« Moore's Law for the blogosphere? | Main | Mr. Keynes and the Classics for the 21st century »
August 11, 2005
Labor market and inflation: then and now
I don't need to tell you that there's been a little dust-up recently concerning the health of the labor market. See Mark Thoma's latest post, for example. Economist (and non-economist) blogs are resounding with shouts of optimism and wails of pessimism.
I've been called a "bad news bear" and been accused of jumping up and down (as if with glee, I suppose). In truth, I tend towards being optimistic when it comes to the macroeconomy, but my tendency for optimism is always tempered by the data. In macroeconomics, data is paramount. But the data does not always speak with a clear voice. Macro data also has the disadvantage of being at a lower frequency than, say, financial data. So you get a bad payroll report and it wears on you for a month. Comes with the territory.
But if you get enough recent data to allow you to compare it to other time periods, things are a little better. Much of the recent debate has centered on economic growth, labor market recovery, and inflation compared to the recovery after the 1990-91 recession. I've tossed in my two cents. Today, I will let the data speak.
To begin with, I took HP filtered (log) real GDP and found the quarter of maximum deviation below trend. For the 1990-91 recession, that was the 4th quarter of 1991 (two quarters after the NBER recession end date). For the most recent recssion, it was the 1st quarter of 2003 (more than a year after the NBER recession date). I would attribute the difference to the fact that the transition both into and out of the most recent recession was more gradual than in 1990-91. Why? That is an open question.
Next, I plotted payroll job growth and core CPI inflation starting from that point of maximum deviation from trend real GDP (October 1991 and January 2003). Here's payroll employment growth:
To make things a little clearer, I took a 3 month moving average centered on the observation.
Oddly, there is a spike in payroll growth about a year and a half after the maximum negative deviation of real GDP from trend. It is pretty clear, however, that the current recovery has been consistently lagging the previous one. And unless things really pick up in a hurry, the cries of those who are more pessimistic are going to get louder. If the job market hums along at 200,000 new jobs per month, that will seem paltry compared to the job growth in 1994.
Please note, however, that the Fed waited longer after that maximum negative deviation of trend real GDP to pull the trigger on interest rates. Indeed, rates were still falling in 1992. All my students should know that this should have led to a faster recovery in 1991 and beyond. (Of course, the Fed's job in the last recession was complicated by the fact that interest rates got so low that we were worried that they might run out of ammunition.)
The fact that rates were still falling in 1992 and didn't start to rise until 1994 should have also meant that inflation would have started to bubble up a bit sooner. The next chart bears this out.
The data is the CPI excluding food and energy taken from FRED and last updated 7-14-2005. For clarity, I only show the 3 month moving average as you can see the trends more easily. After taking the moving average, I converted it to an annual rate for ease in interpretation.
This too, pretty much speaks for itself. Core inflation was very low at the start of 2003, but by the time the Fed started raising interst rates it had climbed. But we are still in a better position than in 1994-95. All in all, not too bad a job of sticking to the mandate of price stability.
On the dual mandates of price stability and full employment, the picture is mixed. Some would say the Fed should have a single mandate of price stability, but that's not the way their mandate reads at present. The real test will be to see if the steady growth we have seen (note: I use the word steady, not booming) will be sustained. If the timing of a soft patch is similar to the timing of the last one, it could still be in the future. I do believe that it can be avoided. It would appear to this observer that the Fed has made a real effort to avoid the mini-boom and mini-bust nature of the last recovery. Does no mini-boom mean no mini-bust? Time will tell.
But the pictures do help put it into perspective.
UPDATE: A misprint in the CPI graph was corrected. The first series starts in October 1991.
One more thing. Most comparisons of the labor market recovery to that after past recessions use the NBER recession end dates. The reason I chose not to use those dates (and use the date of maximum negative devation of real GDP from trend) was to try to allow for the fact that the last recession was more shallow and the below trend growth lasted longer. I was trying to see if the job recovery was delayed by a similar length of time. Not quite, but it might be useful to think about the reasons for the recession being shallow and the deviation from trend lasting longer as at least partially driving the "weak" labor market recovery.
Posted by William Polley at August 11, 2005 3:55 PM
Trackback Pings
TrackBack URL for this entry:
http://www.williampolley.com/cgi-bin/mt-tb.cgi/296
Listed below are links to weblogs that reference Labor market and inflation: then and now:
» foreign exchange brokers from foreign exchange brokers
LONDON, Aug 11 (Reuters)- The euro rallied back above 1. 50 on Monday to trade up on the day after comments from a leading euro zone monetary policymaker… [Read More]
Tracked on August 24, 2008 10:10 PM