I think that's a reasonable inference: that the fact that the weather was bad and that people travel over that period could lead to the increase in advance voting. It doesn't bear on our contention or our finding that of those people who advance-voted, 60% would not have voted otherwise. I think that's a fair estimate of how many people advance voting brings into the system generally. Whether advance voting is done by 10% of voters or 5% of voters, some significant share of those voters would not vote otherwise.
On November 22nd, 2007. See this statement in context.