Ministers still scratching their heads over how to release individual English universities from controls on student numbers while keeping a lid on the cost to the taxpayer have been urged to look at work by leading economists in the US.

Research by academics at the universities of Stanford and Harvard on the placing of doctors in Japanese hospitals could contain the solution to a conundrum that is perplexing policymakers.

The suggestion comes as David Willetts, the universities and science minister, reiterated in a speech last week that finding a solution to freeing up student places was the "big prize" that the government needed to drive competition in the sector.

However, apart from putting forward short-term proposals to allow a proportion of overall places to be contested through a "core and margin model", all he said was that he would "welcome a debate" on a more permanent settlement.

ÖйúAƬ

A radical solution could lie in the kind of research led in the US by Alvin Roth, George Gund professor of economics and business administration at Harvard University, an expert in how to design markets under public sector constraints.

He said the problem faced by the English university system was "somewhat similar" in character to attempts by the Japanese government to allocate enough junior doctors to hospitals in rural areas by using regional "caps" that hold down numbers in cities.

ÖйúAƬ

The issue was investigated by Fuhito Kojima, an assistant professor in economics at Stanford University, and Yuichiro Kamada, a graduate student at Harvard. They designed an algorithm to find the most efficient way to solve the problem.

Dr Kojima said their analysis could be applied to the English university system - given that institutions were in effect operating under one large "regional cap" on numbers - by designing an algorithm to "transfer" places between institutions according to their popularity among students.

"Often, the problem is that these kinds of adjustments go through bureaucratic processes that firstly can take a lot of time and secondly do not result in the correct or efficient outcome. Once you use an algorithm, this kind of adjustment can be automatic," he said.

Mr Willetts asked leading economists in the UK for advice on whether theoretical models could be applied to the student number dilemma, but it is not known if the research in the US has been considered.

ÖйúAƬ

But Ken Binmore, emeritus professor of economics at University College London, who helped design the ?22.5 billion auction of the 3G telecoms network in 2000, said that the government had to realise that any target-driven bureaucratic system would be inefficient.

"They have got to run it as a regulated market. The worst thing they could possibly do is to follow what the previous government did by using targets in areas like the NHS - targets just don't work," he said.

simon.baker@tsleducation.com.