June 04, 2023
The inflation hawks have been getting their way with the Fed for the last 14 months, but it looks like the Fed is prepared to call at least a temporary truce in its war on inflation, foregoing the opportunity to raise interest rates at its meeting this month. That would be good news for tens of millions of workers who are experiencing the strongest labor market in more than half a century.
By any measure, the Fed has been very aggressive with its interest rate hikes over the last fourteen months, going from near zero to a range of 5.00 percent to 5.25 percent. It will be some time before the full impact of these hikes winds its way through the economy. The effect of the rate hikes is compounded by the financial turmoil that followed the collapse of the Silicon Valley Bank. These surely have had the effect of slowing growth from the path we were on at the start of last year.
While some slowing was surely warranted, the question is: how far do we have to go? The big issue, which is still unresolved, is whether we will need to see a substantial rise in the unemployment rate to bring inflation under control. The most recent wage data suggest that we may not.
The annualized rate of wage growth over the last three months was just 4.0 percent. There were several points in 2018-19 when wage growth approached this pace, even as inflation stayed close to the Fed’s 2.0 percent target. Here’s the picture.
Source: Bureau of Labor Statistics and author’s calculations.
As can be seen, the data do fluctuate, but there has clearly been a sharp slowing in wage growth by this measure over the last year and a half. At the end of 2021 and the start of 2022, the annualized rate of wage growth over three-month periods was well over 6.0 percent. It actually has been somewhat under 4.0 percent in prior three-month periods. (By my preferred measure, taking the average for the last three months [March-May] compared with the average of the prior three months [Dec-Feb], the annualized rate has been just under 3.8 percent.)
Whether this pace of wage growth is going to be low enough to bring inflation down to the Fed’s target can be argued, but we are clearly close to a pace consistent with the target. And, it is indisputable that we have seen a sharp drop in the rate of wage growth.
Is the Wage Growth Slowdown Driven by Composition Effects?
Some analysts have argued that it is wrong to see this wage slowdown as a victory over inflation because it is partly driven by a change in the composition of employment. The argument is that the average is being held down by rapid growth in the number of low-paying jobs, as well as layoffs in high-paying sectors. Those making this point note that the Employment Cost Index, which holds composition fixed, does not show as much slowing.
The big problem with this argument is that changes in the mix of jobs do not seem to be slowing wage growth in recent months. While hours growth in the low-paying leisure and hospitality sector was somewhat more rapid than the rate of hours growth in the private sector as a whole, this impact was largely offset by a fall in hours in the low-paying retail sector.
There has been a drop in hours over this period in the high-paying information sector, but this was more than offset by rapid growth in hours in the high-paying financial services sector. Examining the impact of composition changes, using the sectors the Bureau of Labor Statistics publishes in its monthly Employment Report, I find that changes in composition modestly increased the average wage over this period.
The other point about composition effects is that, insofar as they reflect trends and not anomalies (as with the start and end of the lockdowns), we should want to include them in our assessment of wage growth. As a first approximation, we should expect inflation to be roughly equal to the rate of wage growth minus the rate of productivity growth. If we are seeing a massive shift of workers from high-paying sectors to low-paying sectors, it should show up in slower productivity growth.
To see this, imagine an extreme case where millions of software engineers and biotech researchers lost their jobs, while we saw a huge surge of employment in restaurants and convenience stores. Assuming that pay bears some relationship to productivity, we would expect to see a fall in productivity or at least much weaker growth.
To date, we have not seen any clear evidence of a productivity slowdown. (The first quarter was awful, but it looks like the second quarter will show strong growth.) In any case, when we compare wage growth with productivity growth, we should want a measure that includes the effects of changes in composition, since our productivity measure certainly does.
Are Wages the Problem?
Many have argued, correctly, that wages have not been the cause of inflation during the pandemic. This is undeniably true since average wage growth has trailed inflation from the start of the pandemic. There has been a big shift in profits over this period, which has been partly reversed in the most recent quarters.
However, even if wage growth has not been driving inflation, it can be said that barring other economic changes, overly rapid wage growth is inconsistent with moderate inflation. For example, if we had 6.0 percent wage growth, it is hard to see how we can have anything close to 2.0 percent inflation. We can see a further decline in profit shares; they have risen substantially since the start of the century, but that can only go so far. Wages are nearly four times as large as profits as a share of income, so if there is a large gap between wage growth and productivity growth for a sustained period, it almost certainly will be passed on in higher prices.
There are other measures that can reduce inflation. For example, Congress can take steps to weaken patent and copyright monopolies, which would lower prices for prescription drugs, medical equipment, software, and a wide range of other items. However, if we don’t see these policies being implemented, then we should expect that an overly rapid pace of wage growth will eventually be translated into higher prices.
The Fed Should Declare Victory!
But, we don’t have to speculate about a world where we have overly rapid wage growth (although Congress should still weaken patent and copyright monopolies). Wage growth has slowed to a pace that is roughly consistent with the Fed’s inflation target. We did see a modest rise in unemployment in May, but the economy still seems to be creating jobs at a healthy pace. For now, it looks like immaculate disinflation is a reality.
Note that I have imposed a floor and ceiling on the graph that prevent it from showing the full range. This is due to the extraordinary rapid growth reported at the start of the lockdown period, as a huge number of lower paid workers lost their jobs, raising the average pay of those still employed. This effect was reversed in the summer as many of these workers got their jobs back. To allow readers to better focus on the more normal trends in wage growth, I used a scale that leaves out these extremes.
 How much of this shift was due to monopoly power is questionable, since there was a comparable shift in Europe, where anti-trust enforcement has been more vigilant over the last four decades.
 The unemployment rate rounded to 3.7 percent in May, but it was actually 3.65 percent, meaning that the increase from April was somewhat less than a full 0.3 percentage points. It is also worth noting the difference between the strong job growth reported in the establishment survey and the drop in employment reported in the household survey was entirely due to differences in concept (e.g. the establishment survey counts multiple job holders twice and also does not count the self-employed). Using the establishment concept, the household survey actually showed slightly stronger job growth than the establishment survey in May.