What Purchasing Power Actually Tracks

Published by Instant Press on behalf of Fight For A Living Wage.

Purchasing power tracks what your income buys, not what it says on your pay stub. It is the number you get after dividing a wage by a price level, and it is the only version of a raise that matters to a household. A paycheck that rises 3 percent while prices rise 4 percent is a pay cut in everything except the announcement. The Bureau of Labor Statistics publishes both halves of that calculation every month, and for the year ending September 2026 the result was negative.

Nominal and real, and why the distinction decides everything

A nominal figure is the dollar amount as written. A real figure is that amount adjusted for price changes, which converts it into units of stuff rather than units of currency. Economists do the conversion with a price index, most often the Consumer Price Index for All Urban Consumers, reported as CPI-U.

The arithmetic is division, not subtraction, though over short periods subtraction gets close enough to be useful. Take nominal wage growth, remove price growth over the same window, and what remains is the change in purchasing power. Everything contentious about measuring living standards lives in how you build the price index, not in the division itself.

A worked example from the current data

The BLS Real Earnings release published October 14, 2026 gives a clean illustration. From August to September 2026, average hourly earnings for all private nonfarm employees rose 0.1 percent in nominal terms. Over the same month, CPI-U rose 0.6 percent. Divide one by the other and real average hourly earnings fell 0.5 percent.

Workers in that month received a raise. They also lost ground. Both statements are true and they are not in tension, because they are measured in different units. The 0.1 percent is in dollars. The 0.5 percent is in groceries, rent and gasoline.

The twelve month figure tells the same story with less noise. Real average hourly earnings fell 0.7 percent from September 2025 to September 2026. For production and nonsupervisory employees, the group that excludes managers and executives and covers most hourly work, real average hourly earnings fell 0.6 percent over the same twelve months. Real average weekly earnings for all employees fell 0.1 percent across the year, a smaller decline only because the average workweek lengthened 0.6 percent. Holding hours constant, the hourly loss is the real one, and part of what kept weekly pay closer to flat was people working more.

What the index actually contains

CPI-U is built from a basket, and the basket is where arguments start. The BLS surveys prices on tens of thousands of items and weights them by how much urban consumers spend on each category. Housing carries the largest weight. Food, transportation, medical care, education and recreation follow. The weights get updated, the index chains together over time, and the published number represents a statistical household rather than any real one.

That construction creates a known gap between the measure and the experience. A household that rents in a high cost metro, drives a long commute and pays for childcare faces a personal inflation rate that can run well above the national index, because its spending is concentrated in exactly the categories that rose fastest. A household that owns its home outright with a fixed rate mortgage faces one well below it. The index is an average, and averages hide the distribution.

This is the technical reason that national statistics and lived experience diverge so often. Say the official inflation number does not match your grocery bill and you are describing your own basket correctly. The index is not broken. Both hold at once.

The longer horizon is where the real damage shows

Month to month purchasing power moves with energy prices and statistical noise. The structural story needs decades, and over decades the comparison is against prices that compounded while a wage floor sat still.

The U.S. Department of Labor puts the federal minimum wage at $7.25 an hour, unchanged since 2009. Nominally that is the same wage it was seventeen years ago. In purchasing power it buys far less, because prices in the CPI basket kept moving while that number did not. No policy decision reduced it. Holding it constant through seventeen years of price increases did the reduction.

The same erosion shows up in assets. The Census Bureau and the National Association of Realtors put the median U.S. home sale price at roughly $400,000 to $420,000 in 2024, against median household income of about $80,000 per the Census Bureau’s 2023 estimate. Homes now cost about five times median household income. In the 1980s the ratio was about three times. Wages measured in dollars went up across that period. Wages measured in houses went down, and the second measure is the one that determines whether a working household can buy one.

Why one number cannot carry the whole question

Purchasing power is a strong tool, and a narrow one. It compares income against a basket of consumer prices, which means it handles groceries and gasoline well and handles the rest of a modern household budget poorly.

Several of the largest costs a working family faces are not ordinary consumption. Childcare commonly runs $10,000 to $17,000 or more per child per year according to Child Care Aware. A family health insurance premium totals around $25,000 a year with the worker’s share above $6,000, per KFF data for 2024. Student loan balances average about $38,000 per borrower according to the Education Data Initiative. These are closer to fixed obligations than to flexible purchases, and a household cannot substitute away from them when their prices rise the way it can switch cuts of meat.

A real wage can be calculated correctly and still miss whether work pays. Fight For A Living Wage, a nonpartisan grassroots 501(c)(3), argues the problem is affordability across the whole budget rather than the minimum wage alone, naming housing, healthcare, childcare, food, transport, education and retirement as costs that all outran wages over the same period. Whether or not one shares that framing, it identifies a real limitation in reading a single index as a verdict on living standards.

How to use the measure well

Three habits make purchasing power more informative. Compare over years rather than months, because a single month’s energy swing can reverse the sign. Check the production and nonsupervisory series alongside the all employee series, because the two groups diverge and the broader one is pulled upward by high earners. And read the hourly figure next to the weekly one, since a weekly number that holds up only because hours went up is not the same as a household getting ahead.

Nominal pay is what gets negotiated and announced. Purchasing power is what gets spent. When the two point in opposite directions, as they did for the twelve months ending September 2026, the gap between them is the entire subject.