Education: more, please

by John Quiggin on September 10, 2005

In his post on education, Chris

floats a hypothesis for commenters to shoot down if they want to.

However, since most of the commenters agree with Chris, it looks like I’ll have to provide the other side of the debate. I’m also not linking to any evidence, though I discussed a fair bit of it here

I’m going to argue, contrary to Chris and most of the commenters on his post that there’s no reason to suppose that, in aggregate, the proportion of the population undertaking post-secondary education is too high, and every reason to continue trying to remove obstacles to participation in education for students from poor and working class backgrounds. Further, I don’t think credentialism is an important factor in explaining observed changes in participation in education or the labour market.

A basic problem in assessing the argument is the need to get the background conditions right. Roughly speaking, over quite long periods, average education levels have risen greatly, but the wage premium for education hasn’t changed much and neither has social mobility. In a static economy, that would suggest support for a credentialist hypothesis in which the additional education was simply a race for positional goods.

But the economy isn’t static. The proportion of jobs in occupations that traditionally did not require post-secondary education is declining steadily, while the proportion in jobs requiring education is rising. Pretty clearly, this is the result of technological change leads to a steady expansion in demand for skilled/educated labour at the expense of unskilled labour.

Given this trend at the aggregate occupational level, it seems reasonable, in the absence of evidence to the contrary, to impute growth in average educational levels within occupations to changes in the work being undertaken

With this interpretation, we need a steady expansion in education levels just to keep pace with technological change. Where, for one reason or another, participation stagnates as it did in the US in the 1970s, an increase in the wage premium for educated workers is a likely result (this isn’t the only reason for the boom in inequality in the US from the 1980s onwards, but it’s one reason).

If we’re going to get labour market conditions favourable to more social mobility and less inequality, it’s not sufficient to expand access to education. Growth in education has to outpace the rise in relative demand for educated workers, and growth in access to education for students from poor backgrounds has to outpace the growth in participation among the middle and upper classes. On average, we haven’t managed to do this, and it’s not surprising therefore, that the results hoped for by advocates of expanded education haven’t, in general, been delivered.

Turning specifically to credentialism, I’ll start with the labour market I know best, that for academic economists. The requirement for a PhD is now almost universal, which was not the case thirty years ago in Australia or the UK. In my judgement that’s because the knowledge and skills required have increased, and not because of credentialism. If lawyers, nurses and other professions similarly rely more on formal training and less on ‘sitting next to Sally’, I’m willing to believe that similar processes are going on (in addition, the same technological changes that I’ve mentioned already mean that, whereas Sally once found an unskilled trainee useful for a wide range of tasks, she now finds him a burden who just gets in the way)[1].

Among the pieces of evidence against the credentialism hypothesis, one of the most compelling, in my view, is that self-employed people have education levels comparable to, or higher than, those of the workforce as a whole. If education was a matter of getting a credential for employment, people planning on going into business for themselves would get a headstart by skipping it. There are lots of other bits of evidence pointing in the same direction. For example, wage premiums for education tend to be rise with years of experience, when, on the credentialist view, they ought to decay.

I’ve been careful throughout the above to refer to post-secondary education rather than university education. There’s a pretty good case to be made that technical areas of education are in more need of expansion than traditionally ‘academic’ areas, at least in the English-speaking countries. But the idea that expansion of post-secondary education as a whole can or should be reversed is, in my view, mistaken.

fn1. There are some examples I won’t defend, such as most of the journalism/communications courses I’ve seen, but I think these are ‘exceptions that prove the rule’, marketing themselves on the basis of the success of professional qualifications in other fields.



bad Jim 09.10.05 at 4:11 am

There seems to be general agreement that secondary school graduates in most First World countries have attained a level of proficiency which an American acquires after two years of college. To the extent that this is so, we have to allow for some national skew in the terms of this debate. It may be worth noting that American community colleges, which share that two-year horizon, also have a strong occupational orientation.

I’m inclined to quibble with the idea that the pace of technological change implies that workers need to spend more of their youth in post-secondary or post-graduate education. My bias may stem from my experience as a B.A. in math winding up as an electrical engineer, and from my father and my brother (respectively psychology and biology majors) having similar career paths, and from having worked alongside several other creative engineers who got the bulk of their training on the job. It’s my impression that universities lag enough behind the cutting edge that the advantage of a year of study over a year of work isn’t necessarily decisive.

As the first world deindustrializes and manufacturing jobs disappear, it’s incontestable that good jobs for the less educated are already scarce and likely to become more so. We seem to be responding by turning the present generation into salespeople and cosmeticians. I’m of the party that thinks deindustrialization may be the problem.

That entrepreneurs tend to be relatively well-educated isn’t particularly surprising. You need money to make money, and there’s a strong correlation between wealth and educational attainment.


Chris Bertram 09.10.05 at 4:58 am

I’d don’t know what the picture is in Australia or the US, John. But my impression here in the UK is that much of the expansion in the HE sector does not consist of people acquiring the vital skills needed to keep pace with technological change. Rather, an arbitrarily high target for participation has been set and this has in large part been met by the provision of BA degrees of dubious real value which are nevertheless a prerequisite for entry to various white-collar careers.

Taking academic economists as an example strikes me as unfortunate. That’s a case where you can straightforwardly enumerate the skills and knowledge needed and show how the academic courses in question contribute to the person acquiring those skills and that knowledge. I think you’d have a harder time showing how a BA in Media Studies and Culture from the New University of Poppleton provided someone with the skills and knowledge needed for the typical job its graduates ended up in.


Scott Martens 09.10.05 at 5:06 am

Pretty clearly, this is the result of technological change leads to a steady expansion in demand for skilled/educated labour at the expense of unskilled labour.

I don’t think this is true – in fact, I think the opposite is true. Technological change has sharply reduced the number of jobs that actualy demand a higher education. The whole point of technological advancement is to make life easier, not harder, and the whole point of division of labour is to keep the amount of specialised knowledge required by an individual worker as low as possible. Consider car mechanics – a subject my dad taught in high school. Thirty years ago, a car mechanic might not have had a university degree, but he had to have an extensive practical knowledge of automobile engineering in order to sucessfully diagnose and repair car problems. Now, car mechanics is very different. The tech has a manufacturer specific computer that diagnoses the problem and tells him which of a limited array of labeled parts needs to be removed and replaced. Quite a few trades have gone this way. Consider the role of a Toys-R-Us manager – my wife’s former employment. In the old days, a store manager had to keep extensive records of what was and wasn’t selling, and at least advise higher management of what to buy and ship, if not actually participate in the inventory process himself. They handled employment decisions, managed payroll and receipts, did much of their own marketing and took personal responsibility for much of the facility’s outcomes. Now, much of this handled by computers and supply and inventory decisions are made in Soviet style. (And, as a reader of Hayek might have expected, this doesn’t work very well.)

I do agree, though, that some of what motivates the demand for credentials is the inability to support the costs of training an unskilled worker, but I think that is not because work has smartened up, but that it has been dumbed down. Consider programming – modular, object oriented programming was devised so that coders could write sections of programmes without having to see the big picture. Specifications can be made shorter and simpler, so that coders don’t require the level of training needed to see an entire application though. A university diploma means you’re hopefully at least smart enough to read the spec do the tasks that a job might require without further extensive training.

The growth of medical employment might represent a contrary trend, but not because medical professionals require more education than in days past. The field is far more specialised than it once was. Rather, simply the growth of the medical profession with respect to other sources of employment might skew the overall picture in favour of higher demand for post-secondary education.


Matt Daws 09.10.05 at 6:03 am

Chris, Exactly! I think the Time Higher recently noted that Media Studies courses actually had a low graduate employability factor (so many students didn’t get a job quickly after their course) and that it was a rather bad entry into the media: getting a good English/History degree and being prepared to intern etc. is the correct route. I don’t think it’s true that media employers and insisting on Media degrees, but rather that student think they will, and hence spend three years getting what is arguably a “useless” degree.

John, the complaint against Credentialism is not (at least as far as I’m concerned) about highly skilled specific jobs (academia, medicine, engineering) but about the general attitude that to do almost any white-collar job, one needs a degree (does really matter what in). My sister, for various reasons, doesn’t have a degree, and has found it very hard to get simply office jobs (which she is very competant at, and her school record is excellent). I understand why employers do demand a degree, but it makes life very hard for people who don’t have a degree, and means that many people end up with degrees which they don’t really need. Similarly, we find that jobs which maybe do require an unspecific undergrad degree (to demonstrate independence, and a general ability etc.) now require a masters degree.

Balanced against this, of course, is my belief in education for educations sake. I do think, maybe, that the UK (which is what I know about) should have made it easier for mature students/ supported night-schools better etc. rather than pushing for a 50% target of (implicitly) young people getting an undergrad degree.


Matt Daws 09.10.05 at 6:10 am

A university diploma means you’re hopefully at least smart enough to read the spec do the tasks that a job might require without further extensive training.

No, I really disagree here (although I do lack evidence). To me, the employers demands a diploma for two reasons (which are much the same):

a) Reduce the pool of applicants (to make life easier, presuming they have a large number of applicants);

b) If there is a large pool of people with a better qualification, why take the risk of hiring someone without a diploma, if you don’t have to? Obviously there are cases when this would obviously be a bad idea (e.g. someone has glowing references from Microsoft, and a case-iron reason for leaving them, but happens to have no diploma). However, I would guess that in most cases, you have 30 applicants, and there is really little to tell between them, so you discard the 10 without the diploma, as then you need only interview 20 and not 30.

This doesn’t fully explain that complaint one hears of older programmers with good experience, but no formal qualification, being overlooked for younger, qualified people. However, is this age-ism or cost-cutting?


John Quiggin 09.10.05 at 6:17 am

I did pre-emptively concede on Media Studies :-)

But I’d like to see some response to the empirical evidence I’ve cited, rather than restatements of standard credentialist claims. If degrees, on average, add less value than they cost, why is there such a large wage premium for them, and why does the premium increase with experience (the opposite of what a screening model would predict)?

What explanation of employer behavior is being offered here?


Matt Daws 09.10.05 at 6:35 am

John, I concede that I don’t really have an emperical evidence! HOWEVER, my specific complaint is really about the massive increase in undergraduate education in the UK in recent years. As far as I can tell from google quickly, participation was about 30% when Labour came to power, and it’s now touching 50% (vaguely). I would suggest that it’s really far too early to tell if wage premiums will hold up: recent evidence which I’ve seen in the THES suggests that in some “soft” subjects like Media Studies (;->) it does not (i.e. it adds no wage premium compared to others in that age group).

It will also be interesting to see, in the UK, if fees change student behaviour: you say “If degrees, on average, add less value than they cost…?” but in the UK, the cost has been pretty modest until very recently.


Chris Bertram 09.10.05 at 6:47 am

If education was a matter of getting a credential for employment, people planning on going into business for themselves would get a headstart by skipping it.

Not if starting off in employment is a useful platform from which to launch one’s own business and if having a degree is a necessary ticket to that platform. Many people who start their own businesses will have accumulated the necessary contacts, capital (such as their house) and experience in paid employment. If those same people had simply started their own businesses straight from secondary school they’d almost certainly have failed.


Matt Daws 09.10.05 at 6:50 am

Actually, John, I’m not sure why:

why is there such a large wage premium for them, and why does the premium increase with experience (the opposite of what a screening model would predict)?

refutes my argument? I’m arguing that the, say, top 30% of jobs go to graduates because employers demand a degree: we would hence except there to be a correlation between a degree and a wage premium.

As for the experience issue: have you isolated the effect or career path? If a degree helps you get a leg-up early on, then won’t you progress up a career path at the same rate, but at a higher level, than a non-graduate? Is the stated effect then nothing more than the fact that the wage premium for position is non-linear (that is, the jump from manager to senior manager is higher than the jump between supervisor and manager, say?) Maybe that’s not what you mean though?

Again, self-employed people might simply be far more motivated than others, and hence more likely to overcome problems to get to university (again, in a historical situation where getting a degree is seen as “important”). I’m slightly distrustful of arguments like:

people planning on going into business for themselves would get a headstart by skipping it.

Do 18-year-olds really wake up, think “I’m going to be self-employed”, and then plan how best to do that? I doubt it: I would guess they are hard-working, innovative, motivated people who head to university (as everyone, pretty much, tells them this is a good idea). They then later form the idea of starting their own business…

I’m not sure if either of us is right anymore; I’m just trying to offer some other ways in which we would observe the emperical evidence which you cite!


abb1 09.10.05 at 7:11 am

Perhaps one the reasons employers perfer workers with a degree is exactly because of their desire to limit social mobility – in a sense. They are looking for the employee who is a conformist, who’s clearly and firmly accepted and pursued traditional middle-class path to success.

See, any employer will certainly hesitate to hire a high-school drop-out. Why – is it because they worry about the lack of skills? I don’t think so; rather they’re afraid of some kind of eccentric and destructive behavior.

Degree = conformity to middle-class values and dogmas.


Bob B 09.10.05 at 7:37 am

John Q: “the wage premium for education hasn’t changed much”

I don’t have the links to post but my distinct recollection from pieces in The Economist of some years back and from a survey by Greenaway in the Oxford Review of Economic Policy is that in both the US and UK, especially the US, is that compared with the 1970s, there have been marked increases in the premia paid to graduates relative to non-graduates.

On the market evidence, there has been an increase in the relative scarcity of graduates in US and UK job markets over the last c. 30 years. This is consistent with the BBC reports of research into graduate pay I posted in Chris’s thread (# 84) and also with anecdotal accounts of employers who put more trust (faith?) in the significance of degrees over school-leaving exams as indicators of personal aptitude and ability, as well as the expected consequences of the shifts towards globalisation and the “knowledged-based” economy.

“There is a strong consensus among economists that formal education is an important
determinant of individual earnings as well as economic growth. The importance of formal
education has been magnified by recent economic trends underlying U.S. labor market
demand for skilled workers.”


Gabriel 09.10.05 at 7:49 am

This is basically a twist on the of the Invisible Adjunct argument that lower tier Universities are raising false expectations of the employment benefits of their courses to students, when the reality in the UK is that if you want a career in the media (or a lawyer) you’re far better off studying Classics at Cambridge (or philosphy at Bristol) than Media Studies pretty much anywhere. Lawyers in Britain still do mostly learn on the job, compared to most other countries. This quite explicit rejection of credentialism (and the relatively high importance of various class markers/ symbolic capital) has always distinguished Britain from both the US and the rest of Europe. That’s why the higher education expansion in the UK has been so half-baked.


harry b 09.10.05 at 8:00 am

Bob B — I also don’t have a link, but John Goldthorpe has pretty compelling data showing a fall in the premium attaching to a degree since the 80’s — not dramatic, but very very noticeable (in UK). That’s what you would expect, I guess, given dramatic and rapid expansion of HE uptake.


John Emerson 09.10.05 at 8:12 am

“If lawyers, nurses and other professions similarly rely more on formal training and less on ‘sitting next to Sally’, I’m willing to believe that similar processes are going on”

Your willingness to believe doesn’t strike me as an argument. Later you talk about “exceptions that prove the rule” — a dubious concept in almost every case. There are many more exceptions than you seem willing to grant; my alma mater is a third-tier school and has whole buildings full of classes in social work, education, police science, counseling, management, and other areas which have been scientized in the fake kind of way Foucault talked about. (People who study in these areas always complain about the unrealism of what they’re taught, and the lack of actual practical skills learned.)

From the point of view of social mobility, we should think of the bright average student with pretty good grades, from a mediocre or worse HS, and absolutely no family support. A requirement of four or five years of full-time college will always work against someone like that except under the most Utopian conditions, so if the educational requirement is in any way excessive or extraneous (as I think it is), credentialization works against social mobility.

In the American context at least, access to education is greatest at age 18 when people are least mature, and returning to college at a later point is much more difficult. In a credentialized work force, people often end up unable to advance into positions they’re able to perform. In fact, what often happens is that they do the job until someone is hired, often for a considerable period if there’s a hiring freeze or budget problem, and then have the added duty of training some bright young credentialed thing who knows nothing about the work and may well be using the job as a steppingstone to better work. (The people I’ve known to whom something like this has happened have mostly been mothers returning to the work force.)

So I would say that it’s the weakening of on-the-job training and apprenticeship, rather than credentialization per se, which is the villain. (That is, channeling credentialization entirely through the schools is the problem.)

A last point rather peripheral to the main argument is the very weak performance of graduates in the humanities (apparently called “arts” in Britain). Education has traditionally been a thick enterprise, combining job training (of a sort), class validation, cultural enrichment, values training, wild fun, etc. into one big mess. From a social-mobility standpoint, disentangling the job training from this would be the way to go, but of course the other functions are valuable too. In the past I have suggested making cultural enrichment voluntary while subsidizing it as adult education, but actually rather few people are interested in humanities-type-stuff, and there wouldn’t be much constituency for the subsidies.

In scholarly areas such as economics and philosophy, in my opinion, credentialization and methodologization have also led to narrowing and formalization at the cost of weakening their relationship to reality, but that’s for another day. (I’ve been waiting since 1967 for economists to start reading Polanyi again, and it seems that that finaly might happen).

I think that the university is at the kind of peculiar historical crossroads where its problems are fairly widely perceived, but where its institutionalization is unchallenged and no actual alternative has even been imagined. So probably things wil continue with business as usual until some enormous crisis intervenes, at which point everyone will claim to be dumbfounded.


John Emerson 09.10.05 at 8:31 am

Let me throw a bone to the libertarians. Some of the career paths in which credentialization seems most problematic are either forms of government work or government-regulated lines of work. I believe that some kind of credentialization dynamic may be at work in monopolistic and oligopolistic enterprises such as utilities, too.

By and large I think that the more competitive a market is, the more corners enterprises will cut regarding credentialization. Privatization actually is a pretty good test case for the credentialization debate. I really wouldn’t want my medical plan to hire uncredentialed MD’s, but I wouldn’t have minded if my son’s HS had hired English majors instead of Ed majors. (For the record, my son’s public school HS teachers, whatever their degree, were often excellent).

I am not terribly sympathetic to the libertarian point of view, but it should be on the table, and I’m surprised that it hasn’t shown up yet.


Bob B 09.10.05 at 9:12 am

Harry B,

Unfortunately, I have to admit to not following the sociology literature. Compared with John Goldthorpe, we have this:

“A sharp increase in the number of people getting degrees has not reduced the extra earning power of graduates. The annual education report from the Organisation for Economic Co-operation and Development shows graduates are still maintaining a pay gap. . . In the UK, average graduate earnings are 59% above those of non-graduates. ”

Fortunately, I’ve managed to trace the Greenaway reference I alluded to above (it’s much easier for those with current academic affliations and online access to periodical literature to trace and check citations than it is for public service retirees like me):

“Slaughter [The World Economy 1999, vol. 22, pp. 609-30] reports that in 1979 male college graduates in the USA earned on average 30% more than their high-school equivalents. By 1995 that had risen to around 70%. Over the same period, the equivalent premium in the UK grew by around 30%. More modest increases were recorded for Australia, Canada, Japan, and Sweden, but little or no change in France and Italy.” [Oxford Review of Economic Policy Autumn 2000, Vol. 16, No. 3, p. 2]

If so, that establishes what I was saying about market evidence showing an increase in the relative scarcity of graduates, comparing the 1990s with the 1970s. The quote from the recent BBC piece relating to the OECD’s annual report on education, suggests that the graduate premium is holding up well despite the “sharp increase in the number of people getting degrees”.


mythago 09.10.05 at 10:02 am

Some of that pressure is in the other direction–with the death of the US industrial database and the decline of unions, not to mention more stringent requirements on applicants for the military, it is no longer reasonable for a U.S. worker to expect that he can get a market-wage job, or enter the military, without completing high school or further training.

If lawyers, nurses and other professions similarly rely more on formal training and less on ‘sitting next to Sally’

Can’t speak to hurses, but lawyers are actually very much an example of credentialism. Law schools have only very recently started to lean towards teaching practical skills, rather than general background and legal theory. It’s expected that the practical skills are taught from ‘sitting next to Sally,’ i.e. through summer or part-time work as a law clerk, or on the job after graduation and passing the Bar (which is another ‘credential’ issue by itself). Very few states now allow a non-degreed person with real-world legal experience to sit for the bar examination. And ‘credentialism’ most certainly applies to law schools; it’s doubtful that one learns how to write contracts better at Berkeley than at Yale, but guess which one is considered a better choice for getting a job later on?


James Wimberley 09.10.05 at 11:09 am

The corporate welfare state has clearly declined sharply over the last few decades: companies that nurtured talent, like the employers of the young John major and Jacques Delors, in a perpective of long-term work relationships, do so less and less. A hire-and-fire job market externalises training costs and makes credentials and (perhaps even more) personal networking more important.


eudoxis 09.10.05 at 11:15 am

The two models of John and Chris are not exlusive of each other. So the argument is about the size of the component that credentialism or true education play in the premium value and exclusivity of a higher education. On the one hand, (in favor of Chris) is an example such as the replacement of striking air traffic controllers by Reagan. On the other hand (in favor of John) is the increase in education requirements far beyond a step where the type of exclusivity that Chris refers to has taken place. For example, training for medical subspecialties take years longer than training for generalists while the selection that Chris refers to takes place at the entrance to medical school.

There’s a bit of both and it’s difficult to tease apart the premium that employers place on the added knowledge and skill value vs. the type of individual or, as noted above, a proven ability to conform.


Carlos 09.10.05 at 11:24 am

Look at this Krugman paper, where he builds a tentative model trying to explain increasing inequality in salaries with a signalling model involving college education


Christian Johnson 09.10.05 at 11:26 am

I would agree that in the United States, law is perhaps the best example of credentialism run amok – best symbolized by the mid-20th-century movement to rename the standard law degree the “J.D.,” or Juris Doctor, in place of “LL.B.,” Bachelor of Laws. God forbid that law be perceived as of lower academic status than medicine! At the same time, many states ended the ability of non-law-graduates to “read law” and pass the bar. Now, I believe that there may have been at least a colorable issue of quality – I recall seeing arguments that candidates who read law had lower bar success rates, and higher professional complaint rates, than their law-school-educated peers. I’m not persuaded that the differences were worth further gilding the guild, as it were, especially since I think most US lawyers would agree that at least 1/3 of law school education (i.e. the third year) is completely useless.

The more recent attempt by law schools to increase “practical” instruction, which mythago notes above, is unlikely to improve matters much. My law school, NYU, was at the vanguard of this movement back in the late 80s and early 90s, when with much fanfare and expense it launched its “Lawyering” program. “Lawyering” classes attempted to augment the traditional first-year course covering research, writing, and oral argument with a set of somewhat fuzzier skills – negotiation, client counseling, etc. I was, and remain, utterly unpersuaded that these latter units were and are of any use whatsoever, because the skills they covered simply cannot be taught effectively other than through practice. Lots of practice, actually, requiring far more time than a law school could reasonably give.

More unforgiveably, these units effectively reduced the time spent on research and writing, at the same time as the first truly post-literate cohort was moving through law school. The result of this sacrifice became clear late in my second year. I joined the senior editorial board of a journal, a position which required me to help wade through the writing tests that first-year students submitted in their applications for journal membership. I was appalled: students from the best undergraduate schools in America proved incapable of writing a simple persuasive essay in the form of a legal memo. Bad enough that these students had managed to get admitted – but worse that the Lawyering program twiddled around trying to teach client relations while what the students really needed was a sort of writing boot camp.

The sad reality? All of these students left NYU with a degree that gave them virtua carte blanche entry into the most prestigious firms in the US, making the equivalent of over $100K/year from the start. The reason these law firms extract so many billable hours out of their associates is that their work is almost worthless – takes a lot of it to make the salary even remotely economic.

I would much prefer a system that would eliminate the third year of law school and mandate a two to four year clerkship instead – performed with a court, a public prosecutor’s office, a public defender’s office, a legal aid society, etc. This would be the equivalent of a doctor’s clinical training, and would inculcate at least some form of public service, while decreasing the cost of legal education. Chance of it ever happening? Almost none.


Bob B 09.10.05 at 11:39 am

If too many people are getting degrees in Britain, how come that graduates are more likely to be employed than non-graduates?

“Employment rates also vary by the qualifications people have obtained. For both sexes employment rates increase with the level of qualifications (Table 4.12). Ninety per cent of men and 85 per cent of women who had a degree or the equivalent were in employment in spring 2003. This compares with 57 per cent of men and 44 per cent of women who did not have any qualifications.”
– at chapter 4 in:

On the proportional effect on earnings of a degree by degree subject, see the following chapter 5 in the above link, particularly Figure 5.10. It seems that lawyers – and economists – do relatively well compared with graduates in other subjects.

I’ve not had a chance yet to digest all in this link but it looks interesting as well as relevant:
Massimiliano Bratti + Luca Mancini: “Differences in Early Occupational Earnings of UK Male Graduates by Degree Subject: Evidence from the 1980-1993 USR”

As reported, in terms of earnings, male arts graduates may make less money on average than those who left school at 18 so it is perhaps not surprising if they worry about the numbers of new graduates coming onto the job market. I recall an article in The Economist of some 20 years back reflecting on careers in the US of philosophy PhDs where the annual output exceeded the number of academic positions becoming vacant. As reported, a disproportionate number of philosophy PhDs were going into the burgeoning computer industry where logical and analytical skills were – and are – especially valued.

“Business and financial services continue to form the largest single sector of the UK economy, and in 2003 accounted for 31.7 per cent of the total, which was more than double the size of the manufacturing sector. . . In 2003, the contribution of Information and Communication Technologies (ICT) to whole economy GVA at current basic prices accounted for . . 6.7 per cent of the total.”


Matt Daws 09.10.05 at 11:58 am

Bob B, that’s not the argument though! The argument is that for many jobs, one does not *need* a degree in a practical sense, but the employers do demand a degree (for reasons elaborated elsewhere in these two threads). This makes an artifical barrier to entry into higher-paying jobs, one which is harder for those from poorer backgrounds to overcome. If this is true, then your statistics just back up the argument that “too many” degrees are a barrier to social mobility!


Bob B 09.10.05 at 12:15 pm

Those here questioning the case for raising the percentage of the catchment age group going into higher education may be interested in the citations in this useful brief relating to an evidently growing literature on “over-qualification” and “over-education”:


Bob B 09.10.05 at 1:00 pm

“The argument is that for many jobs, one does not need a degree in a practical sense, but the employers do demand a degree.”

If so, it needs to be explained why employers are mostly willing to pay on average a premium for employing graduates – except for arts graduates, that is – when they could readily recruit non-graduates, as and when non-graduates are capable of doing the job.

Are we to seriously believe employers generally do this out of an increasingly pervasive philanthropic inclination to encourage and support the beneficiaries of higher education despite the costs to businesses subject to mounting competitive pressures from globalisation?

FWIW I am more inclined to believe suggestions that where degree-level skills or aptitudes are not actually required for jobs, employers are prone to repose more trust in degrees as signals of personal ability and commitment than the results in school leaving exams – possibly because of recurring debates in Britain about grade inflation in school leaving exams and whether the GCSE exams (usually taken at 16) are “too easy”.

In Britain, there is a manifest public reluctance to address the implications of the fact that historically stay-on rates in full-time education after the age of 17 are low by comparison with our peer group of other affluent countries. As the UK’s then school standards minister put it a few years back:

“It must be one of the most stunning statistics that we are 20th out of 24 OECD countries for staying-on rates at 17.”

Btw apologies for over-lapping posts.


Matt Daws 09.10.05 at 1:59 pm

Bob, okay, a good response: in short, why do employers advertise a job with pay X when they could advertise with pay Y (for some Y


Matt Daws 09.10.05 at 2:03 pm

Oh, blast, it appears to have eaten my post! Erm, I might re-type it in a few minutes…


mythago 09.10.05 at 2:13 pm

christian, by ‘practical education’ I was thinking of exactly the kind of hands-on clinical programs you describe–which, as you say, law schools unfortunately don’t seem interested in promoting. I admit I am baffled as to why.


Matt Daws 09.10.05 at 2:14 pm

The basic argument in my eaten post was this:

a) Employers do not perform 100% accurate searches for employees, as hiring people costs money (advertising, interviewing time, lost time because you want an employee *now* and not at some indefinate point in the future);

b) Hiring the wrong person can be costly: it makes sense to pay a premium to be sure someone is competant, rather than have them perform under-par;

c) Employers consistently over-rate degrees and under-rate lower qualifications, leading them to believe, for reasons (a) and (b), that it’s worth paying extra for a degree that’s not really needed (in the belief, say, that anyone with a degree is basically numerate and literate, and that anyone without is not).

Now, one would expect that someone without a degree, but with experience and references, should be equally attractive. I (as I said above) am not convinced that the statistics show this not to be the case (as you would really need to look at (unexperienced+degree) vs (experiences+no degree) applying for the same job, and not just aggregately over everyone).


John Quiggin 09.10.05 at 3:10 pm

Matt, I don’t think these search costs explain why employers are willing, on average, to pay a large premium for graduates. This is particularly so if (as implied by the fact that the premium grows with experience) they are selecting people for rigidly defined career paths. In these circumstances, employers could save a huge amount by hiring carefully screened high school graduates.

In response to various comments, I’ll restate to the wage premium fluctuates over time, depending on whether growth in the number of educated workers outpaces the increasing demand for such workers.


abb1 09.10.05 at 3:31 pm

It’s not just the matter of screening. College graduates come already housebroken; they already invested tens of thousands of dollars and years of their lives into this socio-economic system and they’re unlikely to rebel against it. IOW, they probably won’t be stealing small sums of money.


Bob B 09.10.05 at 4:13 pm


I agree your items (a) and (b). Your (c) is premised on an increasingly pervasive irrationality in employer hiring practices in America and Britain and I can’t believe that. We need to recognise that service industry employment – especially in business and financial services – has boomed while manufacturing employment has stagnated or shrunk in the affluent market economies.

Notoriously, in Tony Blair’s Britain, manufacturing employment has shrunk by 1 million since 1997 but total employment is currently around record levels and the employment rate among working age people is easily higher than in the other major European economies – France, Germany, Italy. But starting work in the service industries is more likely to stress literacy, numeracy and IT skills. IMO employers take it that job candidates who have a degree – any degree (?) – are more likely to have acceptable minimal competence in such skills than would be signalled by the standard school leaving exams without a degree.

As you suggest, hiring a graduate is taken as a (risk averse) course less likely to incur additional firing and hiring costs downstream if the recruit turns out to lack adequate skills. The cult of amateurism is still well-established in Britain – evidence of prior general ability is widely interpreted as indicating a potential competence to gain the knowledge and skills to do almost anything by learning on the job.

A succession of surveys of adult skills in the UK, extended over many years, has turned up the (alarming) finding that c. 20% of adults are lacking essential literacy and numeracy skills. There is an arguable case, if only in cost-effectiveness terms, for rectifying the literacy and numeracy difficulties of school pupils (and adults) rather than expanding higher education just to meet employers’ demands for basic skills and this government’s target of half the catchment age group going into higher education.

Sadly, for whatever reasons, reducing the percentages of pupils and adults with reported literacy/numeracy problems has proved remarkably intractable, notwithstanding efforts by both this and the previous government to tackle it. Britain’s prison population has recently reached a new record of c. 74 thousand and the prison population per capita is higher than in any other west European country except Portugal. Many prison surveys have reported a relatively high incidence of literacy/numeracy problems among prisoners. The published cost to taxpayers of keeping someone in a closed prison is c. £34k a year but we seem quite unable to break the circle.


Dan Simon 09.10.05 at 4:32 pm

Again, there seem to be a lot of people complaining about “credentialism” the way people complain about “life”–without giving any serious consideration to the available alternatives. In the absence of (possibly superfluous) credentials, how will employers choose their employees? By lottery? By deep exploration of the contents of every applicant’s soul? By giving each applicant a two-week trial?

No, the alternatives to credentialism are nepotism, “old boys’ networks”, “connections”, and the rest–all of which are less efficient, less fair and less inclusive than credentialism.

In the 1980’s, the Canadian postal system (if I may engage in rampant anecdote-mongering) wrested from the postal union the right to control its hiring practices. It then instituted an elaborate barrage of formal tests for evaluating the hordes of applicants that responded to its occasional, highly prized (because extremely cushy and overpaid) job openings. These tests were really quite hard–they had to be, to weed out all but the very top applicants–and measured all sorts of skills that could only vaguely be correlated with postal work–memorization, manipulation and sorting of long lists of words and characters, and the like. The resulting employee pool thus ended up consisting largely of intelligent, talented people with low ambition, who could ace the tests, win their sinecures, and spend the rest of their lives in comfortable stasis. One could scarcely imagine a more flagrant case of pointless credentialism.

The union-controlled system that preceded it? An official policy of granting first right of refusal to relatives of current postal workers.


Matt Daws 09.10.05 at 4:34 pm

John, selecting people for rigidly defined career paths. However, my understanding of the job markets in both the US and the UK is that most jobs are in small companies (it’s amazing how many people own their own business and employ a few extra people). I agree with you that large companies should be better-off looking beyond just graduates; however, this is not the majority of employers.

Bob, Yes, I think you’re arguing me around to your view. I fully agree with your later comments: the real problem is the continued failure of compulsory education. My thesis advisor (as you’d say in the US) was head of the mathematics department for a time, and often met employers to discuss how the university could alter its courses to help them. He said that a consistent issue raised was basic numeracy and literacy: really, this shouldn’t be an issue that universities have to deal with! That said, to my shame, I did spend my PhD learning how to write again, after doing four years of mathematics with little need for writing anything like an essay!


Matt Daws 09.10.05 at 4:39 pm

Dan, Okay, I would argue for the following (in a slightly UK specific sense): Abandon artifically increasing university places (i.e. the fact that undergrad intakes to many departments have, say, doubled over recent history, but the average ability of the student has declines) and ploy extra money and thought into raising school standards.

The hope is that employers would then be more willing to look at people with good high-school results, hence forcing less people to have to go to university and incuring the cost associated with that. I’m arguing for replacing a credentialism based upon an optional university education with a credentialism based upon complusory basic education.


John Emerson 09.10.05 at 5:08 pm

Dan Simon — hiring by test is NOT credentialism. Asking people who can pass a test to take a series of college courses is credentialism.

As I said, one manifestation of credentialism is having an uncredentialed worker, who has actually been doing the job, train an inexperienced new hire who has credentials. This is something that really happens a lot. (If the new hire has been hired to upgrade the job with new skills, that’s not so bad; but in many cases the credentialed new worker’s education has nothing much to do with the job).

Uncredentialed workers can neither change jobs nor be promoted. (In other words, we’re not necessarily talking about the initial hire of untested workers.) Either by company policy or by law, employers are often forbidden to promote or to hire the people they want.

Another case is declaring “education” to be a specialty and hiring education majors to teach HS math and English in preference to English majors and math majors.

All sorts of skills that could only vaguely be correlated with postal work—memorization, manipulation and sorting of long lists of words and characters, and the like.

In the US most mailpersons start as mail sorters who are expected to work very fast. This test is very apropos to their work.


mythago 09.10.05 at 7:31 pm

No, the alternatives to credentialism are nepotism, “old boys’ networks”, “connections”, and the rest

False dilemma. The alternatives to meaningless credentials are meaningful credentials and work experience. (If an applicant has no college degree but spent ten years working at a high-level programming job at Microsoft, does the employer really have no idea of their skill set or reliability?)


Rich Puchalsky 09.10.05 at 7:32 pm

I always come into these threads too late. Once again, Fred Hirsch, the classic _Social Limits To Growth_, entire discussion encapsulated.


Dan Simon 09.10.05 at 7:39 pm

I’m arguing for replacing a credentialism based upon an optional university education with a credentialism based upon complusory basic education.

I’m all in favor of restoring meaning to primary and secondary education credentials, which are today often pretty much meaningless. The key to doing so, of course, is to set and enforce rigorous standards in primary and secondary education–and I’m all for that, as well. Indeed, I doubt too many people who are serious about education would disagree with you. (Sadly, education theorists and ed-school academics are, for the most part, shockingly unserious about education.)

But I was under the impression that the discussion here was less about the time scale of the education system than about society’s goals in providing one, and its effectiveness at achieving those goals. Whether education contributes to social mobility, for example, or inhibits it–as Chris argues–via credentialism, seems to me to be a question that is independent of whether the education in question is a hard-won university degree or an equally-hard-won high school deploma.

Dan Simon—hiring by test is NOT credentialism. Asking people who can pass a test to take a series of college courses is credentialism.

Why on earth should it matter whether or not an employer outsources its job candidate testing to a university?

In the US most mailpersons start as mail sorters who are expected to work very fast. This test is very apropos to their work.

I’m given to understand that the level of performance required at the tasks in the tests was far in excess of what was actually needed for the job. That was the whole point–because the jobs were so desirable, a huge pool of applicants was available, the majority of whom were perfectly well-qualified to do the job. The alternatives were credentialism–awarding the job to the applicant whose test score was most extravagantly above the bar for excellent job performance–or cronyism.

The same holds, I would argue, for most jobs: there being no perfect way of ranking candidates for a job, credentialism at least has the advantage of relative objectivity. Sure, it tends to result in overqualification and credential inflation, but the alternative is still far worse.


Dan Simon 09.10.05 at 7:50 pm

False dilemma. The alternatives to meaningless credentials are meaningful credentials and work experience.

But how does one first obtain work experience?

As for “meaningful credentials”, what happens if–as is usually the case, in practice–more than enough candidates have them? How does an employer distinguish among those in that pool?

I only know of two options: proceed to the “meaningless credentials”, or at least, the somewhat less directly meaningful ones, or start counting “subjective intangibles”. The distance between the latter and rampant cronyism is, I claim, rarely very far.


John Emerson 09.10.05 at 9:20 pm

Dan Simon—hiring by test is NOT credentialism. Asking people who can pass a test to take a series of college courses is credentialism.

Why on earth should it matter whether or not an employer outsources its job candidate testing to a university?

Dan, I’m starting to wonder whether I should take you seriously. I was talking about what people mean when they talk about credentialization. Not “why it should matter”. I was just defining a term and the issue. But the reasons why the difference should matter in terms of social mobility (and the recruitment of talent) are: first, that requiring a class provides a barrier to entry (financial and otherwise) which going by a test does not provide; second, as I have said, that often the classes required are irrelevant make-work; and third, that outsourcing to a university often is hand in glove with attempts by organized groups to restrict entry to a trade.

Perhaps the postal requirements were in excess of the actual requirement, but what you said was that the requirements were “only vaguely related” to the requirements of the job, which is completely false.

You really seem to have an obtuse talent for ignoring things you don’t have an answer for. I did mention an additional difficulty with credentialization, which is that it works to the disadvantage of people who were already doing the job before credentialization was required. In a case I knew of, the requirement of a BS degree ended up causing the hiring of a lot of lab assistants who often had no intention of staying on the job even long enough to learn how to do it.


Matt Daws 09.11.05 at 7:04 am

Dan, there’s a world of difference between high-school and university/college though. High-school is (well, vaguely, up to some age varying by country) compulsory and widely regarded as important. An undergrad education is optional, increasingly costly (c.f. the UK situation) and not widely regarded as important. It’s a *lot* easier for an average middle-class kid to get an average undergrad degree than it is for an average working-class kid; this is not so pronounced in high-school…


Bob B 09.11.05 at 9:23 am

My guess is that in Britain, many parents with school-aged children and almost all employers would gladly cheer on proposals for improving school standards rather than continuing to increase the percentage of young people going to university. The trouble is that even with the very best of intentions all round and disposing lots of taxpayers’ money, that is easier said than done as the following news reports demonstrate:

“Government claims that extra money for England’s schools has boosted GCSE results are not based on fact, MPs say. Real-terms spending rose by 31.6% from 1999 to 2003, while the proportion getting the equivalent of five good GCSEs went up five percentage points. In the previous four years, results rose by 4.4 points, while investment increased by just 3.4%, the Commons education select committee found.”

“The higher-grade GCSE success rate in England appears to have reached a plateau, official figures suggest. Last year the proportion of pupils who achieved at least five GCSEs at grades A* to C – as opposed to other qualifications – was 50.2%.”

“Government policies on education are failing to remedy the UK’s shortage of skilled workers, the director general of the Institute of Directors has said. Around 25,000 16-year-olds a year leave school with no GCSEs, Miles Templeman said in a speech to the Institute of Economic Affairs. He said the skills shortage last year left 135,000 vacancies unfilled.”


Dan Simon 09.11.05 at 11:09 am

But the reasons why the difference should matter in terms of social mobility (and the recruitment of talent) are: first, that requiring a class provides a barrier to entry (financial and otherwise) which going by a test does not provide; second, as I have said, that often the classes required are irrelevant make-work; and third, that outsourcing to a university often is hand in glove with attempts by organized groups to restrict entry to a trade.

Required tests can be every bit as bad in each of these respects as required courses. In fact, if I’m not mistaken, a number of professions limit entry not by requiring a degree or course credential, but rather by requiring their members to pass a set of exams–said exams testing knowledge of a large body of mostly irrelevant material that in practice takes expensive and time-consuming courses to master.

Now, the proliferation of entry-limited “professions” may be a problem in its own right. But even in their complete absence, employers interested in avoiding cronyism would still seek out credentials–even only-minimally-meaningful ones–as an unbiased means of narrowing the job applicant pool.

I did mention an additional difficulty with credentialization, which is that it works to the disadvantage of people who were already doing the job before credentialization was required.

I agree that it’s unfortunate that some employers extend credentialism beyond hiring to evaluation of existing employees. But the fact that some employers misuse it in that way doesn’t make it any less useful or important as a method for avoiding cronyism in evaluating outside job applicants.


John Emerson 09.11.05 at 11:37 am

Dan, the problem with tests vs. credentialization is that requiring class work imposes a financial burden and usually also a waiting period (since not all classes are available at all times; students frequently have to wait an extra year, and sometimes even two, for a single required class. The distinction between tests and credentialization is that academically-routed credentialization excludes people who have learned the material without paying off the schools.

In education and other practically-oriented programs, it’s often clear enough that the schools are selling access and credentials. One top of everything else, the gatekeepers (who often haven’t worked in the field for years, if ever) develop an arrogance based on their “research” and “methodology”.

Dan, the employers I was talking about were not voluntarily extending credentialization beyond hiring. Credentialization is institutionalized and enforced. The employers were forbidden to hire or promote certain employees who were capable of doing the work and who,m in some cases, had already been doing the work satisfactorily on an interim basis for some time.

I’ve known people who qualified for difficult uncredentialed fields (software) by self-study. In ermegent areas that’s the most common way.


mythago 09.11.05 at 1:30 pm

But how does one first obtain work experience?

That’s a problem even for the credentialed. And certainly it’s going to be hard to obtain work experience if credentials are required first.

Trade unions seem to have alleviated this problem somewhat, with apprenticeships, which pay below the level of a fully-trained journeyman but which nonetheless come with wages. The trainee is exchanging labor for training, to a degree, and the credential is a mark not merely of classroom study but of a certain amount of practical work under others’ supervisions.

The alternative to meaningless credentials is meaningful credentials, which can be obtained through work experience. Not many institutions offer bachelors’ degrees in real-world experience, useful or no.


Tracy W 09.11.05 at 7:15 pm

I think that a university education do offer some advantages in the workforce beyond a piece of paper, though my evidence comes from anecdotes:

1) You learn how to do research. My mum, with a university degree, went into business with a friend who was a nurse, trained before that required a degree. My mum states that one difference she notices is that if she doesn’t know something, she generally thinks of doing some research into it, while her partner doesn’t think like that.

2) You make contacts in your area of expertise. This is not just a Oxford/Cambridge/Harvard thing. In NZ there are too few universities to work like that. But as a result of my degrees, I do know people doing things related to my work who I can then call up and ask questions. I do not think this would be so relevant for general office jobs however.

3)You get a broad base of information. One academic I know described it as “a set of tools with which you can jump into any foreign situation and make a complete fool of yourself”. Now, having worked for a number of years I know what he means. But even though I do regularly make a fool of myself, I can also suggest ideas from other areas that might be useful for a new problem that the job is facing.

I do not know how much this applies to arts graduates. And I do think it is a pity to see arts students at university who just sort of wandered into the degree since they got their best marks at school in English, but without any plans for what they will do with that degree in the job market and without the passion for the subject that makes spending 3 years and tens of thousands of dollars worthwhile for its own sake (even if uni is free you face the opportunity cost of the wages you could have earned).


Justus 09.12.05 at 12:41 pm

tracy –

1) I work (as a software engineer) exclusively with people with university degrees. I estimate that in a building with 50 degreed engineers perhaps 5 of them would do research such as your mom does. I don’t think other factors far outweigh the university degrees.

2) This becomes significantly less meaningful in the US if you ever relocate. I spent the first decade of my professional life on the east coast, in Boston. Now I live in Colorado and my east coast contacts are fairly worthless. This is compounded by a slight change in my technical subfield.

3) Anything an academic has to say about the usefulness of academics is suspect from the start :)

Comments on this entry are closed.