Inflation and retirement: your personal rate is not CPI

Healthcare does not inflate like groceries. One blended rate hides the divergence that actually reshapes a retiree's budget over 25 years.

A market crash is loud. Inflation is silent, and it does not arrive as a single rate. It arrives as a dozen rates, and the ones that matter most to a retiree are not the ones in the headline. This spoke sits under the Retirement hub.

One blended rate is the wrong model

Almost every retirement calculator applies a single inflation number to your entire budget. That quietly assumes healthcare behaves like groceries. It does not — and the moment two categories compound at different rates, they diverge, permanently and predictably.

Here is a $60,000 budget, split the way many retirees split theirs, with everything except healthcare inflating at 3% and healthcare at 5%. Both rates are assumptions you supply, not forecasts we make.

Two things fall out of that table, and neither is visible in a single-rate model.

The total is wrong, not just the mix. If healthcare had tracked general inflation at 3%, the year-25 budget would be about $121,968. The two-rate version is $131,506 — roughly $9,500 a year more, every year, forever, on the same starting budget. A blended-rate plan does not merely mislabel that money. It never asks you to fund it.

Healthcare's share grows without you consuming any more of it. It goes from about an eighth of the budget to about a fifth — not because you got sicker, but because its price ran hotter. Widen the gap (7% healthcare against 3% general) and by year 25 healthcare is close to 28% of the budget. Narrow it to zero and the effect disappears entirely. That is the proof it is the gap doing the work, not the level of either rate — test it yourself in the spending planner.

That share matters for a reason beyond arithmetic: discretionary spending is your shock absorber. It is what you cut when markets fall — the single strongest defence against sequence risk. The larger healthcare grows as a share of your budget, the less absorber you have left. The two risks are not independent; inflation eats the tool you use to survive a crash.

Healthcare is the category that runs hotter — with honest caveats

Over the long run, medical prices have outpaced general prices. Since 2000, medical care prices rose about 121% while prices for all consumer goods and services rose about 86% (Peterson-KFF, using CPI data). That is a wide, persistent gap.

Three caveats, because the site's job is to tell you where the evidence is soft:

1. It is not true in every period. From 2021, general inflation ran hotter than medical prices for a stretch. The tendency is persistent; it is not a law of nature, and a plan that assumes a fixed 2-point gap forever is making a forecast.
2. The official index understates what you pay. The CPI's medical care index counts only out-of-pocket spending. That is why its weight in CPI is far smaller than healthcare's share of the economy — the parts paid by Medicare, an employer, or an insurer are largely outside it. The headline "medical inflation" number is not the number a household feels.
3. It is uneven inside the category. Hospital services, physician services and prescription drugs move at very different rates, and in some years drug prices have fallen outright while hospital prices climbed. Your personal exposure depends on which of those you actually consume.

The defensible conclusion is not "healthcare inflates at X%." It is: your personal inflation rate is not CPI, and if healthcare is a bigger share of your basket than it is of the CPI basket, your rate runs above the headline.

Medicare is not free, and it does not track the COLA

Medicare starts at 65, and it is a bill, not a solution.

• Part B premium: $202.90 a month in 2026, per person, deducted straight from your Social Security payment.
• Part B deductible: $283 for the year.
• Part D out-of-pocket cap: $2,100 — a genuine, valuable protection, and one of the few hard ceilings in the system.

The planning point is what those numbers did, not what they are. The standard Part B premium rose more than three times faster between 2025 and 2026 than the 2.8% Social Security COLA that is supposed to cover it (CMS). Because the premium comes out of the benefit before it reaches you, the net raise a retiree receives is smaller than the announced one — sometimes much smaller. Anyone bridging to Medicare before 65 faces a steeper version of the same problem; see health care before Medicare.

Social Security's COLA — and the claim about it that is wrong

The COLA is the most valuable inflation hedge most households will ever own: a lifetime income that resets with prices, backed by the federal government. For 2026 it was 2.8%.

It is also routinely explained wrongly, including in the earlier version of this page. You will read that the COLA is "one reason delaying Social Security is valuable." That reasoning is false. SSA is explicit:

"You're eligible for cost-of-living benefit increases starting with the year you turn age 62. This is true even if you don't get benefits until your full retirement age or even age 70."

COLAs accrue from 62 whether you have claimed or not. They are applied to your primary insurance amount — the base — and your claiming-age factor is then applied to that base. Waiting does not earn you extra COLAs.

Delaying is valuable for a different and better reason: the COLA compounds on a larger base. Wait until 70 and every future cost-of-living increase is a percentage of a permanently bigger number. The two effects multiply. That is the correct argument, and it is stronger than the wrong one. The break-even tool prices the trade-off.

One more honest limit: the COLA is measured on CPI-W, the basket of urban wage earners. It is not a retiree basket. It is a real hedge, and it is an imperfect one.

What hedges inflation, and what does not

Note the pattern. What survives inflation is income or assets that reset with prices. What does not is anything fixed in dollars. This is exactly why a private pension — almost never inflation-adjusted, and explicitly not COLA-protected by the PBGC — is a weaker floor than it looks, and why Social Security is a stronger one.

Key takeaways

• Your personal inflation rate is not CPI. Categories compound at different rates, and the divergence, not the level, is what reshapes a retiree's budget.
• Healthcare has historically run hotter than general prices — about 121% against 86% since 2000 — but not in every year, and the official index only measures out-of-pocket costs.
• A two-rate model changes the total, not just the mix: on a $60,000 budget with a 2-point gap, the year-25 shortfall against a blended-rate plan is roughly $9,500 a year.
• Social Security's COLA is a genuine hedge. Delaying does not earn extra COLAs — it makes each one compound on a larger base.
• Hedge with income that resets: Social Security, TIPS, equities over long horizons, a paid-off house. A flat pension, long nominal bonds and cash do not.

Educational only — not financial advice. Model two rates instead of one in the spending planner.