George Akerlof and Rachel Kranton have gained considerable attention for a series of articles (Reference Akerlof and Kranton2000, Reference Akerlof and Kranton2002, Reference Akerlof and Kranton2005, Reference Akerlof and Kranton2008) which make the concept of identity central to the standard theory of individual behaviour and choice. This book reviews their arguments in a compact, non-technical manner and proclaims ‘identity economics’ a new approach and important advance in economics. Identity economics, as they define it, investigates how social norms associated with expected behaviour in social groups influence individual behaviour and choice. For example, their opening discussion of sexual discrimination at Price Waterhouse tells how people acted in a discriminatory way toward Ann Hopkins (the plaintiff in an important U.S. Supreme Court civil rights case in 1988), because she failed to abide by social norms regarding women. Akerlof and Kranton regard identity economics as an extension of recent development in economics (following Gary Becker) that ‘no longer confines itself to questions about consumption and income: economists today also consider a variety of noneconomic motives’ (p. 4). But they also say that identity economics goes further in showing that social context matters in the following way: since ‘people have a notion of who they are’, that is, a self-image or what their identities are, social group norms influence their ‘beliefs about how they and others are supposed to behave’ (Ibid.). Thus in their view identity economics brings greater realism into economics and allows economists to create better and more useful models.

These improved models, to be sure, are traditional utility function models that now incorporate social group norms that prescribe ideal or exemplary behaviour for individuals associated with being members of certain social groups. For every social group or social category with which a person can be identified, there corresponds an argument in their utility function that motivates choice just as do those arguments associated with ordinary ‘garden-variety’ or ‘individualistic’ tastes. Individuals still seek to maximize utility, but the scope of the tradeoffs they face is now extended to include the effects of their acting in conformity with or contrary to the social norms of the groups with which they identify. That is, individuals gain or lose ‘identity utility’ when their actions conform or depart from social group norms. Further, they may also gain or lose ‘identity utility’ when other individuals’ actions conform or depart from recognized social norms – what Akerlof and Kranton regard as a new type of externality. This can be quite complicated, as in the example of Ann Price's colleagues at Price Waterhouse who acted toward her in a discriminatory way because they lost ‘identity utility’ when she failed to conform to their norms of behaviour for women.

Akerlof and Kranton only outline this identity framework in part 1, and do not duplicate or further explain the formal analysis they produced in their previous papers. Parts 2 and 3 constitute the main body of the book which is devoted to applying this identity framework to four subjects: organizations, education, gender in the labour market and in the home, and race and poverty; part 4 is more methodological, and comments on what makes for ‘good theory’ along lines essentially following Milton Friedman.

It needs to be emphasized from the beginning, then, that Identity Economics is not a scholarly monograph written for a specialist audience but rather a mass-market book presumably meant in its pricing, level of exposition, popular writing style and potential monetary payoff for its authors and its publisher – like Freakonomics and Akerlof and Robert Shiller's Animal Spirits (also priced at $24.95 by Princeton University Press) – to take advantage not only of the new taste for economics writing in the reading public but to also function as a supplementary text in economics teaching. The two main characteristics of this new popular economics genre seem to be that economics should now be thought to be ‘fun’ and also that it tells us something about the ‘real world’ – two matters about which economists have become somewhat defensive in light of the widespread perception on the part of students and many in the public that economics is neither (see Vromen Reference Vromen2009 on the making economics ‘fun’ phenomenon).

Of course it is hardly clear that it ought to be an aim to make economics fun, much less what this might involve. But one thing it appears to imply for Akerlof and Kranton is that economics ought to be made simple and easy to understand. Thus they have pitched their book at a very elementary level that assumes little more knowledge of economics from the reader than that economics has something to do with supply and demand (p. 14). For example, the chapter that introduces the authors’ identity framework begins in an almost patronizing way as if this were a student's first introduction to economics:

It's as if the voice the authors decided to adopt is one that says, ‘now listen carefully kids (but don't worry, we'll go slowly!) . . .’ Aside, then, from the fact that this book was not written for scholars, this level of exposition has odd effects, since the topic of identity and economics is indeed anything but simple. This means that quite subtle issues, which have been at the centre of much recent debate among economists, psychologists and philosophers, such as the relationship between other-regardingness and the traditional view of taste as independent of social context, are either not addressed in the book, or are rather quickly dispensed with in very broad brush stroke theory, and with bold assertions that social context matters and that economists are now coming to appreciate this (10–11).

But of course it is still fair to ask the authors difficult questions, such as: if social context matters in the case of social identity, and our choices are driven by social norms, why doesn't it also matter in the case of all the rest of our more ordinary ‘garden-variety’ tastes, since these could also be norm-driven? If this were allowed, however, social context wouldn't then just matter; it could be altogether determinative of choice and behaviour, so that an ‘individual’ might have many sources and kinds of social identity (not just including group identifications), but evidently not a personal identity that would somehow single them out and hold them together as an independent being. Akerlof and Kranton of course reject this view, since it would undermine the standard view to which they are committed that preferences are in some degree exogenous or independent in the sense of deriving from or belonging to the individual (however they happened to come about). This is indeed what allows them to talk about there being trade-offs between our identity commitments and all the other things we pursue. But this nonetheless leaves them with the position that on one level our choices are endogenous in that they are driven by social norms but on another level they are not. Alarms should be going off here about what the grounds are for this distinction, but that they don't see a problem is no doubt due to the fact that they have assumed from the outset that a person is an independent being in virtue of having an individual utility function, albeit one now ‘enhanced’ or ‘augmented’ by the need to observe certain social norms. That is, what they assume, though to be clear they do not actually say this, is that an individual's personal identity is their utility function. Given this anchor, a person could be influenced by social norms (a step in the direction of realism), so that context matters, but would still ultimately remain an individual. People might thus have many social identities, but as only arguments in the individual's utility function they would be subsumed under their utility function personal identities.

It is unfortunate, then, that having decided that identity matters to economics Akerlof and Kranton didn't simply come out and say that on the standard view an individual's personal identity is their utility function. This would have given more clarity to their discussion of the concept of identity, since people then would be less likely to be led into the confusion that a person's individual identity is paradoxically their social identity, and it would also have made the proposition that an individual's personal identity is their utility function available for evaluation. However, such an evaluation, I believe, would not be favourable to their view for two reasons I have elsewhere set forth (Davis Reference Davis2011).

First, it would create an important problem for their explanation of choice behaviour. As they explain choice, individuals face trade-offs between identity utility gains/losses and ordinary consumption utility gains/losses. If I feel bound by a particular social identity norm, I may sacrifice consumption to maximize utility. But while Akerlof and Kranton allow we have many social identities (and a vector of social identity arguments in our utility functions), they do not address how people trade-off their social identities against one another and, more significantly, choose whether they will even accept certain social identity assignments, as when one chooses to ignore a particular identity because one rejects its norms for ethical or practical reasons (cf. Sen Reference Sen2006).

They don't do this because doing so would require that the individuals be able to step outside of their utility function identities to make judgements regarding their utility functions by evaluating or deliberating upon its arguments. This would not only imply that there is something more to the personal identity of the individual than just what is in the utility function, but would also make deliberation over one's motives – an idea entirely contrary to utility function reasoning – part of that personal identity. The upshot, in any event, is that their utility function view of choice cannot really explain an important domain of choice, which is what economics is supposed to be able to do.

But there is a second, even more serious problem associated with the utility function personal identity position, namely, that as an explanation of individuality it is circular and empty. While it is not a goal of Identity Economics to explain individuality, nonetheless they clearly assume that people understood as collections of preferences are distinct and individual. Yet one cannot show a person to be distinct and individual in virtue of having some single set of preferences, since two different people could have identical preferences. A person's preferences only mark them off as distinct individuals when we say that they are their ‘own’ preferences, so that however they came to have those preferences, having them be their ‘own’ satisfies the requirement that they be seen as exogenous. That is, it is not the character of a set of preferences per se – that they somehow constitute an integral set – that individualizes the person, but rather that that set somehow belongs to the person. But then what makes a person a distinct individual is that they have their ‘own’ preferences, which is circular and accordingly tells us nothing about individuality or personal identity. Needless to say, the ‘belongingness’ idea is also entirely untheorized.

Again, to be clear, Akerlof and Kranton restrict themselves to discussing the concept of social identity, and do not attempt to say how an individual's social identities relate to their personal identity. ‘People's identity defines who they are – their social category’ (13). They explain their grounds for focusing only social identity in terms of how they drew on social psychology to elaborate their particular concept of social identity – a matter that gets limited attention in Identity Economics but is explicit in their first paper (Reference Akerlof and Kranton2000). Thus, there they flag John Turner's influential self-categorization theory (e.g. Turner Reference Turner and Lawler1985), an approach known in social psychology as the ‘social identity approach’, which is chiefly associated with the social psychology of identity in the field of psychology (cf. Hogg et al. Reference Hogg, Terry and White1995). They do not recognize or distinguish an important alternative approach to social identity in social psychology, known as the sociological approach to identity, which is chiefly associated with social psychologists in sociology (Stets and Burke Reference Stets and Burke2000). The psychologists’ social identity approach takes the individual as given, largely sets aside personal identity issues, and asks how established social categories impact individuals. In contrast, the sociological approach to social identity derives from symbolic interactionism (Mead Reference Mead1934), and investigates how individuality/agency and social structure mutually influence one another, and thus how social identity and personal identity are related.

That Akerlof and Kranton, then, draw on the psychologists’ social psychology approach to social identity allows them to take individuals as given, while saying that individuals are impacted by social identity assignments that can be made of them. They ignore the sociologists’ approach to social identity which would have raised the issue of how individuals respond to and possibly deliberate over their social identity assignments. As the latter would imply a different conception of the individual than standard economics employs, it was, so to speak, off their navigational charts before they ever left the harbour. This tells us something interesting about economists’ selective borrowing from other sciences, particularly psychology, in a period when economics and rationality theory is virtually being overrun by ‘outsiders’. The best defence in such circumstances, the lesson would seem to be, is to manage rather than expect to restrain other-science conceptual entry, all with an eye to what is most important to preserve of the traditional programme in economics – here it seems, not rationality theory, but Robinson Crusoe himself.

This may all seem to be no more than a set of academic debates and concerns, but arguably something more significant is involved when we think a bit more about how social identity assignments get made through the creation of social group categories, and how Akerlof and Kranton then presuppose this social construction process in their four applications of their analysis.

Consider their Chapter Five on identity in organizations. Part of what is involved here is a response to standard efficiency wage theory formulated in principal-agent terms. On that view, firms pay higher than competitive wages to secure effort from workers by employing financial incentives to solve the principal-agent problem. But wait, say Akerlof and Kranton: if you look at military organizations, loyalty to the social group allows lower wages to be paid, as individuals trade wages off against military social identities. That is, social context matters, and missing this efficiency wage theory has got it backwards. Or do Akerlof and Kranton have it backwards, since what they are effectively arguing is that if we militarize the workplace by promoting firm loyalty type social identities, we can reduce wages while maintaining or even increasing output. Clearly key to this argument is the idea that social identities are instruments wielded by someone to assign, mobilize or somehow construct others’ social identities for them, a framework which happens to comport with Turner's ‘social identity approach’ because that research programme doesn't ask where social identities come from or how they may be mediated by social processes or individuals themselves. In this respect, Akerlof and Kranton's position fits nicely with behavioral economists’ libertarian or asymmetric paternalism, while also going further. Whereas the latter only want to nudge individual choices toward the expert's view of rationality, identity economics may be used by people interested in nudging people's identities, and not in their own interest but in the interest of organizations with which they are to identify. This all seems an important lesson regarding how economics borrows from social psychology, and makes the issue of cross-disciplinary traffic more than just an academic concern.

There are many other interesting issues in Identity Economics application chapters that deserve the reader's attention. A bit worrisome is a tendency some of the discussion exhibits to reduce everything to social identity norms as if there were no other sorts of norms, such as character or virtue norms, which might cause behavior. For example, one successful school discussed in Chapter Six (73) is described as encouraging Five Habits of Mind. But does student success then result from their identification with the school that has such a programme or from prior acceptance of these norms? For Akerlof and Kranton it is always the former. No doubt social identity is often important, but it may also be an effect of other factors. This point sometimes gets lost in many of the book's examples, though Akerlof and Kranton are surely right to say that identity often plays an important role in people's behaviour.