Monday, August 9, 2010

The problem with incentives

Today's solution can be the basis of tomorrow’s problem. 

The Australian Prime Minister has proposed substantial incentive payments to up to 1000 schools which show improvements in student attendance and results (more...). It is generally a good thing to recognise and reward achievements and the intended outcomes of the proposal are highly desirable.

However this is not simply a system-wide recognition and reward arrangement. Offering incentives with a limited number of possible recipients turns it into a competition.  Competition can certainly promote improved performance, as in sport.

At the same time, there can be unintended side effects. In any competition there will be winners and losers. Incentives work best for those who are likely to be rewarded and they tend to become less effective over time unless the rewards are increased. Very few of us commit ourselves to winning Olympic medals despite the ever increasing value of doing so.

To be competitive we need to minimise the factors that reduce our likelihood of success while maximising the factors that increase the likelihood of our success.  Other than luck, our likelihood of success is associated with factors over which we have some control.

The Prime Minister has promised that national criteria will be developed through consultative processes and measurement of performance will be overseen by an independent body. Fairness is important in any incentive scheme. But is this actually possible?

Student attendance and results are closely associated with many factors well outside the school's control including the natural abilities of individual students; family well-being; levels of parental education; and the families’ access to social capital, community resources and so on. But will the selection processes be able to measure the school’s and teacher’s contributions separately from these non-school factors?

If not, it would make sense for schools and teachers to generally avoid students who have additional needs for support and whose families are in distress; whose parents have low levels of education; and who live in communities with limited services, facilities and social capital.

Conversely it would make sense for schools to attract and retain students who have minimal need for additional individual support; whose families are in great shape; whose parents have high levels education and  easy access to material and social resources and services.

The unintended side effect is that the most successful school response to the Prime Minister's proposal could well be to avoid the very students the proposal is intended to assist.