Strand organiser: Mike Murphy, London School of Economics
Identifying cohorts with favourable mortality experience
Mike Murphy
London School of Economics
Attention has recently been given to cohorts in Britain born around 1930 that are described as 'Golden cohorts', in part because they appear to have experienced greater mortality improvements then immediately surrounding cohorts. Similar patterns have been identified in some but not all other countries under the name 'select cohorts'. A number of approaches to identifying such cohorts using data from Britain, France and Japan are discussed, including period to period changes in mortality rates; use of numerical derivatives using super-smoothers; penalised splines within a generalised additive model framework; and APC models. I discuss reasons for the different findings according the methods used and the selection of the particular time and age surface, and how these affect the interpretation of trends.
The implicit (and sometimes explicit) assumption that observed patterns will continue into the future - and therefore may be used to improve forecasts - is discussed in the light of the appropriate methods for quantifying mortality improvements. It is concluded that the assumption that there exists a persistent advantage of certain cohorts with substantial substantive implications has not yet been established conclusively.
Professor Mike Murphy, Department of Social Policy, London School of Economics,
Houghton Street, London WC2A 2AE
Tele: 020-7955-7661 (direct line) (+44-20-7955-7661 from abroad)
E-mail: m.murphy@lse.ac.uk
Jewish population and death rates in England and Wales, 2001
Steven Haberman and Marlena Schmool
City University & Independent Researcher
Since the end of the 19th century, scholars have assessed the size of Anglo-Jewry by relating deaths' experience within the community to national death rates. The method involves collecting details of Jewish deaths from all burial organisations with the community. This has provided population estimates wider than any based on synagogue or community membership lists but these omit some who are Jewish within the definitions of Jewish law but do not wish a Jewish burial.
Data from the religion question in the 2001 census provided a Jewish base population from a different source and with a different definition. It presented an unprecedented opportunity to evaluate indirect estimates of Jewish population and to derive death rates for the Jewish community. These are compared with the England and Wales age/sex specific rates that have been used since the 1970s to extrapolate Jewish population from deaths' data.
The paper reviews and extends the census population base and presents a series of age and sex specific death rates for the Jews in England and Wales in 2001. It does so by relating annual average number of deaths - recorded in the community for the period 1999 to 2003 - to Jewish population distributions that draw on the census data. These bring out the specific nature of an ageing, predominantly middle-class community at the end o9f the 20th century.
Email: marlena@f-ire.com
Non-independent persons: definitions and statistical methodologies for correct identification.
Alessandra Battisti, Alessandro Solipaca
Italian National Institute of Statistics (ISTAT), Central Directorate on Social Institutions Surveys, Health & Care Service.
Population ageing is one of the main factors in the increase of disability in general and also of mental health problems. Alzheimer's disease and other pathologies such as senile dementia [Alzheimer's is a form of senile dementia, rather than a separate pathology] are closely connected with increasing longevity.
Improvements in the health care system has given more people the possibility of living longer but it has also increased the number of people in need of social and health assistance. Family, on the other hand, remains an important source of support to elderly people but it has also reduced its capacity of assistance.
For all these reasons the importance of "non-independent persons", i.e. those who are unable live independent lives because of health-related problems, is growing rapidly especially for policy makers.
This paper compares different statistical approaches to identify non-independent persons. The first step is a review of the definitions of a non-independent person used in Italian and international literature; the second is the identification of statistical methodologies applied to identify non-independent persons.
Three different approaches are implemented in this study. The first verifies a predetermined classification of functioning lost by "non-independent persons". The second approach is the application of a multivariate correspondence analysis to the microdata of a selected population with a certain degree of disability. The third approach is the application of a probit-ordered model to evaluate the impact of the different Activities of Daily Living on health status.
The results of the different approaches will be compared to identify the main dimensions of the phenomenon. The dimensions identified by all the methods are the same: feeding and preparation of food and ability to wash the hands and face without assistance.
Alessandra Battisti, Researcher, Statistical Information System on Disability - ISTAT Viale Liegi, 13 - 00198 Rome, ITALY. Phone: +39-06-46737582 Fax: +39-06-46737601. E-mail: albattis@istat.it
Alessandro Solipaca, Project Director, Statistical Information System on Disability - ISTAT Viale Liegi, 13 - 00198 Rome, ITALY Phone: +39-06-46737581 Fax: +39-06-46737601. . E-mail: solipaca@istat.it