There is a particular kind of frustration that sets in when economists declare victory over inflation while you're standing in a supermarket aisle, staring at an egg carton that costs twice what it did a few years ago. The disconnect isn't imaginary, and it isn't entirely about psychology. It reflects a genuine tension between how statisticians measure price changes and how human beings actually experience them.

The Consumer Price Index, that venerable workhorse of inflation measurement, tracks a weighted basket of goods and services meant to represent the spending patterns of an average urban household. When the index rises by, say, three percent, the official verdict is that prices have increased by three percent. But averages, by their nature, obscure variation. And in inflation, the variation matters enormously.

The frequency problem

Consider the prices you encounter daily versus those you encounter rarely. Gasoline, groceries, coffee—these are the figures burned into your memory because you see them constantly. A mortgage payment, by contrast, might be locked in for decades. Health insurance premiums change once a year. The official index weights these categories by their share of total spending, which is mathematically sensible but psychologically tone-deaf.

Research in behavioral economics has consistently shown that people weight frequent purchases more heavily in their subjective sense of inflation. A twenty percent increase in egg prices registers more viscerally than a five percent increase in imputed rent, even if the latter represents a larger absolute dollar amount. The statisticians aren't wrong; they're just measuring something different from what you feel at the checkout counter.

The base effect illusion

There's another wrinkle. When inflation cools from eight percent to three percent, the headlines celebrate disinflation. What they rarely emphasize is that prices aren't falling—they're simply rising more slowly. If bread cost a dollar in 2019, jumped to a dollar forty by 2023, and now sits at a dollar forty-five, the rate of increase has slowed dramatically. But you're still paying forty-five percent more than you were. The base never resets.

This is why the phrase "prices have stabilized" can feel like a cruel joke. Stabilization means the new, higher prices are here to stay. Deflation—actual price declines—is rare outside of specific categories like electronics and is generally considered economically dangerous. So the relief promised by falling inflation statistics never quite arrives in the form people expect.

What the index misses

The CPI also struggles with quality adjustments. If a laptop costs the same as last year but has a faster processor, statisticians may record this as a price decrease in quality-adjusted terms. This is called hedonic adjustment, and it's intellectually defensible. But it doesn't change the fact that you still handed over the same amount of money. Similarly, shrinkflation—the practice of reducing package sizes while maintaining prices—is notoriously difficult to capture in official statistics, even as consumers notice their cereal boxes growing mysteriously lighter.

Then there's the composition of the basket itself. The official index assumes substitution: if beef becomes expensive, consumers switch to chicken, and the index adjusts accordingly. This is economically rational but emotionally hollow. Being priced out of your preferred choices is itself a form of diminished welfare that the numbers don't fully capture.

Our take

The gap between official inflation and felt inflation isn't a failure of statistics—it's a reminder that statistics serve a particular purpose, and that purpose isn't to validate your grocery store anxiety. The CPI was designed to guide monetary policy and adjust contracts, not to mirror the texture of daily economic life. Both measurements are real; they're just measuring different things. The frustration you feel is legitimate, and so is the three percent figure. The mature response is to hold both truths simultaneously, understanding that economic data is a map, not the territory. The map is useful. But you still have to buy the eggs.