Insight Line—April 28, 2008

Long-time readers of this commentary will recognize a recurring theme—this time it is different— especially when referencing historical economic statistics. After all, inputting dynamic data into static analysis obscures important nuances, leading to incorrect interpretations of incoming information. Take the recent headlines bemoaning the plight of the American worker, whose wages and salaries are purportedly barely outpacing inflation. Statistically speaking and using available data, the conclusion is plausible. But what if the data being used to draw such conclusions is flawed?

Suffice it to say, the problems surrounding the statistical measurement of an economy as large and complex as the United States’ are both monumental and well documented. Nevertheless, the statistical tracking systems for the U.S. economy are nothing short of world-class and, indeed, often set the world standard. However, the models are not set in stone. For example, the underlying constructs to the most widely referenced inflation gauge, the Consumer Price Index (CPI), undergo continual refinement in an attempt to make it a reasonable facsimile of the consumer’s inflation experience. Yet, many acknowledged measurement problems remain and, as such, the accuracy of the CPI is open to wide and varied levels of interpretation.

Wage and salary data are no different. In October 2003, the Division of Current Employment Statistics at the Bureau of Labor Statistics (BLS) announced its intention to update, refine and expand the Current Employment Survey’s average weekly hours and average hourly earnings surveys to reflect the changing nature and scope of the nation’s labor force. The goal is to expand the currently targeted survey to cover all forms of income to all employees and is to be completed by January 2010.

As part of this initiative, last month the BLS began featuring more comprehensive data (state-specific, to be exact,) on the composition of total wages and hours worked. The BLS considers this data series experimental, given the agency’s limited experience in processing such data under the larger and more inclusive categories.

However, researchers at the Federal Reserve Bank of St. Louis got a behind-the-scenes preview¹ of the BLS’ experimental data series. What they found has implications for the seemingly plausible headlines on the plight of the American worker.

The researchers note, “The experimental measure of hours is on average between 1 and 3 percent higher than the existing measure. The experimental measure of average hourly earnings is also consistently higher than the existing measure: As of November 2007, this difference was about 19.5 percent.”

Said differently, if the experimental data from the BLS better reflects the earnings of the entire U.S. labor force, then, in my opinion, it unquestionably casts doubt over whether the American worker is actually doing no better than treading water to keep from economic drowning. Rather, the reality may be more likely that American workers are getting along swimmingly.

Granted, the data is still experimental at this point and could well prove to be a red herring. Then again, as I see it, the BLS’s newly minted data stream does appear to offer at least some outward support for why the Bureau of Economic Analysis characteristically revises total personal income, in aggregate and over the years, higher.

¹ Kliesen, Kevin L., “An Expanded Look at Employment,” National Economic Trends, Federal Reserve Bank of St. Louis, March 2008. Available at http://research.stlouisfed.org/publications/net/past/2008/

This material has been prepared using sources of information generally believed to be reliable. No representation can be made as to its accuracy. The forecasts and opinions in this piece are not necessarily those of Van Kampen, and may not actually come to pass. Information in this report does not pertain to any Van Kampen product and is not a solicitation for any product.