Photo: Hindustan Times
Photo: Hindustan Times

Why incentives fail

Incentives do matter, but how people respond depends as much upon how incentives are designed as on the context

One of the central tenets of undergraduate economics is that “incentives matter". In one of the most famous graduation speeches ever delivered by an economist, Nobel Prize winner Thomas Sargent listed 12 important lessons from economics. Among them, incentives featured prominently. Sargent’s advice to graduating students was simple: “Everyone responds to incentives, including people you want to help."

Incentives do matter in the real world, but how people respond depends as much upon how incentives are designed as on the context in which they are used—institutions, history, geography and culture matter a great deal in determining whether a particular incentive will work or fail.

In June last year, Ahmedabad’s municipal corporation decided to incentivize toilet usage. The corporation planned to pay every user Re1. Such monetary incentives are doomed to fail because they do not recognize that “it pays too little", and also, sadly, that defecating or urinating in the open has much larger “welfare" gains than a rupee won by using a public toilet.

Undergraduate economics teaches us that monetary incentives are usually desirable, and they improve welfare or productivity. However, there are many situations in which reaching for the wallet could backfire. Richard Titmuss, in his 1970 book The Gift Relationship: From Human Blood to Social Policy, famously posited that offering money for blood donation may put some people off and can reduce the number of donors.

In a 2008 paper, economists Carl Mellström of Gothenburg University and Magnus Johannesson of the Stockholm School of Economics ran an experiment to test the Titmuss hypothesis. They offered two sets of incentives—one with a payment of $7 and another with an option of giving that $7 to charity. Blood supply dropped when only the monetary incentive was offered. Doing good without being recompensed, it seems, is more important for many donors.

Martin Sandel, a professor of philosophy at Harvard, wrote a book-length harangue against economic incentives. In What Money Can’t Buy: The Moral Limits of Markets, Sandel argued that offering incentives for certain actions—motivating kids to perform well in tests, monetary fines as substitute for ban—can have moral implications.

“It is not about inequality and fairness but about the corrosive tendency of markets," wrote Sandel. “Putting a price on the good things in life can corrupt them. That’s because markets don’t only allocate goods; they also express and promote certain attitudes toward the goods being exchanged. Paying kids to read books might make them read but also teach them to regard reading as a chore rather than a source of intrinsic satisfaction. Auctioning seats in the freshman class to the highest bidders might raise revenue but also erode the integrity of the college and the value of its diploma."

In an ambitious randomized experiment that spanned four years, Harvard economist Roland J. Fryer offered monetary incentives to school kids in four US cities, including New York and Washington DC. The incentives, it turned out, were ineffective in improving the learning outcomes, and were only moderately successful in increasing class participation and improving conduct in the class. Using incentives in education favours Sandel’s argument that paying kids to study may not produce any desirable result.

Do incentives matter? In his latest book Economics Rules: The Rights and Wrongs of the Dismal Science, economist Dani Rodrik suggests that the answer to this question depends on what is at stake and the kind of incentive that is on offer.

In a recent research paper published in the Quarterly Journal of Economics, behavioural economists Uri Gneezy of the University of California, San Diego, and Aldo Rustichini of University of Minnesota report the results of their field trials, which show that incentives may not always be linearly related to performance.

The duo engaged high school students to collect donations for a charity in a door-to-door fund-raising campaign and found that those who did not receive any compensation fared better than those who received a small compensation. As soon as the payment became large enough, efforts made towards collection did increase.

In another of their widely cited experiments, Gneezy and Rustichini proposed a fine of $3 for late-coming parents in a daycare centre in Israel. Once the fine was enforced, the late-coming incidents actually increased. Since the fine imposed was a small amount, parents considered the fine as a fee to pick their kids late. The “fee" made them morally ambiguous about picking up their kids late.

Viewing monetary incentives as a panacea is an old problem in economics and one that was highlighted by economic genius Albert Hirschman in his 1985 essay Against Parsimony.

“The reason is probably that they (economists) think of citizens as consumers with unchanging or arbitrarily changing tastes in matters of civic as well as commodity oriented behaviour," wrote Hirschman. “This view tends to neglect the possibility that people are capable of changing their values. A principal purpose of publicly proclaimed laws and regulations is to stigmatise antisocial behaviour and thereby to influence citizens’ values and behaviour codes. This educational, value-moulding function of the law is as important as its deterrent and repressive functions."

Today, several economists have come round to the view that altruistic motives can actually shape public policy as well. In a recent interview, World Bank economist Kaushik Basu argued that better pay alone is not going to solve the problem of horrible law enforcement in India.

“It has something to do with values, which will take some time, but we will have to try to understand this and convey this," said Basu. “Human beings are perfectly capable of carrying out their tasks if they take pride in their work even if there are no rewards or penalties.

“I feel, for instance, that much more than shoring up incomes of the police, shoring up their pride in the task that they do is important so that they feel good and honourable when they walk up to break up a fight, instead of going there in a predatory manner. It sounds a bit preachy, but right from the top, if you teach people to take honour and pride in the work that they are doing, they do it much better.

“Fortunately, in India, you can see this in the military. I do not have an easy solution, but thinking entirely in economistic terms, giving more incentives or money alone won’t solve this problem."

A recent experiment on police reforms in Rajasthan by Massachusetts Institute of Technology economists Esther Duflo and Abhijit Banerjee shows that non-monetary incentives like freezing transfers, and better training actually improved police performance. Whether these reform measures can be scaled up nationally is another question.

Research shows that an optimal mix of monetary and non-financial incentives can be effective in designing public programmes. Non-financial rewards can help whenever paying cash doesn’t work. Identifying which one of these works remains a challenge, though. The truth lies somewhere between the more mainstream banal Sargentism of “incentives matter" and extreme Sandelian cynicism that all economic incentives are immoral and must be avoided.

Economics Express runs weekly, and features interesting reads from the world of economics and finance.

Comments are welcome at feedback@livemint.com

Close