Designing the right incentives is an art that must be practised with care

We saw a few weeks ago how the introduction of compulsory vaccination in Germany, for example, would have led to a reduction in the number of citizens willing to be vaccinated instead of an increase. In this case, compulsory vaccination would have deprived many of the citizens willing to be vaccinated of the opportunity to signal to others and even to themselves their high level of civic-mindedness. It would have 'displaced' their intrinsic motivations. If you force me to do something that I think is right to do, you take away the satisfaction of doing it just because it is right to do it.

But the careless use of incentives, rewards and punishments can also act on another front. Staying on the subject of vaccines, during the pandemic many thought of using rewards and punishments to induce people to get vaccinated. In Italy we used fines, which are now being cancelled as if nothing had happened, putting a patch that is worse than the hole. In the US, premiums have been used extensively. In Ohio, for example, anyone who vaccinated acquired the right to participate in a lottery that offered a million dollars as a prize. In New Jersey, one could win, in the same way, a dinner with Governor Phil Murphy and his wife. In California, tickets to Lakers games. In Connecticut, those inoculated received a voucher for a free drink to be used at participating establishments, while in New York, tickets were given to visit the aquarium, botanical garden and other city attractions.

Which of these prizes proved most effective? It is not really possible to say, because if it is true that they all favoured vaccination, we do not know what would have happened if they had not been introduced. We can, however, say that certainly the million dollars is the most attractive prize, while legitimate doubts may arise regarding effectiveness. When dinner with the governor, tickets to basketball games, or a free drink are put up as prizes, various signals are sent out: it is said that vaccinating is right, it is a gesture appreciated by political authorities and fellow citizens, something that facilitates the resumption of normal business and even recreational activities such as sports. When you put up a million dollars to be vaccinated, however, you are sending a different kind of signal. One is 'saying' that getting vaccinated is risky, a risk so high that it must even be compensated with a million dollars.

Let's try to reason by analogy: who among us would be willing to participate in a medical research project in which we are offered to be injected with an experimental drug in exchange for 50 euros? Many? A few? It depends. But if the proposed amount was 5,000 euros, it is likely that even those who would be willing to participate for 50 euros would begin to doubt the safety of the drug and become suspicious of the goodness of the research. What is it that they are injecting me with, one might ask, if they have to offer so much money to convince me?

Here, the relationship between the size of the incentive and the probability of the behaviour to be incentivised is not always direct and positive. Because the incentive does not only act on the relationship between the costs and benefits of a certain choice, but also tells us something about the very nature of that choice. It is when the two effects - the 'price' effect and the psychological 'signalling' effect - are at odds with each other that trouble ensues. It is as if we are paying someone to entice them to do something and simultaneously telling them that that something is not worth doing.

There are other cases in which poorly designed incentives can generate the opposite result to the one we wanted to achieve. Uri Gneezy, in his Mixed Signals. How Incentives Really Work (Yale University Press, 2023), identifies four emblematic categories. The first category refers to the case where we incentivise quantity at the expense of quality. For 130 years, the so-called 'little green train' has connected the inland locations of Sardinia with the coast. The route that runs north from Cagliari towards the heart of the island is unusually curved - we talked about this in a Mind the Economy a few years ago regarding the problem of 'moral hazard'.

We quoted the passage with which in 1921 the English writer D.H. Lawrence, author among others of Lady Chatterley's Lover, describes the experience in his book Sea and Sardinia. 'It was wonderful to run in the bright morning towards the heart of Sardinia. We were always going up. The line made great curves so that when we looked out of the window, more than once we were startled to see a small train running ahead of us, in a divergent direction, amid great puffs of steam (...) It is a strange railway,' the writer continues, 'I would like to know who built it. The railway was designed by a compatriot of Lawrence's and built by an English company too. A company for whom curves were a better alternative to tunnels and bridges that might have made the route less tortuous. Why better? Simply because the English company in charge of building the line was paid per 'kilometre built' instead of per 'kilometre needed'. It is therefore not surprising that the curved option too often prevailed over the straight alternative.

Quantity instead of quality. As in the case of caesarean sections. Many more are performed than are really necessary. This is particularly true in those systems where hospitals are paid per performance, as in the USA. In an interesting article published a few years ago, Erin Johnson and Marit Rehavi demonstrated this effect, noting, however, that the number of caesarean sections prescribed is significantly reduced when the expectant mother has a medical degree. When the mother-to-be is also a doctor, then colleagues tend to avoid prescribing the operation unless it is strictly necessary. The two economists estimate that if all patients were treated like medical school graduates there would be savings of about two billion dollars and health outcomes would also improve ("Physicians Treating Physicians: Information and Incentives in Childbirth". in the American Economic Journal: Economic Policy 8, pp. 115-141, 2016).

A second category of errors includes those that arise when one would like to incentivise innovation but at the same time penalises failures. Think of a university, for example, that would like to persuade its researchers to engage in highly innovative projects but then chooses to promote with career advancement only those scholars who have many publications. This is saying one thing but doing the exact opposite. In fact, it is unlikely that a young researcher will embark on an ambitious and complex research project that will only yield publishable results after a few years, because after a few years that researcher may have seen his contract not renewed. In order to have many publications, therefore, young researchers will be induced to choose to explore safe routes, already established ideas to which they can only add a marginal contribution. Under these conditions, engaging in innovative but at the same time risky projects certainly does not pay off.

Just as it does not pay to promote long-term goals while only incentivising short-term ones - and we are at thethird category of errors. Politics often appears inconclusive and incapable of bringing about real change for this very reason. Short-termism' (short-termism) as some call it is that attitude according to which the only useful policy measures are those that show their benefits by the time of the next election. Long-term measures and reforms, those that the country would need, are constantly postponed because their effects would not have an impact on the likelihood of being re-elected.

The same problem applies to those managers who, instead of focusing on what would really be good for the company they manage and its shareholders, tend to prioritise share value growth in the very short term. After all, their performance evaluation is quarterly and the average tenure of a manager in a listed company, as the Global CEO Turnover Index shows, does not exceed ten years. Finally, there is afourth category where those cases fall in which the mismatch between objectives and incentives stems from the desire to favour teamwork while rewarding only individual results.

Here again, the university is a good example of what should not be done. Researchers in the same department are often encouraged to collaborate with each other, but then are in fact pushed into competition by the incentives involved. Internal promotions, in fact, are few and far between and are not awarded to those who have collaborated more and better, but only to those who have published more. Therefore, in practice, beyond slogans, it does not pay to collaborate with colleagues in one's own department.

The design of an effective incentive scheme, whether it is included in public regulations or in a company remuneration package, must always take into account not only the 'price' effect that alters the relationship between costs and benefits deriving from individual choices, but also the psychological effect that is generated by the signals that incentives convey and that always tell us something about ourselves and the nature of the task we are supposed to perform. It is true, therefore, that people respond to incentives but it is also true that this response is usually much more complex than one might expect.