When Bonuses are Contraindicated

Should companies pay out performance-related bonuses? It turns out that such payments might well be counterproductive:

The logic of bonus payments itself seems straight-forward: by paying a bonus on the condition of success, the successful outcome becomes more attractive to the agent and he puts in more effort to increase the likelihood of success. But do bonus payments necessarily increase effort and the likelihood of success?
Here, I examine this question and find that the answer is “no.”

The logic is quite clever. Let's say I get a fixed bonus if I manage to achieve a certain result, and that working a lot harder will make that result a bit more likely. Then I have two choices. The first choice is that I work a lot harder, in which case the probability of getting a successful result goes up – a bit. The cost of this choice is high: it involves a lot of hard work. The benefit is low: a relatively small increas in the likelihood of me receiving my bonus.

On the other hand, I can choose to work less. The cost of this choice is low: there's now a slighly smaller chance that I will receive my bonus. But the benefit is high: lots more leisure, and a substantial chance of receiving the bonus without actually doing any extra work for it, which is nice.

Chris Dillow applies this theory to the real world of CEOs:

I suspect this might be true for quite a few CEOs with stock options. They might figure:

It’s going to be really hard work to genuinely improve this business. And even if I do so, it might not raise the share price by much, because equity analysts are stupid buggers and the company’s share price depends upon stock market conditions. I’ll just talk a good game and hope the market rises - as it usually does over time - taking the company‘s share price up with it. If I must do something, I’ll increase the gearing of the company by buying back shares. That’s easy to do, and it’ll increase the share’s beta and hence its chances of the share price rising a lot if the market rises.