The Weekly Unemployment Claims -- Still Overdone

For many months, I used to discuss the weekly figures for new unemployment claims that get reported each Thursday. This week, the number is 362,000 and last week's figure got revised upward to 354,000. These numbers are fine. They do not indicate much of anything with regard to the economy other than that things are plodding along as before. There is no increase in unemployment likely to be seen so long as the numbers stay in this range.

What I find very interesting is how the media overdoes its discussion of these weekly figures. Two weeks ago, the number came in at 351,000 and there were stories about how the number hit the lowest level in months. Then last week, that figure got revised up by 2000 and the figure for last week was released at 351,000. What did the stories say? You got it right if you guess that the headline was that the number hit the lowest level in months. It was as if they cut and pasted from the story of a week earlier. And now today, the number has risen by 8000 for the week. The stories today are discussing how this may drive stock prices down, and then after the futures really did not react the story switch to how the market is shrugging off the bad news. All this is pure nonsense, however, written by reporters who do not understand what they are saying. The truth is that the figures for each of these last three weeks are essentially the same insofar as the economy is concerned. A variation of 2000 or 8000 jobs is not significant unless there are a number of week with change in that range which consistently move in the same direction. For example, five weeks of 8000 job loss increases would mean a change of 40,000 which would be significant. Five weeks of up 8000 one week followed by down 8000 the next would mean nothing. Someone at Bloomberg or Dow Jones must understand this. It would be nice if the reporting followed reality.