1The life tables are decennial until 1931-40 and quinquennial beginning with 1941-45. Of these two tables, two versions exist: one based on all deaths and another excluding deaths from war operations. As our main interest in the present study is to describe the mortality transition, we are using the tables from which war deaths are excluded. For earlier wars this distinction was not made and all deaths are included.
2Period life tables have often been criticized, even dismissed, for being hybrids because nobody lives his life under one period regime. With equal justification, a cohort life table can be called a hybrid because it depicts a mortality that never existed. Whatever the effects that may be observed in cohort analysis, there is hardly any doubt that the successive stages of mortality transition originated in period events. The varying interpretations of the causes of the secular transition are all based on period factors. In the antibiotics stage the cause is obvious. An analysis of more than 20 low mortality countries has shown that all changes during the fourth stage were simultaneous in all old age groups proving that the development was not caused by supposedly healthier cohorts advancing in age but by something that happened on the time axis . Cohort effects, if any, have in all these cases been overwhelmed by period factors.
3The rank correlation coefficient between infant mortality and low socio-economic status for Finnish provinces was in 1871-80 an insignificant + 0.059 rose gradually to + 0.87 in 1931-40 and disappeared again into 0.07 in 1979-83 .
|Finnish Life Tables since 1751
Väinö Kannisto, Oiva Turpeinen, and Mauri Nieminen
© 1999 - 2000 Max-Planck-Gesellschaft ISSN 1435-9871