Wednesday, 5 August 2009

A puzzle about incentives

Interesting puzzle from Chris Dillow today: is it possible to genuinely align the incentives of employees and investors?

This in response to Lord Myners' debate about whether long-term shareholders should have more voting rights than short-term traders. I've touched on this question before but the question of whether shareholders have a real understanding of the company's interests is an excellent one.

I have a particular interest in this question because part of my business is designing structured pricing approaches, which are intended to align the interests of supplier and client in a commercial relationship.

Structured pricing generally rewards the supplier for achieving business objectives of the client. But while it's a big improvement over traditional contracts, it is still not perfect in theory.

For instance, if you hire a telemarketing agency the traditional way to pay them is a fee for each day of time. They'll spend a day on the phone and charge £350. If they set up five sales meetings, lucky you; if they set up none, you still have to pay the same price. In this structure there's little formal economic incentive for them to do a good job (not to dismiss the supplier's motivation from pride and their desire to win future business - but this is a distinct issue). In fact, their incentive is to win as few meetings as they can get away with, so that you keep paying them for more days.

To correct this, you may want to pay them not for each day of calls made, but a percentage of the sales you make. This incentivises them to find good leads, set up meetings that are well qualified, and keep working for you as long as they can keep getting meetings.

This type of contract isn't perfect either, however, for two reasons. It incentivises the supplier to set up more meetings than the optimal number, because they do not bear the full cost of the meeting (you have to travel to the meeting, spend time selling and following up). To control this you might subtract some of the sales cost before working out the revenue share.

On the other hand, the supplier's reward depends on factors outside their control: your sales ability, the quality of your products, your negotiating skills. You might construct a formula to take some of this into account, but it's very hard to invent a structure which creates the perfect incentives and is easy to monitor in practice.

I suspect that designing such structures for managing a company would be equally difficult.

However, if you are willing to be radical, I did read one suggestion (I think in one of Tim Harford's or Stephen Landsburg's books) to perfectly align employee and company incentives. The proposal - in the example, I believe the company was General Motors, ironically - is that every one of the company's 50,000 employees receives the entire profits of the firm - let's say $1 billion - as their salary. Yes, that's a total salary bill of $50 trillion.

Of course, this would bankrupt the company and would be completely unreasonable, so as a quid pro quo, the employee has to buy their job by making an advance payment to the company. How much? Just under $1 billion, as it happens. This provides every employee with exactly the right incentives to maximise company profits and not to take excessive risks.

You don't need to be a behavioural economist to think this scheme is impractical. But it illustrates the trouble with finding a theoretically robust way to solve this agency problem. The best we are likely to achieve is a system that provides some scaled-down version of the correct incentives - profit sharing, say, in recognition of the material interest of the employees in the success of the company - and relies on a bit of humanity for the rest. Cultural factors and employees' personal connection to their firm, colleagues and customers will go a long way to aligning the interests of all stakeholders.

No comments: