The paper Dr Who is currently in the process of revision.

 

The latest version (February 2011) is availble at the link: Dr Who (Feb 2011).

 

There is an Observer article (also available as: Observer article pdf) and a BBC article (also availbale as: BBC article pdf) about this paper based on the ERI Bulletin of May 2011.

 

Political debate about tuition fees and the funding of higher education in the UK has raged furiously since October 2010 with the publication of the Browne Review and subsequent government legislation raising the cap on home student tuition fees from £3k to £9k per annum. The major concern generated by government policy centres on the fear that more able students from poorer backgrounds will be deterred from participating in higher education. The Office for Fair Access (OFFA) is to be charged with regulating the sector to ensure that where maximum fees are charged, universities will simultaneously implement policies designed to widen participation.

Running in tandem with the establishment of OFFA, the 2004 Schwartz Report into ‘Fair Admissions to Higher Education’ made a number of recommendations to support the principles which it identified as appropriate to ensuring fairness in university admission. A key recommendation of the Schwartz Report was that the historic ‘pre-qualification’ admissions system be replaced by a Post-Qualification Application (Post-QA) process.

What is a pre-qualification admissions system and why might a Post-QA process be better?

Admission to UK universities for full-time undergraduate courses is administered through the Universities and Colleges Admission Service (UCAS). Applicants for entry in September/October 2011 must apply for their chosen degree course by a deadline of December 2010, when most are in their final year of secondary school: they will be taking their end-of-school public examinations in June 2011 and receiving their results in August 2011. So they apply in conditions of substantial uncertainty about how their school qualifications will match up with the requirements of the courses to which they might apply. Similarly, course selectors make offer/reject decisions in conditions of uncertainty and base these decisions on: the school’s reference letter, the applicant’s personal statement, age-16 qualifications and the school’s predictions of the forthcoming age-18 qualifications. Applicants receiving offers have to commit to an accept/reject decision prior to taking their summer examinations. This is the current pre-qualification admission system and it is easy to see its potential inefficiencies compared to a Post-QA process based on full information.

We identify three potential problems with the current system. The first concerns inter-temporal inconsistencies within the admission round. At the start of the round, course selectors will make a forecast about the expected conditional offer which will match demand to the number of places available. As the flow of applications arrives, course selectors will issue conditional offers at this predicted ‘market-clearing’ level. But in a context of rising demand for places, if demand is higher than predicted, the ‘offer’ (in the sense of the qualification score required for admission) will be raised for later candidates or, in the limit, no further offers will be made. This gives us our first hypothesis: Later applicants will have a lower probability of receiving an offer of a university place.

The second potential problem concerns inconsistencies across applicants or type of applicant. There are a number of possible sources of this – including overt discrimination – but we consider the following cause. Consider the case of an applicant who is a ‘late developer’ with relatively poor age-16 qualifications and a predicted age-18 score which is a downwardly biased (predictions are made at about the half-way stage of the 2-year course leading to the age-18 qualification). Then this applicant is less likely to receive an offer of a place than is a candidate whose final age-18 qualification score will turn out the same but whose predicted score was unbiased. In other words, the pre-qualification system is loaded against the late developer. We also note evidence that learning trajectories are systematically related to characteristics such as family and school background, inter alia. It is worth emphasising that predicted grades are especially important for admission to more competitive courses: candidates are often rejected if their predicted grades are not strictly higher than the market-clearing ‘offer’ level. Our second hypothesis is that under the current system, there will be systematic differences across candidates according to characteristics such as family and school background, ethnicity and gender.

The third problem concerns the interaction of the two issues identified above. Suppose that there is some systematic bias (unwitting or otherwise) against a particular type of applicant. Then one might hypothesise that this bias could be the more acute the greater is the pressure of demand for places. If late applicants are at a disadvantage, as under our first hypothesis, then we might expect any apparent discrimination against some groups to be particularly strong in the case of later applicants.

In the empirical evidence we present in our paper, we use a rare but rich dataset on all undergraduate applicants to UK medical schools in 1996 and 1997 in order to examine whether the evidence on the probability of receiving an offer is consistent with the hypotheses set out above. Among other results, we find that after controlling for age-18 achievement, there is a significantly lower probability of receiving an offer for an applicant who (i) is from a non-white ethnic background, or (ii) is from a lower social class background, or (iii) attended a non-selective or a non-fee-paying secondary school. We also find (iv) that there is a significant and substantial reduction in the likelihood of receiving an offer the later is the application and (v) that the relatively disadvantaged groups suffer greater relative disadvantage the later is the application in the admissions round.

Following the recommendation of a Post-QA system by Schwartz in 2004 and a consultation process on ‘Improving the Higher Education Applications Process’, the Government in 2006 recommended a further review in 2010/11 to ‘faciliate the objective of moving towards Post-QA by 2012.’ From our analysis, we conclude that the current pre-qualification admission system is associated with outcomes which are both unfair and inefficient and that a Post-QA process in which decisions by all parties are made in the context of full information is likely to be a major improvement in both fairness and efficiency.