For love or money

I have been an adjunct and visiting professor at a number of institutions that have attempted to increase staff research-paper output by using monetary rewards. They are nearly all second- or even third-division places aspiring to improve their status and ratings. The question is: does this strategy work?

The theory is simple. It's pure economics. Money motivates: pay people to publish in good journals and they do so. Monetary rewards are the best; money is, in the jargon, a universal reinforcer. Greed, pride and envy will all work to get academics eagerly and enthusiastically publishing in the best journals. This will, in turn, mean that the status of the institution will rise, which will enable it to charge higher fees. Is it a good strategy, then? Recoverable investment, motivated staff, the progress of knowledge and all that?

The practice goes like this. Some body - perhaps government or a university-led committee - classifies the academic journals in every discipline. Perhaps into groups: top, middle, low (and, of course, "don't count"). The higher the classification, the more you get paid if you publish in them. Yes, it's as simple and straightforward and vulgar as that.

But, of course, there is the issue of multiple authors. So perhaps the first author is paid more than the second, and so on. There may also be the issue of the number of pages of the article, because some journals do brief notes or brief reports.

And so the formula begins to get a little complicated. How much is Dr X paid if he wrote five papers last year, the first where he was second of five authors on a long paper in a second-level (B-class) journal, and the second where he was fifth author on a brief note in a top journal, etc.? Someone, no doubt from the publication reward department in central administration, does this calculation. Or it could be done on a national level where some established and venerated professors (nearly emeritus) have a few jolly consultancy days agreeing on the classification.

About halfway through the following year, Dr X gets a cheque, or perhaps the money is deposited in his research account, which has all sorts of restrictions on how it is used. That, of course, is a big issue - constraints on how the money may be spent.

There are other variations on this theme. The money may be shared by the department or the university as a whole, which means it goes to the head of department and the relevant dean. This can really make the senior people sit up and take notice if the reward involves "interesting" sums of money. And so there is another pressure on the researcher.

This strategy could certainly affect recruitment and selection. Should you hire publishing stars and let them off teaching? Or allow them to bring their spouses along as additional professors because they "pay" for themselves? In such an environment, teaching is soon seen as a "research-failure" punishment. Those who can, do; those who can't, teach; those who can't teach, teach teachers; and those who can't teach teachers become educational consultants.

What if the payment for publications is a government-backed national policy, and thus the university, department and individual get a cut? The question is then about percentages. Of course the successful publisher is more likely to be promoted with higher pay, so the reward may be fairly direct. Or not.

Would it backfire if the researcher were offered some paltry or derisory sum for having a paper in a top journal? At what point is the money seen as an insult rather than a welcome reward? Does this depend on the individual (age, stage, values), or the discipline? Indeed, what if there are disciplinary differences? What if there are (by decree, from the most high) only three A-listed journals in philosophy, but 15 in physics? What if the impact factor of the A-listed journals in media studies is in the "doesn't count" range for medicine?

And what if, horror of horrors, we find huge sex and age differences in the monies "won" by researchers over the years? Old, pale males once again having privileges? That really won't do.

Money as a motivator

Let's put aside the theory questions on who decides on journal ranking, the criteria they use and how frequently this is revisited. There are even more pressing issues. The first is the power of money as a motivator. The second is the effect of introducing such a system. The most obvious of all extrinsic rewards is money. But is there a simple relationship between reward, productivity and satisfaction? The idea is that better-paid people are more productive and happy. Or that if you pay for productivity, you get more productivity proportionately and happier staff. Simple, causal ... but naive and essentially evidence-free.

Indeed, there are at least four reasons why business psychologists see money as much more likely to be a cause of dissatisfaction than satisfaction.

The first reason relates to the simple idea that there is a clear correspondence between pay and performance. This is not, nor has it ever been, true. Perceived and actual low pay can and does lead to considerable dissatisfaction and demotivation, but not vice versa.

The effects of a pay rise soon wear off as people adapt to their new conditions. Any improvements are therefore likely to be temporary. Money can be an effective motivator, but you need a great deal of it to stop adaptation effects - too much, in fact, for most organisations to bear.

Second, what leads to pay satisfaction is not so much absolute salary, but comparative salary. So if my salary goes up dramatically, but so does that of my comparison group, there is no change in my behaviour and relative reward. This is crucial, and relates to the whole problem of performance-related pay. No matter what people are paid, if they believe, with or without evidence, that they are not equitably and fairly paid, they become demotivated. If a don believes it is easier to publish in A-class journals in another sub-discipline and this explains a colleague's smugness (and nice house), you couldn't imagine an angrier, more demotivated person.

Third, money is not everything. Many people would be happy with more time off or more job security, or less teaching, than more money. People are prepared to trade off things for money once they have enough or grow weary of the game that is not worth the candle.

Finally, there is the eternal implication of tax and spend. It is all very well to increase pay, but if increased taxes eat heavily into it, where is the benefit? Why earn more when the government takes so much?

Playing the system

Incentives and punishments do, however, influence behaviour. What if you are - dare one admit it - an ambitious, extrinsically motivated academic? Not the passionate, intrinsically motivated traditional academic, but the person who sees various rather interesting perks in the ivory towers. Media attention, in other words, and quite good dosh. But how to get that dosh?

Easy: you publish only in A-grade journals. But that is easier said than done. Often the rejection rate is more than 90 per cent. Better, then, to publish one B-class and three C-class articles than a single A-class. And because the whole system is now monetarised, you could actually do these calculations. Cost-benefit analysis: plot effort, rejection rate and so on. Perhaps you can even try to influence who the reviewers will be. Yes, I know, this happens already - but here the goal is different.

There are other factors that come into play. Top journals are often very conservative. New ideas never start in them, so it is possible that this reward system will help to decrease creativity and innovation.

And what if one's very special academic interest (read: "unpopular area") means that there are no A- or even B-grade journals in the area? Should one jettison years of labour in an obscure vineyard and get on to a fashionable topic? It is no secret that there are fashions in research - topic, theory, methodology, even style. Do we not look back with embarrassment not only on the clothes and hairstyles we adopted but also on the dusty, tatty papers that were the fashion victims of another age?

Furthermore, as fashions change, journals rise and fall in impact. Indeed, their impact factor can change radically from year to year. Is it worth putting effort into having journals recategorised? If so, how often - and by whom, with what criteria?

And will all this affect the review process? Will it increase the envy and bitterness of lowly, less-successful reviewers, leading to the greater rejection of their papers? We all know how unreliable the process is in the first place. It is pretty demotivating to tell somebody to change their focus, to acquire the current fetishism of a specific journal (edited in another country by a cartel wedded to some particular approach), and then to tell them that the rejection rate is 90 per cent. And so helplessness and anger are increased.

Solutions

There are essentially two problems here: in most jobs, productivity cannot be easily (reliably, fairly, comprehensively) measured; and there are few universal reinforcers that suit everybody.

Academics have to make judgements on pay increases and promotion. My clever colleagues who study decision-making say it is not that difficult to come up with a quite sophisticated and subtle approach to making these decisions. Not just a simple regression based on papers (quality of outlet, impact, novelty, page count) but other relevant factors.

But it can get pretty complicated. Is it an averaging or an additive model? Are there hurdles or compensatory considerations? Can the result be overridden by someone if they don't like the result? Who, then, and why? And if these judgements are possible, why try all this pseudoscience in the first place? Surely it would help in the decision-making, even if it at least provided some rank ordering and data for reasoned argument.

The second problem lies in the reward. We psychologists know there are inter- and intra-individual differences. Age and stage count; there are cash-rich and time-poor academics, and vice versa. For some academics, £5,000 could make a very big difference to their lives; for others, such a sum would mean very little.

What if you could choose your reward for publishing? Less teaching or marking, perhaps? More money, a bigger lab? Now you're talking! But another problem looms - how do we work this one out? Perhaps we could say that every lecture is "work X", and 10X = £5,000, which is the reward for publishing in three A-grade journals last year. A tad mechanistic, but motivationally a lot better than simply offering money.

Caveat

So is this article inspired by sour grapes, the bitter whining of a non-research-active, under-published don? Not quite. Last year I published three books and had papers appear in more than 30 peer-reviewed journals. Had I changed my allegiance on the paper from my current university, which does not have this system, to one that does and where I am adjunct professor, I could have made somewhere between £10,000 and £20,000. Cheque payable to me alone!

My chats with people in departments that offer monetary rewards for research-paper output indicate that they are little affected by it and are, overall, unenthusiastic. One does not do research for money. If you want lots of money, get out of the academy. The prize of publishing in the top journals is in itself enough. It brings peer recognition, invitations to conferences, job offers and so on. It also affects one's "h-index ranking", the index that attempts to measure both an academic's productivity and impact.

And surely there's nothing wrong with a little monetary reward - a case of champagne, perhaps? But much more and the consequences are not positive. Intrinsic motivation turns extrinsic. Peer envy, jealousy and back-stabbing increase. Research gets skewed to particular areas, topics, journals and research methods.

Performance-related pay is the application of the equity principle over the equality principle. Its drawback always lies in measurement of performance. It's a good idea to try it. It sharpens the mind. It's not a bad idea to let people know the criteria. But a single criterion applied uncritically can do much more harm than good. For evidence, look at the disenchantment with all those previous government targets.