Wage growth during the current business cycle has been weak compared to the past several economic expansions. The May U.S. employment report implied a growth rate in wages of 3.1%, which was accompanied by a 2.73% three-month average annualized pace, pointing to broader deceleration in the growth of overall wages. Both metrics have declined for the past three months and it’s likely that wage growth has peaked in the current business cycle, which strongly suggests that the U.S. economy has entered the latter stages of the economic expansion.
In the U.S. economy, the consumer sector plays an outsize role in the level of overall economic activity (see Figure 1). Therefore, the health of household balance sheets and the willingness of consumers to purchase goods and services and to invest in homes are key determinants of the direction of the business cycle.
However as illustrated in Figure 1, the pivot points of consumer spending tend to lag the turning points of GDP growth. Expectations of sustained robust growth in the wholesale and retail sectors, as well as strong demand for the provision of services, often result in erroneous business decisions by small and medium-size firms that do not have the resources and data analytics to identify turning points in the business cycle. This is one reason why these businesses lean toward excess in hiring, as well as mergers and acquisitions or over-expansion near the end of business cycles. In our estimation, employment growth and wage growth offer a better real-time assessment of where we are in a business cycle. While sales and new orders appear to remain solid, firms should proceed with caution as hiring, wages and economic growth decelerate.
At the end of a business cycle expansion, we expect the labor force participation rate to increase noticeably and the growth rate of employment to decelerate as the pool of available labor decreases – particularly in the absence of large increases in wages. In the current post-crisis-recovery period, employment growth peaked at 2.25% year-over-year rate at the end of 2014 and has declined in recent months to less than a 1% rate (see Figure 2).
Wages are also showing signs of peaking in recent months. Weekly payrolls in the goods-producing sector decelerated from a 6.0% year-over-year growth rate in mid-2018 to a 4.0% rate in the spring of 2019 (see Figure 3). In the service sector, which exhibits a smaller range of growth, payroll growth has dropped from 5.7% in the third quarter of 2018 to 5.0% in May 2019 (see Figure 4).
So, can we use the growth rate of earnings to signal of the end of a business cycle upswing and the presence of an impending recession? Our research indicates that, yes, we can. Note that Figure 5 is suggesting a shift in structure occurred in 1981, as companies began to substitute technology for labor, labor unions lost their political clout and traditional manufacturing began to migrate to right-to-work, low-wage states as well as lower-cost off-shore locations. The wage growth gains throughout the 1960s and 70s collapsed overnight from 9% increases per year in 1981 to less than 2% a year in 1985.
In the 50 years of available data , and keeping the 1981 structural shift in mind, Figure 6 indicates that the rate of wage growth has tended to decelerate at the onset of a recession and then throughout the recession and beyond (the exceptions occurred in the 1973-1975 recession and the 1980 recession). In the three recessions since the shift of structure, the average of year-over-year growth rates of wages prior to a recession had been a 4.1%, which drops down to 4.0% during the first three months of the recession, and then 3.8% in the following three months, and to an average of 3.3% growth at the end of the recession.
On average, the growth rate of hourly earnings decreases by only 0.7-0.8 percentage points during the recession, but as we showed in Figure 5, that deceleration tends to continue beyond the limits of the recession as employers are able to offer reduced wages to an increased supply of willing employees. It has been shown that the effects of under-employment endure far beyond the duration of the initial recession, as the labor force loses its ability to keep up with technology or lacks the means to move to where the jobs are. This results in workers sliding into lower-paying jobs and households adjusting their spending accordingly. Thus, waiting for a slowdown in wholesale and retail sales, and demand for services, may appear to be a risk-averse method of assessing the end of business cycles; however, doing so runs the risk of businesses over-expanding. It is also one of the primary reasons why small and medium firms tend not be aware of the end of business cycles until six to nine months after they begin.
Figure 6 implies that wages likely peaked at 3.43% in the current cycle, which should be a signal to firms that a broader slowdown is in train. On a three-month, average annualized pace, wage growth slowed to 2.73% in May, implying a much slower trend than the top line, and down for three straight months. With hiring in the household survey used to estimate the unemployment rate having only added 10,000 jobs over the past three months, it is likely that the unemployment rate will begin to rise in coming months.
The policy-sensitive yield curve (10-year yield less three-month yield) has inverted due to a decline of roughly 115 basis points in yields on the U.S. 10-year treasury; it is now pointing to much slower economic activity ahead. While low wage growth is not a perfect predictor of a recession in the immediate future, it does signal there are elevated risks of a recession over the next 12-18 months as employment growth slows, unemployment rises and wage growth decelerates. The time for firms to stress test their balance sheets, engage in more rigorous evaluations of potential acquisitions and increase efficiencies has come.
