Alison Wolf

It is something of a truism, in higher education c 2006, that "unto those who have, it shall be given". League tables seem to show the same universities at the top, year after year, collecting not just the plaudits but also the research money. That makes it easier to give young staff breathing space for scholarship and publication, which further enhances the university's strengths, and so on, in a virtuous (or vicious) circle.

Or does it? I have certainly assumed that being at a "top" institution is good for young staff, not simply for their CVs but as a source of stimulation and effective mentoring and because greater resources boost their research productivity. Yet a new study suggests just the opposite. Being at an elite university does not make academics more productive; in fact, in some high-profile institutions it seems to have a marked negative effect on their production of high-impact research papers.

The work, published by the National Bureau of Economic Research, involved an extremely detailed study of 3,262 members of economics and finance faculties working in more than 800 universities between the 1970s and today. The authors - E. H. Kim, A. Morse and L. Zingales - gave their monograph an intentionally provocative title: Are Elite Universities Losing Their Competitive Edge?

Before disbelief cuts in, let me be clear about the findings. No one is claiming that academics at the likes of Harvard, Stanford and Berkeley universities are less productive on average than those at, for example, large non-selective state universities or small rural colleges. On the contrary.

However, when you look at the productivity of individual academics who move between the top 25 universities and other institutions, the data show that they tend to be either no more, or significantly less, productive when at the elite than the non-elite. And what makes this especially interesting is that this phenomenon is new. In the 1970s and 1980s, being at a top university generally had a positive impact on your productivity. Only in the 1990s did this kick into reverse, with almost no top universities registering a positive effect, and about a third (including Harvard, Columbia and the Massachusetts Institute of Technology) showing a significant negative impact.

Some apparently plausible explanations either are not borne out by the data or fail to explain why this has happened so recently. Young academics who get a job at a top university might relax, take a holiday and hence produce less. But why wouldn't that happen in the 1970s? Or perhaps top universities have taken to hiring more established stars; and as everyone knows (correctly, this time), established academics produce fewer research papers than those starting their careers. But the analysis controls fully for the seniority or rank of the academics in the sample, so that cannot be the answer.

Instead, the authors point to the IT revolution, and I think they are right. Computers and the web may not have revolutionised teaching, but they have certainly transformed other academic activity. They have had a dramatic, well-documented effect on the number of co-authored academic papers and on the numbers with authors from different institutions. Kim, Morse and Zingales show that there has also been a big increase in the number of co-authored papers involving elite and non-elite institutions.

Young scholars are publishing high-impact papers without having to be at a top institution when they do so. Conversely, when and if their publications win them "better" jobs, the move does not affect their performance by transforming their potential collaborators.

So will status hierarchies become more fluid as the top universities'

built-in advantages lessen? Possibly, although rather less than the authors suggest. Their study looked only at the US, which I think is not important, and only at economics and finance, which is. Disciplines that focus on the statistical analysis of large data sets collected by other people are going to find long-distance collaboration a lot easier than, for example, laboratory subjects.

The data also suggest a significant tendency for successful young researchers to move back to where they did their PhDs. This begs the question of how partnerships between elite and non-elite are formed. If productive authors employed outside the old charmed circles work almost exclusively with contacts they made as students within them, then the pulling power of the top departments will hardly be slackened.J Nonetheless, we can expect that non-elite universities will, in the future, be returning more "top" papers to the research assessment exercise and young researchers willing to consider a wider range of employers. That has to be good for quality across the sector.JJJ Alison Wolf is Sir Roy Griffiths professor of public sector management at King's College London.