SEDAP (Social and Economic Dimensions of an Aging Population) is a multidisciplinary research program studying a wide range of aging-related issues and is funded by the Social Sciences and Humanities Research Council of Canada. SEDAP is centred at McMaster University and involves researchers from that institution as well as from the University of British Columbia, Université de Montréal, Queen's and the University of Toronto.
I. SEDAP Conference
SEDAP will host a conference at McMaster University on "Macroeconomic Aspects of an Aging Population". This conference, previously scheduled for November 9, 2000, will now take place December 8, 2000. A preliminary program is given below. Further information is available from Prof. William Scarth, Department of Economics, McMaster University, Hamilton, Ont. L8S 4M4, firstname.lastname@example.org and on our website.
Session 1 - Marcel Merette (University of Ottawa), "International Consequences of Population Aging in a Three-Region Overlapping- Generations Model"
Session 2 - Robin Brooks (IMF), "What Will Happen to Financial Markets When the Baby Boomers Retire?"
Session 3 - Bill Robson (C.D. Howe Institute), "Will the Boomers Bust the Health Budget? Demographics, Service Intensity and Provincial Health Spending"
Panel Discussion 1 - Saving/Retirement Strategies for Individuals and Debt Reduction Strategies for Governments (Participants to be confirmed)
Panel Discussion 2 - Employment Issues: Participation, Work-Sharing and Immigration (Participants to be confirmed)
II. SEDAP Research Papers
SEDAP Research Papers are available on the SEDAP website at no cost. Paper copies may be obtained for a nominal charge. Please contact Mrs. Gail Kalika, Department of Economics, KTH-426, McMaster University, Hamilton, Ont., Canada, L8S 4M4.
Brief descriptions follow of the most recently released papers in the SEDAP series.
SEDAP Research Paper No. 20:
Describing Disability among High and Low Income Status Older Adults in Canada
Parminder Raina (Health Care and Epidemiology, University of British Columbia), Micheline Wong (Center for Community Health and Health Evaluation Research, Vancouver), Larry W. Chambers (Clinical Epidemiology and Biostatistics, McMaster University), Margaret Denton (Gerontological Studies and Sociology, McMaster University), and Amiram Gafni (Clinical Epidemiology and Biostatistics, McMaster University)
The authors report findings on the prevalence and severity of disability among Canadians aged 55 and over in conjunction with income status. Their data on able and disabled non-institutionalized older adults are from the 1986 and 1991 Health and Activity Limitation Survey (HALS). Data from HALS are linked to relevant census data to obtain information on income and other characteristics. The authors' sample sizes are 22,386 (age 55-64) and 38,518 (age 65+) for the 1986 survey and 11,507 (55-64) and 5106 (65+) for the 1991 survey. The authors then weight these samples to reflect the general population.
Disability is defined in the HALS as having at least one "Activities of Daily Living" restriction or a cognitive limitation. Activities of Daily Living include such activities as reading newsprint (with corrective lenses if usually worn), hearing conversations in a group, or walking up a flight of stairs. Severity of disability is also estimated in HALS. The authors use Statistics Canada's Low Income Cut-offs to classify HALS respondents as high or low income.
Among the finding of the paper are:
For ages 55-64
For ages 65+
SEDAP Research Paper No. 21:
Parental Illness and the Labour Supply of Adult Children
Pierre Thomas Léger (Ecole des Hautes Etudes Commerciales de Montréal)
As adult children often assume caregiving responsibilities, the demographic trend of a growing elderly population and increasing life expectancy may have important implications for the labour market activity of the younger generation. In this paper, Léger attempts to quantify the effect that parental illness has on the labour supply of adult children and children-in-law.
The author uses a U.S. longitudinal data set, the Panel Study of Income Dynamics and its Parent Health Supplement. This data set provides detailed information from 1975 to 1991 about one child (and spouse if present) and their elderly parents. The advantage of this data panel is that it includes information on children both before and during a parental illness and across different parental living arrangements. As well, adult children are included in the survey regardless of their caregiving status. The sample consists of 2437 pairs of adult children and elderly parents, of which 704 cases involve parents identified as being too ill to care for themselves without help. In the first year of parental illness, 38% of parents continued to live independently, 39% moved to a nursing home and 23% moved in with an adult child.
Léger finds that women experience negative impacts on their hours worked when they cohabit with a sick elderly parent. However, his results suggest that much of the negative relationship between caregiving and hours worked is due to the fact that women who work less are more likely to cohabit. The results also indicate that men may anticipate parental nursing home stays and increase their labour supply in the years prior to such living arrangements.
Léger finds that the costs associated with parental illness on families are non-trivial. Reductions in hours worked by women when parents are ill and either remain independent or cohabit are not matched by similar increases in men's labour supply, and thus families as a whole experience considerable financial losses when a sick elderly parent moves in. Estimates presented in the paper indicate that women who cohabit with a sick elderly parent lose over $4000 (1987 $U.S.) a year from reduction in hours worked.
SEDAP Research Paper No. 22:
Some Demographic Consequences of Revising the Definition of 'Old' to Reflect Future Changes in Life Table Probabilities
Frank T. Denton (Economics, McMaster University) and Byron G. Spencer (Economics, McMaster University)
Gains in life expectancy experienced in Canada and in most parts of the world would suggest that the marker for "old age" should be revised upward from the widely used "65 and over". One consequence of keeping the marker at 65 for Canada is an exaggeration of the future increase in the older population. This paper thus explores some of the implications of revising (or not revising) the definition of old age in Canada.
The authors examine the implications for the sizes and percentage shares of the "old" and the "oldest old" populations under three different definitions:
They also employ three different sets of demographic assumptions regarding mortality, fertility and migration rates:
Under the authors' Standard demographic assumption, Definition 2 reduces the projected 2041 "old" population by 16% compared to Definition 1, and Definition 3 reduces it by 12%. Definition 2 reduces the projected 2041 "oldest old" population by 37% compared to Definition 1, and Definition 3 reduces it by 32%. The effects of replacing Definition 1 with Definition 2 or 3 are reduced under the Less Rapid Population Aging assumption and increased under the More Rapid Population Aging assumption.
The authors conclude that if declining mortality rates and increased longevity are judged suitable criteria for reassessing age markers, then the probable effect in practice (given the propensity to use ages that are multiples of 5) is a greater likelihood of 70 replacing 65 as a social standard for the old age threshold and of 90 replacing 85 as the marker for oldest old.
SEDAP Research Paper No. 23:
Geographic Dimensions of Aging: The Canadian Experience 1991-1996
Eric G. Moore (Geography, Queen's University), Donald McGuiness (Geography, Queen's University, Michael A. Pacey (Geography, Queen's University) and Mark W. Rosenberg (Geography, Queen's University)
The distribution of the elderly in both absolute and relative terms is far from even across Canada, with the result that jurisdictions at all levels of government face different demands for local goods and services for the elderly in different places. The authors thus feel it is important to understand both the demographic and socio-economic processes which underlie changing patterns in population aging. In this paper, they examine how the proportion of elderly changes from 1991 to 1996 with particular reference to census divisions.
As it is the proportion of the population that is elderly that is being examined, changes which occur both to the segments over 65 as well as under 65 must be studied. The authors focus specifically on two factors which affect the proportions of these segments in a given geographic area: aging-in-place and net migration. Aging-in-place refers to the processes of change which accrue from births, individual aging and deaths in a given area. Net migration is the sum of migration into and migration out of a particular area. Significant outmigration of younger individuals will increase population aging in the originating area while outmigration of older individuals has the opposite effect.
Examining the Canadian population at the census division level, the authors find that almost half of all census divisions in Canada are experiencing significant aging-in-place which is being ameliorated to some extent by migration, while nearly 20% experience both aging-in-place and increased aging due to migration. Areas which are aging due primarily to migration are much more likely to be found in the Atlantic and Prairie regions. Areas in which aging-in-place is alleviated to some extent by migration are found primarily in British Columbia, Quebec and Ontario. Decline in aging is important only in British Columbia.
In further analysis which considers some socio-economic characteristics of the census divisions, the authors find that population aging is concentrated in areas which tend to be less economically advantaged. They find a very strong negative association between the concentration of the elderly and measures such as average income, rates of recent growth, and measures of educational attainment. They also find a significant attraction of older populations to places with warmer and sunnier climates.
SEDAP Research Paper No. 24:
The Correlation Between Husband's and Wife's Education: Canada, 1971-1996
Lonnie Magee (Economics, McMaster University), John Burbidge (Economics, McMaster University) and Les Robb (Economics, McMaster University)
In this primarily technical paper, the authors study the correlation between education levels of husbands and wives using the Canadian Survey of Consumer Finances over the period 1971 to 1996. Four education levels are considered: less than high school, high school, some post-secondary, and university degree.
Among the findings of the paper are:
SEDAP Research Paper No. 25:
The Effect of Marginal Tax Rates on Taxable Income: A Panel Study of the 1988 Tax Flattening in Canada
Mary-Anne Sillamaa (Statistics Canada) and Michael R. Veall (Economics, McMaster University)
In this paper, the authors study the responsiveness of taxable income to changes in tax rates in Canada. This responsiveness, or "tax price elasticity", is examined during the period of major personal income tax reform that occurred at the federal level in 1988. This reform reduced the number of tax brackets from eleven to four and raised marginal tax rates for some levels of income and lowered rates for others.
The data used in this study are from Statistics Canada's Longitudinal Administrative Database (LAD) over the period 1986 to 1989. The LAD is a panel data file of information as provided by taxfilers on their personal income tax returns. The LAD is a 10% sample of all Canadian taxfilers, and since currently about 20 million Canadians file a personal income tax return, the resulting sample is substantial. Two age groups are considered in the paper: those who were always between ages 25 and 64 over the sample period and those who were always over 64.
The authors' estimates in this paper suggest a tax price elasticity of about .25 for working-age individuals. They find some evidence that reported self-employment income is more responsive to marginal tax rates. When the sample includes only the senior population, the tax price elasticity for gross income is .29, not much different from that of the working-age population. There is some evidence, however, that the employment income component of gross income may be more tax sensitive for seniors, perhaps because those seniors who remain in the workforce have more options for changing their labour force hours (including either retiring or postponing retirement).
SEDAP Research Paper No. 26:
The Stability of Self Assessed Health Status
Thomas F. Crossley (Economics, York University) and Steven Kennedy (Economics, Australian National University)
Self assessed health status is an increasingly common measure of health in empirical research. This paper investigates whether self assessed health status is reported differently under different survey settings.
The authors use the 1995 Australian National Health Survey, conducted from January 1995 to January 1996. Data in the survey are representative of the entire Australian adult population. All respondents were asked to rate their health as excellent, very good, good, fair or poor. A randomly chosen "treatment" group answered this as the first question on a written health and well being questionnaire, then again after completing an interviewer administered health and well being questionnaire and answering some other non-health related questions. The "control" group completed only the interviewer administered portion of the survey and answered the health question as the last question of the interview.
The authors find a statistically significant difference between the distribution of responses by the control group and of responses to either question asked of the treatment group. Among their findings are:
The authors suggest that possible explanations for their findings might be that individuals respond more candidly to sensitive questions when self completing a form as opposed to being personally interviewed, or that respondents in the treatment group "learn" about their health status between the first and second assessments as they answer detailed health questions.
SEDAP Research Paper No. 27:
How Do Contribution Limits Affect Contributions to Tax- Preferred Savings Accounts?
Kevin Milligan (Economics, University of Toronto)
A number of countries allow preferential tax treatment for contributions made to special savings accounts, such as 401(k) plans in the U.S., Personal Pensions in the U.K., and Registered Retirement Savings Plans (RRSPs) in Canada. In this paper, Milligan addresses one aspect of the decision to contribute, specifically the influence of contribution limits.
The author begins by constructing a three-period life-cycle model in which a consumer saves for retirement through two assets, each receiving different tax treatment. In this basic model, when future contributions to the tax-preferred asset are constrained by a contribution limit, current contributions increase to make use of the contribution room while it is available. (Milligan calls this the "use-it-or-lose-it" motivation.) An extension to this basic model allows a carry-forward provision under which unused contribution room for the tax-preferred asset in one period may be used in the next period.
Milligan's basic model predicts that current RRSP contributions will move inversely with future contribution limits. That is, as future limits increase, current contributions will decrease. To test the model, he uses panel data on a random sample of Canadian taxfilers from 1987 to 1991. Four groups of taxfilers are defined who receive different treatment under the RRSP reforms of 1990 (pension income rollovers into RRSPs no longer allowed) and 1991 (various changes in contribution limits), and their contributions are compared with the contributions of taxfilers outside these groups.
Regression results with this data set show a strong negative relationship between future contribution limits and current contributions, consistent with the basic model. When the extension to the basic model is incorporated to allow carry-forward, the use-it-or-lose-it effect on contributions diminishes and in some cases can disappear.