Political  Calculations
Recommend this article

Last April, we were asked to contribute a question to Tim Kane's quarterly survey of economics bloggers for 2012-Q2. After reflecting on the potential causes for the very statistically unlikely number of upward revisions to the number of new jobless claims filed each week over the past year, we submitted the following question:

The Wall Street Journal has reported that the advance number of seasonally-adjusted initial unemployment insurance claims filed each week has been revised upward in 56 of the past 57 weeks. If this were a coin toss, the probability of that occurring would be 2,528,336,632,909,751 to 1, which makes the event extremely statistically unlikely. What is the primary reason behind why the BLS' collection and reporting of the initial data would appear to be so deeply flawed?

  1. Seasonal adjustments that no longer reflect reality well enough to account for the difference between initial and later revised counts.
  2. Inadequate counting practices by the 50+ state agencies that report this data weekly to the BLS.
  3. Deliberate delays in reporting the full scale of the worst case increases in the number of claims recorded each week to create a rosier impression of the data.

Here are the survey results:

Hudson Institute Quarterly Survey of Economics Bloggers 2012-Q2, Question 10 Results
Recommend this article

Political Calculations

Political Calculations is a site that develops, applies and presents both established and cutting edge theory to the topics of investing, business and economics.

Be the first to read Political Calculation's column. Sign up today and receive Townhall.com delivered each morning to your inbox.