Monday, November 22, 2010

The "slippery slope" with financial incentives

Econ 101 teaches us that incentives matter and that rational economic man responds predictably to incentives, especially financial ones. However, recent research from behavioural psychology shows wide variations, often contrary to standard model predictions, in the responses of real world human beings to financial incentives. This has naturally generated an interesting debate about the pros and cons of using financial incentives to instill habits.

Health care has been at the forefront of efforts to use incentives to generate socially desirable health outcomes in individuals. Studies show that 50-70% of the health care costs in the US are preventable. Accordingly, realising the financial benefits associated with better preventive care, health insurers have used financial incentives to encourage people to stay healthy - exercise more and eat healthier foods. They have found that even modest financial incentives up-front to encourage people to remain healthy can save the much larger costs of hospitalization (and insurance payouts).

Health service providers and even employers (who pay the health care premiums) incentivize physical activity, which is measured using web-enabled pedometers, accelerometers or heart-rate monitors. Personal activity data is uploaded from the activity-tracking devices into an individual’s account on the Web. Health insurance premium payments drop by some specified amounts if the individual attains his/her activity goals. This trend has even led to the emergence of health-incentive management companies.

Similarly, it has been found that one-third to one-half of all patients in the US do not take medication as prescribed, and such lapses fuel more than $100 billion dollars in health costs annually because those patients often get sicker. In response, insurers, doctors and pharma companies have been toying with paying people money to take medicine or to comply with prescribed treatment.

Behavioural economists like Uri Gneezy have studied the use of financial incentives to promote habit formation in different walks of life. Gneezy and Gary Charness have found that "even though personal incentives to exercise are already in place, it appears that the financial incentive serves as a catalyst to get some people past the threshold of actually getting started with an exercise regimen".

More controversially, similar line of reasoning have been invoked to get children to study or even behave properly, people to exhibit civic responsibility (like to pick up litter or not litter), and so on.

There is a slippery slope with taking the application of incentives to its extremes, especially when it involves getting people to do what is obviously in their own interest and what they ought to be doing in the normal course. Here is a list of examples of such slippery slope outcomes.

1. It has been found that such incentives may create self-doubts about the true motive for which good deeds are performed. This in turn could crowd-out, even destroy, the inherent pro-social attitudes in human beings. Children paid to study are amongst those most vulnerable to such incentive distortions.

2. Incentives could result in market failures or bad market outcomes. For example, paying for organ donations could end up creating a black market in organ transfers.

3. Financial incentives could create hysteresis effects, especially for pro-social behaviours. This could manifest in two forms. One, pro-social attitudes would not be regained once the incentives are withdrawn. Second, it will take much greater effort to generate the same level of outcomes after incentives are removed. Therefore children exposed to financial incentives to study or help in household chores, could end up not responding once those incentives are withdrawn.

4. Financial incentives can dilute pro-social motivations. Accordingly, payment for volunteering under certain conditions end up reducing effort - people do less work than would have been the case without the incentives.

5. The use of financial incentives opens up possibilities of distortions or sub-optimal outcomes if the magnitude of incentive is either too small or too large. For example, failure to strike the right balance on inncentives to reduce health risks can result in worse performance than not paying at all.

On a similar note, financial penalties on deviations from social norms can take the stigma out of non-conformity. Therefore penalties for turning up late for classes will end up increasing late-coming as people rationalize away their late arrivals on the penalty payment.

Update 1 (12/12/2010)

A neuroeconomics study of human brain activity shows that each brain responds differently to incentives, and reward-related brain activity can predict the undermining effect (the person is less likely to voluntarily engage in a task after performing that task for some sort of extrinsic reward) within an individual. This means that removing extrinsic incentives to engage in an activity can have damaging effects on the desire to voluntarily engage in that activity.

This is particularly interesting because it shows that not all individuals should be treated as equal in economic models of decision-making and incentive-driven. behaviour.


Jayan said...

The title says it all! It is even more slippery in Indian situations.

Instead of paying money to expected behavior like driving carefully, taking medicine correctly etc... we should put the money into better schools and teachers. We can guide the learning generation to a better way to live.

Social engineering works when people influence others with actions and thoughts. Using money for these kind of things is futile and bit scary too.

KP said...

Dear Gulzar,

A particular case :

incentivizing self-interested behaviour is dangerous, since this has the added problem of - assessing and certifying - which is part of the rewards mechanism.

An example of this could be ( a hypothetical one) to reward non-smoking - this is a situation where the bureaucracy that assesses-certifies-rewards the non-smoker can make a killing (literally) - as the consequence of non-conformance is not theirs (moral hazard).

In cases where policies (government incentives) reward self-interested behaviour - (in most cases not as directly obvious as my example may suggest, and I could suggest other seemingly complex situations) - the bureaucracy stands to gain a fantastic rent seeking opportunity, with little downside.


kutoos said...

I'am not sure that I agree with the examples. Students go to school not from the love of studying and they are compliant (they study) because of a negative incentive, the possibility of being punished. They figure out (at least some of them do) that studying may not be such a bad thing after all & maybe much later. Being paid to study will be a case of one incentive replacing another.

Payment for volunteering may actually encourage those who have no interest in volunteering but other really interested volunteers will continue to volunteer.

To believe that financial incentives will distort dilute pro-social motivations is to ignore the diversity of motivations that drive people.

But I think, in general to try to inculcate morals in children through financial-incentives would be disastrous.

And on a lighter note, in my time at least, I don't believe there was any stigma attached to late-coming.

gulzar said...

yes jayan and KP, the slippery-slope operates at many levels. the assessment-certification point is an excellent example, and is true in most cases of such incentive based interventions and its evaluation.

my point, kutoos, is that once the incentive is removed, there is a strong possibility that the effort levels will drop. more so since the student has been socialized into studying to access the incentive.