Incentives for Bankers – The Bonus Outrage

In 2008 the world ended. Ok, not quite but it sure felt like it was about to. The financial crisis of 2008 left many angry (angry may be an understatement). People had trusted the banks with their money and the banks hadn’t acted in their best interest. Following government bailouts and a helping hand from the taxpayer, the banks landed on their feet (sort of). But what was it that left the public feeling so outraged? Mainly it was that the people responsible for the crisis were to be rewarded, with 6 and 7 figure bonuses.

Why? What was the point of these bonuses? We know incentives can have a powerful effect on performance. So, these bonuses should help the CEO’s and people with the real say in banks perform to the top of their ability, but that’s not always the case. Rational people, and bankers, would argue that, the bigger the incentive, the better the performance. Yerkes and Dodson discredited this belief more than a century ago though, when stress is involved. In short the relationship between incentives and performance is positive when the incentives used are small, but as the incentive increases it causes stress, which in turn reduces the performance. If you have taken the time to read the Yerkes & Dodson paper, you will now realise that the experiment used involves rats and electric shocks, and you might be wondering what the connection is here with bankers (No, I’m not suggesting putting bankers in a maze and electrocuting them, although this could be a popular TV show).

Yerkes and Dodson, 1908

Dan Ariely suggests in his book The Upside of Irrationality that these results from Yerkes and Dodson can be applied to humans. After his initial thought he decided to design an experiment to test this. In a novel experiment to test raising incentives with performance Ariely and his team comprised a experiment in which participants would be paid in relation to their performance on 6 games. The games all involved skill and thoughtful thinking. The payments were rather large. In the small incentive condition participants could earn as much as 1 days pay, split across 6 games. The payments were split between 2 standards the participants would have to reach. For achieving Very Good on a game, they would receive 1/6^th of the total payment, for achieving a Good rating they were awarded 1/12^th. In the moderate and large conditions participants could earn up to average of 2 weeks pay and 5 months pay respectively. These certainly were large incentives. Before you start to think Ariely must have a never-ending budget, the experiment was outsourced to India, where an average persons monthly expenditure was roughly 500 Rupees ($11).  This made sure the incentives being used were definitely enough to motivate participants, therefore surely resulting in increased performance.

With so much at stake however, how would participants deal with the added stress? Five months pay is a huge amount of money to rest on your ability to play a few games. The results in the graph below speak for themselves. Participants averaged similar earnings in the Small and Medium incentive conditions, but when more money was on the line, earnings fell, even though Very Good performance on just a few games would earn more than the other two conditions. So what happened here?

Ariely showed decreased earnings when incentives were raised. The resulting stress from higher incentives prompted worse performance on the games.

Ariely suggests the reasons for the small and medium conditions to be equal are because both are relatively large incentives, which would motivate greater performance.  When using an extraordinarily large incentive however, participants were over-motivated and could not calm themselves. Hands were shaking, and their subsequent performance on the games was affected. The conclusion here is that high incentives don’t necessarily work for cognitive tasks*, therefore what is the justification in offering bankers such high incentives if it is likely to decrease their ability to do their job?

He Shoots, He Scores!

A counter argument to Ariely’s findings would be that some people perform better under pressure; this is why they have got the job in the first place and why they will argue the bonuses would work for them. Ariely was interested to know whether this is the case. In basketball there are players called Clutch Players, these are the highly paid players who are believed to perform best in the last 5 minutes of a game. The findings looked good initially. Non-clutch players scored roughly the same amount of points regardless of the time in the game, while the clutch players scored notably more during the last 5 minutes of the game. At face value it would seem that during high stress moments their performance increased, on further investigation however, it was found that players shooting accuracy did not improve. The only thing that increased was the player’s tendency to shoot. Their performance/ability didn’t increase, but their shooting behaviour did.

So these highly skilled, professional basketball players don’t improve with bigger incentives, there was most likely no decrease in performance due to the physical rather than cognitive skills involved with basketball. Therefore we must ask ourselves what is the point in these huge bonuses when Ariely, like Yerkes and Dodson, show there is a U-turn relationship between incentives and performance. One theory as to why bankers are given such high bonuses is that it acts as an incentive for employee’s lower down the ladder. They will work harder in order to get promoted, and richer. This could work, and would be a good direction to take this research in. Another reason why these incentives may be failing is because of Adaptation, the bonuses bankers receive are expected and therefore provide less motivation to perform to higher standards. The first time employees received their big bonus, there may have been an increase in their performance and this has deteriorate over time. It would be interesting to further explore this area, and I would love to hear you opinions on the matter.

*Incentives used during cognitive tasks work in a different way compared to incentives for simple physical tasks. If you paid somebody 1pence for every time they jumped in 1 Minute they wouldn’t jump as many times as if you paid them £1 for every jump.