The government announces that inflation has cooled to a manageable figure. You read the headline, then you buy groceries, pay rent, and fill a prescription. The cognitive dissonance is immediate and infuriating. Either the statisticians are lying, or you are losing your mind. Neither is true, but the explanation is more interesting than either accusation.

The Consumer Price Index, the number that dominates headlines, is an average across hundreds of millions of households with wildly different spending patterns. It weights housing at roughly a third of the basket, food at around thirteen percent, transportation at about fifteen percent. But these weights reflect the spending of a statistically average American—a creature as real as the average family with 2.3 children. If you rent in a major city, spend disproportionately on childcare, or drive an older car that demands premium fuel and frequent repairs, your personal inflation rate may diverge sharply from the official figure.

The substitution problem

Statistical agencies adjust for what economists call substitution: if beef prices surge, consumers shift to chicken, and the index partially reflects that behavioral change. This is methodologically defensible—it captures what people actually buy—but it also means the index can understate the pain of being priced out of what you wanted. You did not choose chicken because you discovered a new passion for poultry. You chose it because beef became a luxury. The index records the adaptation; it does not record the loss.

Housing presents an even thornier measurement challenge. The CPI uses a concept called owners' equivalent rent, which estimates what homeowners would pay to rent their own homes. This smooths out the wild swings of actual housing markets and mortgage rates. When home prices spike, the index lags reality by months or years. When mortgage rates double, the index barely notices, because it is measuring an imputed rent, not your actual monthly payment.

The frequency illusion

Psychology compounds the statistical gap. Humans notice price increases on items they buy frequently—milk, gasoline, coffee—far more than they notice stable or falling prices on items they buy rarely, like televisions or airline tickets. A five percent increase in your daily coffee registers emotionally as a daily insult. A twenty percent decline in the price of a laptop you buy once every four years is invisible. The index weights these correctly by expenditure share, but your brain does not.

There is also the quality adjustment problem. If a new car costs more but includes safety features and fuel efficiency that previous models lacked, statisticians adjust the price downward to reflect the added value. This is intellectually honest—you are getting more car for your money—but it offers cold comfort when you simply needed a vehicle and the sticker price emptied your savings.

Our take

The inflation number is not a lie, but it is not your truth either. It is a population-level abstraction designed for macroeconomic policy, not for validating your grocery receipt. The next time a headline announces that inflation is under control, remember that the index is answering a different question than the one your bank account is asking. Both questions are legitimate. Confusing them is where the rage begins.