August 5, 2013 3:43 pm
By Adam Grant
The evidence suggests it is more effective
to dole out rewards in advance
Most companies pay employees for high
performance. But according to some behavioural economists, they are doing it
all wrong.
In a study for the National Bureau of
Economic Research, researchers investigated whether financial incentives to
teachers in publicly funded schools could boost student achievement. The
economists, led by Harvard University’s Roland Fryer, randomly assigned
teachers to receive extra pay based on their students’ performance in
standardised maths tests. The reward to a typical teacher was worth 8 per cent
of salary, but student marks failed to increase significantly after the reward
was introduced.
The intuitive explanation is that
teachers are not motivated by incentives. As Daniel Pink argues convincingly in
his book Drive, if teachers already intrinsically care about student learning
and achievement, pay for performance is not the best approach.
But another group of students did spike
in achievement after teachers were given an incentive. Prof Fryer’s team had
varied the timing of the rewards being paid. At the start of the school year in
September, one group of teachers learnt they would be rewarded for student
performance in June. These teachers achieved few gains with their students.
The other group of teachers received a
cheque worth 8 per cent of the average salary – the same amount as the others –
but they got it upfront. If their students did not meet their targets, the
teachers would have to give the money back in June. Their students improved
significantly, to the tune of nearly 10 percentiles.
Why was it more effective to dole out
the rewards to teachers in advance? The researchers point to loss-aversion: the
prospect of a loss yields greater pain than the comparative pleasure of a gain.
When the bonus is promised for June, it is an abstract possibility. But when it
arrives on the day the promise is made, it becomes a meaningful sum here and
now that is worth working at to avoid losing.
However, I believe that there is a
social side to the story. Doling out a bonus in advance communicates trust and
high expectations, which are known to encourage greater effort and better
results. When teachers receive extra pay for student performance in advance, it
is clear that leaders have confidence in them, and the teachers may be more
likely to rise to the expectations.
What would happen if all organisations
doled out bonuses like this? At first glance, it does not seem very friendly to
hand out a reward then take it back. But the data show that, under such
circumstances, employees are more likely to earn the reward. Ultimately, this
means better performance and more bonuses – both desirable outcomes.
Another concern is unethical behaviour:
employees might cut corners or falsify results. In schools, this could involve
teaching only to the test, or even changing students’ answers. Interestingly,
Prof Fryer’s team found little evidence of cheating: the students whose
teachers received bonuses in September also improved on tests that were not
part of the deal.
Of course, in some cases, unethical
behaviour will be a risk if employees are more motivated to keep a bonus than
to earn one. There is a way around it, though, which was tested by John List
and Tanjim Hossain in a Chinese factory in another NBER study. Instead of
giving the bonus early, the company announced having provisionally allocated
productivity bonuses for the coming week. If employees fell short of targets,
the allocation would be reduced. Simply framing the bonus as allocated in
advance led to 1 per cent higher average productivity over a six-month period.
In a world in which quick results are
paramount, it is understandable that managers still cling to financial
incentives. After all, money matters, and it is easier to quantify, transfer
and exchange than most rewards. It can also be fair: meritocracy requires
rewards for the value contributed by employees.
To assess the merit of incentive schemes,
it makes sense to rely on evidence, not intuition. This evidence is new, and we
still have much to learn about how and where advance rewards are effective, and
if they work for rewards other than pay.
But next time you plan to pay employees
after they have achieved a target level of performance, it is worth asking if
it would be more effective to give the bonus upfront.
adam@giveandtake.com
The writer is a professor at Wharton
business school and the author of ‘Give and Take’
No comments:
Post a Comment