Financial security Poverty Social security

C.D. Howe report on impact of CPP, OAS / GIS policy changes: Convenient over-simplification presented as ‘stylised fact’

Comparing Nest Eggs: How CPP Reform Affects Retirement Choices Reforms (PDF)
Alexandre Laurin, Kevin Milligan, Tammy Schirle, C.D. HOWE Institute commentary NO. 352

Accordingly, we simulate the flow of CPP benefits for a stylized individual retiring at different ages.

Problem right off the top: The primary determinant of total benefits received is life expectancy; a couple years difference in retirement age is secondary.

So how does the report account for differences in life expectancy?

Our simulations make use of a simplified model of an individual to provide a clear demonstration of how the new pension adjustments affect retirement incentives. Our individual is male, from Ontario, and began his career at age 24. This worker enjoyed relatively high earnings during his career, making contributions based on the Year’s Maximum Pensionable Earnings (YMPE) each year. The individual is now age 60 and needs to decide when to retire and take up CPP benefits.4 We consider potential retirement ages between 60 and 70. For ease of discourse, we call him Joe.

Another problem: Joe is assumed to have been continually employed with high earnings from age 24 to 60-70, entitling him to max out his retirement benefits contributions.  That’s more fantasy that stylised fact in the Canadian labour market today.

So how long will Joe live to collect?

We also consider how these benefits add up over Joe’s lifetime, using the pension wealth measure.  Future benefits are discounted and we account for male life expectancy based on 2007 mortality rates.5

Yet another problem:

Mortality rates were taken from the Canadian Human Mortality Database at Note that the use of female life expectancy would change results slightly. In particular, pension wealth would be higher given longer life expectancy among women.

Unfortunately, gender is neither the only nor even most significant determinant of life expectancy.

Relative wealth and income play a far greater role in determining both quality and expected span of life.

Finding data on relative wealth is a challenge for a number of reasons, primary among them is the choice of metric.  Income on the other hand is more straight-forward, and there is some publicly available data to demonstrate strong correlation between wealth/income and life expectancy.

Healthy Canadians (HLT)
-Canadian Institute for Health Information (CIHI)

Clicking on Table 36b-HLT ‘Life expectancy by income’ should give the desired table.  Or not – more mysteriously missing data from Statistics Canada.  There is a CANSIM table that provides some information on life expectancy based on income quintiles (20% groupings).  It does not provide a breakdown by low-income measure, so analysis would be limited to comparing highest and lowest income quintiles in the absence of more detailed income distribution data.







At birth

All income groups



Income group, quintile 1 (lowest)



Income group, quintile 5 (highest)


At age 65

All income groups



Income group, quintile 1 (lowest)



Income group, quintile 5 (highest)








At birth

All income groups



Income group, quintile 1 (lowest)



Income group, quintile 5 (highest)


At age 65

All income groups



Income group, quintile 1 (lowest)



Income group, quintile 5 (highest)

Source: CANSIM Table 102-0122 Health-adjusted life expectancy, at birth and at age 65, by sex and income, Canada and provinces occasional (years), Statistics Canada

It’s apparent from the table that federal government policy pushing back the OAS/GIS full benefits qualifying age to 67 and changing the CPP payout schedule to favour those receiving benefits after age 70 entirely further disadvantage lower-income Canadians.

Men in the bottom 20% income group at birth on average would not live to qualify for full OAS/GIS benefits under the old scheme;  their life expectancy was just under 65 in the 2005/2007 survey.

Women in the bottom 20% income group at birth on average would not live to qualify for full OAS/GIS benefits under the new scheme, as they lived on average to age 67 according to the survey.

The difference in health-adjusted life expectancy for men and women between the top and bottom income quintiles at birth was ~8 years on average.  While neither men nor women in the bottom income quintile would live to collect any increased benefits past age 70, those in the top income quintile would collect ~2.5 and 5 years increased benefits, respectively.

The concept of health-adjusted life expectancy by income quintile for those 65+ is a bit confusing.  If those in the bottom quintile from birth on average did not live to age 65, this effectively provides a snapshot of life expectancy for Canadians 65+ after half those in the lowest income quintile are deceased (assuming normal distribution).  That said, even the life expectancy figures for those 65+ show a substantial gap  between the lowest and highest 20% of income earners, ~3.5 years for both men and women.

The federal OAS/GIS and CPP policy changes are effectively a transfer of wealth from low to high-income Canadians at a stage in life when those in low-income are most vulnerable.  The policy effectively subsidises affluent older Canadians.  Where are Mr. Smart and Mr. Mintz’s comments on how this policy is ill-advised because “more affluent Canadians are actually deriving a greater benefit than the poor”  (the lame argument they ran up the flag-pole to tax food staples).

Given the data on life expectancy and income, sound public policy would provide full OAS/GIS benefits to low-income Canadians starting at age 60 while clawing back OAS/GIS benefits for higher income seniors – instead of increasing the full benefit qualifying age for all to 67, as the new federal government policy proposes.  It would likewise provide CPP benefits to low-income Canadians starting at age 60 applying an actuarially fair discount instead of increasing benefits for those over age 70.  The latter would be funded through increased CPP contributions, as suggested by some economists.

Such a policy would be especially helpful in providing income security for single women age 60+ given the remarkably high incidence of low-income for the demographic,

Given the demographic shift under way which (in theory) will see baby boomers retiring en masse in the coming years, getting this policy right is paramount.  The anecdotal evidence available suggests problems with the current (and recently proposed) policy:

Are bankrupt seniors harbingers of things to come?
CBC News June 29, 2012

Leave a Reply

Your email address will not be published. Required fields are marked *