Tuesday, July 21, 2009

Uncertainty About Reported Economic Numbers

The public and news media generally accept the reported economic numbers as an accurate measurement of the economy. However, many of the common economic numbers are not actual, raw measurements. Some are samples from a few data sources extrapolated to the whole US economy. Others are survey data, which has all the inaccuracy problems associated with surveys. Still other numbers are mathematically manipulated, such as for seasonal adjustments for unemployment, quality adjustments for inflation price comparison, etc. and others are indices that run into theoretical construction issues about what the index represents.

In just about all the economic numbers, there are errors. Some are sampling errors, some are reporting errors, some are statistical errors and some are plan old mathematical errors. Sometimes, the accuracy needed to make a determination about the state of economy, particularly whether it is getting better or worse, is obscured by the magnitude of the errors in the reported figures.

Rebecca Wilder on the News N Economics Blog has run posts about housing index numbers. There are differences among the several reported housing indices that lead to varied conclusions and confusion about the state of housing. See her post, "House price indices: not necessarily the same story."

Likewise, last week there were discussions on the blogs about the accuracy of the improvement in the new claims for unemployment numbers. There were concerns about the correctness of the seasonal unemployment correction to the new claims numbers for autoworkers. In the past, these autoworkers were treated as furloughed employees that would regain their jobs when auto production resumed and recovered. Is that really likely this time? For example, see John Keefe's CBS Money Watch comments, "Underreported Unemployment: Why Those Jobs May Not Come Back."

As the US economy improves and it reaches its turning point from negative to positive economic numbers, the indicators will be confusing and misleading. The inaccuracies will make the economy look like it is showing gains and loses simultaneously and vacillate between recovery and recession over a few reporting periods. This will go on until the economy is well into a recovery and outside the bounds of showing negative results in the relevant economic indicators.

No comments:

Post a Comment