Reporting/Market Focus from the January 2005 Edition of the SGS Newsletter
As the oldest, regularly published government economic series, the Federal Reserve’s Index of Industrial Production at one time was viewed as the equivalent of today’s GDP. The index was started after World War I (data are available back to 1919), and provided the only systematic measure of manufacturing activity that once was the predominant driver of overall U.S. economic activity.
The index provides a monthly measure of real (inflation-adjusted) output of manufacturing, mining (including oil and gas) and gas and electric utilities. Although some analysts may view production as an artifact of an earlier age, production does tend to generate physical wealth, unlike the services sector so heavily favored in today’s politically-adjusted economic reporting.
As published, the index does tend to move with the broad economy, regardless of what is published by the Bureau of Economic Analysis on the GDP. Industrial production is used, along with factors such as payroll employment, in calling turning points in the economic cycle, as determined by the National Bureau of Economic Research. Industrial production still shows a 2000/2001 recession, while the GDP does not.
That said, there are problems with the series. As of the December 22nd benchmark revision, 46% of industrial production was estimated by electricity consumed or by the number of manufacturing employees in a given industry, 50% was based on actual physical production records and 4% was guesstimated. Prior to the revisions, fully 50% was estimated by electricity usage and worker counts and 46% was from direct physical measures of production.
The trend away from estimates based on electricity and workers could be a plus, or not, since the replacement data are coming from the same people who give us the GDP. The problem is that production based on electricity usage can be skewed by unusual weather factors. Also, manufacturing payrolls can be skewed by reporting problems at the Bureau of Labor Statistics, but the bias factors added into the data are minimal in the manufacturing area.
In that the series tends to work as a coincident indicator, it probably is best left alone, and corrected, when possible, in annual revisions. In the latest revision, overall industrial production growth was revised downward by 0.6% through November 2004, downward by 0.3% for all of 2003, upward by 0.1% in 2002, upward by 0.2% in 2001 and downward by 0.4% in 2000. In broad summary, the 2000/2001 recession was deeper than previously reported, and the recovery from same has been weaker.
Published along with industrial production is a measure of capacity utilization, which tends to be a less reliable indicator than production, due to the poor quality of data available on capacity.