Causation Matey, causation

The value of a GCSE has been revealed for the first time, as an official analysis shows that every higher grade is worth an extra £23,000.

For the first time, researchers at the Department for Education (DfE) have quantified the link between better exam results at age 16 and earning potential.

They calculated that pupils who achieve just one grade higher in a GCSE subject will go on to earn an average of £23,000 more over their lifetime.

And students who secure one grade higher than their peers in each of their nine subjects can expect to earn an extra £207,000 before they retire.

It’s rather more likely that those who have the brains and application – the second possibly being the more important – to do the work to get the higher grades will earn more. The higher grades therefore being a signal of the brains and application, not themselves the cause of the higher earnings.

Straining mightily to plop a higher grade into the latrine of life isn’t causing the higher incomes……

15 thoughts on “Causation Matey, causation”

  1. Credentialisation, Matey, credentialisation.
    Back in the 60s, in my O Level year, I was already working part time. Skiving off school to do it. Pain to have to back & sit the exams but passed the ones interested me. In those days you had to physically go to the school to collect the certificates. Never bothered. Maybe they’re still in a filing cabinet somewhere. Never been asked for them. Most of my year went on to do two years in the sixth & on to university. When I was over that part of London I’d pass them on their bicycles in the MG softop. But never saw much of them. Different lifestyles. They living with Mum & Dad. Me with a flat in Kensington & a live in girlfriend.
    Could you do that now? I doubt it. Could you even get the job get you on the first rung of the ladder enable you to climb to the top one. Or if you did, which rung would you attain before lack of credentials blocked further progress? These days the world’s run by people with bits of paper. And they make damn sure no one without those bits of paper gets anywhere. Shame the only qualifications they seem to ave for what they’re doing is those bits of paper.

  2. The pendant in me also wants to point out that the NPV of £23k ‘over a lifetime’ is about three shillings and sixpence.

  3. And students who secure one grade higher than their peers in each of their nine subjects can expect to earn an extra £207,000 before they retire.

    Sounds like horseshit anyway. I was one of the first or second lot to do GCSEs, so we worked with a lot of old O level papers. GCSEs were WAY easier. Bearing in mind continued grade inflation, hard working bright kids will get top grades, but the bang average and slightly lazy ones will probably get them too.

    Although bearing in mind the decreasing competence of the state education sector, perhaps not.

  4. BiS,
    I suspect the whole point of turning universities into left wing madrassas is to stop right wingers from geting the necessary bits of paper to run anything.

  5. This may be true in the aggregate, as an average. But there is no such thing as an average person. Who you know, where you happen to be and sheer dumb luck are larger influences on lifetime earnings.

  6. Piss off Philip. That sort of defeatist predestination is rot.

    If you work hard you will almost always do well. If you are lazy you will almost never do well.

    The rare exceptions, almost always due to illness, are no way to run your life.

  7. Regardless of what bollocks the DT has decided to spaff out, the actual report does say;

    “1. The dataset cannot measure every variable that could possibly affect both grades
    and earnings. If there are important unseen traits that are correlated with both
    grades and earnings, this will add bias to the estimated returns.
    2. We cannot distinguish between earnings returns caused by gains in useful skills and
    knowledge (human capital) and returns caused by possession of the qualification
    itself, which provides admission to further study and employment, irrespective of
    skills and knowledge (signalling).

    4. Cohorts in the sample completed their GCSE’s almost twenty years ago. Using old
    data is unavoidable if we wish to measure long-run outcomes. The implications of
    the twenty-year delay should be considered when making predictions of current
    policy impacts, based on these estimates”

    https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/993202/GCSE_Attainment_and_Lifetime_Earnings_PDF3A.pdf

    The cohort is those taking the exams between 2001-04. Presumably, there’s a reason why those dates were used, as opposed to 1992-1995, or 2010-2013. That 2001-04 cohort would be 37~ years old by now, so future earnings to retirement are the result of a, umm, model.

  8. @ Philip “Who you know, where you happen to be and sheer dumb luck are larger influences on lifetime earnings.”

    The Left’s mantra that the system is rigged and that everyone who has done well has somehow cheated or been lucky or both provides the perfect excuse for failing or not even bothering to try.

  9. I would suggest that how you spend your cash makes as much difference to your life as how much you earn. Not wazzing most of your money away on stupid shit that you don’t need has to worth a fair bit.

  10. BIS,

    “Could you do that now? I doubt it. Could you even get the job get you on the first rung of the ladder enable you to climb to the top one. Or if you did, which rung would you attain before lack of credentials blocked further progress? ”

    This is one of the great myths perpetuated by the establishment of academia, middle classes, media and government. The bits of paper are mostly about government jobs (including those jobs licensed by government like dentists and lawyers) and the sort of large companies that have diversity workshops and shiny head offices.

    Even in those large shiny companies, doing a good job, working hard, seizing opportunities count more once your foot is in the door.

    Outside of that, no-one cares. I want programmers who can build good software. I know that having a comp sci degree doesn’t guarantee that. I know that most other degrees really don’t (math seems to be reliable). You give me someone who has built Roblox widgets after school and is making £10/week from them. I know that kid has enthusiasm and ability to write code. He’ll barely need teaching to do web programming – you just have to tell him to build it and he’ll find out how.

  11. When I was still working we had a client that would only employ graduate engineers. We used to fix the machines that these “engineers” had been attempting to fix.

  12. A third factor for Timmy’s list is that school grades depend heavily on parental education level / income / class – indeed, they often indicate that parents were well-to-do enough to get the kid into a good school in the first place. And coming from a “good family” with appropriate connections is an excellent start to the career ladder, as a separate effect from any intelligence or work ethic that’s been passed (whether by genetic or environmental factors) down the generations.

    @Ducky McDuckface

    Thanks, it’s an interesting paper. I particularly liked appendices I and J, whereby doing well in certain subjects turns out to be associated with a reduction in earnings…

    Due to the nature of their dataset, they didn’t have a whole lot of information on the parents/family background. They were partially able to control for it (via race, deprivation of the area the family lived, and whether the student was eligible for free school meals) but that’s going to miss a lot of subtleties, particularly between the lower and upper end of the middle class.

    Another thing that would have been interesting to see is what happens to your estimate of the value of qualifications if you control for IQ score (or similar measure)… some US studies on the value of education are able to do this, because of the military’s extensive use of intelligence tests during the era of conscription for Vietnam etc.

    It’s nice to see the authors here actually anticipated Timmy’s points, and take e.g. signalling seriously as an idea even if ultimately they don’t do anything about it. They also wrote a very interesting companion piece about how you can apply the kind of reasoning in the Treasury Green Book to the cost-effectiveness of education policies. Lots of very neoclassical economics in it – elasticity of marginal utility, life-course models, social welfare functions, “Following neoclassical theory, profit-maximising firms employ workers up to point where their marginal cost of labour equals their marginal revenue product” and a bit more in-depth discussion about signalling. Plus the necessary declaration of surrender that at the end of the day, decision-makers might prefer to think about “unmeasurable” things, or “broader” conceptions of “equity”, than a bunch of recommendations based on neoclassical economics.

    https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/993200/Schools_Policy_Appraisal_Handbook_PDF3A.pdf

  13. MBE – That other paper looks interesting, and I haven’t yet had the time to fully work through the first!

    It’s interesting that the apparent potential reduction in lifetime earnings for grade outperformance in certain subjects implies that underperformance in those subjects leads to increased wages. Also interesting that they seem to find diminishing returns as you move up the grades.

    What is concerning about the research is that the 2001-04 cohort used is after the introduction of the minimum wage – the ONS found that the wage distribution has changed such that a higher number of jobs (about 9~10%) are now clustered within 3~5% of the MW – which to my mind has the implication that the reported effect is at least partially down to policy completely unrelated to edumaction.

    But I suppose that I’ll have to read things properly RSN.

Leave a Reply

Your email address will not be published. Required fields are marked *