Sunday, March 01, 2009
When you're wrong, you're wrong
Mea culpa. �On Friday I posted a simple graph of state government employees under the title "Growth in the supply of "witnesses" at DFL "listening sessions"". �Late last night Dave Senf, an economist at DEED, where they compute those numbers, set me straight on the data I had posted. �
MnSCU employs 19,734 persons, up from 17,653 in 2005-06 (MnSCU demographic report.) That second report seems pretty clearly not to include work-study. University of Minnesota system employment rose over that period by less, from 24,498 to 25,976. But those data do not account for half of the rise I showed in the graph on Friday, and it is quite likely that work-study student employment accounts for the bulk of the remainder, and had I pulled that data out the graph would seem to have been much less dramatic. �I regret the misstatement.
The data you are using in the graph at SCSU scholars is correct but it is Current Employment Statistics data which includes work-study students at our public universities and colleges. If you dig deeper, you will find that most of the job growth in state government employment in the last eight years has occurred at the state�s public universities and colleges, places like SCSU.I had thought to look at that split, which is available on the site I had linked, but I guess I decided it didn't matter. What I missed entirely is that "work-study" point though. Senf is correct that the amount of students receiving work-study aid (federal and non-federal) has grown recently, from 5.4% in 1999-2000 to 7.2% in 2003-04 (last data available from the National Center for Education Statistics.) �I was unaware that work-study students are counted in the CES data which I used, but not in the QCEW data that Senf provides here (in Excel form.)
MnSCU employs 19,734 persons, up from 17,653 in 2005-06 (MnSCU demographic report.) That second report seems pretty clearly not to include work-study. University of Minnesota system employment rose over that period by less, from 24,498 to 25,976. But those data do not account for half of the rise I showed in the graph on Friday, and it is quite likely that work-study student employment accounts for the bulk of the remainder, and had I pulled that data out the graph would seem to have been much less dramatic. �I regret the misstatement.
