CANADIAN STUDIES IN POPULATION

Canada (Edmonton) 08

1992 - VOLUME 19, NUMBER 1

96.08.1 - English - Fernando RAJULTON, Department of Sociology, The University of Western Ontario, London, Ontario (Canada)

Life history analysis in demography: Implications for teaching and research (p. 1-16)

In demography, a retrospective observation plan has become a common data collection procedure, and almost all surveys done - in recent times collect life history data. To justify the collection of life history information, all three of its aspects - namely, the timing, sequence and number - should be considered in a meaningful analysis. A multivariate analysis of the timing of one or two events using a number of covariates without consideration of the sequence or number of events is not, and cannot be, a life history analysis. By bringing out relevant points regarding the fundamental assumptions in life history analyses, this paper aims at contributing toward developing theories of change and procedures of estimation and testing. In particular, three stochastic models - Markov, semi-Markov and non-Markov - are discussed in detail. Other possible models, including those of diffusion as well as of unobserved heterogeneity, are suggested. (METHODOLOGY, EVENT HISTORY ANALYSIS, STOCHASTIC MODELS)

96.08.2 - English - Michael J. BERGOB, Department of Sociology, University of Victoria, Victoria, British Columbia (Canada) Where have all the old folks gone? Interprovincial migration of the elderly in Canada: 1981-1986 (p. 17-25)

Changes in provincial age structures due to elderly and non-elderly interprovincial migration in Canada were examined for the 1981-86 migration interval. Although elderly and non-elderly migration patterns diverged, non-elderly migration continued to be a better predictor of changes in provincial population age structures than did elderly migration. The rate of change in provincial elderly populations was found to be significantly correlated with net non-elderly migration rates. (CANADA, AGED, INTERNAL MIGRATION, AGE DISTRIBUTION)

96.08.3 - English - S. MITRA, Emory University, Atlanta, GA (U.S.A.) and Statistics Canada, Ottawa, Ontario (Canada)

Below replacement fertility, net international migration and Canada's future population (p. 27-46)

The indefinite continuation of Canada's below replacement fertility and emigration as well as the current policy of admitting a fixed number of immigrants every year, will eventually result in a stationary population. The size and composition of that population will be determined by the same aspects of the immigrant population and its eventual adoption of the host country's reproductive norm. All characteristics of this population may differ from the conventional stationary population in several respects. Among others, the birth rate will be less than the death rate, and, in Canada's case, the ultimate age structure may not be monotonically declining but may show more than one modal age. (CANADA, LONG-TERM PROJECTIONS, STATIONARY POPULATION, AGE DISTRIBUTION, IMMIGRATION)

96.08.4 - English - Frank TROVATO, University of Alberta, Edmonton, Alberta (Canada), and Clifford C. CLOGG, The Pennsylvania State University, University Park, PA (U.S.A.)

General and cause-specific adult mortality among immigrants in Canada, 1971 and 1981 (p. 47-80)

Much emphasis has been devoted to the adjustment and adaptation of immigrants in Canada, but we know very little about their mortality experience. In this study, we examine general and cause-specific mortality for American, English, Scottish, Italian and other foreign-born immigrants in relation to the Canadian-born population for the census periods, 1971 and 1981. Results from a Poisson regression analysis, based on log-rate models for relative mortality risk, reveal limited support for the hypothesis that the life stresses of the migration experience lead to an increased risk of death due to suicide, homicide and accidents among immigrants. 'I'he prediction that discrepancies in socioeconomic status are of importance in explaining mortality differentials gained partial support. In general, immigrants fare better than their host population with respect to the risk of overall mortality, but in 1971 they suffered increased odds of dying from neoplasms. Concerning other causes, the results tend to be mixed, but Italians, followed by the English experience the lowest levels of mortality, while the residual class of migrants (other foreign-born) and the Canadian-born tend to suffer elevated odds. The Scottish and Americans have an intermediate risk. Overall, the findings are consistent with the literature based on other immigrant societies, that the foreign-born generally experience better survival probabilities than their host countries. (CANADA, IMMIGRANTS, DIFFERENTIAL MORTALITY, CAUSES OF DEATH, PLACE OF ORIGIN)

96.08.5 - English - A. K. M. Nurun NABI, Department of Sociology, University of Alberta, Edmonton, Alberta (Canada) Dynamics of internal migration in Bangladesh (p. 81-98)

Inter-regional net migration rates for Bangladesh have data set from the 1974 and 1981 census reports. A multiple regression model of internal migration in Bangladesh has failed to support the argument that internal migratory flows are determined by high population density. The development of the agrarian social structure of this country is rooted in the historical evolution of the land tenure system. The analysis indicates support for the argument that the unequal hierarchical relations of people to the land in Bangladesh condition the process by which migration takes place. (BANGLADESH, INTERNAL MIGRATION, AGRARIAN STRUCTURE, LAND TENURE)

1992 - VOLUME 19, NUMBER 2

96.08.6 - English - Frank TROVATO, Department of Sociology, University of Alberta, Edmonton, Alberta (Canada) Mortality differentials in Canada by marital status (p. 111-143)

The sociological and epidemiological literature suggests that married people enjoy lower death rates as a result of the social and psychological benefits that marriage affords individuals. In this study, I review the literature to delineate how marital status translates into differential mortality and the mechanisms which intervene in this process. I then proceed to investigate the hypothesis that although data are not readily available to test for all the mechanisms implied in the literature, marital status's association with the postulated intervening factors would ultimately translate into observing significant mortality differentials by marital status. The data for the analysis are from the Mortality Data Base and from the 1986 Census of Canada. Eleven causes of death are examined, classified in accordance with Gove's (1973) formulation that certain causes of death reflect etiologies associated with psychological states of mind (suicide), carelessness (motor vehicle accidents), conflict (homicide), and lifestyle (heart disease, lung cancer, cirrhosis of the liver). It is shown that the married generally display the lowest odds of death, providing support for a marriage protection effect. Further evidence indicates that health selection cannot be ruled out as an additional source of mortality discrepancies. Both marriage protection and health selection appear to be important mechanisms explaining the superior survival probabilities of married people; however, the statistical results suggest that marriage protection accounts more for this social fact. (CANADA, DIFFERENTIAL MORTALITY, MARITAL STATUS)

96.08.7 - English - Carl F. GRINDSTAFF, Department of Sociology, The University of Western Ontario, London, Ontario (Canada) A vanishing breed: Women with large families: Canada in the 1980s (p. 145-162)

The purpose of this paper is to examine a selection of socio-demographic variables associated with three different parity distributions (no children, two children and four or more children) of women aged 35-39 (N=9,389) in Canada in 1981. Data are primarily from the 1981 Census of Canada, Public Use Sample Tape. The findings are examined within the context of Caldwell's theory of flow of wealth within families. The large family is clearly a "vanishing breed" in Canada, as only 15% of ever-married women aged 35-39 have four or more children. These women are distinguished by early age at marriage and low levels of human capital accumulation. Employing multiple regression, about 20% of the variance in the fertility outcome of women 35-39 is explained. Implications of the findings are discussed in the context of below replacement fertility, the changing role of women in Canada, and the economic compensation necessary to assist in stemming the flow of wealth from parent to child in the raising of children at the end of the 20th century. (CANADA, LARGE FAMILY, FERTILITY DECLINE)

96.08.8 - English - S. MITRA, Emory University, Atlanta, GA (U.S.A.) Can immigration affect age composition when fertility is below replacement? (p. 163-174)

The effect of immigration on the age composition of a population when fertility is below replacement level has been studied in the paper. Using Canadian example where the population in the working age-group accounts for more than two-thirds of the population, it has been shown that such an age composition is difficult to perpetuate over an extended period of time. Selective immigration may offer some advantage in the short run but the increase in old age dependency created by the process in the following years will render that strategy extremely inefficient in the long run. Thus a sustained pattern of immigration can only be helpful in preventing population decline or even its extinction when fertility continues to remain at below replacement level. The age composition of such a population is determined solely by its level of mortality and cannot be tampered with effectively by an immigration policy. That is to say, the equivalence of the limiting age composition to its own stationary population is virtually assured when the size and age composition of the annual immigrants remain constant. (CANADA, REPLACEMENT FERTILITY, AGE DISTRIBUTION, IMMIGRATION, STATIONARY POPULATION)

1993 - VOLUME 20, NUMBER 1

96.08.9 - English - Frank TROVATO and Dave ODYNAK, The University of Alberta, Edmonton, Alberta (Canada) The seasonality of births in Canada and the provinces, 1881-1989: Theory and analysis (p. 1-41)

Seasonality of births has been observed in virtually all historical and contemporary populations. In general, two distinct patterns have been identified in modern populations: (1) the American pattern, characterized by a trough in April-May, and a peak in September, (2) the European pattern with an excess of births during spring and summer, and a secondary peak in September. The authors analyze Census and Vital Statistics data for Canada and the provinces for 1881 and from 1926-1989. The result indicate that the Canadian pattern closely approximates the European model, but only since the early part of this century. Thus a radical change took place between the late 1880s and the early part of this century. The evolution of birth seasonality since 1926 was examined in detail as well as its explanations on sociocultural and environmental of this phenomenon. (CANADA, BIRTH, SEASONAL FLUCTUATIONS, THEORETICAL MODELS)

96.08.10 - English - Gerry HILL, Laboratory Centre for Disease Control, Health and Welfare Canada, Ottawa, Ontario (Canada) The entropy of the survival curve: An alternative measure (p. 43-57)

The "entropy" of a life table, Hk, was introduced by Keyfitz as the elasticity of the expectation of life, E, with respect to uniform change in age-specific mortality rates. Hk is not a true measure of entropy in the probability sense, though Hs = Hk+In(E) is the entropy of the distribution by age of the corresponding stationary population. A more direct measure is Hf, the entropy of the distribution of age at death. The mathematical relationships between these indexes are reviewed, and their behaviour compared using historical cohort life tables from various countries, with Gompertz functions being fitted to survival from age 30 to age 85 for each sex. It is shown that as E increases and Hk decreases, there is very little change in the "true" entropy measures Hs and Hf. Entropy and variance are conceptually similar and the variance of the distribution of the age at death, V, also changes very little. An interesting feature of the Gompertz distribution is that a simple mathematical relationship exists between Hf and the force of mortality at the expectation of life, m(E). It follows that the latter is also constant for most of the cohorts. A decrease in Hf and V and an increase in m(E) is seen for the most recent male cohorts in England and Wales. This may be due to the increasing prevalence of cigarette smoking in these cohorts. (METHODOLOGY, SURVIVORSHIP FUNCTION, VARIANCE ANALYSIS)

96.08.11 - English - Margaret MICHALOWSKI, Demography Division, Statistics Canada, Ottawa, Ontario (Canada) Redefining the concept of immigration in Canada (p. 59-84)

In Canada, temporary movements of foreigner have so far been excluded from the system of international migration statistics, and consequently, from any demographic and socio-economic analysis of the impact of these movements on the socio-economic situation in the country. It is easy to recognize however, that these foreigners have an impact on society and the economy. The purpose of this paper is to analyze the trends and selectivity among foreign temporary residents in Canada. The study covers the period between 1981 and 1990, and is confined to the following characteristics: sex, age, marital status, occupation, place of destination, length of stay, and type of authorization to remain in Canada. (CANADA, FOREIGNERS, TEMPORARY MIGRATION, SELECTIVE IMMIGRATION)

96.08.12 - English - Harry ROSENBAUM, University of Winnipeg, Winnipeg, Manitoba (Canada) Selectivity among various types of inter-provincial migrants, Canada 1976-1981 (p. 85-106)

This study examines inter-provincial migration in Canada from 1976 to 1981. Using dummy dependent-variable regression analysis and data from the individual file of the 1981 Canadian Census Plublic Use Sample Tape, the author investigates which factors best differentiate return movers from primary and onward movers. The analysis reveals that the destination region and a composite measure of marital status and family size are best able to differentiate return moves from the other types of moves; education, occupation and age are also significant factors affecting the type of move. (CANADA, INTERNAL MIGRATION, RETURN MIGRATION, MIGRATION DETERMINANTS)

96.08.13 - English - Evelyn PETERS, Queen's University, Kingston, Ontario (Canada) The household structure of status Indians in Regina and Saskatoon, 1982 (p. 107-126)

This paper contributes to work on ethnicity and the family by examining patterns of co-residence and financial support among Canadian Indians living in Regina and Saskatoon, Saskatchewan in 1982. The study examined the characteristics of heads of households with and without additional persons to explore whether or not some families were more likely to have extended households. Additional persons in family households were compared to each other in order to assess the effect of their characteristics on co-residence strategies. (CANADA, INDIGENOUS POPULATION, HOUSEHOLD COMPOSITION)

96.08.14 - English - Sylvia T. WARGON, Statistics Canada, Ottawa, Ontario (Canada) The census, demography and Quebec\: Some milestones (p. 127-137)

1991 was a very important year in the history of the Canadian Census: it was the year of the last decennial enumeration of this century, and it marked the 325th anniversary of Canada's first Census, the famous nominal census taken by Jean Talon in 1665-66. At this time, in the last decade of the twentieth century, it is fitting to remember some of the things we are inclined to forget (or perhaps that some of us never knew) about early census-taking activities in Canada. This brief summary links 1666 with 1991, by focussing attention on selected milestones in the history of Canada's feel census, noting, along the way, some related developments in the teaching of demography and in demographic research. Although the contributions of Quebec and Quebeckers are also highlighted, they are presented in appropriate national context. (CANADA, HISTORY, POPULATION CENSUSES, DEMOGRAPHY)

1993 - VOLUME 20, NUMBER 2

96.08.15 - English - Hallie J. KINTNER, General Motors Research Laboratories (U.S.A.), and David A. SWANSON, University of Arkansas at Little Rock (U.S.A.) Towards measuring uncertainty in estimates of intercensal net migration (p. 153-191)

We present a system for generating confidence intervals around estimates of intercensal net migration made using the life table survival method. The life table survival method applies a life table to a census count to project survivors at some past or future time points. Net migration is the difference between the projected number of survivors and the enumerated population. Confidence intervals for net migrants in an age-sex group are based on the probability distribution of deaths. They can be adjusted when a life table is unavailable for the population in question. The technique is illustrated with data from Puerto Rico, New Jersey, and Alaska. (METHODOLOGY, MIGRATION MEASUREMENT, INTERCENSAL MIGRATION, CONFIDENCE INTERVALS)

96.08.16 - English - Arun KUMAR SINHA and Rajiv KUMAR, Patna University, Patna (India) On the distribution of births in human populations (p. 193-205)

Brass (1958) observed that the negative binomial distribution (NBD) describes the distribution of births in human populations fairly well. However, Onyemaekeogum (1978) has contradicted this finding. Both used data from African countries although, collected at different points of time. This paper attempts to investigate the adequacy of NBD to fit data for India. The results indicate that the model both in its complete and truncated forms provides reasonably good fit to most of the observed distributions of births. The maximum likelihood (ML) estimators of the parameters of the model have been used for the fitting. (INDIA, BIRTH, STATISTICAL TABLES, BINOMIAL LAW)

96.08.17 - English - Frank TROVATO, University of Alberta, Edmonton, Alberta (Canada) Mortality differences by nativity during 1985-87 (p. 207-223)

This study extends the Canadian based literature on immigrant mortality to the 1985-87 census period. Previous work in Canada and in other countries has indicated that foreigners tend to experience a lower level of mortality in relation to their host populations. This analysis demonstrates that death rates are lower among immigrants in relation to the Canadian-born. The overall age pattern of mortality is characterized by lower death rates for the immigrants. In a multivariate analysis the main effect of nativity indicates that the Canadian-born's overall death rate is 26% greater than that of foreigners. The analysis also reveals that age interacts with nativity. Between the ages of 10 and 64, the interaction effects show a survival advantage for the immigrants over their host population. Below the age of 10, there is one statistically significant differential (for age groups 5-9), which is in favour of the Canadian-born. The post-retirement effects are significant (except for age group 65-69) and are in favour of the indigenous population. These interaction effects are interpreted in the context of the migration selection hypothesis. (CANADA, IMMIGRANTS, DIFFERENTIAL MORTALITY, PLACE OF BIRTH)

1994 - VOLUME 21, NUMBER 1

96.08.18 - English - Feng HOU and Roderic BEAUJOT, The University of Western Ontario, London, Ontario (Canada) The differentiation of determinants among return, onward and primary migrants in Canada (p. 1-19)

'I'his study examines the differences in the determinants that distinguish interprovincial primary, return and onward migrants in Canada. Using the individual file of the 1986 Census Public Use Sample, analyses are conducted both at macro or aggregate level in comparing migration across provinces, and at micro level in terms of individual propensity to undertake various types of migration. Compared to primary migrants, return migrants tend to be less educated and less economically motivated. However, it is difficult to distinguish returns due to disappointment from those involving a planned life cycle strategy. Onward migrants are more educated than primary migrants and they appear to be responding to a broader set of factors which would imply a more careful consideration of alternatives. (CANADA, INTERNAL MIGRATION, FIRST MIGRATION, RETURN MIGRATION, TEMPORARY MIGRATION, MIGRATION DETERMINANTS)

96.08.19 - English - Zenaida R. RAVANERA, Fernando RAJULTON and Thomas K. BURCH, Population Studies Centre, The University of Western Ontario, London, Ontario (Canada) Tracing the life course of Canadians (p. 21-34)

This paper examines the life courses of Canadians through an event history analysis of data from the 1990 General Social Survey on Family and Friends. The sequences and the timing of transitions into various life course stages, and the durations of stay in those stages are analyzed through multiple-decrement life tables. Changes in the life courses over 10-year birth cohorts from 1910 to 1970 and differentials by gender are highlighted in the paper. The analysis reveals that what is traditionally thought of as a "typical" life course is experienced only by one-fourth to one-third of a cohort. With the increasing diversity of life course stages among younger cohorts, it would no longer be adequate to consider only the "typical" life stages in future analyses. (CANADA, EVENT HISTORY ANALYSIS, LIFE CYCLE)

96.08.20 - English - Gugulethu GULE, University of Swaziland (Swaziland) Socio-cultural constraints to family planning in Swaziland (p. 35-49)

Swaziland's population growth rate (3.2%) is among the highest in the world. It results from high and stable fertility, and rapidly declining mortality. Strategies aimed at reducing fertility have had little success, evidenced by only a slight decline of fertility from 6.9 children per woman in 1966 to 6.4 in 1986. Contraceptive prevalence has remained low (17% among all women), and compares unfavorably with that of other southern Africa countries. This paper discusses constraints to family planning in Swaziland and proposes the incorporation of these factors into all programs aimed at reducing Swaziland's population growth and fertility rates. (SWAZILAND, POPULATION GROWTH, FAMILY PLANNING POLICY, FERTILITY DETERMINANTS)

96.08.21 - English - Dave ODYNAK, University of Alberta, Edmonton, Alberta (Canada) Age at first intercourse in Canada: Some recent findings (p. 51-70)

Age at first intercourse in Canada is investigated using national data from the 1990 Health Promotion Survey. The onset of sexual intercourse is examined at ages 15, 16 and less than 20 by current age, regional residence, gender and language spoken at home. A multivariate logistic regression analysis shows that gender differences in the onset of sexual intercourse have eroded over time in Canada. Little support was found for the hypotheses that differences in the onset of coitus in adolescence are reflected along regional lines or by subcultural groups. (CANADA, ADOLESCENCE, SEXUAL RELATIONSHIPS, AGE)

1994 - VOLUME 21, NUMBER 2

96.08.22 - English - Roderic BEAUJOT, Paul S. MAXIM and John Z. ZHAO, University of Western Ontario, London, Ontario (Canada) Self-employment among immigrants: A test of the blocked mobility hypothesis (p. 81-96)

The blocked mobility or relative disadvantage thesis posits that because of relative disadvantages experienced by immigrants in the labour market in the host society, many will turn to self-employment as an alternative to wage-labour. Part of the difficulty with previous research in this area is the failure to distinguish between self-employment in professional and non-professional occupations. On the basis of the 1986 Canadian census data, our findings show that the blocked mobility thesis is supported in the case of immigrants with high educational credentials obtained from home countries, who have a higher likelihood of turning to self-employment in nonprofessional occupations, compared to those with similar education levels among either the nativeborn or immigrants with Canadian education. (CANADA, IMMIGRANTS, LABOUR MARKET, SELF-EMPLOYED, LEVELS OF EDUCATION)

96.08.23 - English - Yihua LIN-YUAN and Leszek A. KOSINSKI, Department of Geography, University of Alberta, Edmonton, Alberta (Canada) Why emigration occurred: Chinese immigrants to Edmonton (p. 97-116)

Findings from a 1991 sample survey of recent Chinese immigrants in Edmonton, Alberta, Canada, are presented in this paper. Politics and children's education topped the list of various factors that had pushed the Chinese away from their home countries or regions. They were also the most important goals expected to be achieved through emigration. Canada's perceived multiculturalism, lower levels of racial discrimination, and good educational opportunities were reasons for more than 75% of the respondents who chose to migrate to Canada. The study also revealed the constraints and uncertainties that had been perceived by the immigrants in the process of deciding to emigrate. (CANADA, CHINA, IMMIGRANTS, MIGRATION DETERMINANTS)

96.08.24 - English - S. MITRA and Clark DENNY, Department of Sociology, Emory University, Atlanta, GA (U.S.A.) On the application of a model of mortality (p. 117-132)

A two parameter model of mortality is presented which is a simplification of an earlier three parameter model formulated by Mitra (1983). It express ln(-l(x)) as a linear function of ln(x) and ln(a-x) where x is age, a is the upper limit of life and l(x) is the probability of surviving from birth to age x. This is accomplished by constraining the model to reproduce the given value of infant mortality. The parameters measure two dimensions: the level and the pattern of mortality. The model is applied to real and model life tables and compared with Brass's logit model. The model is shown to fit as well as the Brass model without the difficulty of choosing a standard table. (LIFE TABLES, MATHEMATICAL MODELS, COMPARATIVE ANALYSIS)

96.08.25 - English - Ira WASSERMAN, Eastern Michigan University, Ypsilanti, Michigan (U.S.A.), and Steven STACK, Wayne State University (U.S.A.) Testing the deathdip and deathrise hypothesis: Ohio mortality results, 1989-1991 (p. 133-148)

In a number of social science studies Phillips hypothesizes that individuals can control the timing of their deaths prior to such ceremonial events as birthdays, religious holidays and secular holidays. He also asserts that individuals who are more integrated into the social order, such as married people, are more likely to engage in this type of behavior. This study reanalyzes the hypothesis for birthdays. The study develops a multivariate model that controls for age, gender, marital status, and education. In order to determine whether there is a death-dip before birthdays, and/or a death-rise after birthdays, the research uses dummy variables that determine whether an individual is more likely to die a week before their birthday, and/or a week after their birthday. Using Ohio mortality data from 1989 to 1991 for individuals 18 years of age and older who died of natural causes, the study estimates the coefficients in the multivariate model. Age and seasonal effects, as well as interactions between age and gender and seasonal effects, are significant explanators of the difference between deathday and birthday. However, the study also found that individuals were more likely to die in the week after the birthday. The results lend support to the death rise hypothesis. (UNITED STATES, DEATH, SEASONAL FLUCTUATIONS)

96.08.26 - English - Dilip C. NATH and Kenneth C. LAND, Duke University, Durham, NC (U.S.A.), Kaushalendra K. SINGH, Banaras Hindu University, Varanasi (India), and Pijush K. TALUKDAR, Karimganj College, Assam (India) Most recent birth intervals in a traditional society: A life table and hazards regression analysis (p. 149-164)

The fertility of the Indian rural population is close to natural. High levels of fertility exist in spite of low levels of fecundability associated with longer birth intervals. The birth rate of a female generally is equal to the inverse of her mean closed birth interval. In the present research the most recent birth interval (MRB) - which is defined as the interval between last and next-to-last birth prior to the survey date - is analyzed for the state of India. Single-decrement life table techniques and multivariate hazards regression models are applied to the data. The hazards analyses show that covariates like age at first marriage, present age of women, parity of mother, age difference between spouses, survival status of next-to-last birth, and family income have significant effects on the length of the most recent birth interval. The interactions of age of mother with the level of education, and sib size of the family have highly significant effects on the duration of MRB intervals. The level of natural fertility in this traditional society can be considerably affected by policies that foster an increased bride's age at marriage, a reduction in the infant death rate through better health care programs, improved literacy levels of mothers of the community, and increasing family incomes. (INDIA, NATURAL FERTILITY, BIRTH INTERVALS, FERTILITY DETERMINANTS)

Back to home page