programmers are jobbing programmers.

Is this expectation referring to large firms? I’ve been writing software to pay the bills for 11 years, but never at large companies. At the beginning of my career I did get an offer from a Fortune 100 company whose core business isn’t software-related. That was actually the easiest interview I’ve gone through in my working life. It was more like an hour’s chat with 5 other people and they didn’t ask any algorithmic, systems design, or brainteaser problems. That job posting did have a credential requirement. But I ended up taking a different offer at a small company.

I haven’t worked anywhere where our job postings mentioned any educational requirements. Nor have I been in an interview, on either side of the table, where academic credentials came up. About the closest I’ve seen to credentialism is challenging candidates with questions that are easier if academic coursework problems are still fresh in your mind (write a function to find the Euclidean distance between two vectors, write a function to search a binary tree, write an efficient function to find the intersection of two lists…) There’s a lot of interview prep advice about handling these whiteboarding challenges precisely because you’re expected to demonstrate on the spot; it’s neither necessary nor sufficient that you did the same thing in a classroom a few years ago and got a credential for it.

If the GE grapevine is accurate, Jack Welch questioned the existence of the long-time GE Advanced Engineering Course and other in-house courses, with the straight-from-the-gut reaction: “Why don’t we outsource the training to the universities, and only hire engineers with Masters Degrees?”

At the working level, we knew that experienced technicians and specialists (with 2-year degrees) were worth twice their weight in inexperienced engineers (4 or more year degrees), but in the semi-annual layoffs, the subsection and section managers often started with the lower ranks, assuming an engineer could do anything a technician or specialist could do, but not vice-versa. The best engineer I ever knew was a specialist with a two-year degree (Bob). One day in the midst of a big layoff, our secretary told Bob and me, “Al [our manager’s manager] wants to see you in his office – now.” As we walked side-by-side up the stairs nervously a comforting thought hit me: they might lay me off, but they would never lay off Bob! Sure enough, Al had a job for us.

Two weeks later, while Al was on vacation, his boss had to make some quick decisions to meet Welch’s quota, and laid off Bob.

The rapid rise of the favored few into upper management, where the decisions are made, often means they don’t know or care how things are at the working level. As they used to say, nobody ever got fired for buying from IBM. (I’m not sure that is still the case, but the same kind of thinking goes on.) And don’t get me started on “Human Resources”.

Peter says: Not Germany or Austria, nor (I think) Japan.

Germany is an odd example against credentialism… It’s notoriously full of Herr Doktors in high positions.

Keep in mind that Germany’s apprenticeship programmes are a form of credentialism. They’re on the job training with an added certificate from a school

I’ve seen such programmes lauded by Americas as an alternative to college or classrooms, but in practice it’s often the opposite – it’s adding classes and credentials to a training trajectory to would otherwise be purely on-the-job in entry level jobs.

No you’re exactly wrong about the racism/sexism issue because you’re over trusting “demonstrate its necessity”. Third party credentials almost always end up covering your ass so long as the business can show some reason for believing that the credential could be related to some of the tasks of the job. It is an extremely effective way of masking discrimination from legal scrutiny. Until a few years ago that was my area of practice (in California one of the most employee friendly states). Your idea of how the law is supposed to work is correct. Your idea of how the various legal tests actually get applied is not.

Re small employers I can only offer anecdotal evidence. I worked with and against a broad number of employers–it’s my impression that if you are a skilled worker without a credential you either are a long time legacy (15+ years with a company) or you work for a small company. From a statistical point of view, I wouldn’t be surprised if as the number of college grads went up, the number of college grad employees working at small firms went up purely as a function of population sampling. But I’d suspect in that causal order.

Do you have any clear stats on it?

Re specific type of employment structure I’m not sure what you mean. I mean: any private firm large enough to have an influential HR dept. So most firms over 30 employees or so? Maybe 40? That’s not everyone. But enough to make a huge difference.

This despite the contrast between the obvious prevalence of racism and sexism and the evidence (for example, this thread) that hardly anyone supports credentialism.

This is pretty clearly wrong. Few will say ‘I support credentialism’, but, until the recent rise of the alt-right, few would have said ‘I support racism’.

There’s a purely linguistic thing where supporters of an idea take ownership of a negative label for it. Whether or not that has happened yet changes nothing. The map is not the territory.

Otherwise, if your article did persuade lots of people to change their mind and say ‘credentialism is good’, that would be a self-refutation.

T@75:
Actually, law schools have an excellent way of discriminating between Ivy undergrads with 4.0 GPAs. They strongly favor the “tough majors:” physics, math, philosophy, EE. A 3.6 in physics is worth a lot more than a 4.0 semiotics with the same LSAT.

The thing is, I’m not claiming that the CompSci folks have never done anything useful, what I am saying is that the idea that they’re the sole gate-keepers of what’s useful is very dubious. (Analogy: “Where would your surgeons be without us mechanical engineers? Clearly you should come to *us* for medical advice!”)

That’s a terrible analogy. A better one would be – where would you mechanics be without us mechanical engineers. Should engineers be the gatekeepers for mechanics. No. Would mechanics even have a job without the work of engineers. Clearly not. Could a mechanic do the work of an engineer. Well no.

Exploring an alternate reality scenario where computers were developed without much guiding academic expertise (steam punk?) is interesting, but probably beyond the scope of blog commentary. It strikes me as possible, but it could be I’m missing something.

A few of the things that Computer Science has given us:
+ Databases
+ Pretty much every aspect of Operating Systems
+ Almost every innovation in programming languages
+ Distributed Computing, and the tools to make it possible
+ Real Time Computing
+ Advances in processors
+ Computer Graphics
+ Artificial Intelligence techniques
+ Computer Vision
+ The algorithms and datastructure in the libraries that you rely upon, and the tools to reason about them
+ Networking
+ CPU advances (shared with Electronic Engineering)

Yes some of these advances happened in academia, and some happened in industry, but they were found by people trained in the theory of Computer Science, and who then applied that theory to advance the field. Your ignorance does not define the field.

Your average jobbing programmer doesn’t need to know this stuff, because your average programmer is not doing anything innovative. But that doesn’t mean these things aren’t important. Oracle hires Computer Scientists who work on their database engine because they have the background to understand that stuff, and just as importantly they’ve demonstrated that they have the capacity to deal with complex intellectual problems that most people cannot do. You need very little of that if you’re hired to be a DBA for a fortune 500 company. The skills you do need are not taught well in a university, and probably cannot be taught well in a university.

Sure, definitely [most programmers are bad]. Fundamentally none of us know what we’re doing, because we don’t have any solid way of knowing what we’re doing, because the “Scientists” in this field aren’t interested in doing any science to find out, they just assert that they know already, and you’re supposed to trust them.

The PHP example that someone cited above is the kind of thing that I’m talking about. A competent programmer doesn’t do stuff like that. I’ve recently been working with a database that is not normalized, and where IDs are stored as strings, among other bizarre stuff. Again, competent programmers don’t do that. The inudstry is littered with programmers who don’t understand how to effectively profile code, or how to avoid deadlocks, or to minimize garbage collection/memory usage. Who don’t understand how to write clean maintainable code, basic techniques that help keep code bug free (such as minimizing mutable code). Programmers who rely upon inheritance, rather than composition. Knowing how to write loosely coupled code. I’m talking fundamental best practices here. Stuff where any reflective experienced programmer knows the right way to do something.

Programming is a craft (and one that academic Computer Science has very little influence over, despite your claims). It’s an embodied experiential practice like wood working. You learn by doing, observing your peers, and reflecting on your mistakes.

Yes, PHP is a botched job– and yet, wikipedia. So what do we really know?

That Wikipedia would be a lot easier to build and maintain in a better language. It would also be more performant (as in it would be cheaper to run). Facebook has wasted the best minds of a generation (as well as a small fortune) trying to make PHP suck less.

And against PHP, I give you Pascal, which was once the Latest Thing that was going to save us all– but it was an attempt at creating an idealized, isolated world of perfection, where even any form of I/O required a non-standard extension.

Pascal was developed as a teaching language to teach students structured programming and data structures. That was its entire purpose, and it was extremely successful. So successful that it was turned into a ‘real’ programming language by others. Turbo Pascal and Delphi were both very successful (and very good) programming/development environments based upon Pascal.

Niklaus Wirth, it’s designer, invented modern structured programming. He created procedures, and techniques for structuring data, that we now take for granted. He developed (and popularized) basic ideas in programming techniques (the craft of programming) that are now taken for granted.

And against PHP, I give you Pascal, which was once the Latest Thing that was going to save us all– but it was an attempt at creating an idealized, isolated world of perfection, where even any form of I/O required a non-standard extension.

Sure, but that was the done thing at the time: IO frameworks differed pretty significantly between OSs before the spread of C forced everyone to use Unix’s radically-minimalist approach.

… which means, of course, that the problem you mention was actually fixed by computer scientists after being caused by software engineers. But what did the computer scientists ever do for us?

Fortran (officially no longer shouted) is doing quite a lot of hard work in the middle of a most big numerical calculations and simulations. It’s probably still the best tool for niche it was made for (although painfully unfun to code in)

JQ — A a general matter, you must be correct. The wage premium foe a college degree is huge and increasing. I don’t think it’s a mass delusion that produces this result across countries and cultures. I think the actual in increase in knowledge and expertise, especially in technical fields, is enormous despite the inherent weirdness of that specialized knowledge/skills in programing. The easiest way for someone to get from a bottom 20% income family to get into the top 20% is an engineering degree. Given the enormous disadvantages faced by those students it would be folly to argue that credentialism was not he driving force behind their success.

That said, there are a lot of anomalies as expressed in many of the comments. And given the costs of higher education, it has become much less of a leveling force in the US — compare to the GI Bill years which really did get the masses into good colleges and was an impetus for creating good colleges such as the university system in California.

I have a Fortran compiler on my PC. Fortran still works, and is the best language for scientific and engineering models. Other languages may be better for sorting and filing data, but are much harder to learn than Fortran.

Two Fortran classes and a physics degree got me a job as a computer programmer at GE out of college. Then I took the GE Advanced Engineering course (one 20-hour problem a week for 2-1/2 years, each requiring a computer program to solve) and they told me I was an engineer. For the next 35 years I averaged about one Fortran program a week to solve engineering problems, such as heat transfer in a 50-ton steel casting, or the vibration modes and frequencies of a 52″ turbine vane.

These days there are general-purpose programs to solve engineering models, such as ANSYS, and most engineers don’t do much custom programming. I’ll bet NASA and other places that do leading-edge work still have some custom programming, though.

When I started hearing about languages like C++, I was told it eliminated “spaghetti code” which was considered rife in Fortran, by getting rid of the GOTO statement. I got a book “C++ Complete” from a library, and lo and behold, in chapter ten, I found … the GO TO statement – “not recommended, but in some cases the most efficient way to exit from a complicated procedure, such as nested FOR-loops”.

Having worked with Brownian dynamics and boundary element method codes over the past few years, I can say: Fortran lives to torment grad students and postdocs!

JQ: “My takeway: A bright high school graduate in the 1970s would have been far better off doing a Comp Sci degree than learning COBOL by sitting next to Sally.”

The skill involved in most programming has almost nothing to do with a particular programming language, and people who are good programmers easily pick up new languages as they emerge. I worked with people who were writing COBOL code on IBM mainframes in the late 80s, and then writing C code on UNIX platforms in the late 90s, and then writing multi-platform Java apps in the late 2000s. Personally, I spent 30 years in the business, the last 15 as a CIO, and I was a history major who dropped out after one year of college. If you’re not actually designing chips, a comp sci or engineering degree is irrelevant. Of course, now, no one without an MBA could get the jobs I had.

JimV:

(Continuing the side discussion of programming languages….)

When I started hearing about languages like C++, I was told it eliminated “spaghetti code” which was considered rife in Fortran, by getting rid of the GOTO statement. I got a book “C++ Complete” from a library, and lo and behold, in chapter ten, I found … the GO TO statement – “not recommended, but in some cases the most efficient way to exit from a complicated procedure, such as nested FOR-loops”.

Whoever told you that was insufficiently knowledgable about FORTRAN and C++.

In 1977, FORTRAN 77 added IF statements that could have more than one statement in an IF clause and also provided ELSE IF and ELSE, obviating the need for a huge number of GO TO statements. C++ came out in 1983, so FORTRAN 77 had already been out for 8 years or so.

And C++, of course, has a “goto” statement, and always did, as you discovered.

So it’s really a case of “Switching to FORTRAN 77 from older FORTRANs lets you get rid of a lot of spaghetti code previously required in FORTRAN, because it supports block IFs, although F77, like C++, still lets you write goto-ridden spaghetti code if that’s your inclination”.

And these days, Fortran is apparently an object-oriented language, along with COBOL.

But, with some effort, you can write structured code in FORTRAN IV, and, without too much effort, you can write unstructured code in just about any programming language that still has gotos, so it’s more a question of “what did you learn about constructing software?” rather than “what language did you learn?”, as Layman points out.

(There are probably ways this can be tied to the broader question of credentials, expertise, etc.)

I’d be interested to hear about the situations with regard to formal qualifications in China and India. Not so much the degrees, but the mass of certificates that everyone in Australia and the UK is expected to pick up.

So after 20 years as a mediocre programmer, one of my few consolations is to come on CT and see all the Humanities academics who probably think anyone who can write a “Hello World!” app and solve a quadratic equation is a technical whizzkid.

Now it turns out they don’t. Turns out that *even the CT crowd* are fully briefed on how mediocre people like me are. That’s how bad it is. That’s how useless my pathetic “skill set” has really sunk.

I’d be interested to hear about the situations with regard to formal qualifications in China and India. Not so much the degrees, but the mass of certificates that everyone in Australia and the UK is expected to pick up.

In India it’s worse I think.

Those Java/.Net certificates are worse than useless. In my darker moments I used to think they were a way of identifying people who couldn’t program.

There’s a lot of discussion of programming in this thread. I wonder if that’s because the CT demographic tends heavily towards computer programmers (and so see this as a example that they know about), or because it’s a particularly good example of the academic qualification (a degree in computer science) not being a reliable indicator of competence to do a particular job (e.g. work as a programmer).

Probably a bit of both. My experience of hiring programmers is that credentials are useless. I’ve seen people with first class honors (or whatever the US equivalent is) who couldn’t program their way out of a paper bag. I’ve note noticed a particularly strong correlation between test taking ability and programming ability.

Does anyone here have experience with Brainfuck?

Back late… I’m qualified as an electrical engineer with a masters in software engineering because at the time a PhD made one less hireable in more places than it helped, once you stepped outside universities. Credentials, I has them. Also, I am old in programming job terms – over 40! I have wandered aimlessly round the industry for a long time doing contract work and taking holidays. Also, skilled at lying to potential employers about what I’ve been doing lately, because industry have a strong bias against people who don’t have jobs.

In my experience, from companies ranging from “me and the boss in his dining room” to a major bank (did not enjoy!):

* larger companies usually have HR departments who behave as described above
* small companies will more readily hire school leavers than large ones, generally if they know them (as they prefer to hire in general)
* small companies have much greater risks – if their one employee is useless, they are likely to be in trouble.
* but the rewards… if a small company hires a school leaver on $40k and they’re great, that’s a big win. I do have one such friend who was getting $100k when his friends were finishing university. Sadly, Ruby On Rails is not so popular any more and he’s back on $60k while trying to teach himself a useful skill. The company that hired him out of school couldn’t afford to keep him on once their niche shrank.
* many companies in Australia and Aotearoa will not hire a PhD, full stop. They’re useless academics. Master’s Degrees are iffy, but an Engineerings Master’s is less awful than a Computer Science one. I have left the ME off applications at times. Now I mostly leave my age off.
* on that note, age discrimination is *huge* in IT. There’s a huge bias towards the 2-10 years experience (2-5 in many case) cohort, and after 40-ish it becomes noticeably hard to get work, even through contacts (and the bigger the company the less your contacts matter, because there’s more bureaucrazy between “want that one” and “hire that one”).

After having it done to me, I started requiring a programming test involving actual programming when I have control of the hiring process. It weeds out duds like nobody’s business. Sure, they *claim* to have two years experience with ProgrammingTool Version 3, but sit them in front of it and say “here’s an outline. Add this functionality” and a disturbing number head straight for google. I’m talking “add a button and make it add some text when clicked” type problem. It takes me a couple of days to build a set of 10 problems like that and test them on my coworkers. But it very effectively filters out the few applicants that can actually program at all. Especially with “senior software engineers” or whatever, where you’re asking for five years experience, not being able to operate the basic tools is a huge give-away that you’re not going to be any use on the job.

The flip side was the candidates who gave us the “are you kidding” look and flew through all 10 problems in minutes, mostly slowed by a tendency to say “that’s not a good way to solve this problem, I would do … because …”. Which is actually a good sign, IMO. We hired those ones at whatever pay they asked for (they hired me that way). I am frankly amazed that so few companies do that type of practical test. The online tests don’t work because you can’t see them using the tools, all you get is code pasted into the web form. You don’t even know that it’s actually the candidate writing the code. Get them right in front of you, where you can see how they react to getting stuck, being given stupid tasks, what questions they ask, how they solve problems and so on. There’s a whole lot of “what are they like to work with” being examined as well as “can they program at all”.

In one workplace we used to joke about “failing the lift test” because we had so many applicants who just didn’t show up for interviews. The test was “can you operate the lift”, a snide explanation for people who for one reason or another couldn’t get to the interview but were too ashamed to tell us, they just vanished.

There’s a lot of discussion of programming in this thread. I wonder if that’s because the CT demographic tends heavily towards computer programmers

It’s probably also because a lot of programmers feel strongly about credentialism, especially since most have probably worked with both “really smart self-taught” bucketheads and “first class honours from a really good university” doofuses. There is not, AFAIK, a good way to distinguish good from bad programmers using only a CV. Let us now vent about how awful the hiring process is.

Also, the field has a lot of “recreational thinkers” in it – people who constantly think about ways to improve processes, especially processes that don’t work very well. Like, say, hiring programmers.

It looks like it’s all programming talk here now so I won’t feel bad about continuing that thread. The first software job I actually took* was with a small Bay Area startup. I had no prior experience in industry and the job posting didn’t ask about education. The posting did challenge applicants to code solutions for a few not-too-hard problems in the style of Project Euler. My solutions got me a phone interview, then an in person interview and a job offer which I took.

For people who already live near a high concentration of software jobs, and/or have an impressive CV, it probably isn’t worth the extra time and effort to solve challenge questions. But that particular screening mechanism let me get my foot in the door without a prior software job, and in exchange the company got someone (me) who was willing to work significantly below market-norm salary.

I work remotely, living in one of the cheapest cost of living areas in the US. With that first job I felt like I was making out like a bandit over my in-office colleagues even if I was taking home 30% less pay. It was still more than double what I was paid as a researcher, and my mortgage was half the cost of their two bedroom apartment. I’m still bad at negotiating “market” pay — when initially hired in my most recent job I asked for what I thought was a cheekily high salary, and they countered with a higher number — but I feel affluent to the point of guilt. The salary someone needs in the SF Bay Area just to afford an unremarkable 3 bedroom house will put you on easy street in any place where real estate remains affordable. For a future economics post perhaps John can take a crack at this mystery: why do software companies continue to cluster around the SF Bay Area, where real estate prices are ridiculously high and local tech workers expect pay high enough to compensate?

*As an adult. After I wrote the above I remembered that I wrote software for medical data extraction and visualization (paid internship at a research hospital, not much money) the summer between my junior and senior years of high school. It paid for a trip to Italy and a computer I could take with me to college.

Story I heard (late 90s) was that if a graduate had a job offer from IBM, and decided not to take it, they sent out a form for you to fill in explaining why. Because it was so baffling that anyone could turn down the opportunity.

In those days there were quite a lot of Humanities grads getting in to graduate training courses to become programmers in C++. The company I joined also assigned 1 guy to learn COBOL, wonder what he’s doing now.

Another reason that there might be so much discussion of programming, is that programmers see examples of ‘market failure’ every day in hiring decisions (and are often on the wrong side of them).

Consequently they might push back strongly at John Quiggin’s implication that such things are impossible because markets.

I mean seriously, do you know how many top notch programmers there are out there over 40 who struggle to get hired. Or how awful most programmers with 5-10 years experience are. But it’s the latter who get hired.

So for those of you who are professional programmers (I’m not, but I write some tools for my research), is it true that many CompSci graduates couldn’t deal with something as simple as the FizzBuzz test?

I’ve heard the legend, but find it hard to believe.

Maybe I’m being dumb but I’m not totally sure why using google to do your job is a mortal sin

“is it true that many CompSci graduates couldn’t deal with something as simple as the FizzBuzz test?”

Sounds like self important bullshit. It might take a couple of minutes for the interviewee to recover from the shock of being asked an idiotic question – a question which can at worst be answered by aid of a one minute google lookup to refresh oneself since most normal human beings don’t see the need to store every single bit of trivia in their heads, life and work not being a series of interviews – but that’s about it.
I think a more serious problem is that some/many software engineers take a very cavalier attitude to com-sci theory. Case in point would be the commenter above who seems to think that it is unimportant to understand the science behind “searching & sorting” because there are products out there which supply commodity algorithms to do the work. This is incorrect and dangerous. It’s the kind of thinking that leads people to build the unnormalized databases that another commenter testified to encountering – the engineer who did this obviously didn’t understand relational database theory, query optimization, information retrieval, disk storage management etc – which are at bottom all about “searching & sorting” algorithms.

I’ll also add that while software engineers don’t need to be mathematicians they had better be versed in (or be prepared to learn) what Donald Knuth calls “Concrete Mathematics” ie number theory, linear algebra, calculus, real analysis, probability, measure theory, combinatorics etc. Sooner or later you are going to need it.

None@124 software engineers don’t need to be mathematicians they had better be versed in (or be prepared to learn) what Donald Knuth calls “Concrete Mathematics”

Very much this. I once had the experience of working with someone who had failed both the programming design courses in their degree – they kept the transcript/certificate framed on their desk to prove it. At one stage I was asked to help with a program they’d written that was a bit slow. Thanks to a lot of linear searches and very bad design choices they’d produced a tree-view that was slower than O(N^30) (I stopped counting). For the non-programmers: to draw an item on the screen they looked it up in a list, by starting at the beginning of the list and checking each item “is this the one I want”. That’s a linear search. But they had a tree, so they needed to know what the parent/owner of the current item was… they did another linear search. Then they needed to know how many children there were. Another linear search. And so on, for a range of other attributes.

The know solution to this, well, ok, some of the many known solutions to this include: smarter searches, keeping the information once you have it if you’ll need it again, … not being a complete idiot. Thinking about that experience makes my head hurt. It’s almost as though the person had actively tried to make the worst possible program.

JQ@128: the OP clearly rejects the idea that you need specific credentials, as opposed to post-school education in general, for most professional jobs.

I almost think my anecdote suggests you can go the other way, and say “specific qualifications are not sufficient”. Whether they’re necessary is another question.

To almost argue against myself, I have taught programming to engineers and later experience with programs written by engineers helped convince me that they desperately needed formal training in the field. “some college” does not cut it when you’re writing large and complex software systems.

I don’t know about other fields, but software is very prone to the Peter Principle – people keep extending their program until it exceeds their ability to understand it, at which point they struggle to maintain it. So that program lingers around being maintained by the original author until it is completely replaced by a new program written by someone else. I think that’s actually the definition of a mature piece of software: so complex that the current maintainers struggle to do so. If you find a more able programmer to maintain it they will simple add complexity until they reach their personal limit. You may get extra functionality as a result, or not.

John Quiggin 03.08.17 at 2:23 am @ 128 — The kind of people who are likely to be reading and writing in the comments section of Crooked Timber are likely to have had at least a modest brush with college-level education. Also, after around 1980 it was difficult to get a job in the field without having or being able to simulate a degree that HR departments would accept, regardless of one’s abilities, proven or not.

@129 — although my examples were anecdotal and I can’t document them, I strongly believe they would reflect actual data, if there is any reliable data. On the other hand, I suppose there might be equally high levels of incompetence in many other fields.

“you need specific credentials, as opposed to post-school education in general, for most professional jobs.”

Not sure what you mean by credentials here. Paper attesting to expertise in some specific domain? Some credentials are that. Others are remarkably non-specific (MBAs? PPE degrees?) but obviously open a lot of doors.

Surely in most professions you need theoretical and practical knowledge, and the latter is always in some specific area. An accountant with no ability to use a spreadsheet is not much use (but Excel is not the same as Lotus), nor an engineer without detailed knowledge of the practical idiosyncrasies of materials (but steel is not stone)….

Software points this up, since there are so many languages, many use-specific. As a manager overseeing IT, I found out the hard way that people may have computer science degrees but be helpless if the problems are outside their specific craft skills.

JQ: “So, there are a lot of computer programmers (and related professionals) commenting here, but as far as I can tell, none without at least some college. I’ll count that as support for the OP.”

If two semesters as a history major, with a failing GPA, count as ‘credentials’ for a job as a programmer, then ‘credentials’ is meaningless.

For several years, I ran a quant (mathematical programming) group in the City. I hired mainly PhDs in maths, physics or engineering. What they knew about their speciality was mostly useless to me: their ability to learn it was what I wanted.

Latterly, various universities started offering Master of Finance courses. There wasn’t much wrong with the courses themselves, but the people who took them were paying up in the hope that the credential would get them a job they wouldn’t otherwise be offered. I never hired one of them.

One obvious possibility is that such arguments could be used to justify changes in hiring policies, rather than in educational policy – I assume at least some HR managers are amenable to rational persuasion…

So obvious, indeed, that you’re pointlessly repeating an observation I already made.

However, as far back as I can remember I’ve always been hearing Western governments call for the number of university graduates to be increased, usually for explicitly economic reasons. Tony Blair’s goal was “50% of young adults progressing to higher education by 2010”. Obama’s was that “by 2020 the United States should once again have the highest proportion of college graduates in the world” in order to have the “most competitive workforce in the world”. If “institutional education is systematically over-valued” by companies, that would suggest that this widely shared educational policy goal is misguided, and should be abandoned.

So what change in policy does that translate into? That politicians should say different things in their speeches?

John Quiggin:
So, there are a lot of computer programmers (and related professionals) commenting here, but as far as I can tell, none without at least some college. I’ll count that as support for the OP.

Say what? This is a forum run by academics with a heavy bias towards topics that appeal to academics. So yeah, selection bias will mean most people who read this have at least some college… I do know some good programmers who did not go to college. One of them is actually quite brilliant.

The comparison is not between people today who have been to college, and those who don’t. Because in general people who go to college are those who are more likely to succeed in life at professional jobs. They’re motivated, they’re better at studying, they have socio-economic advantages. Plus in most fields it’s very hard to get a job if you don’t have at least a four year degree. It’s easier in computing (I’ve worked with them), but it still limits your options.

The comparison is between the person who 40 years ago would not have gone to college, but who today does. That the office worker of today with a four year degree is better trained than the office worker of 40 years ago with only a high school/trade school education. That goals such as Tony Blairs of increasing college entry to 50% will result in a more skilled workforce.

As for credentialism – it’s a thing. We’ve seen plenty of examples on this thread, including from people who think it’s a good thing. There are certain jobs that you will not get unless you have the requisite pieces of paper. It’s so common that people joke about it. In IT the jokes are pretty bitter.

J-D: I think I’m probably more prepared than John Quiggin is to entertain arguments that institutional education is systematically over-valued; but even supposing such arguments are correct, what changes in educational policy are they supposed to justify?

For vocational areas – less focus on (expensive) four year education run by researchers with little practical experience. More focus on practical training, with as much on the job experience as possible – with theory supporting practical training.

engels 03.08.17 at 12:23 pm @ 135 — I’m just diagnosing widely observed phenomena. I would think anyone could learn to write computer programs, do auto mechanics, read maps, learn new languages, draw, play the piano, and so on, but that’s not what we observe. Hence the speculation has arisen in some quarters that there are many different forms of intelligence which are different from one another and unevenly distributed, just as physical attributes are.

Cian 03.08.17 at 9:37 pm @ 141 — I have (or had) most of those skills and never used any of them, except maybe in play. If I needed sophisticated mathematics, I went and found a mathematician, or temporarily became one in a very limited way. That kind of thing seldom occurred, however.

It hasn’t been much noted that credentialing apparatuses are often used not only to filter out the incompetent but the competent as well, in order to confine the numbers of the laborers available, limit them to certain ethnicities, locales, cultures, and so on.

Cian

Tony Blair massively increased the number of people attending university without any noticeable affects on the economy, or productivity. So the policy changes in that instance seem pretty obvious.

No, it’s precisely the policy changes that aren’t obvious. I don’t know how education policy works in the UK, and I don’t know how Tony Blair changed it. If you are telling me that the number of people attending university increased massively while Tony Blair was Prime Minister, I can understand that much, but what changes did Blair’s government make to produce that result?

and here

For vocational areas – less focus on (expensive) four year education run by researchers with little practical experience. More focus on practical training, with as much on the job experience as possible – with theory supporting practical training.

I likewise don’t know what kind of focus you’re talking about, so I don’t know how you’re envisaging having that focus changed. It’s easy enough to describe things governments might favour (or not favour), and it’s easy enough for governments to signal that they favour (or don’t favour) those things through their words, but things don’t happen just because governments say they favour them (or stop happening because governments say they don’t favour them). I don’t know what the reason is that university attendance increased under the Blair government, but I do know that statements by ministers announcing that they were focussing on increased university attendance wouldn’t have been enough to do the trick. So what kind of government actions are you referring to?

Harry

It was driven almost entirely by central government policy, mainly by large changes to the funding system, the introduction of student loans, and freeing the polytechnics from local government control (most of the increase was achieved through the expansion of former polytechnics which were less risk averse, and more open to expansion, than the pre-92 institutions).

Now ‘changes to the funding system’ and ‘student loans’ I understand, although ‘freeing the polytechnics from local government control’ is obviously more specific to one country’s institutional arrangements.

So, if there are good arguments that institutional education is systematically over-valued, what changes to educational funding (possibly including student loans) are they supposed to justify? Global cuts, or what? I think I can make a fair guess at some likely effects of global cuts to education funding, mostly harmful, but I doubt that ending a systematic over-valuation of institutional education is likely to be among them.

I kind of regret jumping on the credentialism bandwagon a bit now as the thread now seems to be full people saying higher education should be scaled back because it doesn’t supply Wall Street / Silicon Valley with the labour force it wants. Which is every bit as annoying in its own way as the Panglossian/efficient-markets-esque defence.

I kind of regret jumping on the credentialism bandwagon a bit now as the thread now seems to be full people saying higher education should be scaled back because it doesn’t supply Wall Street / Silicon Valley with the labour force it wants.

My position on this is that anyone who wants to go to be able to college for the fun of it should be able to. Study English, Philosophy, Physics, whatever. If you have demonstrated the capability to study these things at an advanced level, I feel you should be able to. I’m skeptical that these things necessarily result in a better ‘trained’ workforce (at least en masse), but I also don’t care. I want people to be able to study Merleau-Ponty and leave college debt free.

However, increasingly college is being seen as a place of vocational learning. More and more students are being ‘practical’ and going to college to study things like hospitality, business. I don’t think that universities are good places to teach these things, particularly in countries (the UK being particularly bad in this regard) where academics are forced to justify themselves solely through research output.

What’s worse, due to good old fashioned snobbery, vocational subjects are now being taught in four year colleges, because if you don’t have a Bachelors in these societies, you have more limited job options. Now I will accept that most of these subjects have some theoretical knowledge that you need to know (accounting say), there’s not four years worth of theoretical knowledge that needs to be crammed in before you get a job. Sure, there are exceptions. Engineering is one, medicine is another. So is Computing if you pursue it at an engineering level (which few graduates will). But these are disciplines that require a lot of maths and science knowledge as background. They’re unusual.

In the US things have got so bad now, that in many disciplines you now need a masters and possibly a PhD to advance in areas where this makes little sense (teaching, social work, ‘management’). Because otherwise your CV is going straight to the slush pile. I don’t think is healthy, particularly in countries like the UK/US where students get heavily into debt in order to acquire these credentials.

OTOH I found some computer science courses very hard indeed because I’m not particularly inclined to hard theoretical stuff for fun. I need a reason to slog through it and “it’s in the curriculum” isn’t enough.

I had a similar experience also, though I’ve also found I need a practical example to hang the theory on. The worst example of this at my (very theoretical) university was that we had a course on DB theory, but never actually touched a database… I barely passed that course. A couple of years out of college when (under the tutelage of an experienced DBA) I was starting to do serious database work I went back and looked at my textbook and foud it all pretty trivial to understand. I’ve had similar experiences recently with a lot of the theoretical stuff now that functional programming is becoming a thing. Lambda calculus is a lot easier for me to understand (along with things like category theory) now that I’ve written code using it.

I fear that I was lucky enough to latch onto “as simple as it can possible be, and no simpler” as a rule very early on.

I think that’s probably the most important thing a programmer can learn. The other useful thing is an obsession with elegance. If your solution is elegant, it will probably be relatively easy to maintain (and over time you will get faster at programming). I’ve maintained code that was barely documented, but because it was elegant was relatively easy to understand and mantain. Whereas it doesn’t matter how well you document bad code, it’s still impossible to grok.

Are you implying it’s only legitimate to take a course if you DON’T think it will improve your job opportunities?

Sorry, I didn’t mean to imply that. My poorly expressed point was that many of the Masters students are likely to be disappointed by how little the credential does for their job opportunities. Because the employers they want to get hired by are not very interested in a little financial learning, they’re looking for a lot of aptitude.

Revisiting my #73 and your response at #74 “This seems specific to a very particular kind of employment structure.”

I would say that the defensive HR structure I describe in #73 applies to nearly all companies over about 100 employees and a majority or near majority of companies over 50 employees. According to the census, 51.6% of private employment is with large enterprises (more than 500 employees), 14% at medium (100-499 employees) and 16.7% at small enterprises (20-99 employees).

So even if it is just medium and large enterprises we are talking about 65% of the private workforce, and I would estimate that at least half of the small enterprises operate under the same lines. That means that between 65%-75% of the private workforce is implicated under what I discussed (where I showed that in some cases even if a credentialed person was less able to do the job they might get hired preferentially).

So it is a large enough part of the workforce to make a difference and requires at least a bit response.

You suggest that it would have to be something as powerful as racial or gender discrimination to continue that without it being self correcting in the marketplace (due to competitive pressure). How about class? College (and as college became more commonplace, masters degrees) form an excellent class marker.

Anecdotally my brother in law is a project manager at a huge contractor. He says he has made his career much more successful by using minority/gay/female employees who are underappreciated but do great work (which he has mixed feelings about–he’s happy to help them, but unhappy that he needs to). He is unable to do similar things with people who don’t have degrees because they can’t get in the door.

Yep it’s class in Weber’s sense of opportunity-hoarding.

PB, that makes sense (I probably interpreted what you wrote a bit unsympathetically).

Again, supposing for the sake of argument that the preoccupation with the vocational advantages of a university education is a harmful thing, what change in educational policy (not hiring policy, but educational policy) is that supposed to justify?

1) the costs of job-training could be taken up by the private sector, or by the non-profit educational sector. I’d rather see the former, freeing up public and philanthropic funds for other uses, including other educational uses.

2) making job-training central to a university’s mission effectively puts curriculum and research under the indirect control of corporations.

Cian 03.09.17 at 2:20 pm @ 155 — I think what you said is about right, although I would add the substantial benefit of a kind of dumbness or simple-mindedness which is forced to break any problem or question down into little tiny pieces on a conscious level. Maybe this is what you mean by OCD. The engineeringish stuff comes in because of the things that are done with computers, which connects them to the physical, financial, political, etc. worlds, just as a civil engineer eventually has to deal with the stubborn actuality of physical objects, money questions, politics, and the like. One starts out innocently coding and winds up in meetings, sometimes of the heavy-duty sort.

M Caswell

The question I asked was ‘What changes in educational policy?’; you don’t seem to have answered it.

“All that means is that the employers for those good jobs are requiring the credentials. It doesn’t mean the credentials are actually necessary to do the jobs well. Maybe they are, maybe they aren’t, but this observation says nothing about ‘credentialism’.”

Yeah, sorry, that wasn’t supposed to be evidence that John is right, just an observation in response to the comments people are making about coding. And the observation is completely anecdotal and full of selection bias. The 4 CS students I’ve been teaching recently are among the smartest abstract thinkers I’ve ever met; one ended up in CS mainly because he was so horrified by the quality of teaching in a different department where he had intended to major; and 3 of them could, I am sure, have gone straight from high school into high paying jobs. (I conjecture that the 4th wouldn’t but only for labor market discrimination reasons). They’re all officially residents, so paying low tuition). They all, also, have completely unrelated remarkable talents. So, not at all evidence for John’s claim, which I nevertheless agree with….

Harry: Those remarkable starting salaries are partly a reflection of two things:
1) The DotCom/VC bubble (which is probably close to bursting)
2) The insane cost of living in Silicon Valley (a high salary doesn’t go far, particularly if you have a family)

Back on the original Quiggin post, I find the Ashenden quote very odd. Apparently higher education has “exploded like an airbag”. An airbag that expands explosively is doing precisely what it was designed to do, in order to protect the user from a nasty road accident.

He might be thinking of a bubble, but the only way a bubble explodes is when it bursts, and higher education hasn’t burst. If he really means it with the airbag, rather than bursting, it will deflate slowly once the emergency is over. So is higher education like a airbag, inflating rapidly to protect us from some sort of crash and then deflating in a controlled fashion, or a bubble, or something else? If it is in fact like an airbag, perhaps it was triggered to cushion some sort of violent economic shock, in which case it is surely better than the alternative?

Ashenden’s unclear prose betrays his unclear thinking.

(Off-Topic Notice: only tangentially related to the main topic, if at all; moderator’s choice)

” I wonder if this is true in other kinds of engineering — build a bridge, watch it fall down, build another bridge, and so on”

Approximately all mechanical engineers are at some time shown a film of the Tacoma Narrows Bridge twisting itself to destruction due to wind-vortex-induced vibration. Calculations for that effect were henceforth included in the “Standard Handbook for Mechanical Engineers” (one of the books that should be included in all repositories that are intended for use in rebuilding civilization after the next fall).

Both nature and engineering work by the same basic principle: trial-and-error plus memory, otherwise known as evolution. I suspect all human problem-solving uses that principle, but after memory (books, lecture notes, films of bridge collapses, etc.) has expanded to make most tasks routine, we tend to forget how that memory was created.

On QA, when I first joined GE out of college as an engineering applications programmer, to complete each project we had to write a data folder (bound report, signed off by the manager) with a verbal description of the program, its flow chart, the input and output of a sample run, and a glossary of all variable and routine names, in alphabetical order, describing what they where used for and where they were used. The reports were kept in a Program Library with a full-time librarian, in both paper and microfiche form. Backups were sent periodically to Iron Mountain.

Under Welch, such procedures gradually melted away in the glare of his budget-cuts and layoffs.