The estimation of total family size of occupation groups from the distribution of births by order and duration of marriage

Abstract

The legitimate maternities registered in England and Wales in I939 were tabulated by the General Register Office by occupation of father in conjunction with age of mother, order of birth and duration of marriage. This material has been described in two previous papers.' It was shown that a picture of the pattern of fertility among occupation groups can be derived from the distribution of births by order. The methods used, however, do not yield absolute measures of occupational fertility. Thus it was shown that the fertility of a 'Skilled Manual Worker' was higher than that of 'Foremen' but lower than that of 'Unskilled Workers'; but whether the fertility of 'Foremen' was 6o or 8o % of that of the unskilled it was not possible to say. It is clear, of course, that it is not possible from an analysis of the distribution of legitimate births by birth order and duration of marriage to derive any measure of fertility which takes account of differences between occupation groups in nuptiality or in childlessness. The only thing we can hope to achieve is some index relating to the average number of births in families in which there is at least one child. The term 'family size' will be used here to denote the mean number of births to fertile marriages. If there were in all occupation groups a constant number of marriages every year, and if these marriages were subject to constant rates of childbearing and rates of dissolution by death or divorce, and if, when once married no one could change his occupation, then the ratio of all the births2 occurring in a given year to first births in that year would give the average family size. For the total number of births of all orders in such a hypothetical population would be equal to all the births produced throughout their existence by the marriages occurring in one year. Similarly, the number of first births in a year would equal the total of first births produced throughout their existence by the marriages occurring in one year, and the total of first births is the same as the number of marriages which are fertile. Unfortunately, as pointed out in a previous paper, the period before I939 (the