A 10% pay rise does not necessarily mean 10% more purchasing power. To understand a raise, compare the nominal wage with the change in consumer prices over the same period. In one worked example, a 10% nominal hourly increase alongside 8% CPI inflation produces approximately 1.85% real hourly growth, not an exact 2% increase.
This guide shows the full calculation and then changes paid hours, the price index and the composition of workers to reveal how apparently contradictory wage headlines can all be true. We distinguish real hourly earnings from weekly income and explain why a national wage average is not necessarily the experience of the individual employee.
eduKateSG · The Gold Standard of Economics
Choose your reading route
Calculate purchasing power · Chapters 1–3: nominal wage, CPI and exact growth.
Interrogate the headline · Chapters 4–6: paid hours, price indices and worker mix.
Test the evidence · Chapters 7–8: FAQs, linked economics and sources.
Chapter index
Calculate purchasing power · Chapters 1–3
Interrogate the headline · Chapters 4–6
Test the evidence · Chapters 7–8
CHAPTER 1 OF 8
1. A larger pay packet can buy less—or more than the headline suggests
A salary increase is often welcomed because the number printed on a payslip is larger. Yet what that money can purchase depends on consumer prices. Economists distinguish nominal pay, measured in current money, from real pay, which adjusts that money value for movements in an appropriate price index.
Suppose the hourly wage shown in a fictional worker’s payslip rises from 25 currency units to 27.50. That is a nominal increase of 10%. It does not by itself establish that the worker can purchase 10% more goods and services. If the general cost of the chosen reference consumption basket also rises, part or all of the extra money is needed to purchase the same items as before.
The U.S. Bureau of Labor Statistics explains that the Consumer Price Index measures the average change in prices paid for a representative basket of consumer goods and services, and that it is commonly used to express income in real, purchasing-power terms. The OECD likewise reports real hourly wages by deflating nominal wages using a consumer price index.
But ‘real’ is not shorthand for one person’s complete financial situation. A broad CPI is designed for a defined population and representative spending weights. Different households spend different shares on food, housing, transport, childcare and medical needs. Their experienced price changes need not match the overall index exactly.
The first question is therefore precise: how did nominal hourly pay change, how did the relevant consumer price index change over the same period, and what happened to the ratio? The broader question about living standards additionally needs hours worked, taxes, transfers, costs faced by that household and other circumstances.
Use two invented annual observations expressed in the same currency and comparable hourly-work definition. In the first year the nominal hourly wage is 25.00 and the CPI index is 100. In the second year the nominal wage is 27.50 and CPI is 108. The worker’s nominal hourly pay grew by (27.50/25.00)−1 = 10%. The representative price index grew by (108/100)−1 = 8%.
To express the second year’s wage in first-year purchasing-power units, divide by the price ratio: 27.50/(108/100) = 25.46296, or approximately 25.46 in first-year money. The worker began at 25.00 in first-year money, so the corresponding real gain is about 0.46 per hour on that purchasing-power basis.
The exact real growth rate is (27.50/25.00)/(108/100)−1 = 1.10/1.08−1 ≈ 1.85%. It is common to approximate real wage growth by nominal wage growth minus inflation, giving 10%−8%=2 percentage points. The approximation is close here, but it is not the exact percentage change.
Why does the tiny difference matter? Because growth rates are multiplicative. A pay increase applies to one base and a price change applies to another. The exact comparison divides their gross growth factors, rather than treating the two percentage changes as quantities simply to subtract.
Check the units. The CPI is an index without currency units, so its year-to-year ratio is dimensionless. Dividing a nominal wage by that ratio leaves a wage in the base year’s purchasing-power units. A CPI level of 108 is not itself a price or a wage; it only has meaning with its index definition and reference period.
| Measure | Year 1 | Year 2 |
|---|---|---|
| Nominal hourly wage | 25.00 | 27.50 |
| CPI index | 100 | 108 |
| Price ratio relative to Year 1 | 1.00 | 1.08 |
| Hourly wage in Year 1 prices | 25.00 | ≈25.46 |
| Real hourly change relative to Year 1 | — | +1.85% |
CHAPTER 3 OF 8
3. The exact growth formula explains when a raise loses purchasing power
Let g_w denote the nominal wage growth rate and π denote the CPI inflation rate over the same interval, both expressed as decimal fractions. The exact real wage growth rate is (1+g_w)/(1+π)−1. This follows directly from calculating the wage ratio and dividing by the price-index ratio.
If nominal pay grows 10% but consumer prices rise 15%, the calculation is 1.10/1.15−1 ≈ −4.35%. The worker’s nominal wage is higher, but its purchasing power relative to that basket is lower. The rate of change of prices, not the presence or absence of a pay rise alone, decides the sign.
Conversely, if pay grows 10% while prices grow only 2%, the result is 1.10/1.02−1≈+7.84%. Both examples use the same nominal raise but produce very different real outcomes. It is more informative to compare the two ratios than to say ‘inflation has eaten the raise’ without calculation.
Now consider nominal pay unchanged while consumer prices fall by 2%. Real wages would rise by 1/0.98−1≈+2.04% in the simple model. Deflation can increase purchasing power for a given pay amount, though it may also be associated with wider economic stresses not captured by this arithmetic.
An annual inflation rate falling from 8% to 3% does not mean the overall price level has fallen. It means the level is growing more slowly in that period. Similarly, a real wage still below its earlier level can nevertheless be recovering if wages recently grow faster than prices.
Use a period-consistent index, and be alert to revisions and seasonal adjustment when drawing short-term conclusions. An annual average, a December-to-December comparison and a seasonally adjusted monthly comparison are not interchangeable data windows.
| Nominal wage growth | CPI price growth | Exact real wage growth |
|---|---|---|
| +10% | +8% | ≈+1.85% |
| +10% | +15% | ≈−4.35% |
| +10% | +2% | ≈+7.84% |
| 0% | −2% | ≈+2.04% |
An hourly wage is not a weekly pay packet. The amount a worker earns in a week also depends on how many hours are paid or worked under the relevant pay rules. A person can experience a real hourly wage gain and still receive less real weekly pay if hours fall sufficiently.
Return to the toy worker who earns 25.00 per hour in Year 1 and 27.50 in Year 2, with CPI rising from 100 to 108. Suppose the worker works 40 paid hours weekly in Year 1 but 36 in Year 2. Weekly nominal earnings then move from 25×40=1,000 to 27.50×36=990.
Nominal weekly earnings fall by 1%, even though the hourly rate rose 10%. In first-year prices the second week’s 990 has purchasing power 990/1.08≈916.67. Relative to 1,000 in the first year, real weekly earnings have declined by about 8.33%.
That does not prove that shorter hours are necessarily harmful. The worker may have voluntarily reduced hours, taken paid leave accounted for elsewhere or gained time for study, caregiving or recovery. Income, total compensation, nonwage benefits and quality of life are separate dimensions. The arithmetic simply demonstrates that an hourly wage statistic and a weekly income statistic measure different things.
Official statistics often publish both real average hourly earnings and real average weekly earnings. The BLS Real Earnings release explicitly relates changes in weekly earnings to both real hourly earnings and the average workweek. Readers should pay attention to the measure used in a headline.
More generally, a worker may receive bonuses, overtime premiums or irregular commissions. Comparing base hourly pay without total earned compensation may be useful for contract analysis but incomplete for household cash-flow analysis. Name the numerator, denominator and period before drawing a living-standards conclusion.
| Quantity | Year 1 | Year 2 |
|---|---|---|
| Hourly nominal wage | 25.00 | 27.50 |
| Paid hours per week | 40 | 36 |
| Nominal weekly pay | 1,000 | 990 |
| CPI ratio against Year 1 | 1.00 | 1.08 |
| Real weekly pay in Year 1 prices | 1,000 | ≈916.67 |
| Real weekly change | — | ≈−8.33% |
Real wages are commonly deflated with a consumer price index because the reader asks about household purchasing power. The CPI tracks prices associated with consumer expenditure, whereas other indices such as the GDP deflator describe a different production-based scope. The two can move differently, especially when import or export prices change.
A household with substantial spending on imported fuel or food may be affected in ways not captured by a domestic-output price index. Using a GDP deflator to claim that a particular worker’s consumer purchasing power rose is therefore not automatically an appropriate substitution for using a consumer price measure.
Read the index coverage carefully. In the United States the BLS uses CPI-U to deflate certain all-employee earnings series and CPI-W for production and nonsupervisory employee series. Other economies may publish national CPIs, harmonised consumer price indices or different target populations. These labels are not mere stylistic differences.
For comparisons between countries, converting two wages into a common currency at a market exchange rate does not automatically produce comparable local purchasing power. Price levels vary; purchasing power parity measures may be needed for certain international comparisons, with their own assumptions and limitations.
Different household budgets can have very different personal price experiences. Renters, owner-occupiers, students and households with high transport or caregiving expenses may face spending patterns unlike the index’s representative weights. A CPI-deflated wage is a useful population-level indicator, not a promise that every household’s exact bundle grew by the same percentage.
The connected CPI, GDP deflator and import-price guide explains this distinction more fully. This article concentrates on the wage calculation and the labour-market interpretation rather than repeating the whole theory of price indices.
An economy-wide average hourly wage can rise even when some workers receive no raise. The measure may change because wages rose within jobs, the composition of employment changed, high-wage sectors added more workers or low-wage positions disappeared. This is why a national average series cannot be read as the personal pay history of a typical identified worker.
Suppose a fictional workforce has 100 employees: 50 earn 20 per hour and 50 earn 40. The simple headcount-weighted mean is (50×20+50×40)/100=30 per hour. If 25 of the 20-per-hour jobs disappear and nobody’s own wage changes, the new 75-person mean becomes (25×20+50×40)/75=33.33 per hour.
The average has increased by about 11.1% while every remaining individual’s nominal rate is unchanged. That is not evidence that all workers got an 11.1% raise; it reflects who remains in the observed group. Real-wage deflation cannot undo the employment-composition effect.
For distributional questions, examine median wages, wage percentiles, sector composition, employment status and hours alongside an average. A median answers a different question from the arithmetic mean. Even the median can shift when the workforce composition changes, so careful longitudinal evidence is valuable.
Similarly, an average real wage increase does not establish that every worker’s purchasing power improved. Some earn less, some face different price changes, and some have fewer paid hours. The appropriate conclusion is a statement about the measured aggregate population under the data source’s definitions.
When interpreting wages internationally, consider informal employment, self-employment, compensation in kind, unequal survey coverage and differences between employees and all people who work. Narrowly defined payroll data are powerful within their intended scope but should not be presented as a census of all livelihoods.
| Workforce scenario | Count at 20/hour | Count at 40/hour | Average wage |
|---|---|---|---|
| Before job losses | 50 | 50 | 30.00/hour |
| After 25 low-wage jobs leave | 25 | 50 | ≈33.33/hour |
What does a nominal wage raise measure?
It measures a change in current-money pay under a specified pay definition. It does not adjust for changing consumer prices.
How is real wage growth calculated exactly?
Divide the gross nominal wage growth factor by the corresponding consumer-price growth factor and subtract one.
Why is 10% pay growth minus 8% inflation not precisely 2% real growth?
Because exact growth is multiplicative: 1.10/1.08−1≈1.85%. The 2-point difference is a useful first-order approximation.
Does falling inflation mean grocery and rent prices have fallen?
No. Falling inflation usually means prices are increasing more slowly, unless the price-index growth rate itself becomes negative.
Can real hourly earnings rise while real weekly earnings fall?
Yes. Paid hours can decline enough to outweigh an hourly purchasing-power gain.
Does CPI equal the personal inflation rate of every household?
No. CPI represents a defined population and basket weights; household spending patterns differ.
Can the average wage rise even if nobody remaining got a raise?
Yes. Changing the mix of people and jobs can change an observed average.
Can I compare wage purchasing power using the GDP deflator?
Not as a direct substitute without justification. The GDP deflator and consumer price indices refer to different baskets and economic scopes.
Does a positive real wage figure guarantee improving living standards?
No. Work hours, employment, taxes, transfers, nonwage benefits, household composition and distribution also matter.
An accurate wage story has four parts: the monetary pay rate, the consumer-price index, the hours that determine total earnings and the population whose wages are being averaged. The first two determine a real hourly wage measure. The third determines what arrives in weekly or monthly pay. The fourth determines whether an aggregate statistic resembles any particular person’s experience.
Test the original example with only the CPI changed. If the wage still rises from 25 to 27.50 but the CPI rises from 100 to 115 rather than 108, exact real hourly growth becomes 1.10/1.15−1≈−4.35%. No wage data were changed. One changed price assumption reversed the purchasing-power claim.
Subject and linked economics routes
The Gold Standard of Economics — parent subject guide and related evidence.
CPI, the GDP deflator and import prices — which prices belong to which economic question.
GDP per worker versus GDP per hour — another denominator distinction that changes an economic headline.
The Gold Standard of Work — connect paid work with wider social conditions.
Learning Transfers into Life — apply ratios, units and definitions to new evidence.
How X Works — connect to the larger explanatory ecosystem.
Authoritative sources and scope
Sources reviewed on 22 October 2026. All wage levels, price-index values, hours and workforce examples in this article are fictional. The calculations illustrate indicator definitions; they are not a claim about current wage growth in any particular country.
U.S. Bureau of Labor Statistics: Income and the CPI — nominal versus real income and the consumer-price purchasing-power interpretation.
BLS Handbook of Methods: Compensation calculation — consumer-price deflation and real compensation.
OECD Employment Outlook 2026 — real hourly wage definitions and evidence interpretation.
BLS Real Earnings, June 2026 — distinct real hourly and weekly earnings measures and the workweek.