Preliminary Programme

While the 1911 Census was known as the ‘Fertility Census’ in the UK, because special questions were asked concerning the number of children born to current marriages, couples were also asked how many children they had lost through death by the date of the census. With the recent release of the enumerators’ forms from 1911 for individual households it will become possible to map the residences of families who had suffered infant mortality, and to relate elements of the social and physical environment in which they lived to their experience of child loss. This paper will describe an exploratory study of this kind, using data from selected areas of Belfast, Ireland and consider the influence of religion, occupation, literacy, fertility and proximity to environmental hazards on infant mortality, both at the level of individual couples, and more familiar aggregate levels. (Show less)

Convictions of women for murdering their young children were rare in 19th century England. In fact, after 1849 capital punishment was not imposed for a mother's murder of an infant of under one year . According to historian Ellen Ross, some of the "not guilty" verdicts reached at Old Baily in the latter part of the 19th century for women whose newborn babies had died are indeed astonishing. Historian Judith Knelman asserts that the legal system tended to address deaths occurring within the home, in the private sphere as accidental or the will of god. Acquittal was highly likely for women prosecuted for infanticide. Where the murder charge failed, the jury could return a verdict of concealment of birth, a lesser offence punishable by up to two years’ imprisonment. However, the majority of women who killed their offspring avoided prosecution altogether.
There were legitimate reasons for discreetly condoning infanticide: many more babies were born than could be provided for; Judges and juries tended to think it preferable for a child to die immediately after birth than slowly, of starvation or neglect. Since the New Poor Law of 1834 relieved the parish of providing an allowance to unwed mothers, judges and juries apparently felt it only fitting that penalties for infanticide be fairly light. Historians claim that as long as the law required the execution of mothers who had murdered their offspring, judges and juries were reluctant to convict them.
One of the motivations for this unofficial legal policy was the good-mother myth consolidated during the 19th century, according to which maternal instinct compelled mothers to sacrifice themselves for their children. Given this belief, mothers who committed infanticide were seen by middle class Victorians as insane, victims of the biology and the psychology of their unstable natures. Thus they were frequently acquitted on grounds of insanity and sent to mental asylums rather than prisons or the scaffold.
Researchers agree that judges in courtrooms and doctors in mental institutions supported the idea that the effects of giving birth left women in a state of mental excitement sufficient to explain their crime. Historian Hilary Marland points out careful readings of the accounts of trials detect signs of relief among judges, juries and witnesses when women who committed infanticide could be declared insane. Medical and legal authorities inside and outside the courtroom were willing to accept the interpretation of insanity even when they remained unconvinced of its validity.
The present paper follows the realization of these legal and medical attitudes and policies in the case of Ellen Harper, a poor married woman aged 17 who killed her newborn daughter in October 1877 at Langely Marsh, Somerset. The first doctor who examined Ellen for the inquest declared that she had been sane while committing infanticide. The inquest verdict was "guilty". The verdict was changed, however, in the course of her trial in Exeter and on January 15 1878 she was acquitted on grounds of insanity and sent to Broadmoor hospital for the criminally insane. The doctors at Broadmoor, who described her case in psychological terms, established her insanity. And yet, during two years of hospitalization, Ellen showed no signs whatsoever of madness.
My paper will attempt to demonstrate how Ellen's trial and hospitalization manifest policies of the 19th century English legal and medical systems towards cases of infanticide, rather than reflecting her mental state. (Show less)

This paper aims to contribute to the literature on the role
of morbidity in early adult life on the subsequent risk of death in England during the late-nineteenth and early-twentieth centuries, and more generally to our knowledge about the relationship between morbidity and mortality during the demographic transition.

Gorsky et al. (2006) showed that, in a small sample of working-class men in southern England, the number of episodes of sickness reported at ages under 50 years was positively associated with the risk of death at ages over 50 years. They suggested that this might be the result of ‘insult accumulation’ whereby progressive debilitation caused by frequent sickness leads to a weakening of the constitution, but that it might also be consequence of differential ‘frailty’ whereby some individuals manifested unobservable characteristics which made them both more susceptible to sickness and more likely to die (Alter and Riley, 1989). With the volume of data available to them, however, they were unable to distinguish between these two possibilities.

This paper analyses a much larger data set than was available to Gorsky et al. (2006). Specifically, it analyses more than 5,000 morbidity histories, which have been extracted using a sample of the records of the members of the Hampshire Friendly Society (HFS) who joined between 1825 and 1939. The HFS was a local mutual assurance society based in the south of England. It was founded in 1825 and lasted until the penultimate decade of the twentieth century. Our sample consists of about 10 per cent of those who joined the assurance scheme, but rather more than 10 per cent of those who contracted for sickness benefit. Most of the data relate to the period 1850-1950. The members of the HFS were mainly males – indeed for most of the period sickness benefit contracts were not offered to females.

Our data are subdivided into periods (yearly periods until 1895 and quarterly periods after that). We know whether each man received sickness benefit in each period and, if he did, for how many days. We also know dates of death for each man, and for those who ceased to be members of the society while still alive, their dates of cessation.

In addition to information about each man’s sickness history, we have data on a limited number of covariates, including year of birth, place of birth, place of residence on joining the Society, age on joining the Society and occupation on joining the Society. The values of these covariates are fixed for each man at all ages.

References

Alter, G. and Riley, J. (1989) ‘Frailty, sickness and death: models of morbidity and mortality in historical populations’, Population Studies 43, 25-45.

Gorsky, M., Harris, B. and Hinde, A. (2006) ‘Age, sickness and longevity in the late nineteenth and the early twentieth centuries: evidence from the Hampshire Friendly Society’, Social Science History 30, 571-600. (Show less)

Techniques for studying dynamic heterogeneity in mortality frailty have not been developed in demography, although the subject has important implications for understanding the health transition and for health interventions. A method of ‘scoring’ correlated frailty for a household is defined by comparing observed mortality between panels with the expected mortality of the household members as determined from an aggregate life-table. Changes in average frailty over time at the population level are decomposed into three components: change experienced by households measured on two consecutive occasions; a selection effect if the frailty of new households is consistently different from the established group; and a corresponding effect if the households that drop out differ from the ones that continue. The decomposition method is exact and seems to be unknown in the social sciences. Frailty scoring and decomposition are applied to longitudinal data on households in the China Multi-Generational Panel Dataset, 1789-1885. The results suggest that the apparent long-run stability of mortality in the population conceals a consistent evolutionary dynamic at the household level. (Show less)

Objective:
This paper compares the impact of early parental loss on the survival of children in two different historical populations in Krummhoern [East Frisia, Germany] and Quebec [Canada]. The Krummhoern can be characterized as a saturated habitat with little expansion possibilities while in Quebec expansion was virtually unlimited for the European settlers. Because of these different socio-economic and geographic conditions, expectations are that early parental loss affects the survival of children differently.

Methods:
We used event history analysis methods with time-varying specifications of family structure on a sample of 7,077 male and 6,906 female children from the Krummhoern born between 1720 and 1859 and 31,490 male and 33,109 female children from Quebec whose parents married between 1670 and 1720.

Results:
Results indicate that in both populations parental loss is associated with increased child mortality. Maternal loss in both populations is associated with a universal and consistent effect on boy’s and girl’s survival, while the impact of paternal loss is less easy to establish and to interpret. The effect of the remarriage of the surviving spouse is population-specific: In Krummhoern father’s remarriage reduces dramatically the survival rate of his children of former marriage, while such an effect in Quebec is missing. In contrast, mother’s remarriage has no effect in Krummhoern, while children in Quebec tend to benefit from mother’s remarriage.

Conclusions:
These population specific effects of parental loss and its consequences (e.g. remarriage of the surviving spouse) seem to be driven by the socio-economic living conditions (e.g. availability of resources). (Show less)