Average Hourly Earnings: Deciphering Historical Trends

Yesterday I received an email from Ted Kavadas, the mastermind of EconomicGreenfield, regarding the most recent BLS update on Average Hourly Earnings. The St. Louis Federal Reserve data repository (FRED) has a staggering amount of data on hourly wages, over 90 series with some reaching back to the Civil War era.

Ted called my attention to a widely circulated FRED chart showing the year-over-year percent change in hourly earnings for the total private sector (first chart below). The chart looks grim, but the total private earnings data is only available from March 2006. Thus the first YoY data point is for March 2007.

There is, however, another broad series that dates from 1964 -- one tracks the hourly earnings for Production and Nonsupervisory private employees. Before we look at that complete series, here is an overlay of the two (Total Private and Production/Nonsupervisory) for the overlapping timeframe. The objective is to see the relative correlation between the two.

Now let's step back and view the complete Production/Nonsupervisory series, which dates from January 1964 (giving us a January 1965 start for the YoY data). At first glance the chart looks horrible. We are currently at a level below the start of all recessions.

But the chart above is totally misleading. The single biggest force in hourly wages over the last half century has been inflation. So let's have a look at the same chart, but this time with the hourly wages adjusted for inflation using the Consumer Price Index as the deflator and chained to today's dollar.

The "real" picture is far less grim. Yes, hourly wages, YoY, are currently just fractionally above zero (0.15% to be precise), but that's higher than the YoY number at the start of all but one of the recessions in this timeframe.

But wait. Before you write me off as a hopeless optimist on private wages, let's look at two more snapshots: nominal and real wage growth using dollar values for the vertical axis. Remember, the series that starts in the 1960s is for Production and Nonsupervisory employees. To give us a clue about the spread between the longer series and the total private series from early 2006, I've included it as well (the red line).

The first of this pair has an obvious look about it. Hourly wages have grown dramatically.

But when we adjust for inflation using the CPI, we get a very different picture. Real hourly earnings peaked in January of 1973 and hit a trough in May of 1995. Note that the peak was followed later in the year by the Oil Embargo and the onset of the era of stagflation.

There are many underlying forces at work in the chart above: the growth of women in the labor force (which began accelerating in 1973 and began peaking in the mid-1990s), gender differences in wages, the impact of automation on labor tasks, the secular trend away from manufacturing toward services, etc.

Suffice to say that, in real terms, the average hourly earnings for Production and Nonsupervisory employees, as the latest update, comes to about $39,500 for a 2000-hour work year. The average for all private employees (the series that dates from 2006) is about $7,500 more at $47,040.

Are these dramatic advances over the mid-1960s? No. When we adjust for inflation, chained in today's dollar, the average annual figure in 1964 would have been $37,072 (the comparable purchasing power of a 1964 annual salary of $5000 for Production/Nonsupervisory employees).

At the household level other trends were afoot -- not least of which was the growth of two-income households as more women entered the workforce. But the percentage of two-income households has been eroded by the Great Recession and the demographics of aging boomers entering their retirement years.

If we look at the Census Bureau's historical data, we can calculate that the ratio of two-income households to single-income households peaked in 1994 at 51.5% (see Table H-12). The ratio of two-income households to all households peaked the following year at 35.5%. As of 2010, the most recent year in the CB's published data, the ratios have dropped to 45.5% and 31.6%, respectively.

I'll close this commentary with a longer look back. The main series illustrated above was for Production and Nonsupervisory employees dating from 1964. The BLS has a series dating from 1939 for Manufacturing employees. Here it is, adjusted for inflation.

From a wages standpoint, the manufacturing-driven economy in the US peaked in January 1978. The greatest acceleration was during the WWII years, followed by a dip during the adjustment period bracketed by the first two post-war recessions. The acceleration resumed, at a reduced pace, to the peak in January 1978.

The year of the peak resonates with me personally in one respect. It was the year I had my first encounter with a personal computer, the Apple II, which was released in 1977. That fledgling Apple quickly demolished its early competition, Radio Shack's TRS-80 and the Commodore PET. Flash forward to the present: Last month we learned that Apple has set a new record for market capitalization.

The Apple II, of course, was a manufactured product, but one that introduced the masses to radically new ways to approach business and pleasure. It marked a turning point in the transition from the Era of Manufacturing to the Information Age.

I remain skeptical of average hourly earnings, nominal or real, as a leading or coincident indicator of business cycle peaks and troughs. But in the larger historical context, a chart of hourly wages tells us a great deal about once-in-a-lifetime secular trends in our economy and culture.