The Department of Labor releases data on initial jobless claims every Thursday morning for the previous full week.  I track this data in my own database in order to analyze trends, but usually don’t write about it unless the data itself shows a significant change.  The Associated Press offers a somewhat misleading report on the data, however, which becomes apparent when one sees the numbers for the entire quarter:

New claims for jobless benefits fell for the third straight week but remain elevated, suggesting the labor market is still sluggish.

Initial claims fell by 3,000 to a seasonally adjusted 456,000, the Labor Department said Thursday. That’s nearly the same level as in January.

At the same time, the tally of laid-off workers continuing to claim jobless benefits fell by the largest amount in almost a year. That could be because more people are finding work. But it may simply mean that they have exhausted their initial state benefits.

Continuing claims fell by 255,000 to 4.5 million, the lowest level since December 2008, the department said. A Labor Department analystsaid state agencies didn’t provide any explanation for the drop.

First, let’s take the issue of jobless claims dropping for three straight weeks.  That’s technically true, but only because this week’s report increased the previous week’s initial jobless claims by 6,000 (a normal adjustment).  Had that adjustment not occurred, this week’s number would have actually been an increase of 3,000.

But that’s just nibbling at the margins anyway.  The three straight decreases total up to a change of 8,000 in their entirety, which is arguably statistical noise when looking at the total numbers in play.  It’s less than a third of the 28,000 jump in claims that occurred four weeks ago, and in any case doesn’t change the trajectory of the near flat line in 2010Q2:

The green line is the seasonally-adjusted number, the data used for comparison.  Note that it was actually declining at a slightly sharper rate in April before the month’s final week had that big jump.  The “three straight declines” haven’t yet brought us back to the level at the end of March or the middle of April.

For that matter, the AP’s note that the current level is “nearly the same level as in January” is also a little misleading.  While technically true, it’s higher than many of the data points since January.   Either way, it doesn’t connote any sort of improvement, as the AP’s reporting tends to suggest, and we’re far away from the 325K floor (shown in gray on the graph) that we need to hit to get real private-sector job growth.

This is a picture of stagnation at a particularly high level of unemployment for two quarters in a row, which means that we won’t see job growth for quite some time.  Now that people have begun to run out of benefits, we may see the unemployment rate decline, but we won’t see jobs increase unless these numbers start improving in real terms very quickly.