he stories were heartbreaking; the sense of injustice utterly profound. The government had little choice but to U-turn on its decision to use an algorithm to dish out A-level “results” to thousands of young people that had bafflingly little connection to their ability or the quality of their work. But in announcing this without setting out a plan for what it means for university admissions – some institutions will now be massively oversubscribed in terms of places, others undersubscribed – it has simply pushed the problem elsewhere. Step back, and this whole fiasco exposes some uncomfortable truths about how fair the system is normally, and the true extent of and limits to meritocracy. We take it as given that dropping a couple of grades in a one-off exam should amount to the be-all and end-all in determining which university you go to. Have a good exam day, and you could be attending the university of your dreams; have a bad day and the anxious cycle of clearing starts. All this is predicated on the crazy idea that we need to avoid AAA students studying with ABB students, BBC students studying with BCD students at all costs or … what? Rankings allow illusions of meritocracy and simple choices to prevail – obviously we should go for the AAA student over the ABB one – when the reality is they may be covering up a choice that is more random and arbitrary than we may like to think. A-levels may be great at ranking the ability of students to take a particular exam on a given day. But how useful is that for predicting how well they will do at university, sometimes in an entirely unrelated subject? Or in any given job? There will be some link, but is it strong enough to justify a university system in which dropping a single grade can close off entry to your first choice? If exams aren’t perfect, perhaps there are other ways of assessing students. But the other options all come with their own problems. Teacher-based assessment tends to be biased against young people from disadvantaged backgrounds; coursework can be a more accurate assessment of a teacher’s willingness to coach the work out rather than a student’s true ability. There simply is no such thing as a perfectly accurate method for ranking young people’s abilities that works for everything from university entry to job recruitment. That’s only bad news if we consider a failsafe assessment of a young person’s future abilities at the end of school as the alpha and omega of the education system. But why should it be? Why do universities and employers even need this in the first place? It’s only important for universities if we believe that it is paramount that the young person who gets BBC in their A-levels should not study alongside the young person who gets BCC: that they must to be sorted into different institutions. To see how bizarre this is look at the school system, where experts have shunned the use of academic selection in light of evidence that creaming off the most able children into separate schools does barely anything for their learning and simply worsens outcomes for everyone else. Yet for some reason, when it comes to post-18 education, we turn this on its head and go for extreme levels of academic stratification – which, because children from more affluent backgrounds are much more likely to attend good-quality schools, also produces a highly socially stratified university system. You get tutors worrying about whether a student has the ability to complete a course at their university if they’ve dropped a couple of grades, without ever wondering whether this says a whole lot more about the quality of teaching at their institution than the student in question. (It absolutely does.) This system works brilliantly for the most selective universities, who get to select the highest A-level performers who hail disproportionately from affluent backgrounds, then claim the kudos when – surprise, surprise! – those students go on to do well. But it also fuels unjustified elitism. Because universities essentially mark their own homework – a first from one university is not equivalent to a first from another – employers tend to use the university someone attended as a shorthand for their labour market potential, rather than the skills they actually developed there. The prophecy becomes self-fulfilling. We couldn’t design a better way to entrench privilege if we tried. Employers also deserve more scrutiny. Traditional recruitment includes screening CVs – with A-levels and degrees a key filter for entry-level jobs – followed by an interview. Not only is this method ineffective at selecting the best people for the job, it can be downright counterproductive. Experimental studies have shown that interviewers are notoriously unreliable at predicting someone’s capabilities. And interviews are a reliable way to smuggle bias into the process: interviewers tend to go for candidates that look and think like them, leading to less diverse, more groupthink-dominated – and less successful – workplaces. One real-world example comes from Texas in the late 1970s, where a doctor shortage pushed politicians to instruct the state medical school to increase its admissions after it had already selected 150 applicants after interview. It took another 50 candidates who had been rejected at interview – after much of that pool had already been snapped up by other universities. Those 50 went on to perform just as well clinically and academically as the original crop: their success at interview made no difference at all; they might as well have been picked from the shortlist at random. We don’t need A-level results to provide a high-stakes ranking. If we were more honest about the limits of meritocracy, we would move to a more comprehensive-style university system – where, like at school, young people of different abilities learn alongside each other, with real academic benefit. More large employers would disregard A-level grades and degrees, running aptitude tests and assessment centres to assess the skills relevant to their workplaces, randomly selecting from those candidates who make a cut-off to ensure a more diverse intake than interviewing could ever allow. We have extensively debated the fairness of an algorithm, but let’s not fail to ask why we even needed one in the first place. The real reason is that the system as it stands shores up elitism and maintains the grip that the upper middle class – for no good reason – has on influential jobs. But this weak attempt to fake meritocracy is a harmful pretence that denies too many young people a chance in the first place. • Sonia Sodha is chief leader writer at the Observer and a Guardian and Observer columnist • This article was amended on 18 August 2020. An earlier version said that the head of one Oxford College told the writer that she supported the idea of experimenting with ditching interviews and allocating places by lottery. This has been removed because after publication, the college’s principal said she had misread a Twitter thread and had not intended to support those ideas.