This chapter develops Le Grand’s argument about the need to recognize that those who choose to work in public services are not wholly ‘knights’ (Le Grand, 2003) driven by altruism, but are a mix of ‘knights’ (altruistic) and ‘knaves’ (selfish). He argued that choice and competition between schools and hospitals through quasi markets in which money follows the pupil or patient is the primary policy instrument to combat government failure. This is because quasi markets appeal to both knightly and knavish motives. They create pressure to improve services to respond to threats to market shares, providers’ incomes and hence jobs (Le Grand, 2007). The reason why we disagree with Le Grand in his proposed remedy to his diagnosis is that the choice and competition model assumes that people as users and providers of public services act as ‘econs’, as described by Thaler and Sunstein (2008), i.e. behave as in conventional microeconomics. We know that ‘humans’ behave differently, as has been demonstrated by carefully designed psychological experiments. Thus, although, if asked, people will say they desire more choice and believe that the greater the choice the better, there is ample evidence that in health care, just as in more mundane consumer choices (Iyengar and Lepper, 2000), users do not behave like econs and do not use information on provider performance to switch from those that are poor to those that are good. Instead of the choice and competition model, we see scope for developing the more psychologically plausible model of governance based on users and providers as ‘humans’, where users do not use public information on provider performance to switch between providers, but providers respond to threats to their reputation via public reporting of rankings of performance (Hibbard et al., 2003).

In this first section we present the four alternative models of governance: altruism, hierarchy and targets, reputation, and choice and competition.