A bigger paycheck can still buy less

What happened

A paycheck can grow while the shopping it pays for shrinks. The Bureau of Labor Statistics put that tension into numbers on October 21. Median usual weekly earnings for full-time wage and salary workers were $1,232 in the third quarter of 2026, up 1.5 percent from a year earlier. Consumer prices, measured by CPI-U, rose 3.5 percent over the same period. These quarterly earnings figures are not seasonally adjusted.

The report describes a population midpoint, not a raise received by every worker. Still, the comparison answers an important question: did the reported median keep pace with the broad price benchmark? This time, it did not. To understand why that matters, separate the dollars on a paycheck from what those dollars can purchase.

Sources: [current]

The economic concept

Nominal earnings are expressed in current dollars. Real earnings adjust those dollars for price changes. If earnings increase more slowly than prices, the purchasing power represented by that earnings measure declines. Subtracting the inflation rate from earnings growth gives a useful approximation; dividing the growth factors gives a more accurate calculation.

Using the rounded rates in the release, 1.015 divided by 1.035, minus one, is about minus 1.9 percent. That is our calculation from published rounded inputs, not a separately reported BLS estimate. It describes the inflation-adjusted change in the median, rather than the experience of an unchanged group of individual workers.

Consider a fictional worker earning $1,000 per week who receives a 2 percent raise. Pay becomes $1,020. If the same basket of purchases that previously cost $1,000 now costs $1,040, the larger paycheck cannot buy the old basket. The worker has more dollars but less purchasing power. This example deliberately holds the basket constant so the arithmetic is easy to see.

For a household budget, the relevant spending mix also matters. A renter, a homeowner with a fixed mortgage, and a household facing substantial medical bills can experience different price pressures. CPI summarizes a broad population’s purchases. It cannot tell any one family exactly how its bills changed.

Sources: [current][prices]

The historical parallel

The third quarter of 2021 provides a clear comparison. In its October 19 release that year, BLS reported median weekly earnings of $1,001, an increase of 0.7 percent from a year earlier. CPI-U had increased 5.3 percent over the same period. Nominal earnings were rising, yet prices were rising substantially faster.

The shared mechanism is the gap between money earnings and purchasing power. Using those rounded historical rates gives an inflation-adjusted change of roughly minus 4.4 percent. The 2021 gap was larger than the current one. Matching the sign of the two calculations does not establish that the causes, severity, or duration are identical.

This is why a headline saying wages rose can coexist with reports of financial strain. A higher dollar amount does not settle whether an income can cover the same purchases. Equally, the median alone cannot prove that all workers lost ground. It is a summary of a distribution, and the workers represented by that distribution can change.

Sources: [history][definitions]

What happened afterward

Two years later, the comparison had changed direction. BLS’s October 18, 2023 release reported third-quarter median weekly earnings of $1,118, up 4.5 percent from a year earlier, while CPI-U rose 3.5 percent. The same calculation gives approximately 1.0 percent growth in inflation-adjusted median earnings over that particular year.

That is a documented improvement in the annual growth comparison by the third quarter of 2023. It does not show that every household recovered earlier losses. Nor does a positive one-year change establish that the purchasing-power level regained a previous peak. A measure can begin climbing while remaining below an earlier position. The observation also does not isolate which policy or economic force caused the turnaround.

Sources: [outcome]

What is different today

The current report is a new observation, not the opening scene of a replay. The numerical gap is smaller than in the 2021 comparison. The historical outcome took us to a specific quarter two years later; it supplies no deadline for a similar outcome now. Future earnings, prices, employment, and the mix of workers remain uncertain.

Coverage is another limit. CPS earnings measure pay before taxes and other deductions and include usual overtime, commissions, and tips. Self-employed people are excluded. A change in this statistic is therefore not a change in every household’s disposable income. Taxes, transfers, the number of earners, and working hours can alter a family’s circumstances in ways the headline does not capture.

Composition can matter even without changes in individual pay. Imagine five workers earning 400, 500, 600, 700, and 800 dollars. The middle value is 600. If the lowest-paid worker leaves and a worker earning 900 enters, the middle becomes 700 even though none of the remaining workers received a raise. This is an illustration of a statistical possibility, not an explanation established for this quarter.

Sources: [current][history][definitions]

What to watch next

Read earnings and prices over matching periods, and label whether figures are seasonally adjusted. Keep the population definition beside the number. Then distinguish three questions: are dollar earnings rising, is the inflation-adjusted measure rising, and has its level recovered an earlier loss? They are different questions and can have different answers.

For the next comparison, look for sustained evidence rather than projecting one quarter forward. The historical record shows that the growth gap can reverse. It does not promise when, why, or for whom that will happen again.

Sources: [current][outcome][prices]

Discuss it

What evidence would show that purchasing power recovered, rather than merely stopped falling?

Keep learning

• Inflation basics
• Economic growth and cycles

Sources

1. BLS: Third-quarter 2026 usual weekly earnings, October 21, 2026 · Accessed October 26, 2026
2. BLS: Third-quarter 2021 usual weekly earnings, October 19, 2021 · Accessed October 26, 2026
3. BLS: Third-quarter 2023 usual weekly earnings, October 18, 2023 · Accessed October 26, 2026
4. BLS: CPS earnings concepts and coverage · Accessed October 26, 2026
5. BLS: CPI frequently asked questions · Accessed October 26, 2026