Principal-agent problems among Sicilian firefighters

In the first week of ECON110, we talk about unintended consequences. One of the examples I ask students to consider in class is what would happen if firefighters were paid for each fire they start, rather than bulk funded. Inevitably, someone always suggests that the firefighters might start fires themselves, in order to increase their funding. I always point out that seems unlikely, and that firefighters would more likely simply spend less effort on prevention activities. It turns out though, maybe those students were right. The Guardian reported back in August:

Fifteen volunteer firefighters have been arrested in Sicily on suspicion of starting wildfires and reporting non-existent blazes so they could earn €10 (£9) an hour for putting them out.

Police in Ragusa province, in the south of the Mediterranean island, said the fire department became suspicious when it emerged that the auxiliary brigade had responded to 120 incidents compared with just 40 tackled by other volunteer teams over the same period...

Most of the team were under investigation for fraud, with several also suspected of arson, Ansa said. The island is plagued by fires in summer and auxiliary firefighters are paid €10 an hour by the state to help extinguish them.

This is also a good example of a principal-agent problem (a type of moral hazard). In this case, the fire department is the principal, and it engages the volunteers (its' agents) to fight fires. However, the fire department probably doesn't want more fires, so the incentives of the fire department (less fires) and the volunteers (more fires, so that they get paid more) are in conflict. The agents will take advantage of the fact that they are not being closely monitored by the principal, in this case by lighting fires themselves.

There are four main ways to deal with moral hazard problems (including principal-agent problems):

1. Better monitoring of the agent (just keeping an eye on the volunteers to make sure they aren't up to no good, I guess - this appears to be what has happened in this case)
2. Efficiency wages (paying a wage that is higher than the equilibrium wage, so that if the agent is up to no good, then have an incentive to behave themselves or else they lose the higher-paying job and have to work somewhere else for a much lower wage)
3. Performance-based pay (paying based on some metric of output - probably this is where things went wrong in this case, since the volunteers were paid for each fire they fought!)
4. Delayed payment (holding back some part of the agent's wages until it is clear whether they did a good job - this works for contractors, but it might not be appropriate in this case).

Performance-based pay is not always the right solution to every moral hazard problem. In this case, it might well have been the cause of the problem in the first place!

[HT: Marginal Revolution]