The final reports are nearly in, and Curtis Gans and Michael McDonald agree on at least one thing: turnout in 2008 did not set records. Somewhere between 128 and 130 million votes will be recorded, according to Gans and McDonald.
Record registration numbers dd not lead to record turnout, nor did record early voting numbers.
And the relatively modest election day turnout meant that the early voting figures loom even larger–certainly exceeding 30%, and might even hit that one-third estimate.
Most interestingly, though, is a comment made by McDonald in the Globe story that I think bears some reflection.
“It became more evident to voters at the end that Barack Obama was going to win. That probably tamped down the turnout and disproportionately affected the Republicans.”
It bears repeating: did early voting returns (particularly the strong Democratic and African American numbers) influence–and potentially depress–election day turnout? If true, this would be particularly worrisome because, as was pointed out on this blog and on earlyvoting.net, and by Michael in some early news reports, early voting states are not representative of all states, nor are early voters representative of all voters.
Using the partisan coloration of early voters as some early warning signal of election day trends is probably a mistake–a lot of caveats and “what ifs” are necessary and these are very difficult to convey in a news story.
This issue may come up again as we consider extensions to early voting in the coming congressional session.