The Broadbent Blog

Rising Life Expectancy of the Rich Should Not Threaten Pensions for the Poor


As expected, the federal budget delivered on  the Liberal promise to leave the age of eligibility for Old Age Security (OAS) and Guaranteed Income Supplement (GIS) retirement benefits at age sixty five. The Harper government had previously decided to phase in an increase to age sixty seven.

Many pundits have argued that the eligibility age should rise in line with longer life expectancy. They say that a higher retirement age would reduce the growing cost of the OAS/GIS program, and will boost the economy by pushing seniors to work longer.

These arguments ignore wide variations in the economic circumstances of seniors which would bring about highly inequitable outcomes.

With respect to rising life expectancy, policy-makers should beware the mean. Averages hide significant differences among seniors based upon socio-economic status.

In all countries, including Canada, the rich tend to live much longer than the poor and the middle-class, and there is evidence that these differences may be increasing.

In the United States, Barry P. Bosworth of the Brookings Institution has recently shown that there is a very strong relationship between life expectancy and a person's ranking in the income distribution.

For the cohort of men born in 1940, average life expectancy at age 55 is is much longer (34.9 years) for those who were in the top 10% of earners at mid career than for for those who were in the bottom 10% of earners (24.2 years.) Thus there is a huge ten year gap in life expectancy between the top and bottom of the income scale.

For women in the United States, there is also a big gap of about ten years in life expectancy at age 55 between the top and bottom 10% of earners (35.3 years compared to 25.8 years.)

Dr. Bosworth also shows that rising average life expectancy in the United States has mainly been experienced by higher income earners. The increase in average life expectancy at age 55 between the cohort born in 1920 and that born in 1940 was about four years on average for men. But the increase was 5.9 years for those in the top 10% compared to just 1.7 years for those in the bottom 10%, a big difference of 4.2 years.

The difference in increased life expectancy between the high and low income groups was even greater for women, at 5.2 years. And life expectancy has actually declined for American women in the bottom one third of earners.

A highly publicized recent US study by Nobel economics prize winner Angus Deaton further suggests that life expectancy has been declining among lower income, middle-aged white men (a phenomenon which has been linked by the media to the Donald Trump campaign.)

Many commentators in the United States have argued that it is unfair to raise the age of eligibility for social security retirement benefits based on rising average life expectancy if lower income persons are not, in fact, living significantly longer than in the past.

Further, an increase in the eligibility age for retirement programs is likely to have a disproportionate impact upon lower income earners who are most likely to be in poor health and to have worked for a lifetime in physically demanding jobs. Such a change would force them to work longer given a lack of other sources of income such as employer pensions and investments.

Here in Canada, we have much less data than in the United States due to difficulties linking mortality records to data on incomes. But life expectancy as calculated by Statistics Canada is also highly unequal.

The average remaining life expectancy at age 25 for men in the bottom 10% ranked by income is 48.6 years, 7.1 years less than the 56.0 years of remaining life expectancy for those in the top 10%. Women age 25 in the bottom 10% will live 4.5 years less than women in the top 10% (56.5 years compared to 61.0 years.) As in the US, the relationship is quite linear, meaning that life expectancy rises steadily as income rises.

A study by the Chief Actuary of the OAS/GIS program shows that the life expectancy of male GIS recipients is 2.4 years less than that of recipients of OAS only (16.2 vs 18.6 years.) For women, the difference is 2.1 years (19.8 years compared to 21.9 years.) Receipt of the GIS is a good proxy for lower income since benefits are paid only to the one in three seniors with the lowest incomes.

Unfortunately, we have no Canadian data to show how gains in life expectancy have varied across the income spectrum. But it is clear that an across the board two year increase in the eligibility age for OAS and GIS would have had a much bigger negative impact on lower income seniors for the simple reason that they die earlier and have much more limited retirement options at age sixty-five.

As in so many areas of public policy, full account should be taken of the distribution of economic well-being rather than taking averages at face value.


Andrew Jackson is Adjunct Research Professor in the Institute of Political Economy at Carleton University, and senior policy advisor to the Broadbent Institute.

Photo: Garry Knight. Used under a Creative Commons license.