Don't bank on bonuses

If you want to boost people's performance, don't bank on bonuses

BONUS culture has come under intense scrutiny since the ongoing financial crisis began in 2007. Many people have been outraged by the way some bankers and top executives seem to have been rewarded for failure. Others find the idea of multimillion-dollar bonuses morally abhorrent. Even US President Barack Obama has gone as far as to call large bonuses "obscene".

But few have asked whether performance-related bonuses really do boost performance. The answer seems so obvious that even to ask the question can appear absurd. Indeed, despite all the fuss about them, financial incentives continue to be introduced in more and more areas, from healthcare and public services to teaching and academia.

"Economists and workplace consultants regard it as almost unquestioned dogma that people are motivated by rewards, so they don't feel the need to test this," says Alfie Kohn, a teacher turned writer. "It has the status more of religious truth than scientific hypothesis."

So it may come as a shock to many to learn that a large and growing body of evidence suggests that in many circumstances, paying for results can actually make people perform badly, and that the more you pay, the worse they perform.

No one is disputing that bonuses can help companies and institutions attract and retain the best staff. Nor does anyone deny that you can encourage people to do specific tasks by linking payments to those tasks. Rather, the issue is about how to get the best out of people. Do employees really perform better if you promise to pay them more for getting results?

There are some obvious reasons why such payments can backfire. It has been argued, for instance, that cash bonuses contributed to the financial crash, because traders had little motivation to ensure their companies' long-term survival (New Scientist, 19 March, p 30).

Most bonus schemes are poorly designed, says Malcolm Higgs of the School of Management at the University of Southampton in the UK. He thinks the reason is that organisations try to keep schemes simple. Nevertheless, he thinks bonus schemes can work as long as they align the interests of individual employees with the long-term goals of a business.

Bonuses can also encourage cheating. "Once you start making people's rewards dependent on outcomes rather than behaviours, the evidence is people will take the shortest route to those outcomes," says psychologist Edward Deci of the University of Rochester in New York state.

Deci blames stock-option bonuses for the collapse of Enron and other corporate scandals. "In many cases the top executives simply lied and cheated to make the stock price go up so they got huge bonuses."

But the work of Deci and others suggests the problem with bonuses runs far deeper than poor scheme design or cheating. In 1971, he asked students to solve puzzles, with some receiving cash prizes for doing well and others getting nothing. Deci found those offered cash were less likely to keep working on puzzles after they had done enough to get paid.

Two years later, a team led by Mark Lepper of Stanford University, California, asked children aged between 3 and 5 years old to draw with felt-tip pens. Some were told they would receive a special ribbon as a prize for doing so, and duly received it. These children were less likely to choose to draw with felt-tip pens when they were later given a free choice of activities. No such effect was seen with children who were not offered a reward, whether they subsequently received an unexpected one or not (Journal of Personality and Social Psychology, vol 28, p 129).

These studies suggest that offering rewards can stop people doing things for the sheer joy of it, an idea known as the overjustification effect. This was the basis for a series of books by Kohn in which he argues that rewarding children, students and workers with grades, incentives and other "bribes" leads to inferior work in the long run.

Those who believe in the power of bonuses fail to distinguish between intrinsic and extrinsic motivation - wanting to do something because you like it in its own right versus doing something because you want the reward, Kohn says. "It's not just that these two are different, it's that they are often inversely related. The more you reward people for doing something, the more their intrinsic motivation tends to decline."

A "do this and get that" approach might improve performance in the short term, but over longer periods it will always fail, Kohn says, as it turns play into work and work into drudgery. Bonus recipients inevitably play safe, become less creative, collaborate less and feel less valued, he adds.

The existence of the overjustification effect has been disputed. However, a 1999 meta-analysis by Deci and colleagues of 128 studies strongly suggests it is real (Psychological Bulletin, vol 125, p 627). "The facts are absolutely clear," says Deci. "There is no question that in virtually all circumstances in which people are doing things in order to get rewards, extrinsic tangible rewards undermine intrinsic motivation."

What's more, the studies suggest that the greater surveillance, evaluation and competition that tend to accompany performance-related rewards further undermine intrinsic motivation, and that offering rewards can also stop people taking responsibility.

These findings suggest that in the kind of jobs many people do as much for love as for money - from healthcare to science journalism - any incentives specifically tied to performance, as opposed to a normal salary or unanticipated bonuses, may backfire. However, this clearly doesn't apply to someone who gains no satisfaction from their job and does it only for the money. "If you are doing a boring, stupid task, rewards cannot undermine intrinsic motivation that you don't have," Deci says.

In such cases, surely, the more you pay the more you get? Not when tasks require brain rather than brawn, according to work by Dan Ariely of Duke University in Durham, North Carolina. To allow him to pay relatively huge sums, Ariely recruited people living in villages in India. The volunteers played six games testing creativity, memory and motor skills, and were offered 4, 40 or 400 rupees for achieving high scores in each of them. The maximum possible reward was equivalent to the amount an average person in rural India spends in five months.

Backfiring bonuses

Surprisingly, the big incentives did not produce big improvements in performance. Quite the opposite, in fact. Those offered the biggest incentives earned just 20 per cent of the maximum possible on average, compared with around 36 per cent for those in the moderate and low-incentive groups.

In a follow-up experiment, 24 students in the US were offered cash rewards if they performed well enough in a mathematical task and a repetitive key-pressing task. Some students were offered from $15 to $30, others from $150 to $300. In the key-pressing task, the higher reward did indeed lead to a better performance: on average, the volunteers earned 78 per cent of the maximum possible, compared with just 40 per cent for those offered lower incentives.

In the mathematics task, however, offering bigger rewards backfired. Participants earned an average of 63 per cent of their maximum potential earnings when offered the lower incentives, but only 43 per cent when offered the higher incentives. This suggests that while bigger bonuses boost the performance of people doing simple manual labour, for other kinds of work employers might actually get less when they offer large sums (Review of Economic Studies, vol 76, p 451).

Why? Many explanations have been proposed. One suggestion is that big rewards make people try too hard. "Everyone understands that money can have a motivating effect, but it can also affect stress levels," says Ariely.

For simple physical tasks this doesn't matter, but for complex mental tasks it clearly does. Work by Sian Beilock, a psychologist at the University of Chicago who studies how pressure affects performance, suggests stress interferes with our working memory.

"Our working memory is a kind of mental scratchpad that allows us to work with the information we have in our consciousness," Beilock says. "However, when people are doing demanding tasks in a stressful situation, we see this system malfunction. They allocate attention to other things, they worry about the consequences of, say, not getting the bonus."

Another reason may simply be that people spend too much time thinking about the rewards instead of the task in hand. "Bankers don't admit to me that they get stressed, but they do admit that in October, November and December they spend a substantial part of every day thinking about their bonuses and recalculating them on a spreadsheet," says Ariely. "If 80 per cent of your earnings came in the form of a bonus, I expect you would do the same."

Then again, people will work far longer hours to get big bonuses. For such reasons, some argue that experiments like Ariely's do not apply to the real world. However, a few researchers have now looked at the effects of performance-related pay in the real world.

In 2006, Laura Petersen of the Baylor College of Medicine, Texas, reviewed 17 studies of healthcare pay-for-performance schemes. Some of these studies did point to positive effects, with one in particular suggesting such schemes could boost childhood vaccination rates. However, there were also some unintended consequences. For example, in one case researchers found that offering incentives to try to improve the treatment of people with substance-abuse problems made doctors more reluctant to take on the most serious cases (Annals of Internal Medicine, vol 145, p 265).

Last year, Thomas Gavagan, also at Baylor, published a study comparing rates of childhood immunisation, smear test take-up and mammography in 11 primary care clinics in and around Houston. Six of the clinics offered doctors up to $4000 in extra payments if they reached specified targets, whereas five continued paying a normal wage. There were no differences in performance, Gavagan found (The Journal of the American Board of Family Medicine, vol 23, p 622).

The largest and most comprehensive study of the effect of bonuses on healthcare comes from the UK. In 2004, the National Health Service introduced a scheme through which family doctors could earn up to a quarter extra on top of their basic salary by hitting some or all of 136 targets.

Brian Serumaga of the University of Nottingham and colleagues looked at the effect on 470,000 hypertension patients diagnosed in the four years before and three years after the introduction of the scheme. They found the financial incentives had no effect on the proportion of patients who had their blood pressure under control, who were being monitored and treated, and who suffered adverse outcomes such as heart attacks, stroke and renal failure. Neither did they have any impact on how patients' actual blood pressure changed over time (BMJ, DOI: 10.1136/bmj.d108).

Full of testosterone

"Having spent three years looking at the evidence of payment-for-performance, I am astonished at how weak the evidence is," Serumaga says.

"An assumption is being made that doctors are substantially influenced by their income, or even primarily attracted to healthcare because of the financial rewards. But if you look at the literature on what actually motivates doctors to become doctors, very little of it is related to financial reward."

Politicians and planners of healthcare reforms appear to be unaware of these findings. For instance, President Obama's healthcare reforms, signed into law a year ago, include the introduction of the first two federal healthcare treatment incentive payment schemes in the US. "For too long quality and payment have been separate in our system," Jean Moody-Williams, director of the Centers for Medicare & Medicaid Services quality improvement group, told New Scientist. "Our goal is to align them for the purpose of value." But she admitted that she was unaware of Serumaga's study, and requested a copy.

"President Obama has said he wants his government to be based on science and on what is effective," says one of Serumaga's co-authors, Stephen Soumerai of Harvard Medical School, Boston. "A lot of economists believe incentives are the whole story, however the evidence tells us they are wrong."

Are these findings relevant to other sectors, such as banking? "Doctors and those in the other great professions have a completely different work ethic because they have a vocation," says David Buik of inter-dealer broker BGC Partners, who has worked in the City of London for 44 years.

"Bankers and brokers are full of testosterone, full of adrenalin and fiercely competitive. Without the financial incentives there would be no need to be competitive. Psychologists may have done plenty of research on this but they're making generalisations that don't necessarily apply in the City," he says.

In the absence of any rigorous trials, many will no doubt agree with Buik. Perhaps they are right. Recently, however, bankers and economists have turned out to be wrong about an awful lot of things.

Nic Fleming is a freelance journalist based in London