Thursday, 13 September 2012

Anchoring and pay

I am very pleased. I've just managed to demonstrate the well-established cognitive bias known as 'anchoring' using a sample of well-educated people who work in corporate governance in one way or another (policy, asset management etc).

We gave all audience members a set of questions, one of which was to estimate the average fees paid to a supervisory board member in Germany. Before asking them for their estimate we exposed them to a figure which we explicitly told them was false. One group got a very high figure, the other got a low figure.

The result, as you might expect, was that those exposed to the very high (but irrelevant) figure gave higher estimates than those exposed to the low figure. The size of difference was dramatic - the average estimate with the high prompt was more than six times larger than the estimate with the low prompt.

I'm going to run this on a bigger sample next time. Would be good to use a sample of corp gov people who regularly analyse and/or engage on pay.

No comments: