A crack in the higher education model? (hi future…!)

One of the UK’s biggest graduate recruiters is to remove degree classification from the entry criteria for its hiring programmes, having found “no evidence” that success at university was correlated with achievement in professional qualifications.

Accountancy firm Ernst and Young, known as EY, will no longer require students to have a 2:1 degree and the equivalent of three B grades at A level to be considered for its graduate programmes.

Instead, the company will use numerical tests and online “strength” assessments to assess the potential of applicants.

There is more here, via the excellent Jake Seliger.


It is called in common business parlance "pricing yourself out of the market".

That's jolly handy for anyone with friends prepared to do the online tests for them.

The proof of the pudding is in the eating. It's not a school test where you are done once you've passed it. It will be clear soon enough if you underperform, and if you don't, who cares about the test.

Quite right. If I were Ernst and Young, I would expect that my elaborate interview process would weed out people who wouldn't perform, regardless of degree classification. However, it can't be ignored that grade inflation is part of this story. There was a time when II.1 was a really high-class achievement and recruiters used it as a filter to reduce the number of paper applications they had to handle. Now there are online tests and CV-algorithms, they can skip it. Still, I'd be surprised if there is ZERO correlation between degree class and success at the firm.

Remember that it is their clientele ultimately who are deciding this. Accounting is a necessary evil, a pure cost where beyond getting your taxes paid and filling regulatory requirements, you don't see value added beyond that. Most of the trend stuff for business and operational decisions is in house.

Maybe the real story here is the choice of off shoring or hiring local. The choice of an inexpensive degreed worker from Mumbai or a non degreed inexpensive worker from Manchester.

It looks like they train internally, so the skills to do the job are not necessary. Maybe this is also a matter of efficiency. It costs far less to train someone with basic literacy and numeracy out of high school than take someone who ostensibly is trained, accompanied with debt and high wage expectations, and still needs to be trained how to do the job.

The crisis in education where costs and debt are increasing will have some evidence in the colleges themselves, but the real evidence that there is a problem will show up when all those costs are represented in the wage premium. At one point it isn't worth it, except for the very few who are exceptionally intelligent and capable.

All the literature I am aware of says that interviews are all but worthless. At the very bottom of the list of things that correlate with job success.

From my time in the hiring process of a "big six", lots of data was tracked.

Interviews weren't worthless, but clearly mistakes were made and recruiting errors did happen. The easiest to know about are the people a firm hires that don't work out. The data about those you don't hire though, that's more difficult to track.

These professional services firms are quite smart about hiring, imo.

I think the E&Y new rules are to attract minority hires, so that they can be easily hired within the formal corporate rules (otherwise it requires an exception that has to travel up the chain of command, maybe even to the board, which is a hassle). I could be wrong but it seems that way.

If they're basing the hiring on written tests, then I guess they could use that to select for minorities (Asians in the US) whose performance is better than average on tests. In a US context, if someone is trying to hire more minorities (blacks and hispanics), they tend to de-emphasize test scores in favor of grades, class-rank, interviewing, etc., because blacks and hispanics have lower average test scores than the population average.

I thought that this wasn't allowed in the U.S. because of disparate-impact laws, ever since the famous Griggs v. Duke Power in 1971.

That is, you can't have an actual test because different ethnic groups perform differently on it, unless you can prove that the test is DIRECTLY related to the job they will be performing, not just testing general intellectual ability or IQ or anything. So, companies use educational attainment as a sort-of-okay proxy for their performance on the test.

I bet that companies like Google, Amazon, and Uber would love to just use IQ and coding tests instead of asking for certain educational certifications. In fact, aren't their interviews basically moving in that direction anyway? How has there not been a disparate-impact case against those kinds of companies?

My understanding is that it isn't illegal but there is uncertainty about what you can and can't use. And of course in the US, you're always one unlucky lawsuit away from failure even IF you conform to the law.

Not illegal in the US. It is fairly easy to design a test that covers a lot of "basic education skills" but is also job specific.

I've taken them, because employers no longer trust a prospective employee to have competent basic education skills, regardless of degree.

It's not illegal but the companies have to show that the exams are testing for skills relevant for the job. Most don't want to go through the hassle but some big companies still utilize IQ like tests in the hiring process. P&G and Capital One come to mind. The founder of Capital One is big on hiring for IQ.

Even some tests that are specifically based on job knowledge have been thrown out as discriminatory, the EOC courts have a lot of arbitrary power, so it can come down to a dice-roll basically.

One case I read about in a business text was a mortgage company that employed people to do some basic online research. They gave the candidates a timed test with factual answers they had to find online. The EOC court found there was disparate impact and it wasn't a business necessity because of the timed aspect. In other words, they could just hire more people to get the work done more quickly. You aren't allowed to "discriminate" in favor of more productive employees--that's racist.

Whiteboard based coding tests are pretty common for people going for software development jobs.

Yeah, good luck doing anything close to IQ testing in the US after Griggs. Unless E&Y is very careful in dressing the test up the government will quickly shut them down. The audacity of some companies to think that they can decide how to run their business know no bounds. They didn't build that!

I really do not believe the American government is going to shut them down for their hiring practices in another country.

I thought that this wasn’t allowed in the U.S. because of disparate-impact laws ...

The story is about "One of the UK’s biggest graduate recruiters."

E&Y operates in the U.S. as well

So they basically discount a summary mark that reflects three years of steady demonstrable effort, and replace it with a three-hour do-or-die test under extreme conditions. Brilliant. That's sure to give them the dynamic up-and-at-em 'storm-the-trenches!' world of erm... accounting and auditing.

Isn't that even greater evidence that people don't trust the integrity of that credential? Especially if grade inflation is quite different across departments and across universities.

Not sure this would work for Ernst, at least not for client facing roles. However I hope more companies adopt it.

One day the interview will be DNA sequencing test?

That's basically the plot of Gattaca

Given MBA and other business programs teach amoral economic theory that prefers rentiers and monopolists to serve only getting rich by any means possible, it's not hard to understand why a firm that needs to attest under risk of prosecution to conforming to laws trying to legislate morality would see not relationship to education performance and the qualities of the employees they need.

In fact, a student that got C's because he refused to answer "the only responsibility of the CEO is maximizing profits any way possible" is a far better candidate for EY than the many who fully believed that killing workers or customers if that increased profits was required. Look at the penalties that GM is paying with many people outraged that no workers are being tried for murder reflecting the expectation of We the People corporations conform to the same moral code as all individuals.

I see you've finally reached the "word salad" stage of whatever disorder you have.

Why do some people of a certain part of the ideological spectrum immediately become highly insulting when someone raises legitimate critiques of unfettered capitalism?

The lack of criminal liability for white collar workers who cover up data and/or act too slowly to prevent more deaths/injuries is appalling.

It is, indeed. And the lack of malaria nets and vaccines for preventable childhood diseases in poor tropical countries is even more appalling. Neither, however, is the topic of this post.

Nathan, it's from the mulp.

Finally? He's been there for years

Grade inflation creates a problem (in distinguishing between graduates) in the U.S. I am a lawyer. My law school (which I attended many, many years ago) used a bell curve for grading, a blind anonymous grading system (BAGS), and kept track of class rank from the first semester to the last. Many students complained that professors cheated on the BAGS because students had a high probability of performing at or near their class rank (that they performed at their potential didn't seem to register). Hence, the top firms would seek out the students with the highest class rank. Today, many if not most of the elite law schools don't keep track of class rank (or don't share it with recruiters), believing as they do that all of their students rank near the top of the class. I understand this creates some tension between the law schools and firms that wish to recruit only the top students. I'm not sure how this is resolved. I suspect by subjective means, such as letters of recommendation from professors. When law graduates were in high demand, recruitment was accomplished primarily with summer clerkships, so the firms could review the students up close (although summer clerkships were often less about work than about recruitment). Now that law students aren't in such high demand, clerkships aren't as popular among the firms. I'm curious if the elite law schools have reinstated class rank, or if they continue to maintain the fiction that every student is a top student.

"having found “no evidence” that success at university was correlated with achievement in professional qualifications": is the claim simply that their degree and A-level grades don't predict how well they'll do in their accountancy exams? If so the obvious question is how well do their grades in their accountancy exams predict how well they'll do in their careers?

I imagine this will help them to screen large numbers of applicants who can be met with low-ball offers ... lower-ball offers than they would have been able to make to graduates, whe presumably should internalize the cost/time of their education in their wage expectations.

The accounting courses I took were online courses, and it seems to me that the material in accounting is highly amenable to self learning. Moreover, in most jurisdictions, accountants will eventually have to pass formal credential requirements.

I expect this will help the firm to get a lot of simple work done at a lower price, while still ensuring that they can train for longer term growth/abilities over a career.

It has nothing to do with that. E&Y doesn't succeed by low cost/price.

It is a way to find potential great recruits from a pool of people (likely lower socio-economic status) that don't get into the best schools, but are still very bright.

btw, nathan, you're about 60% as intelligent as you think you are.

I suppose I fail to pass your rest of ideological rigour? Or is it my non-American spelling that bothers you?

It seems like the most obvious explanation here is the one they give--their data in tracking their students doesn't show much effect of grades. Now, I assume they're smart enough to know what they're doing, but there are a couple obvious ways they could have gone wrong here:

a. If they had restrictions so that only people with very good grades from top schools got hired, and they assessed those employees' performance against their grades, they may fall prey to range restriction. That is, maybe within a population of people who all got pretty good grades, the straight-A students aren't much different from the mostly-As-occasional-Bs students, but both may give you very different performance from the guys who got all Cs and barely passed.

b. If you had to be a better candidate otherwise to get hired when your grades were good but not perfect, then their hiring process may have selected a batch of employees where the perfect students were less impressive than the less-perfect ones who could only get hired because of their other really impressive traits.

To see why both of these might have happened, imagine looking at NBA players by height to decide who's going to be the most valuable player. The correlation between height and value as a player is probably relatively small within the NBA, where there are essentially no short people, and which is competitive enought that a 6'3" guy who makes it into the NBA at all must have all kinds of amazing abilities that make up for his not being four inches taller. But you might want to be careful using this insight to predict whether height predicts value as a basketball player in the wider world.

A friend told me that the big insurance companies that he worked for liked to hire English majors in the accounting department because they could be taught to do just as good job as the accounting majors and they had far fewer options and so where less likely to jump to another firm and you could pay them less. Ernst and Young could make this work the same for them.

Too bad that thanks to a stupid Supreme Court decision in the 1970s, it is illegal in the US for firms to require intelligence testing for potential hires.

Comments for this post are closed