As John Welshman so persuasively shows, the idea of an ‘underclass’ – a dangerous class of the morally blameworthy and undeserving poor cut off at the bottom of society – has a long and chequered history. Welshman suggests that since 1880 there have been ‘at least nine reconstructions’ of the basic underclass idea and his research delves deep to unearth the intellectual connections and conceptual linkages between: ‘the social residuum’, ‘the unemployable’, ‘the social problem group’, ‘the problem family’, ‘the culture of poverty’, ‘the cycle of deprivation’, ‘the underclass’, ‘social exclusion’ and ‘troubled families’. Together with an introduction and conclusion, ten chapters are chronologically organised to cover these ‘reconstructions’ (underclass theory per se gets two chapters – the US version and the UK version).

Underclass ideas can be regarded as what Paul Spicker has called ‘zombie arguments’; ones that ‘no matter how many times they are shot down in flames or have a stake driven through their heart . . . seem to get up again afterwards’ (2007: 102). Welshman, as an historian rather than a sociologist, is not in the business of trying to impale zombies (for example by close scrutiny of the evidence for and against underclass ideas), he is more interested in exploring how, albeit with occasional discontinuities, an ‘underclass has been successively re-invented over the past 132 years in Britain and the US’ (p. 2) or how ‘the social residuum of the 1880s is the troubled family of the present day’ (p. 14). Thus, the book is a history of discourses and ideas rather than a sociological investigation of class and poverty. This perhaps explains why some empirical research that engages directly with underclass theory is not referenced (including some by this reviewer).

This is the second edition of Underclass. It has been updated since the first in 2006 with two new chapters (those about the evolution of the concept of social exclusion under the New Labour governments and the ‘Troubled Families’ of the current Coalition government). Although there are other texts that deal with similar questions and issues, none, to my knowledge, is so detailed, thorough and up-to-date (even though it is difficult to keep up-to-date as new history is being so speedily written by a Coalition government keen to widen the net from the originally alleged 120,000 so-called Troubled Families to the now 400,000 ‘at risk families’). This undoubtedly will become the key reference on the history of the underclass idea. It is a fantastically useful resource − erudite, scholarly, reliable. The historian's skill is evident throughout, and very neatly shown, for instance, in the tracing and dissection of intellectual origins and pragmatic outcomes of Sir Keith Joseph's ‘cycles of deprivation’/ ‘cycles of disadvantage’ programme.

For those already acquainted with the history of the underclass, there is much that is familiar. For me, though, there were many valuable reminders – and new nuggets. For instance, I had not known that, apparently, members of the Victorian social residuum were, according to one observer, ‘unable to remember street names, and house numbers, had a poor sense of direction and . . . trouble in distinguishing left and right’ (p. 32). Another new reference, personally valuable for my own work on the persistence of poverty in deeply ‘Troubled Families’, is the line Welshman offers from the 1973 SSRC/DHSS Working Party on Transmitted Deprivation, about how causal processes in continuity over generations may not be ‘familial but the influence of a common environment on successive generations’ (p.127). This is an intellectual furrow that I am ploughing, with colleagues, and the tip-off back to this report has been extremely helpful.

Of course, reminding sociologists and policymakers of the historical dimensions to these debates is one of Welshman's aims. One example is that throughout the story there has been a tendency to confuse description of the effects of poverty with the causes of poverty. This is neatly exemplified in Welshman's discussion of the ‘culture of poverty’ thesis, but many present-day examples could be found in the pronouncements of influential, government-friendly think-tanks. Historical continuities are multiple. I did know that Blair's government in the 2000s concluded that roughly 2.5% of every generation were ‘deeply excluded’ (p. 200) but I had not known that in the 1950s the same figure (2−3%) was said to comprise the rate of ‘problem families’ in any given area. Nowadays, moral opprobrium is dumped on Britain's ‘underclass’ because of their large-screen plasma TVs and inappropriate consumption of cigarettes and alcohol. In the 1960s, impoverished ‘problem families’ were criticised for their possession of ‘tape recorders, radiograms, cocktail cabinets, and often one or more large breed pedigree dogs’ (p. 92). More serious examples of the deadening sense of historical amnesia that infects contemporary politics and policy is the way that politicians, from Tony Blair to Eric Pickles (p. 214), trot out old ‘cycle of deprivation’ ideas as a self-evident truths when, as the book shows, these are nothing of the sort. There is a sinister parallel as well in the way that the current ‘Troubled Families’ programme and the Problem Families Committee of the Eugenics Society of the late 1940s have gone about identifying and counting the families under their surveillance.

As well as tracing this history, Welshman also seeks to understand why these ideas have been so ‘strikingly resilient’ (p. 56), so fleetingly in abeyance over the past century and more, and so resistant to sociological rebuttal. In my view, this question is not satisfactorily answered. Direct consideration of it is left mainly to the final paragraphs of the book where Welshman makes the argument that, contrary to a common view, these are not just ideas that arise in times of particular economic or social malaise. He raises several suggestions, including, firstly, that the relative importance of behavioural versus structural factors in the causation of poverty remains unresolved. I am not sure that all readers would agree with him on that. Or more to the point, what can happen with debates about poverty and ‘the poor’ is that behaviourally based, underclass-type ideas are asserted as true with little support from rigorous, extensive social scientific evidence. In British politics and policy currently, there is certainly plenty of evidence of exactly this – and, as Welshman shows, of the persistent power of pernicious ideas about the so-called underclass.