The principal-agent problem

I was listening to Evan Davis’s business talk show The Bottom Line on R4 yesterday evening.  I almost always learn something from  this and Peter Day’s In Business.

Last night’s Bottom Line taught me something about the principal-agent problem, which I had been thinking about but  didn’t realise I had been thinking about.

The principal-agent problem, recognised in the fields of economics, management  and political science, runs as follows. The principal (in this example a business) has a set of standards which must be met if it is to operate profitably; the agent (its staff, a third-party seller) must act in ways that protect and promote these interests. To what extent do these interests converge? Can they be made to converge? Or must other inevitably costly and time-consuming procedures be put in place (layers of management, monitoring) to ensure that the agent is acting in the principal’s interests?

In conventional economic and management terms, the principal-agent problem is typically addressed by incentivising good performance by the agent (bonuses or commissions, for example) and punishing poor performance (short-term contracts, swift and easy dismissal, loss of earnings).

The reason that this principal-agent problem resonated so strongly with me I think is because recently I have been thinking about the problems inherent in selling anything that is complex or specialised and in which the seller has far more information than the buyer but often struggles to convey that information. For example, a customer — let’s call her Jane — wants to commission a website designer, or use the services of a lawyer, or buy a care package for an elderly relative. Someone out there has the service Jane wants, and Jane knows that some providers will better meet her needs than others — if only she could differentiate the good from the indifferent.

But — in the current jargon — the available information is asymmetrical: the provider talks in jargon and specialist, technical terms; while Jane knows what she wants but can’t describe it in language the provider recognises.

Well, when you think about it, this too is part of the principal-agent problem. In  fact, it’s a very good example: the agent (salespeople, for example) is not protecting the principal’s interests (the business) because it is not doing a good, customer-friendly job of selling.

Perhaps it’s time for us all to think again about what the principal-agent problem means for social enterprises. After all, social enterprises are trying to align the interests of principals, agents and customers. Indeed, social enterprise models that use employee- or co-operative-ownership are quite explicitly seeking to develop employee-owners who are both principals and agents. Looked at in this light, the government’s current interest in the ‘John Lewis model’ in public services becomes only too readily explainable: do away with layers of costly management and invigilation of ‘agents’ by ensuring that principals and agents are to a large degree one and the same.

This principal-agent problem opens up some interesting ways of thinking about what we do in social enterprises, especially in terms of customer service and marketing. For example, do all of your staff have the best information (and the right attitudes, aptitudes and training) to get across the selling messages of the business? Are they equipped to act as principals-and-agents? Are they able to act in the interests of the business, its employees and its customers? Do they see that that is what they are doing?

The principal-agents problem: a problem which we sometimes forget social enterprise models were devised at least in part to address…


Leave a Reply