Here we have, once again, our favorite meaningless non-correlation plot. Blue dots are points for Democratic presidents; red dots for Republicans. X axis is Tax receipts/GDP in a given year, Y axis is GDP growth for that same year. Operating on the theory that if it's good enough for Econ. Prof. Lazear, it's good enough for an old trombonist, let's twist this to serve our purposes.
The red and blue horizontal lines are averages for the 36 years of Republican presidents and 24 years of Democratic presidents, respectively, from 1950 through 2009. The slanty lines are best fits for Repug and Demo data sets. The vertical yellow line is Lazear's average of Tax/GDP, the magical 18%.
Note the Inconvenient 2009 point, labeled "What the Hell?!? Low Tax/GDP along with NEGATIVE GDP Growth. Don't you love it when a lie crashes and burns?
Some mathechistic stuff:
Some iron-clad, indisputable conclusions:
1) The average growth of GDP under a Democratic President is 1.4% higher than under a Republican.
2) In 25 years, the level of GDP will be about 30% higher if we elect Democrats than it would be if we were to be stupid and evil (I'm quoting O'Really here) enough to elect Republicans.
3) Never mind that the Standard Deviation of either data set is larger than the difference between their averages. Shhhhhh!
4) Lazear's point is that high Taxes/GDP causes lower GDP growth. Well, the correlation coefficient for Repugs is -0.44. Sadly, this is a weak correlation, but it's more than twice as large as the -0.20 correlation coefficient for Demos. So: if his premise has any validity (Remember, it doesn't) it is valid only when a Repug is in the White House. When a Demo is in residence, the meaningless non-correlation, such as it isn't, crumbles. Damn.
5) Oh. Slopes. For Repugs, -1.17. For Demos, -0.33. That's right, folks, the slope is 3.5 times greater with a Repug. See point 4.
Have I forgotten anything? If so, please don't tell me. I'm sick to the teeth of the whole damned thing.
UPDATE: Here is part 7.2.