Performance-related pay isn’t performing as well as it should.
It’s a simple idea – you get paid more when you perform well. Surely it can’t fail to work? Well it arguably contributed to the banking crisis (although that may have been more because pay was more closely related to short-term rather than long-term performance). And a recent paper suggests that higher levels of performance related pay may have caused lower stock returns.
An interesting paper on this topic “Corporate Governance Going Astray“, discusses some potential behavioural effects that are likely to limit the effectiveness of performance related pay. A few days ago I posted a discussion about incentives in the workplace and described how the relationship between work and reward really isn’t very straightforward: being paid to do something makes the task less enjoyable and reduces creativity. So it’s not illogical to think that linking wages even more closely with performance could be counter-productive.
However, even if these “behavioural” side-effects aren’t an issue, performance-related pay can fail due to information problems. How do you measure performance? How do we separate the effect of effort from luck? The paper notes that executives tend to be paid for good luck but not punished for bad luck – when things are going well it’s assumed that it’s down to high performing executives. But when things aren’t going well, it’s put down to bad luck.
Perhaps there’s a link here between the problems with separating effort and luck and the high proportion of men in executive positions. Some research certainly suggests that men have a greater tendency to attribute failure to bad luck…