In their recent account of the relationship between social capital and the political economy of public health,1 Simon Szreter and Michael Woolcock distinguish between two different forms of social capital—‘bonding’ social capital and ‘bridging’ social capital—and identify a specialized form of bridging social capital called ‘linking social capital’, which is distinguished from other manifestations of bridging social capital by its emphasis on the construction of relationships between acknowledged social unequals. Their theoretical argument is buttressed by a historical case-study of the relationship between social action and the decline of mortality in nineteenth- and early-twentieth-century Britain.
Although Szreter and Woolcock have sought to emphasize the fact that social capital is ‘emphatically not … the sole (or always primary) variable that explains (or should be used to try to explain) public health outcomes’,2 it is at least arguable that their enthusiasm for the concept has nevertheless led them to present a somewhat one-sided account of the history of public health reform and mortality change in nineteenth-century Britain. I would like to develop this argument further by considering three main points: (i) the causes of mortality decline in the late-eighteenth and early-nineteenth centuries; (ii) the origins and nature of changes in nineteenth-century social policy; and (iii) the timing of mortality changes.
The causes of mortality decline in the late-eighteenth and early-nineteenth centuries
At the present time, it is undoubtedly difficult to offer any categorical explanations for the decline of mortality in England and Wales between c. 1750 and 1820, but any plausible explanation is likely to include some acknowledgement of the impact of inoculation and vaccination against smallpox, the draining of mosquito-infested marshlands, and the beneficial effects of efforts at civic improvement and what James Riley has called ‘the campaign to avoid disease’.3–9 However, there are also grounds for believing that there was a moderate improvement in real wages and an increase in per capita food consumption.10,11 In the past, economic and social historians have argued that real wages declined in England and Wales between 1780 and 1820, and Simon Szreter contends that ‘aggregate real wages failed to show any trend rise’ between 1750 and 1800.12 However, this is not strictly true, as Charles Feinstein's reworking of the real wage data (which is not cited in Szreter and Woolcock's original paper or in either of their responses) suggests that real earnings rose, on average, by 12.5% between 1770/1702 and 1818/1822, and by 23.1% between 1818/1822 and 1848/1852.13
The origins and nature of changes in nineteenth-century social policy
As Robert Putnam has shown in his perceptive comment on Szreter and Woolcock's original article,14 the utility of their use of the concept of ‘linking social capital’ is compromised by their failure to distinguish more clearly between the causes of policy changes and the policies themselves. This failure is compounded by an oversimplified account of the determinants of certain changes, including the New Poor Law.
In their account of the origins of the New Poor Law, Szreter and Woolcock appear to link the decline of the Old Poor Law to the emergence of a new ‘urbanising propertied class [which] increasingly embraced the morality of the market and voted for a meaner ‘new’ Poor Law … in 1834’,2 but, as Anthony Brundage has shown, much of the impetus behind the introduction of the New Poor Law was generated by the landed aristocracy, and much of the early opposition to its implementation came from both middle- and working-class interests in manufacturing areas.15,16 Szreter and Woolcock also present what many historians would still regard as an oversimplified view of the impact of the 1834 ‘reforms’. They state that ‘no longer were there to be cash handouts for the families of the unemployed. Now they were to be segregated by sex and compelled to repay their meagre social security allowances by arduous labour inside workhouses’,1 but, as many historians have shown, outdoor relief continued to be paid to the dependants of unemployed men and even though the total number of poor law claimants undoubtedly declined, the vast majority of those who continued to receive poor relief did so outside the workhouse.16 However, the really interesting question is how far the abrogation of welfare rights affected living standards. Szreter and Woolcock imply that the main lesson to be drawn from their account is what it tells us about social capital, but this question is arguably rather less important, in terms of social epidemiology, than its impact on material resources.
In addition to examining changes in poor relief, Szreter and Woolcock also pay considerable attention to the factors which impelled Joseph Chamberlain along the road of ‘gas and water socialism’ in the 1870s. In their original article, the two authors highlighted the importance of Chamberlain's social and political networks and his religious beliefs, allied to ‘the opportunities for a more democratic and participatory urban politics opened up by the British state's belated enfranchisement … of a large section of the working classes’, and Szreter himself returns to this theme in his response to Davey Smith and Lynch.1,2,17 However, unless the changing political climate of mid- to late-nineteenth century Britain is given its proper weightage, the reader is left with little sense of the reasons why the ‘civic gospel’ might have developed. This may well have been a response to changes in the orientation of religious belief and, for that matter, to the visitation of the fourth great cholera epidemic in 1866–1867, but it was also fuelled by a changing appreciation of political realities. As Chamberlain himself noted, in a phrase not quoted by Szreter and Woolcock but which might have been used by their critics, ‘the foundations of property are made more secure when no real grievance is felt by the poor against the rich’.16,18,19
The timing of mortality changes
The exchange between Szreter, Davey Smith, and Lynch highlights the significance of the debate over the onset of the late-nineteenth-century mortality decline, and this debate gets carried over into a debate about the impact of early-life conditions on subsequent mortality. Although these issues are clearly interlinked, they can also be dealt with separately.
Szreter and Woolcock argue that there was no significant improvement in British mortality rates before the 1870s, and they claim that this reinforces their view that the creation of linking social capital, as pioneered by Joseph Chamberlain, was central to the process of mortality decline. However, whether or not one accepts the validity of a cohort-based interpretation of mortality change, it is clear that childhood mortality did begin to decline before the 1870s, and Szreter and Mooney's figures suggest that the average level of life expectancy at birth among people living in Britain's largest provincial towns rose from 30 to 34 during the course of the 1850s.18,20 If one accepts the view that there was little effective public health intervention in the vast majority of Britain's cities during this period, it is difficult to avoid the conclusion that the most important determinant of mortality decline at this particular point in time was the acceleration in the rate at which real wages were rising after 1850. However, it is important to add that this observation in itself does nothing to diminish the influence of environmental factors on the poor state of health experienced by the residents of nineteenth-century cities or the importance, at some later date, of effective social intervention to remedy these conditions.13
In addition to querying the timing of mortality decline, Szreter also raises the larger question of how far the overall pattern of mortality change in the years after 1850 can be attributed to cohort effects.12 He accepts that ‘conditions in early life must be important’, but suggests that these are likely to have left a much smaller imprint on the demographic record than events closer to the time of death. One of the main reasons for his scepticism is the apparent absence of any relationship between mortality at ages 0–4 years in England and Wales and mortality at ages five and upwards. However, as I showed in my analysis of Kermack et al.'s pathbreaking article in this journal in 2001,21 it is important to distinguish between mortality at ages 0–1 and 1–4. The English and Welsh data do not permit us to do this, and this means that the relationship between young child (1–4) mortality and older child (5–14) mortality is obscured by the statistical preponderance of deaths at ages 0–1. We are able to obtain separate data for 1- 4-year-old mortality in case of Scotland, and the relationship between these data and mortality rates at older ages is much closer to the one which the cohort model would predict.22
There is insufficient space here for a more detailed discussion of the relationship between early-life conditions and later-life mortality, but such a discussion is probably unnecessary anyway, in view of the widespread currency which such arguments have already received in this journal and elsewhere in the epidemiological literature.23
Although I have taken issue with some of the historical aspects of Szreter and Woolcock's article, I also think that they are right to draw attention to the factors which are likely to influence relationships between the advantaged and the disadvantaged, and it seems clear from the reactions they have already provoked that the concept of ‘linking social capital’ is likely to play an important part in subsequent debates. However, it was surprising that they should devote so little attention, in their original article, to the role of philanthropy in the nineteenth and twentieth centuries. Nineteenth-century philanthropy was often denominational, self-interested, and misdirected, and organizations such as the Charity Organisation Society were often more interested in ‘reforming’ the poor than in developing a common interest with them.16 However, during the early-1900s, many middle-class philanthropists abandoned the Charity Organisation Society in favour of what became known as ‘Guilds of Help’,24 and in 1934 the leading theoretician of voluntary action, Elizabeth Macadam, called for a new kind of philanthropy, in which voluntary organizations would take responsibility for pioneering new social schemes and working with disadvantaged groups to bring ‘pressure to bear on the state to amend existing and introduce fresh provision for social needs’.25 An examination of the role played by groups such as the Society of Friends and the National Council of Social Service during this period could provide a useful historical test-case for the value of the concept of ‘linking social capital’, of the kind which Kawachi et al. rightly recommend.26