Next year British universities go under the microscope for the fifth Research Assessment Exercise (RAE), an attempt to help the government invest wisely in academic infrastructure. And even some of its most persistent critics acknowledge that the RAE, which has drawn international attention, has overcome a rocky start and is working increasingly well. "Each successive RAE is moving closer to the consensus [of what constitutes high-class research]," says Paul Cottrell, the assistant general secretary of the Association of University Teachers, which opposed the RAE's introduction in 1986.

The RAE began as an attempt to funnel dwindling resources into the best research programmes at a time when severe cuts in public spending evoked fears of a major brain drain. Unlike the recently introduced U.S. system of National Research Council (NRC) ratings, a university's RAE rankings have a direct impact on government funding. "The better you do [in the RAE], the more money you get," says John Rogers, RAE manager at the Higher Education Funding Council for England (HEFCE), which oversees the exercise.

The process, which covers 68 fields, puts every university department or programme under scrutiny by an independent panel of peers. Each department receives a score, from 1 to 5* that is based on four pieces of work submitted by every participating researcher, and such information as prizes, outside funding, and research plans. "The gold standard is always international excellence," says Rogers. That score, adjusted for the number of participating researchers in the department, determines funding levels.

Despite its narrower focus and deliberate elitism--last year about 75% of HEFCE's $1.3 billion budget went to just 26 of England's 128 Higher Education institutions--the RAE has raised some of the same concerns as the NRC in the U.S. "Teaching is not esteemed as highly as research and always gets a back seat," Cottrell argues. Although teaching skills are evaluated in a separate exercise, the Teaching Quality Assessment (TQA), the outcome is not linked directly to funding. The question of how much panels are affected by a researcher's reputation also remains an issue, although Rogers says attitudes will play a smaller role next year than in earlier exercises.

Rogers calls RAE "the longest standing assessment process on this scale worldwide," and its influence may soon be spreading beyond its borders. In Japan, where university funding has been based on precedent and enrolment and there is little oversight of programmes and performance, the government is moving slowly towards greater accountability. This year the Ministry of Education, Science, Sports and Culture (Monbusho) hopes to create an evaluation organisation that will serve initially as an accreditation board to review curricula and to prod universities to raise education standards. But an advisory panel recommended that Monbusho begin evaluating university research efforts on a departmental level, with the results somehow tied to funding for new buildings and large-scale equipment.

The RAE approach has also found a home in Eastern Europe, where this spring the Czech Republic hopes to begin a long-awaited review of academic research at its 27 universities. The reports from the visiting panels, which will include foreign scientists, are expected to lead to a two-tier system that favours a handful of elite universities.