Average Hourly Earnings YoY

Average Hourly Earnings YoY is the 12 month percent change in gross pay per hour for private nonfarm payrolls after seasonal adjustment. It is published by BLS, the federal labor statistics agency, from the Current Employment Statistics survey of employers. The survey covers about 121000 businesses representing about 631000 worksites for the pay period including day 12 of the reference month.

It covers employees who worked or received pay for any part of that pay period, with each job counted separately. It is not a wage rate or a measure of employer labor cost. It excludes benefits, irregular bonuses, retroactive pay, tips and payments in kind, and it excludes farm workers and the self employed.

Responses come from payroll records for the pay period including day 12 of the reference month. Results appear monthly in the Employment Situation release at 8:30 Eastern Time under embargo. Timing shifts around federal holidays.

The 2 months before each release are revised with late reports and new seasonal factors. Each year the sample is aligned to near complete tax records and seasonal factors for the 5 years before each release are recalculated. The release also brings household survey results and establishment tables, while a separate real earnings release deflates nominal pay with consumer prices.

The index has no fixed basket. It is an hours weighted average across industries. Hours heavy industries shape the headline and shifts between high paid and low paid jobs can move it with no change in pay.

• Total private: the headline average across all private industries weighted by worker hours.
• Mining and logging: pay in extraction and logging jobs where hours swings can shift the average.
• Construction: pay for building site workers where hours and seasonal patterns matter.
• Manufacturing: factory pay with overtime detail and the only series excluding overtime.
• Private services: pay across trade, transport, utilities, information, finance, business, health, education, hospitality and other services.
• Production and nonsupervisory employees: a narrower worker group showing whether gains are broad or concentrated.

In one published example, average hourly earnings rose by 13 cents to 31.73 dollars, up 5.6 percent over 12 months. That 5.6 percent used revised history at 30.06 dollars, not the first reported 29.96 dollars, which would have implied 5.9 percent. That release also showed the workweek fell by 0.1 hour to 34.6 hours while earnings rose. Narrow group pay rose by only 11 cents to 27.06 dollars, showing gains were uneven across workers.

• Workplace rate: for each matched workplace reporting in both months, divide gross payroll by worker hours with sample weights. For example, payroll of 115000 dollars over 3800 hours gives 30.26 dollars per hour.
• Monthly blending: combine the earlier estimate with weight 0.9 and the matched sample change with weight 0.1 for continuity, while down weighting unusual reports. For example, an earlier 30.00 dollars with an implied 30.30 dollars gives about 30.03 dollars.
• Aggregation: multiply each cell average by hours to get payroll aggregates, then divide summed payroll by summed hours. For example, 115000 dollars plus 30000 dollars from a second cell of 1200 hours at 25.00 dollars gives 145000 dollars over 5000 hours, which is 29.00 dollars.
• Seasonal adjustment: use X-12-ARIMA with new factors each month and corrections for reference intervals and moving holidays, building broad groups from adjusted components.
• Year over year rate: compute 100 times current minus year ago divided by year ago on adjusted levels. For example, 31.73 dollars against 30.06 dollars gives 100 times 1.67 divided by 30.06 to give 5.6 percent over 12 months.

Key point: a rising average does not mean every worker received a raise. Layoffs or rehires among lower paid workers can raise or lower the average with no change in pay.