The businesses surveyed by the NY Fed (analysis here) and the Philadelphia Fed reside in different universes. The NY Fed’s survey opinion returned to negative levels last seen in the early days of the Great Recession.
For the respondents to the Philadelphia Fed’s business survey for November 2010, conditions are pushing all time highs. The headlines:
Results from the Business Outlook Survey suggest that regional manufacturing activity showed improvement in November. All of the survey’s broad indicators of economic performance showed improvement from their reading in October, and firms reported an increase in employment and work hours. More firms reported increases in input prices this month, although downward pressure on the prices of firms’ manufactured goods was still evident. The survey’s broad indicators of future activity suggest that firms remain optimistic about growth over the next six months.
Indicators Suggest Improvement
The survey’s broadest measure of manufacturing conditions, the diffusion index of current activity, increased from a reading of 1.0 in October to 22.5 in November (see Chart). This is the highest reading in the index since last December. Indexes for new orders and shipments also improved this month, and each index increased 15 points. Indexes for both delivery times and unfilled orders changed from negative to positive this
month, suggesting improvement. Labor market conditions also showed some improvement this month, paralleling the improvement in other broad indicators. This month, firms also reported some growth in employment and a longer workweek. The percentage of firms reporting increases in employment (27 percent) was greater than the percentage of firms reporting decreases (14 percent). The index for employment was positive for the third consecutive month and increased 11 points. The average workweek index increased significantly, from ‐6.0 to 10.9.
Is it possible to make sense out of two surveys suggesting opposite conditions? Surveys, even survey’s from the well run Fed – are still surveys. These are not based on data, but the feelings of the administrative assistant who fills out the forms for his boss.
My guess is that the seasonal adjustment factors or weightings of respondents vary between these two Fed outposts. When looking at the unadjusted data for new orders, a slightly different picture emerges.
Looking at backlog (unfilled orders), which is a good indicator of the direction of the economy – the unadjusted data says backlog is declining. With backlog declining, we are getting no indication yet the economy is improving. From my experience, backlog has very little seasonal adjustment characteristics. This seems to validate the feeling that the seasonal adjustment factors are likely off.
To answer my question whether it is possible to make sense of two surveys suggesting opposite conditions – the answer is no. These are surveys, and one group had a bad day while the other had a very good one. The two surveys nullified each other, and now we can only wait for hard data to learn the facts.