So much rubbish has been written about multiculturalism in recent years it is difficult to know where to begin. Politicians, journalists and other commentators who should know better have made currency out of treating multiculturalism as synonymous with ‘separatism’ and ‘relativism’. There are many reasons why diverse communities may not integrate effectively and, to be sure, government policies designed to embrace diverse cultural identities and needs may sometimes exacerbate social tensions. But why then condemn multiculturalism per se? Democratic institutions sometimes malfunction. Should we therefore abandon democracy as well? Balancing the principles of equality and diversity, universalism and particularism, the material and the cultural is far from easy; yet the attempt is also preferable to asking us to sign up to reductive lists of national British character. Identity has to be conceived as dynamic, deliberative and open. To yearn for anything else is to risk invoking a Golden Age that never was.

This ‘retrospective utopianism’ has shaped publications by Goodhart (Reference Goodhart2006) and Dench et al (Reference Dench, Gavron and Young2006) – admittedly, the latter does not discuss multiculturalism as such but its genesis and purpose are clear. Both are infected by wishful thinking, superficial analysis, a lack of data and context. Ranged against this are defences of multiculturalism (or at least certain types of multiculturalism) by Sen (Reference Sen2006) and Phillips (Reference Phillips2007), though their direct relevance to social policy issues is weak. The additional contribution of Banting and Kymlicka to these voices is therefore welcome. Both have spent years critiquing some of the wilder claims within this debate and in this text, Multiculturalism and the Welfare State, they turn their attention specifically to, well, multiculturalism and the welfare state.

The book includes several chapters offering conceptual and theoretical reflections. For the most part, though, it presents analyses based upon a wide range of data and research, some of it cross-national and some pertaining to particular countries; there are individual chapters on the USA, Britain, the Netherlands, Germany, Canada, Latin America and Bolivia. The authors do what social scientists are meant to do: shoot down misconceptions and myths, propose hypotheses and provisional conclusions, while identifying questions for further research. For this and many other reasons they provide a much-needed antidote to the sillier depths of recent public debate.

These analyses pretty much follow the framework laid down by the editors and co-authors in two long introductory chapters. They outline two hypotheses. The ‘heterogeneity/redistribution trade-off’ hypothesis maintains that ethnic diversity undermines redistributive policies. If people don't feel connected to one another, the reasoning goes, they will be reluctant to fund services for each other. The ‘recognition/redistribution trade-off’ hypothesis states that a concern with the politics of recognition and difference drains attention and energy away from a politics of redistribution (Fraser and Honneth, 2003). To explore whether these hypotheses are valid, Banting and Kymlicka assemble an array of theoretical arguments and empirical data. Their argument is not so much that the hypotheses are incorrect, more that they are over-generalised. Redistribution can be undermined by diversity and a politics of recognition but more importantly there is no inherent logic at work here. For instance, multiculturalism may threaten trust and solidarity but it may equally support them. There are too many factors to justify simplistic generalisations, and much depends on how multiculturalism is made to work. Ironically, it is an anti-multiculturalist rhetoric that could make it harder to preserve social trust and solidarity.

The philosopher David Miller is offered a platform for some counter-arguments in a penultimate chapter. He agrees that we should not condemn multiculturalism per se but is also willing to defend critics like Brian Barry (Reference Barry2001; cf. Kelly, Reference Kelly2002) when they worry about the implications for egalitarianism. The answer is to define and defend a ‘moderate multiculturalism’ consistent with the continued salience of nationality as the site for state, culture and citizenship. There are problems with Miller's approach (see Fitzpatrick, Reference Fitzpatrick2008: Ch.11), but it too challenges the simplicities of much political debate.

The book is offered as the beginning of a process and not its final word. It invites challenge, revision and addition rather than resisting them. Those with a particular axe to grind can make their own case elsewhere. My view is that, while it is scattered throughout the chapters, the book would have benefited from a more systematic analysis of migration policies. True, this would have added another layer of complexity to an already complex field, but given that the Netherlands chapter deals with migration at length its absence elsewhere is notable.

It's a minor point, though. The book deserves to be a standard point of reference and is already on its way to becoming one. It does not offer facile answers and knocks down the facile answers proposed by others. The idea that there is no final, determinate way of balancing equality and diversity, universalism and particularism, the material and the cultural is troubling to some. But Multiculturalism and the Welfare State suggests that the monocultural alternatives are only superficially attractive. How can a humble book reviewer disagree?