The paycheck for life that quietly shrinks

Two checks that look identical

Imagine two retirees, both 67. One gets $3,000 a month from Social Security. The other gets $3,000 a month from a corporate pension. On the day they retire, those are the same thing — same deposit, same date, same feeling of security.

Thirty years later they are not remotely the same thing. Social Security has been adjusted for inflation every year along the way, so it still buys roughly what it bought at 67. The pension has paid exactly $3,000 every single month, which means that by 95 it buys about $1,311 of what it originally did. Same check. Less than half the groceries.

This is the least dramatic risk in retirement planning and one of the most consequential. There's no crash, no bad year, no moment where anything visibly goes wrong. The statement always says $3,000. The erosion happens so slowly that it never announces itself — and by the time it's obvious, the decades you'd have needed to plan around it are gone.

Nothing about a fixed pension ever looks like it's failing. That's precisely what makes it easy to plan badly around.

The arithmetic nobody runs

The math itself is not complicated. At 3% annual inflation, here's what one fixed dollar is actually worth as the years pass:

3% inflation, the long-run planning assumption used throughout this piece. Higher inflation makes every one of these numbers worse, and the effect is not linear.

Most people, shown this table, nod and move on. It feels like a technicality. It isn't, and the reason is what happens next: the shortfall doesn't just sit there — your portfolio has to cover it. Every dollar of purchasing power the pension loses is a dollar your savings must now produce instead, for the rest of your life, on top of everything it was already doing.

Three households, one difference

To measure that properly, we built three households that are identical in every respect we could hold constant. Both spouses 67, retiring now, planning to 95. The same $1.45 million invested. The same $8,000 a month of spending. The same expected returns, the same state tax, the same everything.

And critically: all three start with exactly $6,000 a month of guaranteed income. On day one, these three retirements look indistinguishable on paper. The only thing we varied is where that $6,000 comes from.

All figures in today's dollars per month, so they're directly comparable. The nominal check never falls in any of these households — the pension-heavy couple still receives $6,000 in nominal terms at 95. It simply buys 42% less.

Notice the arithmetic reconciles exactly with the simple table above. The half-and-half household's floor at 95 is $3,000 of Social Security (which held its value) plus $3,000 of pension at 44 cents on the dollar, which is $1,311 — a total of $4,311. That's not an approximation; it's the same number the full simulation produced. The mechanism really is that simple. It's the consequence that's easy to miss.

Why the damage compounds

Here's what the portfolio has to pull each year to keep spending flat, once the eroding floor stops covering its share:

Today's dollars per year. The rise in the first row is ordinary — healthcare and Medicare costs climb with age in every plan. The rise in the third row is that plus a floor collapsing underneath it.

The pension-heavy household starts out drawing about $5,600 a year more than the all-Social-Security household. By 95 that difference has grown to over $31,000 a year. And the timing is the cruel part: the extra draw arrives late, when the balance is at its smallest and has the fewest years left to recover. This is the same sequence-of-returns problem that makes a crash early in retirement so much worse than the same crash later — except here it's entirely predictable, and it runs in the opposite direction.

Run all three through 1,000 simulations and the outcome separates sharply:

Same money. Same spending. Same starting income. A 34-point spread in whether the plan holds — from a single line item most people have never checked.

What it would cost to fix

The cleanest way to size a risk is to ask what it would take to make it go away. We held the pension-heavy household's plan fixed and solved for how much more it would need to reach the same 88.9% as the all-Social-Security household.

The answer is $867,000 in additional savings — roughly a 60% larger portfolio. Alternatively, that household could keep its savings exactly as they are and cut spending permanently from $8,000 to $6,350 a month, a 21% reduction for life.

That's the real price of the difference between an inflation-adjusted floor and a fixed one. Not a technicality. Most of a second portfolio.

This is also why the risk deserves to be named rather than absorbed. Nobody would casually accept an $867,000 hole in their plan. Plenty of people accept this one, because it never appears as a number anywhere — it's distributed invisibly across thirty years of statements that all say the same reassuring figure.

If you're married, there's a second layer

Inflation is the slow erosion. The first death is the sudden one, and the two stack.

When one spouse dies, Social Security does not continue paying both checks — the survivor keeps the larger of the two and the smaller one stops. What happens to the pension depends entirely on an election that was made, often years earlier, at retirement: a single-life election pays the biggest monthly check but stops completely at death, while a joint-and-survivor election pays less now in exchange for continuing some percentage to the survivor.

Here's the floor at 80 for our three households — already eroded by thirteen years of inflation — under each election:

Today's dollars per month, at age 80. The pension-heavy household with a single-life election is the case worth staring at.

That last cell is the one to sit with. A couple who retired on $6,000 a month of guaranteed income can leave a survivor with $1,000 a month in real terms — not because anything went wrong, but because inflation ran for thirteen years and a single-life election was made at retirement. The survivor's expenses, meanwhile, don't fall to a sixth of the couple's. They fall by maybe a quarter.

Two things follow. If a pension election is still ahead of you, that decision deserves far more weight than the monthly difference makes it look — we cover the tradeoff in lump sum vs. the monthly check. And if it's already been made, that's not a failure, it's just a fact your plan needs to know about, because the portfolio is what has to absorb it. The survivor tax trap compounds this further: the surviving spouse also moves to single filer brackets, so a smaller income can face a higher rate.

The one lever that buys more inflation-adjusted income

You generally cannot add a COLA to a pension that doesn't have one, and inflation-adjusted annuities are expensive and increasingly rare. But there is one lever that directly increases the inflation-protected share of your floor, and it's available to almost everyone: delaying Social Security.

Every month you wait past full retirement age permanently increases the one income stream that does keep up with inflation. You're not just buying a bigger check — you're buying a bigger check in the currency that holds its value. Here's what moving from claiming at 67 to claiming at 70 did for each household:

And here the honest answer is uncomfortable, so it's worth stating plainly rather than selling the lever harder than it deserves: the household that most needs more inflation-adjusted income gets the least from this. Delaying gains the all-Social-Security household ten points and the pension-heavy household under two, because you can only grow the piece you already have. If Social Security is a third of your floor, tripling your effort on it moves a third of the problem.

Which means for a pension-heavy household the real levers are the ordinary ones — spending flexibility, how the portfolio is invested for a longer real drawdown, and knowing the shortfall is coming decades before it arrives. That's a less satisfying answer than a single fix. It's the true one.

What this doesn't capture

A few honest limits. Everything here assumes 3% inflation held steady for 28 years; real inflation arrives in bursts, and a high-inflation decade early in retirement is materially worse than this models. We assumed a fully fixed pension, but some plans carry a capped COLA — 2%, or CPI-limited — which lands between our first and second households rather than at the extreme. We modeled a specific household at a specific portfolio size, and while the direction of the effect holds broadly, the magnitude depends on your own numbers.

We also haven't touched the argument for fixed guaranteed income, which is real: it removes investment and longevity risk from that slice entirely, and that protection is worth something no success-rate percentage fully captures. This isn't a case against pensions or annuities. A fixed pension is a genuinely valuable asset. The case is against modeling a fixed payment and an inflation-adjusted one as though they're the same asset — because they aren't, and the plan that assumes they are will be wrong slowly, in the direction of optimism, for a very long time before anyone notices.

How this is calculated

Every figure in this article was produced by the same engine that runs the full app — no estimates and no hand-arithmetic. The three households were run through 1,000 Monte Carlo simulations each, with a fixed random seed so the results are reproducible, and the deterministic year-by-year projection supplied the income and drawdown figures. The $867,000 and $6,350/mo answers come from a bisection search against the target success rate, and we verified the underlying curve is smooth and monotonic so those aren't artifacts of the search. Real-dollar figures are nominal amounts discounted at the same 3% inflation assumption used in the simulation. Full assumptions live in the trust framework.

Common questions