When should you take CPP and OAS?

Two decisions sit near the centre of most Canadian retirement plans, and both are usually made by accident. CPP can begin in any month from 60 to 70. OAS can begin in any month from 65 to 70. Start either one early and the monthly amount is permanently smaller; wait, and it is permanently larger. Nothing else about the benefit changes — not the indexation, not the tax treatment, not the fact that it arrives every month until you die.

Because the adjustment is fixed and mechanical, the decision comes down to a handful of knowable things: how long you expect the payments to run, what other income you have available in the meantime, whether you are still working, and what the extra taxable income does to everything else on your return. The rules themselves are simple enough to state in a paragraph.

The adjustment is mechanical

CPP is reduced by 0.6% for each month you start it before your 65th birthday, and increased by 0.7% for each month you start it after. Sixty months either side of 65 means a 36% permanent reduction at 60 and a 42% permanent increase at 70. Waiting past 70 adds nothing — the increase stops, and every month of delay after that is a pure loss.

OAS has no early option. It cannot begin before 65, and it grows by 0.6% for each month you defer, to a maximum of 36% at 70. It is also residency-based rather than contribution-based, so a partial pension is common for anyone who spent working years outside the country, and the amount steps up again at 75 regardless of when you started.

One administrative detail undoes more deferral plans than any argument about break-even: many people are enrolled in OAS automatically and receive a letter telling them so. If you intend to wait, deferral has to be requested — otherwise payments simply start, and the increase you were counting on never accrues. Neither pension arrives on the strength of a decision you made privately; both require an application with a chosen start month, and neither backdates well.

It is also worth being precise about what the increase is. Waiting a year past 65 adds 8.4% to the CPP base, but it is 8.4% of the age-65 amount rather than a return compounding on a balance, and it is paid in the form of higher monthly income rather than a larger asset. That distinction matters when someone frames deferral as "an 8% guaranteed return" — the guarantee is real, the shape is not the same as a portfolio's.

The spread on CPP is wider than most people picture: the payment at 70 is more than double the payment at 60, for life, indexed. That is the whole decision in one line.

Monthly CPP by start age

	Monthly CPP by start age
Age 60	64%
Age 62	78.4%
Age 65	100%
Age 67	116.8%
Age 70	142%

Share of the amount payable at 65. Reduced 0.6% per month before 65, increased 0.7% per month after.

Break-even, and what it leaves out

The usual next step is to ask when the deferred pension catches up on cumulative dollars. That arithmetic is easy, and the answers cluster tightly.

Those figures ignore investment return on the early payments and treat indexation as cancelling out on both sides, which it broadly does in real terms. Add a meaningful real return on the money you took early and every crossing point moves later — a few years later at modest returns, further at aggressive ones. But that comparison is not fair as usually stated: the deferral increase is guaranteed, indexed, and independent of markets, while the return you would need to beat it is neither. You are comparing a certainty to a hope.

The deeper problem with break-even is that it answers the wrong question. It asks whether you will beat an average lifespan, when the thing that actually damages a retirement is the tail. Median remaining life at 65 already reaches into the mid-eighties, and a substantial minority live well past 90 — which means the scenarios where deferral "loses" are the ones where you did not need the money anyway, and the scenario where it wins is the one where you are 91 and your portfolio is thin. Deferring is not a bet. It is buying more indexed lifetime income with savings you would otherwise have to make last, which is why it is best understood as insurance.

The clawback runs in both directions

OAS is subject to a recovery tax: once net income passes a threshold, 15 cents of every additional dollar claws back the benefit, stacking on top of your marginal rate and producing an effective rate most people would not recognise from any bracket table. Deferring OAS raises the monthly amount and therefore raises what is exposed.

That sounds like an argument against waiting, and taken alone it is. In practice the years between 65 and 70 are also the best chance you will get to lower the income that triggers the clawback later — a window with no employment income, no OAS yet, and full control over how much you draw from registered accounts. Emptying registered savings deliberately in that window, before minimum withdrawals take over, is often worth more than the clawback exposure it creates. How the RRIF schedule constrains later years is the piece of this most people discover too late, and the decumulation planner will show what your own withdrawal order does to taxable income at each age.

Taking CPP early while still working

Starting CPP at 60 while you are still earning is the version of this decision that most reliably goes wrong. The pension lands on top of employment income and is taxed at your top marginal rate, in the years your rate is highest, in exchange for a permanent 36% cut to a benefit you will draw for decades. Contributions remain mandatory until 65, which does generate small post-retirement additions each year — real, but not remotely a match for what the reduction removes.

There are honest reasons to start early anyway. Poor health or a family history that argues against a long life is the clearest. So is needing the money: a pension you use at 62 to avoid selling investments in a bad market, or to avoid borrowing, is doing a job that no break-even table measures. Cash flow you actually need beats optimisation you cannot afford.

Run your numbersRetirement projection→

The bridge, which is the common case

For most people who retire in their early sixties with savings, the strongest structure is not "take it early" or "take it late" but a bridge: retire, live on the portfolio for a few years, and start CPP as late as the cash flow allows. Three things happen at once. Registered accounts get drawn down while your income is otherwise low, so less is left to be forced out later. The taxable income of those gap years stays low enough to make room for other planning. And the portfolio ends up supporting a shorter, more predictable stretch of retirement, because from 70 onward a much larger indexed pension covers a bigger share of the fixed costs.

That last effect is worth sitting with. A retirement where guaranteed income covers most of the essential spending is a fundamentally different problem from one where the portfolio covers everything — the sustainable withdrawal rate question softens considerably when the floor beneath it is higher. Model the two shapes against your own numbers with the retirement planner rather than reasoning about them in the abstract.

Deciding as a couple

Run it per person, then look at the pair. The higher-earning spouse usually has the stronger case for deferring CPP, because their larger pension is the one that best protects the survivor — subject to the cap on a combined survivor's and own pension, which limits how much of a deferred amount actually carries over. The lower-earning spouse often has the better case for starting earlier, funding the household while the other defers.

Ages matter too. A meaningful age gap changes which pension is likely to be paid longest, and therefore which one is worth growing. And because splitting eligible pension income between spouses can move a couple below a clawback threshold that either alone would cross, what qualifies for income splitting belongs in this decision rather than after it.