Wage Inflation and Expected Price Inflation
Yesterday, I noted that an expectations un-augmented Phillips curve fits data from the past two decades rather well. Here is an even simpler illustration of the point that lagged price inflation has not correlated with wage inflation for the past two decades.
The figure is a scatter of lgwinfl — the percent annual rate of increase of
Business Sector: Compensation Per Hour (HCOMPBS), Index 2009=100, Semiannual, Seasonally Adjusted from FRED on pcecinf the lagged annual percent rate of JCXFE “Personal Consumption Expenditures: Chain-type Price Index Less Food and Energy (JCXFE), Index 2009=100, Semiannual, Seasonally Adjusted.”
Note that the data are of semiannual frequency, so each raw data point from FRED affects two overlapping intervals and therefore two dots on the graph. I look at data for intervals starting from the second half of 1992 through the second half of 2012 (the most recent available). The labels refer to the start of the intervals.
There is no sign of an effect of lagged price inflation on wage inflation. As noted by Nick Rowe in comments, this is what one might expect if the Fed were successfully targeting inflation. In that case the best estimate of future price inflation would be the target no matter what inflation had recently been. (Actually now that I think about it, current new Keynesian models still suggest a correalation as sticky wages catch up with other than expected price growth. But that’s not the point of this post.)
This is what central bankers and other people mean when they say that inflation expectations are anchored — they mean changes in price inflation don’t persist, because they don’t feed over into wages. However, anchored expectations can also be a statement about expectations. A lack of feedback to wages might be due to something else — Krugman argues that it can occur if there is downward nominal wage rigidity.
I have noted that lagged inflation is correlated with the breakeven inflation rates which would make the return on Treasury Inflation Protected Securities (TIPS) equal to the return on ordinary nominal Treasury Securities. This post looks at 5 year breakevens which should reflect expected inflation over the next five years. Bond traders’ expectations sure don’t seem to be anchored.
Surveys of people believed to be expert are another source of data. I have been playing with the Livingston Expectations urvey. This is a semi-annual survey of forecasts of many variables including the consumer price index. The base year index level is included in the publicly available data set, so it is possible to calculate forecasts for inflation over intervals. Again I look at 12 month intervals (so intervals overlap). The format of the Livinston Survey was changed for 2004. I use data only from the old version. My excuse is that I am keeping the new data (which are publicly available in files with a different format) for out of sample forecasting with no risk of data snooping. The real reason is that I am lazy.
Here is the scatter of the median Livinston CPI inflation forecast for CPI inflation over the following and PCECINF
Expectations sure don’t look anchored. Expected CPI inflation seems to increase about one for one with lagged core PCE deflator inflation.
Finally, I cut out the middleman. Here are hourly compensation growth and expected CPI inflation for the same intervals.
There is plenty of variation in the median Livinston Survey forecast of inflation, but there is no sign of correlation with wage inflation.
It remains possible that the expectations of people who set wages are anchored. Bond traders and Livingston survey participants don’t set most wages. Managers and other employees of the same firms bargain over wages somehow.
In the Fed’s dual mandate of maximum employment and price stability, maximum employment isn’t defined. However, price stability is defined as 2% annual inflation in the long-run.
So, it seems, the Fed leans towards price stability, more than maximum employment, which may be more difficult to define, and we can expect annual inflation around 2%.
Also, I may add, ten year expected inflation has been in a steady decline, since the early 1980s. Chart:
http://www.clevelandfed.org/research/data/inflation_expectations/2014/March/image1.gif
Maximum unemployment may not be defined by the current Fed but that is an alloyed advance over previous regimes that slavishly adhered to NAIRU often enough pegged to 5.5% and so treated ANY god news on the employment or real wage front as ample reason to ‘slam the brakes’.
The empirical basis for NAIRU in general and specific levels of such was shaky at best but through most of its history the Fed treated its “dual mandate” as instead one that preserved ROI for bond holders at whatever costs to labor.
Now the Fed under Bernanke and now Yellin seem content to not handcuff themselves to a specific NAIRU that would have them squelch inflation before it actual eventuates and instead let decisions be made by close monitoring of actual inflation rather than lazy reliance on purported Confidence Fairy reaction to good news on the wage/employment front.
I can’t quite tell from your comment whether you approve or disapprove of this departure from the Chicago/Uncle Miltie “Screw the Worker, We Can’t Risk bad Effects on Coupon Clippers” reliance on a fairly arbitrary setting of NAIRU. Me I thoroughly approve. Because at some point you have to give SOME credence to “Greatest Good” metrics rather than totally theoretical “Trickle Down from the 1%” that ha s been dominant in monetary policy my entire life.
Bruce:
🙂
In the mid-1990s, Greenspan believed the economy could grow faster without accelerating inflation. The result was actual output exceeded potential output and the country was beyond full employment (the annual unemployment rate fell to 4.0% in 2000). Moreover, in the mid-2000s, actual output reached potential output and the country reached full employment (the unemployment rate fell to 4.6% in 2006 and stayed at 4.6% in 2007).
The Fed takes into account many factors that influence inflation and employment. Moreover, the Fed works in the future economy, because of lags in the adjustment process.
If you feel the 99% didn’t benefit enough, it’s ridiculous to blame the Fed. You need to look elsewhere. You can praise the Fed later.
and Participation Rate was what in 2006/7?
I stated before, The Fed creates and destroys money to achieve sustainable growth, which is optimal growth. and the Fed doesn’t micromanage the economy.
Since the easing cycle began in late 2007, the Fed stimulated growth by lowering interest rates for households and firms, along with creating a “wealth effect” (in asset markets) to induce demand and reduce saving, in the attempt to generate a self-sustaining cycle of consumption-employment (where consumption generates employment and employment generates consumption, etc.).
Moreover, the Fed raised “animal spirits” to facilitate nominal growth, which facilitates real growth.
However, a virtuous cycle of consumption-employment failed to take hold, because fiscal policy, and policies out of Washington, basically resulted in an economy that has one foot on the accelerator and the other foot on the brake, causing an expensive recoverless expansion.
The Fed has raised living standards for the masses substantially, ceteris paribus, because smoothing-out business cycles results in a higher level of growth. Economic boom/bust cycles are inefficient both in the boom and bust phases.
The Fed ain’t done shit except help the banks rip off mortgage buyers.
The roads are full of potholes, bridges are unsafe, it takes two jobs to eat and pay rent, kids go to school hungry, education has gotten shabby, college is way too expensive with NO jobs afterward, 26 year olds STILL have to be on Mom&Pops insurance.
Peak, what planet do YOU live on?
PK the very fact that you can with a straight face state that actual employment was “beyond full employment” reveals that you are just as theory bound as any other NAIRU “screw the worker because we can’t take the risk that (fill in the inflationary blank).”
Can you give an empirically based argument as to why we shouldn’t target 4.5% unemployment as perfectly achievable consistent with 2-3% inflation? Because out seems a lot of this “potential output” is more theory than data driven.
PT,
You said this about the 90s…
“The result was actual output exceeded potential output and the country was beyond full employment (the annual unemployment rate fell to 4.0% in 2000). Moreover, in the mid-2000s, actual output reached potential output and the country reached full employment (the unemployment rate fell to 4.6% in 2006 and stayed at 4.6% in 2007).
Let me respond… In my research in effective demand, the full employment level depends on the level of capacity utilization. Now when you say that unemployment at 4% was below the NAIRU, that is technically not true from my standpoint. Capacity utilization had been dropping, while employment dropped. So actually employment stayed at the full employment level. It did not actually go beyond it. My numbers showed that the utilization of labor and capital danced right along the edge of full employment. While employment rose, employment of capital fell to compensate.
As far as 2006 and 2007, again the combination of emplyment and capacity utilization danced along the full employment level of effective demand for a couple of years. So again, employment actually did not go beyond a full employment level, because capacity utilization adjusted so full employment was not surpassed.
PT,
Sorry, I meant to say…
“Capacity utilization had been dropping, while unemployment dropped.”