The toughest trick

For performance pay to work, it must reflect directly the goals we want to achieve - and there lies the rub, says David Marsden

In the past two decades, Britain's universities have become much more like businesses than they were in the past. It is hard to imagine David Lodge's University of Rummidge amid the current wave of mergers and acquisitions, rationalisation, restructuring and redundancies in our universities. Less spectacularly, university managements now look much more systematically at the income streams from different kinds of activity, and seek to identify new "products" and new "customers". So if universities are now run much more like businesses, why not, as has been mooted, adopt incentive pay schemes that are widely used in the private sector?

Our research into performance-related pay in the public services at the Centre for Economic Performance at the London School of Economics found that many employees agree with the principle of pay for performance. Although its financial incentive has been modest, when administered well, performance pay can help staff and their line managers better identify work objectives and link these with their organisation's goals.

Like many other public servants, university academics enjoy a good deal of discretion in their work, and the job of management (whether hierarchical or collegial) in such organisations is to ensure that it is used to the benefit of the organisation and wider society, and not just enjoyed as "on-the-job leisure". However, the same research also found that there was widespread disenchantment among public servants with the operation of their performance-pay schemes. This led the government's Makinson report to conclude that such schemes were "ineffective and discredited" because of the way they were operated. The record of public-service performance pay to date has been like the curate's egg: good in parts.

Performance pay would be only one of several other performance incentives already in operation. Promotion and competitive recruitment are powerful incentives for research publications, as is the quest for scientific prestige and consultancy work. The "up-or-out" system for the increasingly large number of fixed-term contract research staff provides similar, if less rewarding, drives. Assessments of university research and teaching provide collective incentives for institutions and generate peer pressures on colleagues to do their bit. Thus, universities are not lacking in incentives to perform at their core activities. Leaving aside the vexed question of pay levels, do university teachers need any additional incentives? To answer this, we need to know what they are expected to achieve.

In 1975, Steven Kerr warned in the Academy of Management Journal about the "folly of rewarding A and hoping for B". Are there activities that the university wishes to encourage, and which would be supported by financial incentives? The biggest difficulty with any performance-pay scheme lies in deciding what "B" is, and persuading an organisation's members that it is worth pursuing. That done, one can worry about how to determine whether "B" has been achieved and what kind of reward should be applied.

According to the CEP's research, the major cause of the disenchantment with existing public-service schemes lay in the quality and top-down nature of the goal-setting and performance-appraisal process. Top management can criticise inexperienced line managers for being nervous or overbearing when discussing individual employees' performance with them. But if the organisation's priorities are not clear, or there is no consensus as to what they are, then line managers are out on a limb.

In these circumstances, it is tempting for line managers to reward those who are cooperative but who may be seen by their colleagues as the "blue-eyed" girls or boys. If middle managers are unclear about the performance objectives and what is being rewarded, fear of being penalised for being lax will lead them to ration performance payments even if everyone does well. Ordinary staff believe that there is a quota on good appraisals, making them doubt that they will be rewarded even if they perform well. Many also suspect that management restricts performance ratings to save money, and if the goals are imposed by top management, lower down they often appear irrelevant and cause resentment. Such lack of transparency and of employee involvement makes performance-pay schemes appear arbitrary and undermines their role as performance incentives and as aids to goal-setting.

The increased diversity of the higher education sector means that a "one size fits all" approach to performance pay is unlikely to work. Different universities have different objectives and will wish to prioritise different kinds of performance. A sector-wide scheme will almost inevitably end up rewarding "A", which would most likely be "more" of the same, whatever that is. However, as Britain's universities seek to identify distinctive niches within an increasingly global market, their needs are likely to become more diverse. Both Adam Smith and Emile Durkheim agree that increased specialisation is the natural and beneficial outcome of increased competition.

Consider an example. To achieve such diversity in higher education depends on individual universities and their staff having the autonomy to experiment with new initiatives. Some of these will lead to new niches, although many will fail. Some may be akin to developing new lines of "business" that generate extra income. Others might seek to broaden the social vocation of universities, by widening access for example.

What they all have in common is that individual institutions and their staff need the autonomy to experiment. Such initiatives often go beyond the capacities of particular individuals and involve team working. These teams commonly extend across disciplines and involve not just academics but all those providing infrastructure services in the libraries, in information technology and in administration.

If this is the kind of performance that our society and our government want from its universities, then performance pay, if we are to have it, should reflect these goals. In other words, performance-pay schemes should be devised by individual institutions and not imposed nationally, and they should accommodate teams of staff, which could be academic departments, project teams or the whole institution, depending on the goals chosen. Our research included two National Health Service trust hospitals - one operating individual performance pay, one operating a trust-wide bonus. The former appeared to do better on incentives, the latter did better on cooperation and work relations.

At present, the dominant pay incentives are overwhelmingly individualistic and discipline-based. If one were to add more individual performance incentives to these, as has been the dominant pattern across the public services under the influence of the Treasury, it is very likely that more "A" activities will be the result and "B" will not be forthcoming.

David Marsden is professor of industrial relations at the London School of Economics.