Average Hourly Earnings vs. the Employment Cost Index

AHE is an average of earnings currently reported by payroll establishments. The ECI is an index of how hourly labor costs change over time for a defined, fixed employment mix. AHE is stated in dollars per hour. ECI is stated as an index level and as percentage changes, not as the average dollar wage received by a worker.

That difference matters when headlines appear to conflict. AHE can fall even if no individual pay rate falls, because the jobs represented in the monthly payroll average changed. ECI is designed to reduce the influence of shifts among occupations and industries, but its total-compensation measure includes employer benefits that AHE excludes. Different movements can therefore be consistent with both series being correctly calculated.

Neither series is an individual raise, a typical worker's paycheck, or a complete account of what every employer spends on labor. To interpret a comparison, identify the worker population, earnings or compensation component, time period, and adjustment basis for each series.

The Current Employment Statistics (CES) program collects monthly employment, hours, and payroll data from a sample of nonfarm establishments. CES earnings and hours estimates are for private-sector workers. The all-employees series summarizes the private-sector employees covered by the release; it is not a survey of every worker or every household. CES employment counts jobs rather than unique people, so a person with jobs at multiple worksites can be counted at each worksite. {source:blsCesConcepts}

BLS calculates AHE as aggregate weekly payroll divided by aggregate weekly hours. In a simplified group where everyone works the same number of hours, that is equivalent to averaging hourly payroll earnings across workers. In the actual aggregate, hours worked affect the weights: someone who contributes more hours contributes more to aggregate payroll and aggregate hours.

AHE is a gross earnings measure, not a posted wage rate or take-home pay. It can reflect changes in base and incentive pay, overtime and late-shift premiums, and the relative amounts of higher-paid and lower-paid work. The series excludes employer-paid benefits, irregular bonuses, retroactive pay items, and employer payroll taxes. Those exclusions are one reason AHE and an ECI total-compensation series do not measure the same cost. {source:blsCesConcepts}

The ECI comes from the National Compensation Survey and measures changes in hourly labor costs to employers. It is published quarterly. Its reference period is the pay period that includes the 12th day of March, June, September, or December. The index reports wage-and-salary, benefit, and total-compensation measures, with percentage changes over three and twelve months. An index level tracks change from its base period; it is not a dollar amount. {source:blsEmploymentCostIndex}

The ECI applies fixed employment weights across occupation and industry groups between periodic weight updates. This design limits the effect on the index of workers moving between higher- and lower-paid occupations or industries. The weights do not make the ECI a count of the same people from quarter to quarter, nor do they turn it into an average paycheck. They define the labor mix used to summarize cost changes. BLS periodically updates the weights so the fixed mix does not become permanently detached from the economy's employment structure. {source:blsEciWeightUpdates}

The ECI's target population includes private-industry and state-and-local-government establishments. It excludes federal government, agriculture, and private households, among other exclusions. CES hours and earnings, by contrast, cover private-sector workers. A comparison of broad civilian ECI with CES AHE therefore also differs in coverage; a private-industry ECI component may be a closer scope match, though its construction still differs. {source:blsEciDataSources} {source:blsEmploymentCostIndex}

Four differences explain many gaps:

• Worker mix: AHE reflects the current payroll composition and the hours worked by those employees. The ECI holds occupation and industry employment weights fixed between updates, so shifts in the current mix have less direct influence on its change.
• What is counted as pay: AHE is gross payroll earnings per hour and can move with overtime or other premium pay. ECI wage-and-salary indexes measure compensation rates; ECI total compensation also includes employer benefit costs.
• Coverage: CES earnings are for private-sector workers, while ECI series may cover private industry, state and local government, or a broader civilian aggregate. The chosen ECI series matters.
• Timing and form: CES publishes AHE monthly as a dollar average. ECI is quarterly and publishes index levels and three- or twelve-month changes. Comparing one monthly move with a quarterly or year-over-year rate can create a difference that is only about the interval.

These are not reasons to prefer one measure in every situation. AHE is useful for current average payroll earnings per hour in its covered group. ECI is useful for tracking employer compensation-cost changes while limiting the effect of industry and occupation mix shifts. Neither one alone identifies why compensation changed.

AHE is a ratio of aggregate payroll to aggregate hours, not a simple average of posted wage rates. Changes in overtime hours or in which industries contribute more payroll and hours can affect the broad average even when rates within particular jobs are steady. Conversely, if hourly wage rates are steady but the employer cost of benefits rises, AHE does not record that benefit-cost increase while ECI total compensation may. Each measure filters a different part of the labor-cost picture.

Suppose a private-sector group has two kinds of jobs. Eighty employees earn $20 per hour and 20 employees earn $40 per hour. Assume everyone works the same number of hours and that hourly rates and benefits stay unchanged.

The simplified AHE falls by about 7.4%, calculated as $22.22 ÷ $24.00 − 1. Yet no remaining employee's hourly rate fell. The result changed because fewer higher-paid jobs were represented in the current mix. Equal hours make the headcount-weighted arithmetic match the payroll-divided-by-hours formula; if hours differed, the hours would also affect the average.

In this deliberately simple example, if the ECI's sampled wage and benefit costs in both job groups stayed unchanged, the fixed-weight ECI would show no cost increase from the mix shift. In real data, ECI also reflects cost changes in sampled occupations and industries, and its fixed weights are periodically updated. The example isolates one reason the monthly average and the fixed-mix index can move differently; it is not an estimate of either BLS series.

<!-- learn:illustration -->

First choose the question. For average gross payroll earnings per hour among private-sector employees, use CES AHE. For employer cost changes that include benefits and reduce the influence of occupation and industry mix shifts, use ECI total compensation. For a cash-pay comparison, use an ECI wage-and-salary measure and state that the fixed weights and survey coverage still differ from CES.

Then align the interval. An ECI three-month change compares consecutive quarterly reference periods; a twelve-month change compares with the same quarter a year earlier. For an AHE comparison, calculate the percentage change between monthly observations that span the same endpoints. Do not compare a one-month AHE change with an ECI twelve-month rate, or an ECI index level directly with dollars per hour.

If a release places a monthly AHE change beside a quarterly ECI change, their proximity in a table does not make the rates directly comparable. The ECI percentage covers three months, while the AHE figure may cover only one. For a like-period comparison, use the AHE observations at the corresponding quarter endpoints and label the result as a three-month change.

Also keep seasonal adjustment consistent. If one rate is seasonally adjusted and the other is not, say so or choose matching series. And specify which ECI population and component you used: private industry or civilian workers, wages and salaries or total compensation. The CES and CPS guide explains a related scope issue: payroll and household surveys count different things.

CES estimates of employment, hours, and earnings are preliminary when first released. BLS revises them twice as additional and corrected respondent data arrive, and seasonal factors can also change seasonally adjusted estimates. Annual employment benchmarking reanchors employment estimates; when broad summary earnings series are reaggregated using revised employment weights, some summary-level earnings estimates can change. At basic industry levels, BLS says average hourly earnings themselves do not change from the benchmark. Keep the data vintage and adjustment status with any comparison. {source:blsCesRevisions} {source:blsCesCalculation}

ECI seasonal factors are recalculated annually, and seasonally adjusted ECI history for the most recent five years is revised in that process. The fixed employment weights also change periodically. A long comparison that crosses a weight or classification update should be read with the relevant BLS methods notes, rather than assuming the fixed basket has been identical forever. {source:blsEciSeasonalAdjustment} {source:blsEciWeightUpdates}

Both are sample-based statistical estimates. AHE can be influenced by hour mix and changes in the industries and workers represented in the current payroll data. ECI controls for specified occupation and industry shifts, but it is not a record of each worker's wage, hours, take-home pay, or benefit use. Revisions, coverage differences, and sampling uncertainty are part of interpreting either series, not proof that one is wrong.

Use AHE when the question is how average gross payroll earnings per hour changed in the CES private-sector group. Use ECI when the question is how employer hourly compensation costs changed for its specified fixed employment mix. If the headline says “wages,” name whether it means AHE, ECI wages and salaries, or total compensation; those labels prevent a reader from mistaking a mix-sensitive average for a fixed-mix cost index.

For the difference between nominal and inflation-adjusted pay, see nominal vs. real wages. For how labor cost, productivity, and output combine, see unit labor costs and productivity. Those comparisons use related concepts but do not make AHE and ECI interchangeable.