I'm generally a fan of empirical models like yours, but Kevin Drum posted an interesting response to your model earlier today. Basically, he notes that we can predict more elections (13 of 16) if we only consider the number of years the incumbent party has been in office -- the incumbent wins if it's been 4 years and loses if it's been 8. Any thoughts on why this pattern seems to occur?
Thanks for doing this, and for all your hard work. My morning doesn't feel complete if I don't have a chance to read Wonkbook.
indyguy36 karma
I'm generally a fan of empirical models like yours, but Kevin Drum posted an interesting response to your model earlier today. Basically, he notes that we can predict more elections (13 of 16) if we only consider the number of years the incumbent party has been in office -- the incumbent wins if it's been 4 years and loses if it's been 8. Any thoughts on why this pattern seems to occur?
Thanks for doing this, and for all your hard work. My morning doesn't feel complete if I don't have a chance to read Wonkbook.
View HistoryShare Link