Historical demography abstracts

Strand organisers: Romola Davenport, Alice Reid, University of Cambridge

Historical demography: Mortality in historical populations - Tuesday 9 September 4.45pm

Disease and the certification of child deaths in urban and rural Scotland, 1861-1901
Alice Reid, University of Cambridge, Eilidh Garrett, University of St Andrews

Before the certification of deaths by a doctor was universally adhered to, it was common to find differences in the likelihood of medical certification according to the age of the deceased: in particular infant deaths were much less likely to have been medically certified than those of older children and adults. This could be ascribed to a more fatalistic attitude towards the likelihood of death in infancy and childhood, which itself has been attributed to infant lives being valued less than those of older people. Alternatively, it is possible that parents may have been differently inclined to call the doctor according to the illness manifested – some diseases or conditions might have been viewed as more treatable than others, thus age patterns might have emerged as a result of the different age pattern of disease. The speed at which the disease became critical might also have affected the likelihood of summoning a doctor. This paper uses individual registrations of death for two Scottish communities between 1861 and 1901 to investigate investigates the differences in the likelihood of medical certification by age, sex, geography, distance of nearest doctor’s residence, season, length of last illness. It will use the results to comment on the value put on the lives of children of different ages and genders.

Email: amr1001@cam.ac.uk

Social class and mortality in the first stages of the epidemiological transition: infant mortality by social status in Georgian London 1752-1812.
Romola Davenport, Cambridge Group for the History of Population and Social Structure, Jeremy Boulton, University of Newcastle

While it is now the norm that urban dwellers and higher income groups enjoy higher life expectancies than their rural or poorer counterparts, this was not the case in historical populations. In England these phenomena appear to have emerged in the late eighteenth century, and coincided with rapid population growth and urbanization. Here we present new evidence regarding both mortality decline and class differences in infant mortality in eighteenth-century London using the extraordinarily rich records of the Westminster parish of St. Martin in the Fields for the period 1752-1812. We used a family reconstitution methodology to partially reconstitute families and to classify these by social status according to fees paid for baptism. Analysis of mortality rates and birth intervals indicated an inverse social gradient in mortality in the first month of life in the mid-eighteenth century that paralleled to some extent the prevalence of maternal breastfeeding by social status. A reduction in the proportion of short birth intervals amongst wealthier families in the last quarter of the eighteenth century suggested that there was some convergence to a pattern of relatively lengthy maternal breastfeeding in this period and this was accompanied by falls in neonatal mortality. Post-neonatal mortality varied little by social status and declined only slightly before c.1800. Declines after 1800 were associated with a specific fall in smallpox death rates that coincided with the advent of vaccination, and were undifferentiated by social status. The disease environment of eighteenth-century London appears to have overwhelmed any advantages of wealth in infancy.

Email: rjd23@cam.ac.uk

The transmissibility of influenza pandemics of 1889-1890 and 1918-1920 in a large Urban environment: Madrid, Spain.
Diego Ramiro Fariñas, Sara García Ferrero, Institute of Economics, Geography and Demography, Center for Humanities and Social Sciences, CCHS- CSIC, Spanish National Research Council

The 1918–1919 pandemic influenza, the “Spanish” flu, killed about 50 million people worldwide. There have been many studies of the transmissibility of the 1918 Spanish flu virus. Many analyses have involved fitting transmission models to the observed epidemic curves based on published data from cities in Europe or América. These attempts to estimate the rate of transmissibility of influenza among people have the objective of planning mitigation strategies and control of infectious diseases from potential new pandemics. Quite often these estimations relays on historical published data from where parameters that model the transmission of the disease are estimated. The transmissibility of the influenza can be quantified by the basic reproductive number (R0), an epidemic that represents the number of secondary infections arising from each of the primary cases of infection in a susceptible population. There is abundant literature contributing to the estimation of R0 from total counts of deaths, using sometimes cause of deaths and not that often other socioeconomic variables (See bibliography below). Other pandemics, like the influenza pandemic during the winter of 1889-1890, the "Russian Flu", was one of the most important pandemics during the XIXth century and it was the first influenza pandemic in an interconnected world. There have been several studies in recent years which have estimated the rate at which influenza moved worldwide as an attempt to know the diffusion and speed at which an influenza epidemic can spread. An example of this is found in the research carried out by the Institut National de la Santé et Recherche Médicale in Paris, in which they estimated that the diffusion of the 1889-1890 pandemic was truly amazing, moving around great part of the World in just four months. The average speed was estimated at 394 km/week for the European continent and 1,015 km/week within the United States. The transmissibility rate for the 1889-1890 pandemic in Europe was estimated to be R0=2.1.

While the transmissibility rate of the 1918-1919 pandemic was estimated approximately to be R0=2 and 3 for 45 cities in the United States. For a more recent Pandemic influenza, the 2009 A (H1N1) in Spain, the R0 value was estimated to be 1.29. Therefore, the estimation of this parameter and the patterns of geographical distribution within a big urban environment are of great interest because it will allow determining the potential diffusion of an epidemic and how that epidemic could be tackled and controlled. Therefore, the scientific goal of this contribution is to estimate transmissibility rates and the geographical distribution of two influenza pandemics 1889-1890 and 1918-1920 in the City of Madrid, which had a population around 500.000 inhabitants in 1900 and c700.000 in 1920. We will use the Longitudinal Historical Population Register of the City of Madrid which uses individual level information for all the individuals who lived and died in Madrid. The Register includes c20.000 deaths per year, during the period 1889-1923. The information included is, beside other variables, age, sex, causes of death, place and day of death, civil status.

Email: diego.ramiro@cchs.csic.es   

Fertility and child mortality transition. Alghero, Sardinia (1866-1935)
Lucia Pozzi, Marco Breschi, Stanislao Mazzoni, University of Sassari, Italy, Department of Economics and Business


The interrelationship between infant/child mortality is complex since involves multiple dimensions that are difficult to disentangle and despite its relevance, has received up to now a very limited attention. Furthermore research on this topic has not permitted to measure adequately the interaction mechanisms involved, since it has mostly relied on aggregate data. This paper aims to explore this complex relationship on the basis of an individual level data set reconstructed for the north western community of Alghero for the years 1866-1935, based on civil and parish records of birth, death and marriage. This topic represents a crucial question for understanding the mechanisms of reproductive behaviours transition particularly in a region, like Sardinia, which experienced a very slow and delayed fertility decline as well as well a prolonged and very moderate reduction in infant and child mortality. However our previous research has permitted to highlight significant fertility and infant/child mortality differentials according to socio-economic status. One of the main question of this paper, undoubtedly intriguing, is therefore to examine to what extent the decline in fertility observed for the most advantaged social group can be attributed to the simultaneous increase in survival experienced by the children belonging to the well of families.

Email: lpozzi@uniss.it

Migration in historical populations - Wednesday 10 September 9.00am

The determinants of nineteenth-century migration streams: Evidence from the 1881 CEBs
Joe Day, University of Cambridge

Using the 1881 CEBs as the root source, this paper has enriched the birthplace records by linking them into a GIS. Analysing lifetime migration streams in the context of the leaving home process, allows the determinants of migration to be disentangled from those where migration is determined by individuals’ utility-maximising strategies, to those where migration has been forced on individuals by parents to relieve pressure on household resources. Analysing those migration streams which can be inferred to have been the result of individuals’ choices, clear patterns begin to emerge when these phenomena are mapped.

Specifically, this paper analyses how migration might have been a subsistence response to the agricultural depression of the late 1870s and early 1880s. By combining migration streams taken from census evidence with wage data, it becomes clear that migrants were split between movers in the south-east that did so in response to economic shocks and movers and the north that were responsive to the perceived risk and return of migration. This paper is intended to go beyond a descriptive analysis of migration trends in late nineteenth-century England and Wales, and instead move towards a determination of their causes. Analysing the incentives which migrants had to exit and enter regions ‘x’ and ‘y’ in the context of risk produces a fuller model than one that describes the move simply in terms of push and pull. Rather, by interrogating the context and the perceived risks and return from migration, a more nuanced picture of migration streams and their determinants arises.

Email: jd466@cam.ac.uk

Do migration and population structure predict variation in vernacular culture? Explaining the distribution of patterns in fishermen's ganseys
Malcolm Smith, Evolutionary Anthropology Research Group, Department of Anthropology, Durham University

The coastal fishing communities of the British Isles have a rich vernacular material culture, including the tradition of knitting ganseys – seamless woollen sweaters handmade as work-wear for fishermen –which flourished from the 19C to the mid-20C. These garments, knitted in the round on four needles, share a common basic structure but contain a wealth of pattern variation created from different combinations of plain and purl stitches, which may decorate the yoke, sleeves and body. Ganseys were usually knitted by women, and patterns were transmitted orally, typically from mother to daughter. In this paper, observed nineteenth-century migration, surname distributions and formal models of population structure (Wright's island model and a one-dimensional stepping-stone model) were used to predict relationships between Yorkshire coast fishing communities. These predictions were compared with the distribution of patterns recorded from 113 ganseys from Filey, Scarborough, Robin Hood's Bay, Whitby, Runswick and Staithes in order to test ideas about the generation and distribution of variation. Phylogenetic inference software and Mantel tests were used to test hypotheses of origin and dispersal. Migration fitted the island model better than the stepping-stone, with little evidence of isolation-by-distance. Surnames, however, followed geography closely, but with Scarborough and Filey as outliers to the other settlements. Gansey patterns showed some geographical distribution, but it was not consistent with the predictions from migration or surnames. Alternative explanations for pattern distribution – including the “community identifier” narrative, (alleging the need to identify the bodies of drowned fishermen), and the rhetoric of individual creativity – have also been explored.

Email: malcolm.smith@durham.ac.uk

Hidden communities: a quantitative assessment of international mogration to Edinburgh at the turn of the twentieth century
Marc Di Tommasi, University of Edinburgh

Migration to UK before the First World War is one of the most important phenomena of our age. The early migrants’ communities went through the full spectrum of migration experiences, from being ostracized in the beginning to then become an exemplar model of integration. Despite the significance of the migratory phenomenon the migrants’ role in shaping British society and their contributions have been often under-estimated, both by the public opinion and by the mainstream historical writing. Even when the migrants’ contribution has been acknowledged, the studies have mainly been circumscribed to London and a few industrial cities. Moreover, the exact mechanisms of their integration have not been explored. Focusing on the city of Edinburgh this paper will re-evaluate the impact of the migrants using a

comparative, quantitative and spatial approach. Starting from census data, 1911, the migrants’ households have been reconstructed entirely providing a new estimation of the effective number of migrants. Rather than focusing on a single community, this research has compared all the different migrants’ groups giving context and weight to the analysis. Moreover, using a Geographic Information System (GIS), the patterns of settlement have been reconstructed and compared to the migrants’ social condition. The resulting data paint a surprising picture of the period, very much different from what previously known, where the migrants form an important and integral part of the social context. Although limited to only one city the results have deep implications, both from a methodological and from a social point of view.

Email: mtommasi@staffmail.ed.ac.uk

'Over the sea from Skye': a record linkage approach to migration using Scottish civil register and micro-census data, 1861-1881
Eilidh Garrett, University of St. Andrews/Cambridge Group, Alice Reid, Ros Davies, Cambridge Group, University of Cambridge

Parish-based family reconstitution studies have traditionally been restricted to analysis of those who remain in the parish from birth to death. Only a few studies have undertaken the time consuming work of following individuals into neighbouring parishes. Now with the release of individual-level census data for whole countries it may prove possible to link migrants back to their places of birth, allowing their experiences as a ‘mover’ to be compared with those who have ‘stayed’ in their joint place of origin. This paper builds on work originally undertaken to link the events recorded in the civil registers of births, marriages and deaths for the 7 parishes on the Isle of Skye to the individuals enumerated in census returns of the five enumerations as part of the Census of Scotland between 1861 and 1901. The individual 1881 census records of Scotland, along with those of England and Wales, are available via the UK Data Archive. Those claiming to have been born on Skye between 1861 and 1881 but living elsewhere in Scotland in 1881 will be identified and linked to their birth certificates. In addition the 1881 addresses of those born on the island during the 1860s and 1870s who appear to have left the island by 1881 will be sought. The paper thus considers issues of linkage methodology, investigates the accuracy of individual level census micro-data and also addresses some longstanding questions concerning the biases introduced into both prospective and retrospective historical demographic studies by migration.

Email: eilidh.garrett@btinternet.com