Allegory wrote:
Yes, Republicans had more voters in 2004 than they did in 2008, but that because they won the swing voters in 2004 and lost them in 2008.
Yes. Hence my use of the word "relative" in my original statement about this.
Quote:
If you compare votes they earned in 2008 to the past few decades you'll see they are beating their median in votes earned.
So are the Dems. Voter turnout is higher overall over the last decade than the previous one (in terms of partisan voting at least). What is your point?
Quote:
Only in 4 out of the 13 elections shown here did they get a higher percent of the population to vote for them.
And only 2 out of that same 13 elections shown did the Dems get a higher percentage. You really suck at analysis if you thought that your statement above is significant within the context of what we're talking about here (turnout within a party and swing voters switching from one party to the other).
Quote:
2008 was a good year for the GOP, they got more people to vote for them than in 2000, 1996, 1992, 1988. They've won 4 elections in the last 13 elections with fewer votes than they got in 2008. And one election with fewer votes than the Democrats. It was a good year for both parties, but it just happened to be much better for the Democrats.
Ok. But if we're asking the question "why did the GOP win in 2004, but lose in 2008", isn't the fact that basically 4-5% of the total votes shifted from one to the other kinda the most important (or "primary" even!) thing? The total percentage of eligible voters who voted only increased by about 1.5% (and we'd need more data to see if that was actually more people voting, or less people eligible, or some combination of the two). In any case, by far the most significant factor wasn't new voters but voters shifting from one party to the other.
How the hell can you look at that data and not see this? When you shift from 28/30 in one election to 31/28 in the other, you're clearly looking at voter shift. That should be the first and most immediate result you conclude from that data, right? It's by far the most straight forward explanation. You could assume that a whole bunch of people who voted GOP last time around just didn't vote this year and at the same time, a whole bunch of people who'd never voted before ever all signed up and voted Dem, but there's no data to support this. It's certainly reasonable to assume that *some* people didn't vote who voted before, and clearly *some* people did vote who hadn't voted in the previous election, but to assume that is a more significant factor that people changing their party votes is a pretty incredible stretch.
Occam's razor sort of applies here. The simplest model assumes that mostly the same people voted (plus/minus some small number), but that a couple percent of them switched from one party to the other.