“Wages have kept up with inflation.” “No, they haven’t.”

You’ve heard both. Both sides have data. Both sides are right — and that’s the problem.

Pew Research recently ran the math four ways, using four different inflation gauges the federal government produces. The verdict on your buying power since 1999 ranges anywhere from a measly 11.5% to a respectable 22.1% real growth.

Same 26 years. Same paychecks. Same prices. Four wildly different answers.

Here’s why that matters: Depending on which number the government picks, your Social Security check, your tax bracket, and your retirement math all shift. And one of those four answers should scare you more than the others.

The 4 measures and the 4 verdicts

Pew analyzed median weekly wages from December 1999 through December 2025, using federal data. The headline number is impressive: The median weekly wage more than doubled, climbing from $482 to $1,040.

But nominal pay doesn’t pay your bills. Inflation-adjusted pay does. So here’s what that $482 from 1999 actually bought in 2025 dollars, depending on which inflation index you use and ranked from worst news for your paycheck to best:

1. The Consumer Price Index (CPI) Retroactive Series: +11.5%. This it the best estimate of what past inflation would look like under today’s methods, according to the Bureau of Labor Statistics (BLS). Your $482 in 1999 bought roughly what $933 buys today — about 0.4% real growth per year. Treading water for 26 years.
2. The Main CPI (CPI-U): +12.1%. The number in every news headline, around in some form since 1913 and updated monthly. Under CPI-U, your $482 from 1999 would buy what $928 buys today — almost identical to the Retroactive Series. The number politicians fight about barely moved your standard of living.
3. The Chained CPI: +20.1%. Things look brighter here. This measure adjusts for substitution — when steak gets pricey, you buy chicken instead. That tweak makes inflation look slightly tamer. Under chained CPI, your $482 grew to the equivalent of about $866 today. Friendlier ruler, better story.
4. The PCE: +22.1%. The Personal Consumption Expenditures index is the Federal Reserve’s favorite. The Fed uses it to set interest rate policy. Under PCE, your $482 in 1999 buys what $852 buys today — about 0.8% real wage growth per year. The rosiest answer, and still nothing to brag about.

Quick gut-check — if your money advice is coming from random online influencers, you’re playing a dangerous game. I’ve been a CPA since 1981 and writing about money since before the internet existed. Sign up for the free Money Talks Newsletter and get expert advice that’s been tested by time.

The answer that should scare you

Even on the best measure, real wages crept up at less than 1% per year for a quarter century. In a country that calls itself the world’s economic engine, that’s a slow walk.

But here’s the part Pew tucked into a chart and politicians of every stripe would prefer you skip: Zoom into the last five years — December 2020 through December 2025 — and every single measure shows real wages have fallen.

• CPI-U: down 3.2%
• CPI retroactive: down 3.5%
• Chained CPI: down 1.9%
• PCE: down 1.0%

That’s why you feel poorer at the grocery store. You are poorer. The post-pandemic inflation surge wiped out years of real wage gains.

The long-term arc is positive but feeble. The recent reality is brutal. That’s the answer that should worry you.

Why this affects your wallet

This isn’t an academic argument. Which inflation gauge the government uses determines actual money in your pocket — and you probably didn’t realize you were paying for it.

Take your tax brackets. Until 2018, the Internal Revenue Service (IRS) adjusted those thresholds using the main CPI. Then the Tax Cuts and Jobs Act permanently switched the calculation to chained CPI — the measure that runs lower.

According to the Bipartisan Policy Center, from January 2018 through February 2025, the Chained CPI rose 26.3% versus 28.5% for the traditional CPI. That gap sounds small. It isn’t.

It means bracket thresholds creep up slower than your raises, more of your income gets taxed at higher rates, and the IRS pockets the difference every year.

Then there’s Social Security. Your cost-of-living adjustment isn’t based on any of the four measures Pew analyzed. The Social Security Administration uses yet another variant called CPI-W, which tracks urban wage earners. For 2026, that produced a 2.8% bump — about $56 a month for the average retiree.

Critics argue CPI-W understates what older Americans actually pay for housing and medical care. I’ve written about why your COLA never feels like enough, and the math gets uglier when you factor in Medicare premium hikes.

When you plan retirement, most calculators assume inflation of 2.5% to 3% per year. If real inflation runs higher — as it has lately — your savings won’t last as long as the projections promised.

The good news? Even when inflation flares, there are practical truths that can ease your worry, especially if you’re already retired.

The bottom line

After 35-plus years of writing about money, here’s what I’ve learned: Any time someone hands you a single number to settle a complicated argument, ask what they left out.

The wages-versus-inflation fight is a textbook example. Pick the right index, and you can prove almost anything. The truth lives between the answers.

Yes, real wages have grown since 1999. Yes, they’ve grown slowly. And yes, they’ve recently gone backward. All three statements are true at the same time.

What should you do with this?

First, don’t believe anyone — politician, pundit, or in-law — who tells you “wages have kept up with inflation” as if it’s a settled fact. Ask which measure they’re using and which years they picked. Almost always, they’re cherry-picking.

Second, don’t depend on government inflation numbers to plan your own life. Track your own cost of living — your grocery bills, rent, insurance premiums, gas prices. Those numbers are personal — and they’re the only ones that matter for your budget.

The government measures inflation for the country. You need to measure it for yourself.