Amartya Sen is a great economist and social philosopher whose willingness to recognise a central role for market institutions in securing economic development and individual freedom shows considerable commonality with the classical liberal tradition. Sen’s commitment to the values that underpin a free society is, however, equivocal and indeed often points towards what classical liberals would see as a dangerous form of paternalism. This tendency is particularly evident in the discussion of ‘adaptive preferences’ and their relationship to freedom.
Sen questions the traditional liberal view that we should accept that the choices people make in their lives are a reasonable indicator of the subjective value they attach to the choices concerned. The notion of adaptive preference emphasises how the values that people express can be conditioned by the social environment and more specifically by oppressive structures – such as, for example, those that emphasise asymmetric gender relationships within the household. According to Sen, we cannot always trust in the revealed preferences of individual agents as reflecting their ‘true’ best interests because faced with structures that narrow their range of options people may ‘adapt’ to their environment by ‘accepting’ their lot and lowering expectations of what life has to offer. * Just as ‘libertarian paternalists’ attribute the difference between the ‘revealed preferences’ of people for fatty foods or a low savings rate and their ‘real’ preferences to various cognitive biases, so Sen focuses on the role of ‘adaptation’ in accounting for the difference between the subjectively expressed beliefs of actors and their underlying ‘objective’ interests. The assumption in both of these cases is that these biases can be addressed via appropriate policy interventions either in the form of ’nudges’ a la Sunstein and Thaler or via public education in the case of Sen.
From a classical liberal standpoint there are two fundamental problems with this line of analysis. First, it is far from evident that there is anything ‘wrong’ with people adapting their preferences to their particular circumstances. One might adapt one’s desires to be an Olympic athlete, to make the England football team, or to be a champion boxer to the realities of one’s physical limitations. Most people would consider adaptation of this nature entirely rational and indeed essential to having any hope in finding satisfaction from life. Of course, it might be argued that there is a fundamental difference between the constraints set by physical limitations and ‘socially constructed’ constraints which might be thought more malleable. Yet, it is far from evident that this is so. Suppose that most people have a preference for greater social equality and support Rawls’ view that the ‘basic structure’ of society should operate to ‘maximise the position of the worst off’. Suppose also, however, that evidence indicates (as it does) that only societies that tolerate substantial inequalities in income and ownership do in fact raise the absolute condition of the worst- off. In these circumstances, adapting to social inequality would be the rational course to pursue. I do not believe for a minute that the recent London riots were a response to social inequality, but suppose for a moment that they were. One way in which the risk of future social disturbances could be reduced would be for potential rioters to learn to ‘adapt’ their preferences to the necessary inequality.
Second, one can accept the point that cognitive biases or ‘adaptations’ may prevent people from getting ‘what they really want’, but it is another matter to suggest that public policy – or at least policy which is compatible with liberal principles – should play a major role in dealing with these biases. In order for there to be any chance of success in this regard policy-makers need to be able to distinguish between the ‘real preferences’ or the ‘objective interests’ of people and those preferences that result from cognitive biases. In the case of libertarian paternalists they need to distinguish preferences for fatty food that are ‘genuine’ from those that are distorted by ‘weakness of will’. In the case of Sen’ s adaptive preferences, policy makers need to judge whether a woman’s endorsement of asymmetric gender roles reflects her ‘real’ beliefs or whether these beliefs are a reflection of an overly-constrained social environment. In addition, they would need to decide whether this adaptation is rational in the face of social circumstances which cannot realistically be changed, or whether it results from a more malleable set of conditions. There is little or no reason to suppose that policy makers are able to make such distinctions competently and yet in the absence of this competence the danger is that public policy will collapse into a more conventional form of paternalism – which claims to know what preferences people should have.
The suspicion is, of course, that ‘old fashioned paternalism’ is the value system that really lies behind this recent behavioural theory and Sen’s approach in particular. Sen’s commitment to interventionist methods to widen the horizons of women in developing countries reflects a view that traditional gender roles are almost by definition morally problematic and must therefore be ‘explained’ by the existence of oppressive structures which prevent the ‘politically correct’ outcome. I have considerable sympathy with Sen’s distaste for asymmetric family relations, but I do not believe they are a matter for public policy. Better to rely on trade, economic growth and the cross-cultural contact this brings as the best route to unintentionally expose women to the wider world and what it has to offer. To argue for deliberate intervention to transform what are deemed ‘adaptive’ preferences is deeply paternalistic and at root, fundamentally illiberal.
*On this, see, for example, Sen, A. (2002) Rationality and Freedom, Cambridge, Mass: Belknap.