Fertility of Canadian Men: Levels, Trends, and Correlates

P C 1183-7284 Fertility of Canadian Men: Levels, Trends, and Correlates Zenaida R. Ravanera 0 Rajulton Fernando 0 0 University of Western Ontario Recommended Citation - Article 1 Follow this and additional works at: https://ir.lib.uwo.ca/pscpapers by Zenaida R. Ravanera Fernando Rajulton Discussion Paper no. 03-06 May 2003 On the web in PDF format: http://www.ssc.uwo.ca/sociology/popstudies/dp/dp03-06.pdf Paper presented at the 2003Annual Meeting of Population Association of America Minneapolis, Minnesota, May 1-3 Zenaida R. Ravanera and Fernando Rajulton I. Introduction In Canada, as in many Western countries, men’s fertility has not been extensively studied as fertility or reproduction has been mainly taken as women’s concern (Goldscheider and Kaufman, 1996). However, recent socio-demographic changes have put men’s fertility to the fore. These changes include high divorce and cohabitation rates, increased participation of women in the labour force, and the shift from family to individual wage rates, each affecting gender division of labour that places more responsibility on men for children rearing. This paper aims at a better understanding of the roles that men play in reproduction and consists mainly of two parts. The first describes the levels and timing of men’s fertility and the second explores factors affecting fertility. Before these two parts, we discuss the data and methodology. The final section mentions further research on men’s fertility that we intend to do. II. Data and Methodology The study uses the 1995 General Social Survey of the Family that collected retrospective data on individuals and their families including marital and fertility histories from 10,750 men and women aged 15 and older residing in Canada, excluding residents of institutions and the Territories. In most of our analysis, we focus on the sample of men born in 1921 to 1970 (aged 25 to 74 as of the survey date) consisting of 3,930 respondents. We make use of both the main and children’s public use micro-data files. The former contains information on the respondents and the latter has information on their children, in particular, the age of respondent at the birth of each child, the information necessary for estimating the levels and timing of fertility. For estimating levels of fertility, we computed the age-specific fertility rates for each of the 10-year birth cohorts as follows: From the children’s file, we obtained the number of births by 5-year age groups (15-19, 20-24 ..., 60-64) for each cohort of men and women. Since each respondent, regardless of marital status, was asked about his/her children, we divided the number of births in each age group by the number of respondents in each cohort (obtained from the main file) to get estimates of age-specific fertility rates for both men and women. The discussion of results focuses mainly on fertility of men with women’s fertility used mainly for comparison of completeness of reported births. For summary measures of the timing of fertility, we made life tables of age at birth of first child for each of the 10-year birth cohorts of men, which provided the median ages at first birth and the cumulative proportion surviving the risk of first birth. Sample weights were used for this procedure1. 1 Weights were not used for the estimation of the age-specific fertility rates as no weights were included in the children’s file. To explore the effects of certain factors on the timing of first birth, we did hazards analysis (Cox regression procedure) of age at first birth using a set of explanatory variables, all of which are categorical, except for the factor score used to indicate familyoriented values. The theoretical framework, models and the variables are discussed below. III. Fertility Levels and Timing Gender Difference in Fertility Figures 1 and 2 present estimates of age-specific fertility rates. These estimates are most likely under-estimated with under-estimation more serious for men than for women. About 5% of children had missing information on the age of respondent at birth of child. Moreover, like the survey data on fertility in Britain and the United States, the1995 General Social Survey (GSS) suffers from under-count particularly of non-marital births and births in unions subsequently dissolved (Rendall et al, 1997). Juby and Le Bourdais (1999) contend that the under-reporting in the 1995 GSS is greater than the 2% usually cited in the literature and that this is more severe in the case of fathers who do not have frequent contacts with non-(co)resident children. To get better estimates, we await the release of the more recent survey on the family, the 2001 GSS, which we expect to have more complete data on children. In spite of this limitation, the results of our analysis are adequate for interpreting general trends. Figure 1 shows the age-specific fertility rates of men and women born in 1931-40 and the total fertility rates for three 10-year birth c (...truncated)