OCTOBER SURPRISES DON'T WIN ELECTIONS

National Journal's Chuck Todd makes a case against October surprises swinging political races.

Big news in the last month does matter, but October stories aren't singularly important. If they were, campaigns wouldn't spend a dollar in any month other than October.

Campaigns are rarely won or lost in October; the decisive moment happens much earlier in the cycle.

October incidents can have an effect on turnout, but they usually do little to add or change the perception of a candidate. A candidate's image is built months earlier.

In the two weeks since the Foley scandal unfolded, I've noticed an interesting pattern in key races: Republicans seeing the biggest drops in support are those who didn't think they were in big trouble three months ago. Now they have little framework with which to define themselves or their opponents in this awful climate.

The most embattled Republicans, many of whom we identified 18 months ago, seem to be weathering "Hurricane Mark" better than expected. The reason: Incumbents in places like Connecticut, Missouri or Pennsylvania have been preparing for what they feared would be a rough 2006 for nearly 18 months.

What this proves is that campaigns are rarely won or lost in October; the decisive moment happens much earlier in the cycle.