This article is more than 5 years old

Why we should select students by lottery

Lee Elliot Major and Stephen Machin introduce their forthcoming book, and propose an interesting idea that could help social mobility.
This article is more than 5 years old

Stephen Machin is professor of economics at the London School of Economics, director of the Centre for Economic Performance and is a fellow of the British Academy and Society of Labor Economists.


Lee Elliot Major is Professor of Social Mobility at the University of Exeter

We cling to the belief that education is the great social leveller, enabling children from poorer backgrounds to overcome the circumstances into which they are born.

But in our book Social Mobility and Its Enemies we discuss evidence, gathered over a number of decades and from a range of countries, that shows for most children, education has failed to live up to these lofty expectations. It is a near-impossible task when the income and wealth gap between the haves and have-nots is so wide outside the school and university gates.

In 1961 Lord Robbins established the ‘Robbins principle’ – declaring university places “should be available to all who are qualified by ability and attainment to pursue them and who wish to do so”. It is an aspiration that has remained at the heart of HE policy ever since. Yet the march of Britain’s universities from elite to an expanded system has remained a highly exclusive endeavour.

A middle-class affair

In recent decades university expansion has predominantly been a middle-class affair. The proportion of the poorest young people earning degrees grew from 6% in 1981 to 18% in 2013. But, at the same time, the proportion of richest young people earning degrees went up from 20% to 55%. In one generation, the graduation gap nearly trebled; and because graduates make more money than non-graduates, the middle classes have got even richer. Inequality in education begets inequality in life and this transmits further across generations to cause intergenerational inequalities.

What we have observed has been an ever-escalating education arms race, in which the poorest children have been hopelessly ill-equipped to fight. The signs are all around us: the booming billion pound industry of private tutors paid to boost pupils’ grades; the sharp-elbowed tiger parents stopping at nothing to get their children into the best schools, and the stressed-out students trying to navigate a complex and often baffling university admissions system.

Middle-class advantage

In the past the middle-class advantage was found through A levels and university degrees. Today it is achieved through postgraduate degrees and exclusive internships. A degree is no longer the automatic passport to a well-paid job it once was; now a particularly exclusive degree plus a masters qualification is required. Just as the education system expands to equalise opportunities, so a new frontier emerges enabling the well-off to climb one step up again.

Bombarded by thousands of A-grade candidates, the most sought-after universities are resorting to hyper selectivity – ever more refined, but increasingly unreliable, ways of selecting the very best academic talent. The battery of admissions criteria designed to distinguish between equally well-qualified candidates expands with each year: personal statements, teacher recommendations, school exam grades, university admissions tests, interviews, ‘contextual offers’, and much more. In the last year the murky practice of unconditional offers has come onto the market – as universities strive to fill their degree courses.

It comes as little surprise – given the escalating competition and the wide gaps in school attainment – that university student enrolments are so skewed to those from privileged backgrounds. We now spend millions of pounds on outreach work to raise the aspirations of low-income students and engage in endless debates about the fees and funding for students. But one simple reform could wipe away middle-class advantage in university admissions in one fell swoop.

Time to select students by lottery

A university selection system not only fair, but fit for purpose, would identify a threshold of academic excellence and then, as long as it is oversubscribed, select students through a ballot. Universities could identify the minimal grades that are good enough to get onto a degree course. Undeniably, the most equitable way to then allocate places to equally deserving candidates would be to pick them randomly. Those who don’t win a place could, perhaps, be guaranteed a place at a similar university.

This would not compromise academic standards – no one wants to set young people up to fail in an intense and demanding academic environment. At the same time we must also remind ourselves what universities are selecting for with most graduates pursuing non-academic careers after graduation, requiring only a certain level of academic achievement. Universities could develop a ballot system that suited their needs.

Dutch medical schools for example selected the very highest academic performers on academic grades, and then entered lower achievers into a lottery.  A progressive lottery system meanwhile might allow poorer students to be entered with slightly lower grades than their more privileged counterparts.

Deploying random allocation alongside simple and transparent academic criteria would create more diverse intakes of students. Universities could help rather than hinder social mobility. And it would have the added benefit of cutting down on the escalating costs of admissions.

7 responses to “Why we should select students by lottery

  1. Why we should not:

    1) lotteries guarantee (pretty much) that the most deserving students do not win places;
    2) unless the “threshold” is set at a very low level indeed (eg functional literacy & numeracy), the consequences for admission of social/ethnic groups in proportion to their numbers are hard to model and may in fact generate better outcomes for some groups at the cost of worse outcomes for others;
    3) setting the threshold at or close to baseline would result in an application stampede akin to the California gold rush (the only participants profiting from the prospectors were the sellers of shovels and food) aggravating 1);
    4) there is no evidence that the general public / taxpayer views lotteries as a fair allocation mechanism (eg nobody has yet proposed replacing the House of Lords with a 2nd chamber where voters are allocated by lot and asked if they wish to serve or not; eg Sutton Trust do not recruit their staff by lottery after sifting applicants by means of a threshold test);
    5) there are no examples in global HE where a lottery system has been adopted and shown to work.

  2. PS the one example given in the post – admission Dutch medical schools – is actually being discontinued after repeated controversy and operational difficulties. There is a summary of the saga here: http://benwilbrink.nl/projecten/lottery.htm

    Back in 2007, Sutton Trust commissioned Rand to produce an independent survey of lotteries and admissions.
    https://www.rand.org/pubs/working_papers/WR460.html

    It found that ‘The evidence gaps in evaluations of existing lottery schemes mean that this study cannot answer what the impact of the introduction of lottery schemes on equitable outcomes is likely to be and what types of lottery schemes might make school admission more socially equitable in the UK context.’

  3. Its a great idea. The ‘low threshold’ complaint is easily dealt with. For example, Oxbridge should allocate by lottery amongst 3 xA* students; and then, if that doesn’t provide enough, at the next tariff point down – the author’s ‘minimal grades’ point.
    There is no doubt that the Oxbridge interview system is about nothing other than social selection. It is gamed to the point of corruption (eg pre meetings with admissions tutors, but only if you are private school connected). More generally, there are all kinds of curricular discriminations. Want to do Classics ? But given no state schools do Latin or Greek, to speak of, then all state school students are automatically disadvantaged, and are so far ‘behind’ they never get to be Classics scholars. And so it reproduces.

  4. @William Cooke

    Because of the immense attainment gap by household income, selecting by lottery from only those students achieving A*A*A* (and so on down the grade gradient) would in fact generate a far more socially stratified intake than the one Oxbridge has!

    For evidence, see Crawford, Dearden, Vignoles, & Micklewright, Family Background & University Success (2017), p.92

  5. I welcome this debate and, broadly, agree that such a system should be considered – although I think the probability of any university being prepared to try it is low! – but have some concerns. There are qualities other than academic achievement needed for success at university and in graduates’ future careers. For example, a good doctor or nurse needs a range of inter-personal skills.

Leave a Reply