End piece: Behavioural change and ‘nudge’

While this handbook has been dedicated to the psychology of economic behaviour, a number of the contributions have been about behaviour change from a broadly social science perspective, such as Jackson and Smith's work on sustainable lifestyles; encouraging sustainable energy use in the home (Perlaviciute et al.); and increasing tax compliance (Olsen et al.). Contributors are generally agreed that simple economic models based on incentives and disincentives have their limitations, in, for example, curbing tax evasion and encouraging environmentally friendly behaviour. Even worse, financial incentives may actually ‘crowd out’ desirable behaviour when driven by implicit rather than explicit motives. In the past, social psychologists believe (and some still do feel this way) that if you could decipher a person's attitude you could predict his or her behaviour. This was in an era where consistency, congruity, balance and consonance (dissonance) theories prevailed (e.g., Heider,1946; Festinger,1957). It was asserted that people preferred balanced states of mind over dissonant or discordant ones. The task of government then might be to change attitudes through advertising and information campaigns with the idea that changes in behaviour will follow. However, campaigns of this sort have not been especially successful, and within the academic literature the attitude/behaviour link has been heavily criticised from as long ago as 1969 (Wicker, 1969) as well as in more contemporary writings (e.g., Potter and Wetherell, 1987). A reaction to this, in various forms, was the theory of reasoned action and the theory of planned behaviour (Fishbein and Ajzen, 1975; Ajzen, 1991), which took into account subjective norms and behavioural intentions. The idea was that the approval or disapproval of others matters to us, and our behaviour is not only dependent on our attitudes but our intention to act in one particular way or another; for example, I have a negative attitude towards smoking, other people believe I should stop and I intend to stop. When all these ticks are in place, prediction of behaviour is indeed improved as literature in the area of health psychology has revealed (Conner and Norman, 2005). However, these models remain rather reductionist in nature.