Alain Bélanger

Active life among the elderly in the United States: multistate life-table estimates and population projections.

Calculations of multistate life expectancy not only measure how long a population may live beyond a certain age, but also what fractions of this continuing lifetime will be spent in an independent or dependent status. Many Americans aged 70 and over are leading long, active lives; large numbers of individuals who become dependent, moreover, do so temporarily and return to independent status. Men and women have disparate total and active life expectancies, however, reflecting differential survival patterns and varying rates of transition among statuses. Policy makers must consider the increased size of the future elderly population, and changes in its age composition and functional status, when planning relevant health services.

Disability-Free Life among the Elderly in the United States: Sociodemographic Correlates of Functional Health

This article examines the sociodemographic factors associated not only with the length but also the quality of life. Although much research has examined dependency or the reliance on other individuals or institutions, this analysis instead focuses on difficulty in performing activities of daily living. The authors employ logit analysis and the Longitudinal Study of Aging data set to examine the sociodemographic factors-age, sex, race, education, poverty status, and marital status-that affect the prevalence of disability and the transitions that arise between the able and disabled states. Findings show that age is linked with disability: Older respondents are more likely to be or to become disabled and less likely to recover from their disabilities. Education is also strongly linked with disability: More highly educated elderly consistently enjoy less disability; if disabled, they generally have greater chances to regain their abilities. Race and sex are key variables: Disability varies by race and sex groups; further, within race/sex subpopulations, disability is differentially affected by other covariates, especially marital status and poverty. Therefore, it is crucial to examine disability not only for the total elderly population but to appreciate different disability characteristics among race- and sex-specific subpopulations. These social, demographic, and economic factors identify characteristics that place the elderly at risk of disability, and they also identify those characteristics that enhance an individuals chances of successful aging.

Gender differences in disability-free life expectancy for selected risk factors and chronic conditions in Canada.

SUMMARY This article shows how mortality and morbidity patterns differ for women and men 45 years of age and older. The impact on disability-free life expectancy was calculated for selected risk factors and chronic conditions: low income, low education, abnormal body mass index, lack of physical activity, smoking, cancer, diabetes, and arthritis. For each factor, the expected number of years free of disability was calculated for men and women using multi-state life tables. In terms of disability-free life expectancy, the greatest impacts on affected women were for diabetes (14.1 years), arthritis (8.8 years), and physical inactivity (6.0 years), while for affected men, the greatest impacts were for diabetes (10.5 years), smoking (6.9 years), arthritis (6.5 years), and cancer (6.4 years). The implications of these results are discussed from the perspective of developing programs designed to improve population health status.

New approach for constructing Canadian working life tables, 1986-1987*

Working life tables are a useful analytical tool in estimating indices such as the frequency of entries and exits from the labour force and the average length of individual working life. The latest Canadian working life tables have been published by Statistics Canada and were based on 1971 data. Since then important transformations have occurred in the pattern of Canadian labour force participation rates. Females have substantially increased their participation in the labour force and earlier retirement has reduced the participation rates of older male workers. This paper presents results from the first application of multi state methodology to Canadian labour force data, making use of a new data set - the Labour Market Activity Survey. Comparison between conventional and multistate approaches reveals that the latter approach, besides providing estimates of female work patterns, gives more reliable estimates than the conventional model.

Using Microsimulation to Reassess Aging Trends in Canada

Le vieillissement de la population est l’enjeux démographique du XXI ième siècle et plusieurs indicateurs sont utilisés pour en mesurer le niveau et les tendances. Dans Science (2010), Sanderson et Scherbov ont suggéré des améliorations à la mesure du rapport de dépendance des personnes âgées. Ils ont identifié plusieurs limites à l’utilisation de l’âge chronologique comme la principale variable et ont proposé un nouvel indice, le rapport de dépendance des adultes avec incapacités, défini comme le nombre d’adultes ayant une incapacité qui ont au moins 20 ans, divisé par le nombre d’adultes du même âge sans incapacité. Ils ont utilisé la méthode de Sullivan, basée sur la prévalence, en multipliant des taux d’incapacité dérivés à des projections démographiques de niveau macro. Les résultats ont été calculés pour plusieurs pays de la CEE et de l’OCDE. Les résultats pour le Canada (voir l’annexe en ligne) ont été calculés en utilisant les coefficients de l’Italie. Cependant, l’incapacité est un processus complexe et multidimensionnel (voir Carrière et al, 2007; Légaré et Décarie, 2011), et la microsimulation peut tenir compte de cette complexité implicite. Nos résultats pour le Canada, présentés ici, sont supérieurs à ceux de Science, et indiquent comment des projections plus élaborées des personnes âgées avec incapacités peuvent améliorer l’analyse. Nous avons utilisé LifePaths, un modèle de microsimulation de Statistique Canada, pour fournir une perspective du phénomène du vieillissement impossible à obtenir en utilisant des méthodes basées sur la prévalence.Population aging is the population issue of the XXI century and many indices are used to measure its level and pace. In Science (2010), Sanderson and Scherbov suggested improvements to the measure of elderly dependency ratio. They identified several limitations to the use of chronological age as the main variable and proposed a new index, the Adult Disability Dependency Ratio, defined as the number of adults at least 20 years old with disabilities divided by the number of similarly aged adults without disabilities. They used the Sullivan prevalence-based method by multiplying derived disability rates to macro population projections. They showed results for several ECE and OECD countries; results for Canada were derived using coefficients of Italy. However, disability is a complex multidimensional process (see Carrière, Keefe, Légaré, Lin, & Rowe, 2007; Légaré and Décarie, 2011), and microsimulation can take into account its implied complexity. Our results for Canada, presented here, exceed those in Science to show how more-sophisticated projections of disabled older adults can improve the analysis. We used LifePaths, a Statistics Canada’s microsimulation model, to provide a perspective of the phenomena unobtainable with prevalence-based methods.

A Microsimulation Model to Study the Interaction between Fertility and Union Formation and Dissolution: An Application to Canada and Quebec

Union formation and dissolution are among the main determinants explaining variations in fertility. Compared to the rest of Canada, Quebec’s marital histories are more complex and its prevalence of common-law unions much higher. The objective of this article is to examine the role of marital behaviours on fertility by comparing different indicators of fertility and conjugal life that were obtained through microsimulation. Parameters of the microsimulation model were estimated from hazard regressions performed on the marital and fertility histories collected in two retrospective longitudinal surveys: the Canadian General Social Survey (GSS) 2001 and 2006. Results show that the more complex marital histories of Quebecers can explain more than one-quarter of their fertility differences with the rest of the country.

The Life Table

With the help of the Modgen wizard, we have developed and successfully run our first microsimulation program. This simple model was able to calculate life expectancy for a cohort whose risk of death remains constant throughout the simulation. For demographers, and probably for most users, this unrealistic mortality hypothesis is unsatisfactory. As a minimum, even in a simple cohort model, mortality should vary with age and sex, thus enabling us to calculate all the components of a conventional life table.

The Base Population

This fifth chapter will deal with the integration of a base population into a Modgen microsimulation model, and as such will be more technical than conceptual or demographic. Up to now, the simulated population has been synthetic; all the actors and their characteristics have been generated by the model itself. A synthetic population requires statistical distributions to be created, describing the distribution of characteristics in the population. An example of this is the distribution of births by province which we included in Chap. 3. However, it may be more practical to use microdata taken directly from a real population as a starting point for the simulation. Rather than generating random characteristics for the actors, such as birth province or sex, we will be using information contained in a database representing an actual population for our simulation. In this chapter, data from the National Household Survey (NHS) will be used to make a projection of the population of Canada starting from 2011 (the year of the survey). We will start by giving a brief description of how to prepare and format the data file to be used for the base population. Next, we will see how to incorporate an external class into the Modgen project in order to read the datafile and import its content. Finally, we will modify the main file and the Start function so as to incorporate the import functions into the model code and initialize actor states using the imported data.

Creating a Basic Cohort Model

In this chapter our main aim is to introduce you to the Modgen software and to the Visual Studio programming software, guiding you through the creation of a first basic model. This model will be based on one single type of event: death. Although this will yield only a small number of results, such as life expectancy at birth, it will allow us to demonstrate and to practise the use of several of the basic ideas in Modgen. We will move on to the calculation and presentation of other elements of the life table in the next chapter.

The Multiple Increment-Decrement Life Table

In the mid 1970s, Professor Andrei Rogers (1975) proposed adapting the life table methodology for regional analysis. The classic demographic approach often abstracted migration, which was considered to be a disrupting phenomenon. Rogers had the idea that migration could be taken into account explicitly by using a multi-regional table which would make it possible to calculate, for example, the share of life expectancy spent in a specific region (or country). The multi-regional model he developed is now an integral part of the demographer’s toolkit.