Why "Average" Wage Rates May Be Misleading

Much has been made of Fed Chair Janet Yellen's stance that the labor market still has a lot of slack.  That message remained in the Fed’s September communique.  As part of her justification, the Chairwoman points to the fact that the average hourly wage rate has remained at or near the "official" rate of inflation, and, therefore, real wage rates have been stagnant.  The September jobs report, while showing considerable labor market strength. 

Yet, the financial markets attached an unusual amount of emphasis to the penny decline in the September “average” hourly earnings report (from $24.54 to $24.53 for all employees in the private sector).  So despite the report’s underlying strength showing net new jobs of 248,000 (July and August’s jobs were also revised upward), the market’s are hoping that this one indicator will keep the Fed on hold with regard to interest rates.

Labor Market Indicators

While we do not know all of the dashboard indicators of labor market conditions used by Yellen and the FOMC, here is what we do know:

·      The traditional unemployment rate, U3, is down from 10.0% in October '09, and 7.2% a year ago to 5.9% in September, the lowest level since July ‘08;

·      The U6 unemployment rate which includes "marginally attached' workers and those employed part-time for economic reasons has fallen from 17.2% in April '10 and 13.6% a year ago to 11.8% in September; this is the lowest this index has been since October '08;

·      In the last year the number working part-time for economic reasons has fallen by 811,000 to 7.103 million. In September alone, the decline was 174,000.  Its peak was 9.216 million in March '10.  While part-time for economic reasons still appears high by pre-recession standards (fluctuated around 4 million in '06), significant structural changes have occurred since then, including the need for a more educated workforce in the new technological age.

·      Meanwhile, over the past year, the number employed has risen 2.635 million to 139.435 million. 

·      The "quit rate" (take this job and shove it) is up to levels not seen since '08 (only briefly), and ’04 before that, and is an indicator of the growing confidence that those with skills can find employment elsewhere;

·      Job openings, at 4.673 million (July) are now higher than they were at the peak of the former cycle in June '07 (4.639 million) and are higher than at any time since February '01;

·      The rate of layoffs for 2014 are lower than in any year this century indicating that employers are hanging on to employees because skilled workers are hard to find;

·      Jobs in the all important 25-54 age cohort are up by 1.4 million so far in 2014.  These are the primary working years for most people. 

The Yellen Fed View

The Yellen Fed has convinced the markets that the issues of long-term joblessness, part-time for economic reasons, marginally attached workers, the participation rate, and average wages are cyclical in nature and that low nominal and negative real interest rates will induce economic growth and return these indicators to pre-recession levels.  This position appears to have been reaffirmed at the September FOMC meeting.

When the "Average" is Misleading: A Thought Experiment

Given all of the evidence that the labor markets are wound tight, why are average wage rates only moving at the "official" rate of inflation (real wages stagnant), and why has this been the case for the past 5 years?  Is this an anomaly?

The issue here may be, at least partially, the statistical nature of an "average."  Given that a large segment of the population, the baby boomers, are now retiring (10,000 per day beginning in 2011 and running until 2030) and these people are likely at the peak of the wage scale, it could be that the "average" wage measurement is misleading.  Here is a thought experiment to illustrate the point.

Let's postulate a simple economy with 1,000 employees with their ages equally distributed, their pay scales as shown in Table 1, and that for every retiring employee there is a replacement employee just entering the work force.  If wage rates stay constant, and as workers age their pay rises, then, in 10 years nothing will have changed, i.e., the table will look exactly the same.

Table 1

Pay/hour

$10

$20

$30

$40

$50

Age cohort

16-24

25-34

35-44

45-54

55-64

Employees

200

200

200

200

200

Average wage: $30/hour

Now change the assumption such that over a 10 year period there is inflation and that the wage rates all advance by exactly 10%.  Table 2 shows the result:

Table 2

Pay/hour

$11

$22

$33

$44

$55

Age cohort

16-24

25-34

35-44

45-54

55-64

Employees

200

200

200

200

200

Average wage: $33/hour

Wages have risen across the board by 10%, and so has the average wage.  So, if the age cohorts have equal numbers of employees, using the change in the average wage is a pretty good indicator of what has transpired throughout the labor market.  But this isn't true if the age cohorts have different percentages of the employees.  Let's postulate the same economy with the same beginning wage rates but with a different distribution of employees in the age cohorts as shown in Table 3:

Table 3

Pay/hour

$10

$20

$30

$40

$50

Age cohort

16-24

25-34

35-44

45-54

55-64

Employees

265

180

100

200

255

Average wage: $30/hour

Here again the average wage is $30 per hour. Now, look at table 4.  After 10 years, wages in each age cohort have increased by 10%, and the 255 workers that retired have been replaced by 255 16-24 year olds. 

Table 4

Pay/hour

$11

$22

$33

$44

$55

Age cohort

16-24

25-34

35-44

45-54

55-64

Employees

255

265

180

100

200

Average wage: $29.98/hour

Note that the actual wages have advanced by 10% for all of the employees who were in the labor force 10 years ago, but the "average" wage has actually remained flat (actually lower by 2 cents) because of the high concentration of those near retirement in the original employee distribution.

This exercise indicates that there are dangers in using the "average" concept.  One has to wonder if the current distribution of baby boomers that are retiring is playing tricks with "average" wages, much like the thought experiment above, especially when almost all other indicators of labor market conditions are indicating something else.