Economic and Social Reports
A cross-cohort comparison of the economic impact of divorce and widowhood on seniors
DOI: https://doi.org/10.25318/36280001202300400003-eng
Skip to text
Text begins
The financial security of seniors has long been a concern for policy makers, especially with respect to seniors who have to deal with the financial aftermath of a divorce or separation (hereby “divorce”) or the death of a spouse later in life. Without sufficient economic resources to offset the loss of spousal income, divorce or widowhood can be devastating, especially at older ages when individuals are likely to be retired and relying to a greater extent on pension benefits and government transfers.
Previous studies have shown that living standards fall following divorce and that the resulting income losses tend to disproportionately affect women (LaRochelle-Coté, Myles and Picot, 2012; Leopold, 2018; Lin and Brown, 2021). However, the increased access of women to employer pensions (Morissette, 2020), and their growing wages and labour force participation rates over the last few decades (Statistics Canada, 2022) suggest that the financial effect of widowhood and divorce might be lower for recent cohorts of retired women than it was for earlier cohorts.
The study by LaRochelle-Coté, Myles and Picot (2012) tracked the permanent after-tax family income adjusted for family size (hereby “family income”) of individuals from ages 54 to 56 until ages 68 to 70 across two cohorts—1983 and 1993—and found no significant difference in outcomes across the cohorts of women who became widowed or divorced before age 68.Note However, this finding does not preclude future cohorts of women from experiencing improved outcomes per se. This study makes two key contributions to the literature. First, it expands the analysis of the impact of divorce and widowhood on seniors’ living standards to a larger set of cohorts: 1984, 1987, 1990, 1993 and 1996 (the 1996 cohort has not been analyzed previously).Note This allows for a more up-to-date assessment of the impact of marital dissolution on seniors’ living standards. Second, building on the method of Mehdi (2023), this study addresses measurement issues by computing family income received over a 10-year period from ages 70 to 80. This provides a more comprehensive measure of living standards following divorce or spousal death than the relatively short period (i.e., ages 68 to 70) used in previous Canadian studies.Note Therefore, this approach enables a more accurate assessment of the long-term effects of marital dissolution on seniors’ living standards.
Overall, recent cohorts of widowed and divorced women had higher family income later in life than earlier cohorts
In all cohorts, widows and divorcees were, overall, financially worse off later in life than individuals who remained married through ages 54 to 80. However, recent cohorts of widowed and divorced women fared somewhat better than their counterparts from earlier cohorts. For instance, median family incomes (2020 dollars) of widowed women, averaged across ages 70 to 80, increased from $30,900 for the 1984 cohort to $33,600 for the 1996 cohort. Median family incomes of divorced women, averaged across ages 70 to 80, increased from $26,300 for the 1984 cohort to $27,800 for the 1996 cohort (Table 1).
Income trajectories varied across the income distribution
Income trajectories differed across income quintiles. In the top quintile—individuals who were in the top 20% of the family income distribution at ages 54 to 56—the 1996 cohort of widowed women aged 70 to 80 years had median family incomes that were $7,500 higher than those of their 1984 counterparts. This was because, in part, of increased income from Registered Retirement Savings Plans (RRSPs) and Registered Pensions Plans (RPPs) and net capital gains. In contrast, divorced women from the 1996 cohort saw their median family incomes decrease by $5,400 relative to their 1984 counterparts, partly because of a reduction in investment income.
In the middle quintile, the 1996 cohort of widowed (+$3,300) and divorced (+$2,500) women saw their median family incomes at ages 70 to 80 increase relative to those of their 1984 counterparts. These improvements were because of increases in RRSP and RPP incomes.
In the bottom quintile, median family incomes remained stable across cohorts as changes over time generally amounted to around $1,000. Median family incomes of widowed women aged 70 to 80 years from the 1996 cohort increased by $600 relative to their 1984 counterparts, while median family incomes of divorced women aged 70 to 80 years decreased by $500. Old Age Security and the Guaranteed Income Supplement were the largest sources of retirement income for the bottom quintile.
Regardless of income quintiles at ages 54 to 56, the cross-cohort income gains (or losses) of divorced women were lower (or higher) compared to single, married and widowed women.
Divorce also had a negative financial impact on men, but they were more likely, on average, to have higher family incomes than women at ages 70 to 80.
| Single from age 54 to 80 | Married by age 54 to 56 | |||
|---|---|---|---|---|
| Married until age 80 | Widowed by age 68 | Divorced or separated by age 65 | ||
| dollars | ||||
| Women at age 54 to 56 | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 37,300 | 48,400 | 42,200 | 40,200 |
| 1987 | 36,500 | 49,300 | 42,400 | 39,900 |
| 1990 | 35,700 | 51,100 | 43,100 | 43,300 |
| 1993 | 34,000 | 49,300 | 41,500 | 41,200 |
| 1996 | 33,900 | 51,300 | 43,000 | 43,600 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 79,300 | 81,600 | 78,700 | 81,600 |
| 1987 | 81,800 | 84,900 | 81,600 | 87,300 |
| 1990 | 85,600 | 87,900 | 84,500 | 87,400 |
| 1993 | 83,300 | 88,400 | 86,200 | 86,400 |
| 1996 | 83,900 | 91,300 | 88,000 | 93,200 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 43,600 | 44,200 | 44,000 | 44,300 |
| 1987 | 44,400 | 44,700 | 44,500 | 44,800 |
| 1990 | 45,200 | 45,900 | 45,300 | 45,700 |
| 1993 | 43,700 | 44,100 | 44,000 | 44,000 |
| 1996 | 44,800 | 45,300 | 44,900 | 45,600 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 21,400 | 23,200 | 22,900 | 21,300 |
| 1987 | 20,900 | 22,800 | 22,300 | 21,400 |
| 1990 | 20,700 | 23,600 | 23,100 | 22,100 |
| 1993 | 18,600 | 22,200 | 22,000 | 20,900 |
| 1996 | 18,100 | 21,700 | 21,400 | 21,300 |
| Men at age 54 to 56 | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 39,500 | 47,000 | 44,500 | 39,100 |
| 1987 | 40,200 | 48,100 | 45,200 | 39,200 |
| 1990 | 38,300 | 50,300 | 45,800 | 38,300 |
| 1993 | 34,600 | 49,000 | 44,200 | 39,200 |
| 1996 | 35,300 | 50,400 | 44,100 | 40,900 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 83,200 | 80,900 | 78,400 | 80,500 |
| 1987 | 86,400 | 85,400 | 83,400 | 83,100 |
| 1990 | 88,900 | 90,000 | 87,800 | 86,100 |
| 1993 | 89,500 | 90,600 | 88,700 | 88,800 |
| 1996 | 91,800 | 94,100 | 89,200 | 90,000 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 45,600 | 45,800 | 45,000 | 45,400 |
| 1987 | 46,700 | 47,000 | 46,600 | 45,900 |
| 1990 | 48,500 | 48,300 | 48,400 | 48,400 |
| 1993 | 46,400 | 47,000 | 46,500 | 46,200 |
| 1996 | 47,300 | 48,100 | 47,500 | 47,100 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 21,300 | 23,800 | 22,400 | 23,000 |
| 1987 | 21,500 | 24,500 | 24,300 | 23,500 |
| 1990 | 20,700 | 25,100 | 24,900 | 24,200 |
| 1993 | 18,500 | 23,400 | 23,900 | 22,000 |
| 1996 | 17,700 | 22,800 | 22,900 | 22,300 |
| Women at age 70 to 80 | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 30,100 | 40,300 | 30,900 | 26,300 |
| 1987 | 30,000 | 40,900 | 31,600 | 27,000 |
| 1990 | 29,300 | 41,900 | 32,400 | 26,600 |
| 1993 | 29,100 | 43,000 | 32,800 | 27,700 |
| 1996 | 30,100 | 44,500 | 33,600 | 27,800 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 55,600 | 62,800 | 52,500 | 47,400 |
| 1987 | 55,500 | 63,700 | 54,100 | 46,500 |
| 1990 | 58,000 | 64,900 | 55,300 | 44,600 |
| 1993 | 59,900 | 66,100 | 56,300 | 45,600 |
| 1996 | 61,900 | 68,200 | 60,000 | 42,000 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 34,100 | 36,400 | 30,600 | 26,200 |
| 1987 | 34,000 | 36,900 | 32,000 | 27,000 |
| 1990 | 34,900 | 37,400 | 32,700 | 26,200 |
| 1993 | 35,500 | 38,200 | 33,400 | 28,000 |
| 1996 | 37,400 | 39,400 | 33,900 | 28,700 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 22,400 | 26,300 | 22,900 | 22,200 |
| 1987 | 22,600 | 26,300 | 23,400 | 22,500 |
| 1990 | 22,000 | 26,200 | 23,200 | 21,500 |
| 1993 | 21,900 | 26,700 | 23,400 | 20,700 |
| 1996 | 22,100 | 26,900 | 23,500 | 21,700 |
| Men at age 70 to 80 | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 32,300 | 39,800 | 38,000 | 32,200 |
| 1987 | 33,300 | 40,600 | 36,600 | 31,400 |
| 1990 | 31,400 | 42,000 | 38,500 | 28,300 |
| 1993 | 31,900 | 43,600 | 38,800 | 28,600 |
| 1996 | 33,000 | 44,800 | 38,600 | 30,500 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 60,100 | 62,200 | 55,400 | 53,900 |
| 1987 | 61,500 | 63,700 | 59,800 | 56,100 |
| 1990 | 61,300 | 65,600 | 59,900 | 53,200 |
| 1993 | 60,500 | 68,000 | 63,000 | 48,500 |
| 1996 | 61,800 | 70,800 | 67,300 | 51,900 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 33,700 | 37,200 | 36,200 | 35,300 |
| 1987 | 34,900 | 38,200 | 35,800 | 37,200 |
| 1990 | 34,800 | 39,000 | 38,400 | 32,900 |
| 1993 | 36,400 | 40,100 | 39,600 | 32,900 |
| 1996 | 38,100 | 41,300 | 40,600 | 33,200 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 22,700 | 26,800 | 25,600 | 23,400 |
| 1987 | 23,000 | 27,000 | 24,700 | 23,300 |
| 1990 | 22,700 | 27,100 | 24,800 | 22,200 |
| 1993 | 22,500 | 27,600 | 25,000 | 22,100 |
| 1996 | 22,600 | 28,200 | 25,400 | 22,900 |
|
Note: The quintiles refer to family income distribution rank at age 54 to 56; top quintile = top 20% of the distribution; middle quintile = middle 20% of the distribution; bottom quintile = bottom 20% of the distribution. Only individuals with at least $13,000 at age 54 to 56 are included. Income quintiles are calculated separately by sex based on permanent after-tax family income adjusted for family size at age 54 to 56. Family income is divided by the square root of family size. Permanent income here is calculated by taking a three-year average of an individual’s annual after-tax family income adjusted for family size. Median incomes at age 70 to 80 are calculated by averaging the median incomes across ages 69 to 71 through 78 to 80. The small group of widows and divorcees who subsequently remarried are excluded. Individuals who divorced or separated after age 65 are also excluded since it is rare among seniors but may be overestimated in administrative data. Source: Statistics Canada, Longitudinal Administrative Databank, 1982 to 2020. |
||||
On average, recent cohorts of widowed men and women experienced smaller decreases in family income later in life than earlier cohorts
On average, widowed women from the 1984 cohort received, at ages 70 to 80, 77% of the income they had received at ages 54 to 56 (Table 2). The income replacement rate of the 1996 cohort was, at 83%, six percentage points higher, thereby indicating a smaller decrease in family income later in life for this cohort. Income replacement rates also increased across cohorts for widowed men.
In contrast, the income replacement rate for divorced women from the 1996 cohort (70%) was slightly lower than that of the 1984 cohort (73%). This drop in the income replacement rate occurred because the cross-cohort gains in family income of divorced women at ages 70 to 80 were not sufficient to offset the gains made at ages 54 to 56 when they were still married (Table 1).
In the top quintile, income replacement rates for widowed women amounted to roughly 64% for the 1984 and the 1996 cohorts. But replacement rates for divorced women decreased from 56% (1984 cohort) to 45% (1996 cohort). This reduction occurred as divorced women in the top quintile saw their income at ages 70 to 80 decrease across cohorts while their income at ages 54 to 56 rose across cohorts.
In the middle quintile, income replacement rates increased from 70% (1984 cohort) to 76% (1996 cohort) for widowed women and from 61% to 63% for divorced women.
In the bottom quintile, both widowed and divorced women from the 1996 cohort had somewhat higher income replacement rates than their counterparts from the 1984 cohort. For all cohorts, income replacement rates exceeded 100%, as income at ages 70 to 80 surpassed that at ages 54 to 56 (Table 1).
Lastly, in the middle and top quintiles, income replacement rates fell over time for divorced men but increased for widowed men.
Conclusion
Using a comprehensive measure of income and longitudinal data that track several cohorts of individuals from ages 54 to 56 onwards, this study shows that the living standards of widowed and divorced women later in life have, in general, improved slightly over time. However, the extent of the improvements varies across income quintiles. The study also shows that divorce reduces living standards during retirement to a greater extent than spousal death. In all the cohorts considered, women had lower family incomes than men later in life, regardless of marital status. Whether the current retirement income trajectories of widows and divorcees will persist in the future remains to be seen.
| Single from age 54 to 80 | Married by age 54 to 56 | |||
|---|---|---|---|---|
| Married until age 80 | Widowed by age 68 | Divorced or separated by age 65 | ||
| percent | ||||
| Women | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 85.5 | 84.3 | 77.4 | 73.4 |
| 1987 | 86.9 | 83.2 | 78.8 | 75.4 |
| 1990 | 86.2 | 81.8 | 78.4 | 67.8 |
| 1993 | 91.1 | 86.2 | 82.1 | 72.0 |
| 1996 | 94.1 | 87.0 | 83.0 | 70.4 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 68.8 | 74.1 | 63.5 | 56.1 |
| 1987 | 66.3 | 72.1 | 63.8 | 50.9 |
| 1990 | 64.5 | 70.4 | 60.8 | 45.3 |
| 1993 | 69.1 | 71.7 | 62.9 | 50.6 |
| 1996 | 70.7 | 72.5 | 64.0 | 45.3 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 77.4 | 82.3 | 69.9 | 61.1 |
| 1987 | 76.8 | 82.4 | 72.2 | 61.0 |
| 1990 | 76.4 | 81.5 | 72.3 | 57.5 |
| 1993 | 80.8 | 86.4 | 75.8 | 64.4 |
| 1996 | 82.8 | 87.3 | 75.5 | 63.3 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 110.1 | 120.8 | 108.7 | 108.4 |
| 1987 | 114.0 | 121.0 | 112.3 | 109.0 |
| 1990 | 111.5 | 118.3 | 107.6 | 104.3 |
| 1993 | 121.0 | 128.9 | 112.1 | 105.4 |
| 1996 | 127.5 | 132.6 | 116.0 | 111.4 |
| Men | ||||
| Overall | ||||
| Cohort year | ||||
| 1984 | 83.4 | 85.4 | 83.1 | 82.5 |
| 1987 | 82.5 | 84.6 | 82.9 | 80.5 |
| 1990 | 82.3 | 82.7 | 81.9 | 78.1 |
| 1993 | 88.9 | 87.5 | 87.8 | 79.5 |
| 1996 | 90.0 | 88.1 | 89.1 | 81.1 |
| Top quintile | ||||
| Cohort year | ||||
| 1984 | 69.9 | 74.2 | 69.6 | 63.0 |
| 1987 | 67.0 | 72.1 | 73.2 | 62.7 |
| 1990 | 65.9 | 70.2 | 65.2 | 56.2 |
| 1993 | 66.2 | 72.2 | 69.0 | 53.3 |
| 1996 | 65.4 | 72.5 | 73.7 | 56.8 |
| Middle quintile | ||||
| Cohort year | ||||
| 1984 | 72.9 | 81.6 | 79.2 | 75.7 |
| 1987 | 74.9 | 81.5 | 75.9 | 80.0 |
| 1990 | 71.7 | 80.6 | 78.1 | 69.7 |
| 1993 | 78.2 | 85.4 | 83.6 | 70.9 |
| 1996 | 79.4 | 85.9 | 86.0 | 67.8 |
| Bottom quintile | ||||
| Cohort year | ||||
| 1984 | 112.3 | 119.9 | 119.2 | 110.2 |
| 1987 | 113.6 | 117.4 | 111.0 | 111.2 |
| 1990 | 112.8 | 115.7 | 112.4 | 99.1 |
| 1993 | 125.2 | 127.7 | 118.6 | 103.5 |
| 1996 | 132.2 | 132.0 | 118.9 | 114.1 |
|
Note: The quintiles refer to family income distribution rank at age 54 to 56; top quintile = top 20% of the distribution; middle quintile = middle 20% of the distribution; bottom quintile = bottom 20% of the distribution. Only individuals with at least $13,000 at age 54 to 56 are included. Income quintiles are calculated separately by sex based on permanent after-tax family income adjusted for family size at age 54 to 56. Family income is divided by the square root of family size. Permanent income here is calculated by taking a three-year average of an individual’s annual after-tax family income adjusted for family size. Replacement rates are calculated in three steps, which involves 1) computing at the individual level, income at all ages between 69 to 71 and 78 to 80 as a share of income at age 54 to 56, 2) taking the median of these shares across all individuals, and 3) averaging the median shares across ages 69 to 71 through 78 to 80. The small group of widows and divorcees who subsequently remarried are excluded. Individuals who divorced or separated after age 65 are also excluded since it is rare among seniors but may be overestimated in administrative data. Source: Statistics Canada, Longitudinal Administrative Databank, 1982 to 2020. |
||||
Author
Tahsin Mehdi is with the Social Analysis and Modelling Division at Statistics Canada.
References
LaRochelle-Coté, S., Myles, J., and Picot, G. 2012. Income Replacement Rates Among Canadian Seniors: The Effect of Widowhood and Divorce . Canadian Public Policy 38(4): 471-495.
Leopold, T. 2018. Gender Differences in the Consequences of Divorce: A Study of Multiple Outcomes . Demography 55(3): 769-797.
Lin, S., I-F., and Brown, S. L. 2021. The Economic Consequences of Gray Divorce for Women and Men . The Journals of Gerontology: Series B 76(10): 2073-2085.
Mehdi, T. 2023. A Cross-Cohort Comparison of Economic Well-being During Retirement, Statistics Canada, Economic and Social Reports. (forthcoming).
Morissette, R. 2020. The Changing Job Landscape, 1981 to 2019. Statistics Canada, Economic Insights.
Statistics Canada. 2022 Table 14-10-0327-01 Labour force characteristics by sex and detailed age group, annual . Ottawa: Statistics Canada.
- Date modified: