Politifact rates the following statement as "half true." "Having an entirely Democrat congressional delegation in 2009, when the [federal stimulus] bill passed, increases the per capita stimulus dollars that the state receives per person by $460." The bottom line is that a chain email that had this quote stripped out the explanation that I gave along with the supporting evidence and the graph that I had. It was thus apparently not clear that I had run a regression. I will give Gene Emery credit for the long discussion that he wrote up. From Politifact
So where did Lott's number come from?
If you create a graph that shows how much a state received per capita on one side and the percentage of Democrats in the congressional delegation on the other, you get a graph that looks like 50 drops of paint spattered against a wall.
But in this case, there is a pattern to that spattering. Lott uncovered it with a statistical analysis that turned the scattershot data into a straight line, indicating the overall trend. The lesson of the line: whenever Democratic representation increased by 10 percentage points, the per-person allotment in the state increased by $46.
Thus, he said, by that measure, going from 0 percent Democratic representation to 100 percent gives you an extra $460.
Plenty of individual states don't conform to this, as demonstrated by the fact that, using Lott's data, the only all-Republican state of Wyoming got more money than two of the five all-Democrat states.
(Lott did not include the District of Columbia, which has all Democrats in its non-voting delegation and has received five times more stimulus money per resident than the national average.)
Lott's analysis on FoxNews.com make it clear that he is looking at trends, noting, for example, that "each one percentage point reduction in the percent of a state's workforce represented by unions saw a $26 drop in stimulus dollars per person."
But the e-mail stripped out that context.
Labels: mediabias, stimulus