Purchasing Power

The engine underneath every purchasing-power claim is the consumer price index. The idea behind it is simple enough to explain over coffee and hard enough to build that entire agencies exist to do it. You pick a basket of goods and services meant to represent what a typical household buys — food, housing, transport, healthcare, clothing, a haircut, a streaming subscription. You track the price of that basket month after month. When the basket costs more, prices have risen; when it costs less, they've fallen. Divide wages by the change in the basket, and you get a rough read on purchasing power.

The trouble is hidden in the words "typical" and "basket." The index does not follow one family's spending; it follows a weighted composite assembled from large consumer-expenditure surveys. Housing gets a big share because households spend a lot on it. A category the average person barely touches gets a tiny share. The weights are meant to mirror aggregate spending across the whole population, which means the basket is a portrait of everyone and therefore of no one in particular.

The single average conceals a fact that decides almost everything about how purchasing power feels: different people buy different baskets, and prices in those baskets rise at wildly different rates. This is where the abstraction and the receipt actually diverge. A renter in a city where rents jumped fifteen percent lives in a different price world than a homeowner with a fixed mortgage locked in years ago. A family with young children spends heavily on food, childcare, and housing — precisely the categories that tend to rise fastest — while a household without those costs spends more on things whose prices are flatter. The same headline inflation rate lands on them as two different numbers.

Economists sometimes call this the difference between headline and personal inflation, and the gap can be large. Lower-income households spend a bigger share of income on essentials — food, energy, rent — that are hard to substitute away from and that have led recent price surges. When the necessities inflate faster than the luxuries, the official average, which blends both, understates the squeeze on the people with the least room to maneuver. The index reports the temperature of the whole room; the people standing next to the radiator feel something the thermometer averages away.

Step back and the recurring quarrel over purchasing power looks less like a failure of measurement and more like a feature of what measurement is. To produce a national figure at all, you must average across millions of households whose experiences genuinely differ, and the act of averaging necessarily erases the differences that any individual cares most about. The index is not a bad description of the country. It is a precise description of an entity — the average household — that does not exist and never sits down to dinner. The gap between the statistic and the receipt is not noise to be cleaned up; it is the unavoidable cost of turning a population into a number.

This has a consequence that statistical agencies themselves increasingly acknowledge. In recent years several have begun publishing supplementary measures — inflation experiences broken out by income group, by age, by household type — precisely because the single headline number, however accurate, fails the people trying to see themselves in it. That move is a quiet admission that a purchasing-power figure is not only a scientific object but a social one. It has to be defensible to statisticians and recognizable to citizens, and those two demands pull in opposite directions.

Come back to that kitchen-table argument, where one side holds the release and the other holds the receipt. Neither is spinning. The release is a faithful average of a country's basket, softened by adjustments that are individually reasonable and collectively estranging. The receipt is a faithful record of one household's necessities, bought at the frequencies that stamp themselves on memory. They diverge because they are measuring two different things and calling both of them purchasing power. The wonder is not that they disagree. It is that we expected a single number to reconcile them.