5 Discussion and conclusions

View the most recent version.

Before discussing the results of this research there are several limits that have to be acknowledged. For instance, when looking at the needs of the disabled elderly population, this research is not measuring unmet needs. For example, if someone is receiving assistance strictly from the informal network, but also stating not receiving enough assistance, one could argue that there is a need for the formal network to provide some assistance to this person. The projections could hardly account for this pressure on the formal network. Among the characteristics that are accounted for when estimating the probability of receiving assistance from different sources, we are not looking at the health status and the geographic proximity of children. The latter would seem to be more of a problem as having a child is not merely enough to assume he or she will be able to provide assistance. Among the many factors that may affect the probability for a child of providing assistance to a disabled parent, geographic proximity is certainly an important one. This information was not collected in the 1996 General Social Survey.

If the health status of the children is more or less important considering that the majority of them have not reached old age yet, it certainly would be important to know the health status of the spouse. The spouse being the main provider of assistance for those living with a partner, information regarding their health status would surely help in understanding why some disabled elderly receive assistance from their spouse while others, living with a partner, do not receive assistance from this source. This is especially true when the person in need of assistance is among the oldest age group. The spouse is likely very old and may not be in a position of providing assistance. However, the information on the health status of the spouse is not collected in the 1996 General Social Survey. This is why "age of spouse" has been included in the model on source of assistance. Although this is just a proxy we can assume that those in the 65 to 74 age group are probably doing better regarding health than those in the 85 and over age group. The results of the analysis give an indication of the importance of knowing the health status of the spouse when studying factors associated with the use of different sources of assistance.

The research is also limited in its interpretation by the fact that the number of hours of services provided by different sources of assistance was not studied. By simply projecting the number of disabled elderly using a given support network, it is impossible to look at the impact on human resources needed to respond to the demand and to the overall cost of such a demand. The results of this research provide valuable information on future trends regarding the availability of children and spouses for providing assistance, the number of disabled elderly in need of assistance, and the potential use of formal and informal networks. Population aging being in good part the result of lower fertility, one would expect that the proportion of elderly without any surviving children will increase in the future. However, the results show that in the near future the proportion of those without any surviving children among those with greater needs for assistance, the oldest old, will be lower than what has been observed in the recent past. This trend shows that until 2021, this proportion will decrease down to 16% (versus 22% in 2001). Even though the proportion will rise after that point, only in 2036 will it raise to the level observed in 2001.

As far as the living arrangement of elderly persons, trends indicate that among those living in private households the proportion living alone should be relatively stable, even decreasing somewhat among females in the longer term. All things being equal, a greater proportion of elderly persons living with a spouse would relieve some of the pressure on the formal network. This positive impact might however be neutralized because of aging spouses not being able to provide adequate assistance because of failing health. This positive effect could be dampened in part when joint survivorship is also meaning joint disability.

Looking at disability, of course the major driver of the number of disabled elderly is the aging of the baby boomers. With constant probabilities of disability over time, the proportion of disabled elderly between 2001 and 2031 will be just below the proportion observed in 1996. This is in part related to the arrival of the baby boomers within the older age group. At first they are among those aged 65 to 74, an age group with lower probability of disability. It is also related to their improved level of education, a characteristic having a positive effect on health. Overall, for the period 2001 to 2031, the average annual growth rate of the number of disabled elderly needing assistance could be about 2.5%. One important result of the projection was to show how significant an improvement in the health of the population could be on this number. According to the hypothesis presented earlier, this annual rate could be as low as 1.9%.

The major advantage of the microsimulation model used in this research was to allow the possibility of accounting for all these changes in the population – disability, family network, personal characteristics – to look at future trends in the use of the formal and informal networks. Among all the trends that were projected, it is clear that the pressure on the formal network will be greater in the future. Of course, assumptions on the future health of the population are very important, so are the assumptions regarding institutionalization rates among the elderly population. This increasing pressure would definitely command an increase in the resources available for home care services, be it from the public or private sector. The question of availability of human resources for home care services will be an issue in the coming years as the baby boomers reach old age.