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 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. 38:
Social-psychological and Structural Factors Influencing the Experience of Chronic Disease: A Focus on Individuals with Severe Arthritis
Peri J. Ballantyne (Pharmacy, University of Toronto), Gillian A. Hawker (Medicine, University of Toronto) and Detelina Radoeva (Institute for Human Development, University of Toronto)
This study is based on interviews with a convenience sample of 29 participants, aged 59 to 86, all of whom were assessed as having severe arthritis, were assessed as appropriate candidates for total joint arthroplasty (tjr) and who rejected the tjr option after being given detailed information on its risks and benefits. (The authors note that of the 29 participants, only one had a physician who recommended tjr.) Interviews were focused around the general question, "how do aging individuals' perceptions and experiences of arthritis influence health care decision-making and management of illness?".
Analysis of the interview transcripts resulted in the formulation of a model integrating micro-, meso- and macro-level processes that contextualize the experience of arthritis and influence decision- making regarding its management. In the discussion of this model, the authors highlight the following points:
SEDAP Research Paper No. 39:
The Extended Self: Illness Experiences of Older Married Arthritis Sufferers
Peri J. Ballantyne (Pharmacy, University of Toronto), Gillian A. Hawker (Medicine, University of Toronto) and Detelina Radoeva (Institute for Human Development, University of Toronto)
In this paper, the authors seek to examine the relationship between the quality of the marital relationship and an individual's perceptions and experience of arthritis as well as the individual's views of total joint arthroplasty (tjr) as a treatment option. The sample is as discussed above for SEDAP Research Paper No. 38, except that it is reduced to include only 16 persons classified as married.
The authors identify three types of marital relationships. Synchronous relationships involve the sharing of activities as well as a shared state of mind. The independence relationship is characterized by separate and distinct emotional spaces for members of the couple. In dependence relationships, there is a limited capacity for reciprocity between the members of the couple due to the functional status of one. It is noted that while synchronous relationships accommodate the physical state of decline and make it manageable so that individuals delay or avoid strategies such as tjr, independence relationships are potentially problematic, as least from the perspective of continued and separate functioning.
The authors find three important implications in this study:
SEDAP Research Paper No. 40:
A Comparison of Alternative Methods to Model Endogeneity in Count Models: An Application to the Demand for Health Care and Health Insurance Choice
Martin Schellhorn (Institute of Economics, University of Bern and Department of Economics, McMaster University)
The effect of the choice of different levels of health insurance deductibles on the demand for physician visits in Switzerland is examined here. In Switzerland, basic health insurance is mandatory. The insurance premium varies by region of residence but is independent of income and risk factors such as age. The insured face a minimum annual deductible for ambulatory health services, but each year are given a choice of higher deductibles in return for reduced premiums. Schellhorn uses recently developed econometric methods to separate out incentive-induced behavioural changes (that is, the effect of higher deductible payments on physician visits) from selection effects (that is, insurees who would make fewer physician visits in any case are also likelier to choose a higher deductible/lower premium health insurance policy).
Data used are from the 1997 Swiss Health Survey. A sample of 4057 men is selected; women are excluded from the analysis here as all pregnancy-related health care costs are exempt from deductible payments. Of this sample, 3199 are classed as having low deductible payments and 858 as high deductible. In general, the probability of being high deductible decreases with age.
The author finds that most of the observed lower utilization of physician visits can be attributed to self-selection of individuals into insurance contracts with higher deductibles. He notes, however, that the results should be interpreted with caution, as the reforms which first mandated the purchase of basic health insurance were introduced in 1996 and the survey data used are from 1997. It is suggested that more time may be required for the insured to understand the new system and to choose an optimal deductible.
SEDAP Research Paper No. 41:
Wealth Accumulation of U.S. Households: What Do We Learn from the SIPP Data?
Vincent Hildebrand (Department of Economics, York University and IRISS, Luxembourg)
The author begins by arguing that the dramatic change in the age distribution of the population in the western world is threatening traditional state-run unfunded pension systems which have been so successful in sharply reducing poverty rates among the elderly. An alternative which seems to appeal to many governments is a system which relies on individuals' own contributions. An assessment of such an alternative requires that we know whether individuals or households are farsighted enough to plan effectively for their own retirement. A forward-looking individual should save during working life and smoothly run down assets upon retirement.
This paper examines the issue of wealth accumulation over the life cycle using a time series of independent cross-sections from the U.S. Survey of Income and Program Participation (SIPP). Unlike other U.S. wealth-holding surveys which oversample the wealthy, the SIPP has the advantage of being representative of the U.S. population. SIPP was conducted in 1984, 1985, 1986, 1987, 1990, 1991, 1992 and 1993.
The author does not find any evidence that households are depleting their assets upon retirement. However, because the definition of household wealth in the SIPP data does not include Social Security and occupational pension wealth, this conclusion must be tempered. Rather than the absence of dissaving by the average household in the data, the author feels it is the lack of substantial wealth holding by a large number of households which is notable. Descriptive results confirm that a large number of households reach retirement having saved very little. Again, it could be argued that once Social Security and occupational pensions are included in the definition of wealth, adequate levels of retirement saving may be found. Hildebrand concludes by noting that about 30 per cent of elderly households rely primarily on Social Security for retirement income and points out that whether these households would be able to respond adequately to future reductions in Social Security payments is the important issue that needs to be addressed.
SEDAP Research Paper No. 42:
Pension Portability and Labour Mobility in the United States: New Evidence from SIPP Data
Vincenzo Andrietti (Department of Economics, Universidad Carlos III, Madrid and CeRP, University of Turin) and Vincent Hildebrand (Department of Economics, York University and CEPS- INSTEAD)
Pension portability may be defined as the capacity of workers covered by an employer-provided pension plan to carry the actuarially fair value of their accrued rights from one job to the next. In a traditional defined benefit plan, each employee's future benefit is determined by a specific formula and the plan provides a nominal level of benefits upon retirement. Alternatively, defined contribution plans provide for periodic contributions into an individual pension account for each worker, and the level of benefits at retirement is determined by the total amount of these contributions plus the rate of return on them. In general, defined contribution plans are portable and workers can change jobs without suffering any loss in future pension benefits. This is generally not so for defined benefit plans.
The authors examine the role of employer-provided pensions on job mobility in the United States using the Survey of Income and Program Participation (SIPP). They use the survey years 1984, 1986, 1990 and 1992, for which detailed information on pensions is available. These survey years also span the time frame of U.S. Tax Reform Act of 1986 (which took effect Jan. 1, 1989), which reduced the maximum vesting period for employer-provided pensions from 10 years to full vesting after 5 years or graded vesting over 3 to 7 years. These changes are seen as affecting defined benefit pension plans as most defined contribution plans allow for immediate vesting of employee contributions.
The authors find that workers covered by defined benefit pensions are significantly less likely to move. However, the potential portability loss arising to workers leaving a defined benefit plan does not seem to play a significant role in explaining job mobility choices. The results also show that defined contribution plans, despite their complete portability, are as important as defined benefit plans in reducing job mobility. From a policy perspective, these results cast doubt on the effectiveness of reforms aimed at improving labour market efficiency through portability measures.
The authors conclude that a more convincing argument in favour of increased pension portability would be to ensure retirement income adequacy for multiple job changers. The effect of the reduction in the vesting period implemented with the Tax Reform Act clearly illustrates the point. Although they found that the reform did not affect mobility, the average pension loss of workers affected by the reform was reduced by 46 per cent. On the other hand, they note that one may question the need to increase pension portability since pension-covered jobs are also associated with higher remuneration levels. If the concern is the adequacy of pension income after retirement, a more equitable policy goal may be to address the observed decline in pension coverage.
SEDAP Research Paper No. 43:
Exploring the Effects of Population Change on the Costs of Physician Services
Frank T. Denton (Department of Economics, McMaster University), Amiram Gafni (Clinical Epidemiology and Biostatistics, McMaster University) and Byron G. Spencer (Department of Economics, McMaster University)
Population aging is common throughout the industrialized world and every aspect of the health care system is affected in some way by such changes in the population, including the requirements for physicians, nurses and other skilled personnel, the demand for hospital beds and equipment, and the need for long-term institutional care facilities and for home care services for the elderly. In this paper, the authors focus on the effects of population change on the overall costs of physician services.
The data set used in this paper is an unpublished one and was provided by the Canadian Institute for Health Information. It was compiled from records of the Ontario Health Insurance Plan (OHIP) and gives detailed data on billings by fee-for-service physicians in the province of Ontario, classified by age and sex of patient and by category of physicians for the fiscal year 1995-96. (Fee-for-service billings represent 98 per cent of all payments received by physicians in Ontario in 1995-96.) Statistics Canada population data for 1995 are also used. The authors construct profiles representing the per capita costs of services provided to the male and female populations, by five-year age groups, for each physician category, and apply those profiles to study the past and future effects of population change in Ontario. These profiles are also applied to the projected populations of fifteen selected countries and to four theoretical populations.
The authors distinguish two effects on the rate of increase of overall physician costs, acting in opposite directions. On the one hand, population aging tends to raise the rate of increase, while on the other hand, slower population growth tends to lower it. In Ontario as well as in the fifteen industrialized countries examined, the overall rate of increase resulting from population change is projected to be lower in 2020-2040 than in 2000-2020. The calculations done for the four theoretical populations demonstrate that, in the extreme, population aging has the potential to produce very large differences in overall physician costs per capita from one population to another. However, the authors feel the extremes are unlikely to be realized in real-world situations, and moreover, that strong demographic pressure on costs is likely to produce compelling incentives to find alternative, less costly ways of delivering health care services.
SEDAP Research Paper No. 44:
Reflexive Planning for Later Life: A Conceptual Model and Evidence from Canada
Margaret A. Denton (Gerontological Studies and Department of Sociology, McMaster University), Susan French (Faculty of Health Sciences, McMaster University), Amiram Gafni (Faculty of Health Sciences, McMaster University), Anju Joshi (Gerontological Studies, McMaster University), Carolyn Rosenthal (Gerontological Studies and Department of Sociology, McMaster University) and Sharon Webb (Gerontological Studies, McMaster University)
In this paper the authors present a model to describe the resources individuals have for achieving independence and financial security in later life. Their approach is a holistic one, integrating the financial and non-financial, individual and collective preparations made by the individual and the state.
Three types of preparations for later life are outlined. The first, public protection, includes the benefits provided by the state to reduce for individuals the social risks of poverty, homelessness, disease and disability. The second, self insurance, includes the financial preparations that individuals make to insure against poverty and/or to maintain their standard of living in later life. The third, self protection, includes the nonfinancial activities or preparations that individuals make to avoid disease and disability, maintain high physical and cognitive function and sustain engagement in social and productive activities.
To assist in developing their conceptual model, the authors studied a convenience sample of 240 seniors in Canada. Respondents ranged in age from 55 to over 90 and two-thirds were female. Nearly half the sample was made up of retired teachers or educational administrators, although the authors note that others from all walks of life were included.
The authors conclude the paper by developing a set of research questions to be addressed, based on their model. These questions are:
SEDAP Research Paper No. 45:
Time Series Properties and Stochastic Forecasts: Some Econometrics of Mortality from the Canadian Laboratory
Frank T. Denton, Christine H. Feaver and Byron G. Spencer (Department of Economics, McMaster University)
This technical paper is an exploratory study in the time series modeling of mortality rates and the stochastic forecasting of life expectancies. Working with Canadian time series, the authors investigate several methods of introducing probabilistic elements into the modeling of mortality and the quantification of uncertainty about future life expectancies. Their aim is not primarily to produce predictions but to examine possible ways of making predictions.
The principal data set used is a set of annual age-sex group mortality rates for the period 1926 to 1996, based on numbers of deaths and population figures from various Statistics Canada sources. The time series behaviour of the rates is analyzed and the data and analysis are used as a basis for the subsequent forecasts of life expectancies. The paper begins by constructing overall summary measures of mortality, then examines the time series patterns of the age group mortality rates, then moves on to stochastic forecasting.
Even in stochastic forecasting the question arises as to whether the distribution of past rates of change (that is, declining mortality rates) can be viewed as a stable probability distribution for the purpose of making probabalistic statements about the future of mortality and life expectancies. The problem is that mortality rates at all ages will approach zero asymptotically, so sooner or later the pace of decline must be reduced. For the mortality forecaster, the question is then, "how soon is sooner?". In the stochastic forecasts presented in this paper, it is assumed that the possibility of a systematic long-term drift toward zero means in the distributions of rates of mortality decline can be ignored for the 21st century.
The results of seven stochastic forecasts of life expectancies at various ages and over various time horizons are presented, based on seven different methods. For example, starting with a mean life expectancy at age zero in the year 2000 of 78.98 years, comparable results for the year 2100 range from 88.68 to 98.54 years. The authors note that all results should be regarded as experimental.
SEDAP Research Paper No. 46:
Linear Public Goods Experiements: A Meta-Analysis
Jennifer Zelmer (Canadian Institute for Health Information, Toronto)
Roads, police services, and projects to improve air quality are examples of public goods, commodities for which use of the good by one person does not preclude its use by others. With these types of goods, if one person provides a unit of the good, all will benefit. An important question, therefore, is under what conditions individuals will voluntarily contribute to the provision of public goods.
Individuals' willingness to contribute to public goods has been the subject of much study. For more than two decades, experimental economists have examined factors that influence cooperation in laboratory environments. Excellent survey articles have been written, but as the number of studies grows, the accuracy of this technique to represent the results of primary research has been questioned. In addition, although literature reviews summarize what is known, they provide little forecasting power. A review may, for instance, provide a table with price elasticities found in the different studies surveyed, but will generally not provide an estimated elasticity based on the total available research.
Meta-analysis has the potential to address some of these concerns. Meta-analysis is the statistical analysis of a collection of results from individual studies. It systematically identifies as complete a sample of studies pertaining to the issue of interest as possible, rather than highlighting the findings of key research only. The features and results of these studies are then described in a consistent quantitative way.
In this paper, the author examined 349 primary sources using meta-analysis and found that several factors significantly affected mean contributions to public goods. For example, as the marginal payoff to an experiment participant from allocations to a public good relative to a private good increased, significantly larger contribution levels were observed. Similarly, communication among participants improved cooperation as did maintaining a constant group of participants over a session. In contrast, experienced participants tended to make significantly lower contributions than first-time participants.
The author concludes that although the use of meta-analysis continues to be relatively rare in economics, it has the potential to offer significant benefit for policy-makers as well as for future experimental work.