It is one thing to say that inequalities exist in enrolments to the UK¡¯s universities; it is quite another to say that the admissions process is unfair.

This is perhaps more, not less, evident in the wake of the publication for the first time by Ucas of institution-level data on intakes and offer-making.

The figures, broken down by gender, socio-economic background and ethnicity, laid bare the disparities in admissions to ÖйúAƬ more starkly than ever before: they show, for example, that 18-year-old females brought up in the UK are more likely to enter nearly nine out of 10 universities than their male peers, and that applicants from the most advantaged backgrounds are nearly eight times more likely to enrol at a Russell Group institution than teenagers with the least advantaged upbringings.

However, the statistics on offer-making show that, once an applicant¡¯s qualifications and chosen subject are taken into account, these differences shrink significantly.

ÖйúAƬ

For instance, women were only 0.7 percentage points more likely to be offered a place than men last year, once attainment and discipline are controlled for, and the gap for the most disadvantaged students hoping to enter a Russell Group university reduces to just a handful of percentage points in most cases.

This allowed Ucas, when it released the data for 132 institutions, to argue that the admissions process to UK universities was ¡°broadly fair¡±.

ÖйúAƬ

However, a closer look at the figures reveals little cause for complacency, and shows that students from some backgrounds remain significantly less likely to be offered a place at many universities.

Analysis by Times ÖйúAƬ shows that,?compared with what Ucas would expect for students with similar qualifications and course choices, 18-year-old UK-domiciled students from the most disadvantaged backgrounds were offered a lower proportion of places than expected at 103 of the universities in 2015.

In contrast, students from the most advantaged backgrounds received more offers than would have been expected at 110 out of 132 universities.

Biggest gaps in offer-making, controlling for prior attainment and course choice (excludes smallest cohorts)

Least advantaged

Black applicants

Asian applicants

On ethnicity, similarly worrying trends are evident. Although white 18-year-olds are now less likely to enter ÖйúAƬ than teenagers from any other group, and are under-represented at nearly half of UK universities, they are 0.4 percentage points more likely than would be expected to get an offer,?nationwide.

It is ethnic minority students who are less likely to get an offer: applicants from black backgrounds are 2.6 percentage points less likely to get an offer, compared with what would be expected once attainment and subject choice are taken into account. Asian students are 1.2 percentage points less likely to get an offer.

ÖйúAƬ

ÖйúAƬ

At institutional level, this means that black applicants are less likely than would be expected to get an offer at 100 universities, and Asian students are less likely to get an offer at 90.

White students are more likely than would be expected, based on their grades and course choices, to get an offer at 107 institutions.

Mary Curnock Cook, Ucas¡¯ chief executive, said that individual institutions will want to examine areas where their admissions rates are significantly below average.

¡°People will find some little surprises in the data for their own institution and will go away and look at it,¡± Ms Curnock Cook told the annual conference of the ÖйúAƬ Policy Institute. ¡°My experience is that universities that find anything wrong or anything that looks wrong will very quickly go away and do something about it.¡±

There is much that the Ucas data do not tell us. The offer-making analysis controls for applicants¡¯ grades but does not control for which subjects these grades are in, potentially a key issue in admissions. It also cannot take account of issues such as personal statements, academic references or interview performance: all areas in which students from more affluent backgrounds may be able to present themselves more attractively to universities. And it does not give information on whether applicants came from state schools or private schools.

Number of universities where certain groups are over- or under-represented among offers made, controlling for prior attainment and course choice

But some institutions are likely to look at the data very closely indeed. At four Russell Group institutions ¨C the University of Glasgow, Durham University, Imperial College London and King¡¯s College London ¨C black applicants were more than 5 percentage points less likely than would have been expected to get an offer in 2015. At the University of Oxford, students from an Asian background were 7.7 percentage points less likely to get an offer than would have been expected. And arts-focused ÖйúAƬ providers, such as the University of the Arts London, appeared to perform particularly poorly on making offers to students from the most disadvantaged backgrounds.

At the same time, the data could be used to identify institutions that are performing particularly well on widening access and could share their best practice with the rest of the sector. At the London School of Economics, for example, students from the poorest backgrounds were 26.8 percentage points more likely to get an offer than would have been expected in 2015; and black students were 6.1 percentage points more likely to be offered a place.

Vikki Boliver, senior lecturer in sociology and social policy at Durham, has hypothesised that highly selective universities might?unfairly reject ethnic minority applicants from courses that attract a lot of interest from ethnic minority students, in order to recruit cohorts that are more representative of the wider population.

But Dr Boliver said that a full picture of inequality in admissions would not become clear until researchers were able to access data on individual applications, something that is hoped for in 2017.

ÖйúAƬ

¡°If you look at things at the aggregate level, it doesn¡¯t look so bad, but there are pockets where it looks much worse and particular institutions have quite substantial gaps,¡± she said. ¡°I would imagine that, if you drilled down even further to specific courses, you would find things there which were not apparent at institutional level.¡±?

chris.havergal@tesglobal.com