Marc Ambinder has a good write-up of the final numbers on voter turnout for 2008.  Delays in counting and certification of all ballots meant that the total number could only be determined now.  What he found was very interesting.  Read more...

Overall, 131.1 million ballots were cast, and turnout was slightly higher than in 2004.  But what was really interesting was this, combined with the exit poll data about voter contact:

It doesn't automatically follow that a larger-scale voter contact effort will necessarily contact huge numbers of additional voters; indeed, Obama and Kerry contacted about 26% of the electorate in both years.

Now, this relies on exit poll data, so it almost certainly confuses "the Kerry campaign" with the DNC, America Coming Together, and every other progressive group under the sun that was doing voter turnout efforts that year.  But regardless, it's striking that despite all the stories of unprecedented volunteer activity, the percentage contacted stayed about the same.  There are a few ways to think about this:

• Quality of contact: Ambinder mentions this one.  If you're doing a better job at minimizing the touches of people who you don't want to talk to, this could drive down the number of contacts.  For example, talking to a firm McCain supporter is a waste of time--why not avoid it if you can find a good way to predict this outcome?
• Number of touches: The exit polling did not break out the number of times each person was contacted.  More contacts would produce a higher-quality turnout operation, even if the number of people contacted was the same.
• Types of contact: It's also tough to find figures about what types of voter contact were done.  We know that door-to-door from people living in your community is the most effective field tool by far, and if the Obama team substituted more of that for less phonebanking to different states, it might have consumed more labor per contact.

As the kids say, read the whole thing.  There's some interesting information there.