Where the increase sits

Average hourly earnings rose 0.3 per cent from July to August, and real average hourly earnings — the same figure with price change taken out — fell 0.1 per cent. Real average weekly earnings rose 0.2 per cent over the same month, and the reason is the average workweek, which lengthened 0.3 per cent.[1]

The other side of the subtraction that produces the fall is the consumer price index: all items rose 0.4 per cent in August, ahead of the 0.3 per cent nominal gain in hourly pay. A household that worked the same hours in August as in July was paid more and could buy slightly less with it.[1], [2]

Which basket carries it

The 3.4 per cent twelve-month increase is spread unevenly across the basket. Energy was 16.3 per cent higher than a year earlier and gasoline 27.4 per cent higher, while core prices, without food and energy, rose 2.4 per cent. Within August itself, energy rose 2.1 per cent and gasoline 3.9 per cent, against 0.1 per cent for food and 0.3 per cent for shelter.[2]

Whether that 3.4 per cent reaches every household in the same proportion depends on how large a share of a budget goes to fuel, and the release publishes the basket, not that share. A household that drives to a fixed shift and one that can leave the car at home meet the same 27.4 per cent on different terms, and neither of them appears separately in the figure.[2]

The twelve-month account

Over the twelve months to August, real average hourly earnings fell 0.3 per cent and real average weekly earnings rose 0.3 per cent. The two figures are the same size and point in opposite directions, and the only thing between them is time at work.[1]

The same split between the pay rate and the length of the week was the subject of an earlier column here, on second-quarter earnings data where part of a faster headline came from a longer working week. The August figures repeat the pattern with the price adjustment added: there the weekly measure grew, and here it grows while the hourly measure declines.[3], [1]

If the average workweek stops lengthening while prices keep rising at the August pace, the weekly measure loses the component holding it up, and real average weekly earnings turn negative in a monthly release. That is the thing to watch before the end of the year, and it is checkable in the same two lines.[1]