Blog

The Canadian Institute of Actuaries’ recent release of the 2024 Canadian Pensioner Mortality Research Project is an important milestone for Canadian pension plans.  It is only the second time Canadian pensioner mortality tables have been developed using the mortality experience of Canadians, and it follows more than a decade of reliance on the CPM2014 mortality tables.

This is not simply a refresh of numbers, nor a quiet prelude to an inevitable standards change.  Instead, it represents a structural shift in how pensioner longevity is measured, interpreted, and – crucially – how professional judgment is expected in applying it.

This new research adds a new set of mortality tables to a pension actuary’s toolbox.  However, it also shifts greater responsibility onto actuaries and plan sponsors to explain, justify, and defend the mortality assumption they choose.

That theme becomes clearer when the research is read alongside the CIA’s recent Educational Note on the Selection of Mortality Assumptions.

Taken together, these documents don’t force change – but they do raise expectations.

Why CPM2024 is Fundamentally Different from CPM2014

It is tempting to view CPM2024 as simply a “new mortality table”.  But that misses the point.

The most significant development is not the mortality rates themselves, but the depth and credibility of the underlying data.  The CPM2024 research is based on experience from 33 pension plans, covering approximately 1.6 million individual records and more than 300,000 deaths.  That is roughly three times the volume of data which was used in the 2014 study.

The experience period for the study spans 2011 to 2021, which of course includes COVID‑19.  The research did not simply blend pandemic‑era mortality into long‑term expectations.  Instead, mortality spikes, particularly the spring 2020 wave, were identified and isolated.  Moreover, the experience during the study period was trended forward so that the resulting mortality rates are effective for the year beginning 2024.

Moving Beyond Public vs. Private Sector Mortality

One of the more consequential changes in CPM2024 is the abandonment of the public‑sector versus private‑sector split that was prevalent in the CPM2014 tables.

In its place are three table options:

• CPM2024 Light
• CPM2024 Heavy
• CPM2024 Combined

This shift reflects a core conclusion of the research: mortality differences are better explained by socio‑economic and occupational factors, rather than by whether a plan member happens to be in the public or private sector.

For example, a white‑collar, management‑heavy plan in the private sector may plausibly exhibit lower mortality than a blue‑collar plan in the public sector – something the CPM2014 framework could not easily accommodate.

From a conceptual standpoint, this is clearly an improvement.  In practice, it removes a simple rule that many plans had grown comfortable relying upon.

As such, there is no longer a prescribed mapping from “plan type” to a specific “table”.  Selecting an appropriate mortality table requires explicit judgment about the characteristics of a plan’s membership.  That judgment should be reasonable – but also needs to be justifiable.

Survivor Mortality can be Treated Differently

CPM2024 also introduces separate mortality tables for surviving spouses, including distinctions between widows and widowers.  This allows plan sponsors to measure the liability for spouses and surviving spouses with a bit more precision, as their mortality rates tend to be a touch higher than that of plan members.

This aligns more closely with how pension benefits are actually paid, particularly for plans with generous joint‑and‑survivor provisions.  It is also consistent with the CIA’s updated educational guidance, which identifies surviving‑spouse mortality as a distinct consideration.

For some plans, this refinement will have minimal impact.  For others, especially those with material survivor liabilities, it could affect liability measurement, and lead to better replication of annuity pricing.

Where the Real Impact Comes From: Mortality Improvement

When using the new CPM2024 mortality tables alongside the CIA’s CanMI‑2024 mortality improvement scale, the combined impact relative to the older CPM2014 mortality table and CPM‑B mortality improvement scale most often translates into an increase in liability measurements of about 2% to 3%.

However, one of the most important clarifications is that most of the increase does not come from the base CPM2024 tables – interestingly, it is the new improvement scale that ultimately drives the increase in liabilities.

This distinction matters – and needs to be understood by the users of actuarial work.  Previously, when mortality tables were updated, it was the change in the base mortality tables that led to the majority of the liability increases.  This is not the case this time around.  Rather, the increase in liabilities is mainly attributable to a more optimistic view of future mortality improvement.

This raises the prospect that some actuaries and plan sponsors wish to make tweaks to the new improvement scale.  Given that recent mortality experience has basically kept pace with the previous trend, some may question whether there is truly a need to adopt the more generous improvement rates in the newer improvement scales.

Funding vs. Solvency: No Immediate Rule Change

From a practical standpoint, nothing changes immediately.  Specifically:

• the previous CPM2014 mortality table and CPM-B improvement scale remain the prescribed table for solvency valuations – until actuarial standards applicable for commuted value calculations are formally updated (which may happen in the near future); and
• for going‑concern and accounting valuations, the CPM2014 mortality tables and CPM-B improvement scale are still considered acceptable by the actuarial profession, as adoption of the new mortality table and improvement scale is not considered mandatory.

In other words, this is not exactly a “flip the switch” moment.

With that said, the updated CIA Educational Note subtly raises the bar.  While it does not require use of the new CPM2024 mortality table and a new improvement scale, it clearly expects actuaries to consider published mortality research and to document their reasoning – particularly where older mortality assumptions are retained.

Personally, I would expect actuaries and plan sponsors to transition to the new CP2024 mortality tables and a newer mortality improvement scale within the next year or two.  I suspect pension regulators and auditors will be watching for this – and expecting plan sponsors to “modernize” their mortality assumptions.  

I would also point out that the CPM-B improvement scale reaches its ultimate improvement rate in 2031, at which point it transitions to one-dimensional scale – and arguably becomes less relevant.

What This Means for Plan Sponsors

For plan sponsors, the takeaway is not urgency – it’s preparedness.  You should expect:

• more discussions on the mortality assumptions at valuation time,
• clearer explanations for why particular mortality tables and improvement scales are chosen,
• and more explicit documentation of judgment.

This shift will put governance front and centre.

Final Thought

The release of the CPM2024 mortality tables is a step forward for Canadian pension mortality assumptions.  The research is deeper, broader, and reflects more recent mortality experience.

But progress comes with a trade‑off.

The selection of a new mortality table will require more judgment.  And explaining that judgment – to boards, members, auditors, and regulators – will increasingly be part of the job.

In that sense, CPM2024 is not just a new table.  It’s a reminder that longevity risk – like most pension risks – cannot be managed with general rules of thumb, and that assumptions increasingly need to be supported with evidence and clearly documented judgment.