October 11, 2007
THE OUTLOOK ACCORDING TO ONE ECONOMIST
The August employment report initially created quite a stir. More than a few observers of the economic scene thought the recession clouds were finally gathering. But a month later, the government revised its estimate of nonfarm payroll numbers up and the world suddenly looked brighter. One can only guess at how the Fed will react at the October 31 FOMC meeting.
For some thoughts on where the economy's headed, and where it's been, today we turn to Robert Dieli, president and founder of RDLB, Inc., an economic research and management consulting firm in Lombard, Illinois that publishes its reports for subscribers at NoSpinForecast.com. Bob's a valuable source for this reporter, and a fresh visit with him is timely, given all the questions swirling about on matters of macro relevance. What follows is his take on the current economic climate, a view written by Dieli--exclusively for The Capital Spectator. As a preview, he doesn't think a recession's imminent. For the details, read on...
When the August employment figures were first released, we wrote to our subscribers that they should not take the number too seriously. Our first reason for saying that was because close inspection of the August jobs report showed that most of the decline was due to a drop in state and local education employment. That suggested that something was out of its normal seasonal pattern, or that there had been an incomplete response to the employment survey. In any case, the revision in the jobs report for September, which showed employment growth for August and September, confirmed our suspicion.
Our second reason is seen in the chart below. Note that significant revisions are common between the first and second editions of the employment report. It is on those grounds that we publish this chart in our analysis of the jobs figures every month.
The chart above also shows that in September 2006, the initial estimate was quite small, suggesting some weakness might be developing. That impression was dispelled with the revisions. As a general rule, it's best to wait for the second report before trying to draw conclusions. But, in the overheated atmosphere that prevailed in early September 2007, it was easy to get sucked in by the headline.
As for the question about the threat of recession, let's start by looking at the second chart below, which shows the year-over-year percent change in total nonfarm payrolls going back to 1955. We like this chart for a variety of reasons because it helps put the current situation into perspective. The chart also addresses the recession issue directly: the shaded areas are the recession periods, as dated by the National Bureau of Economic Research. As you can see there, the onset of several recessions occurred after the rate of growth of payroll employment dropped below 1% (the red line). We are not there yet, and we may not get there for some time.
For the current expansion, the annual change in payroll employment peaked at 2.14% in March 2006 and it's been trending steadily downward since. The 2.14% pace was the lowest peak rate of growth of any expansion since 1955. The question remains: Is the current slowdown similar to those that occurred in the middle of the expansions during 1982-1990 and 1991-2001, when the Fed adjusted interest rates and moved from a recovery-generation stance to an expansion-maintenance stance? Or, is the current descent associated with a cycle peak?
We have been saying for most of this year that the Fed was going to have to find the right set of circumstances to lower the Fed funds rate target because of what was going on in Chart 2. The sub-prime liquidity flap and the lousy first estimate of August employment forced the Fed's hand. The question now is whether the actions taken to date are enough to get payroll employment to reverse course. We addressed some of those issues in this month’s report to our subscribers.
The employment statistics are useful in trying to assess the threat of recession, but additional analysis is needed and so we developed a forward-looking indicator called the Aggregate Spread. (For a full description of how the indicator works, please visit my website at www.nospinforecast.com and click on The Tour.) This measures the likelihood of a business cycle peak or trough occurring within the next nine months. As you can see in Chart 3 below, the indicator advises that there's little chance of a recession starting any time soon.
Since 1968, recessions have started around the time the Aggregate Spread went negative and ended when it went back to the plus side. The latest reading is +258. As the chart also makes plain, when the number descends to the +200 range we usually start getting news that all is not well in the economy. And so it has been in this expansion. We reached +200 from above around the time Katrina hit. We have been advising our subscribers for about a year that we are in the phase of the cycle where bad news is not uncommon. But we have also told them that it will take more than just a little bad news to tilt the economy into a full-fledged downturn. The black vertical lines on the chart correspond to the last three dates on which the Federal Open Market Committee (FOMC) cut rates in response to a financial market crisis. In all three cases, the Aggregate Spread was well above zero and it was quite some time before the number reached zero.
The last minutes of the FOMC included a policy directive stating that “the FOMC seeks monetary and financial conditions that will foster price stability and promote sustainable growth in output.” Those words, and the level and trend of our leading indicator bring us to the conclusion that the recession threat is not high at this time.
Posted by jp at October 11, 2007 8:06 AM
Very interesting chart. Most months initial estimate is at least reasonably accurate. Perhaps it's more than coincidence that last year in Aug and Sept was the last time (prior to the August debacle this year) that the initial estimates were way off. Maybe there's something about Aug and Sept that make the initial estimates so unreliably understated. Makes me wonder if they had similar errors in '05, '04, etc. and still haven't figured out why and how to compensate.
Posted by: Jeff Partlow at October 12, 2007 10:14 AM
Nonfarm payrolls have risen an average of 122K per month so far this year. If we hold to that average through next May, which is as far as the Aggregate Spread can see, we will not pierce the 1% level, but we will get very close.
The question that was asked of me was whether I thought there was an imminent threat of a recession. My answer was no, and it remains no.
Posted by: R. Dieli at October 11, 2007 2:23 PM
"As you can see there, the onset of several recessions occurred after the rate of growth of payroll employment dropped below 1% (the red line). We are not there yet, and we may not get there for some time."
Yes I can see the graphic proof for your first point, But given a relatively unfocused view of most recent data, it looks a bit optomistic to say we may not get to it for some time. It looks like we are right on the verge of breaching this level, and cannot find any data point that bounced off this level. So would it be unreasonable to say that following trend, we are months away from going below 1% employment growth?
Where are you coming up with "we may not get there for some time"?
Posted by: Alex at October 11, 2007 10:11 AM