Last Friday, I went to hear Charlie Cook, publisher of the Cook Political Report speak about the 2008 Elections at Fordham University’s Center for Electoral Politics and Democracy.

Among the smart and interesting things Cook said was that voters download information about candidates at different times – that people who follow politics learned a lot about the 2008 candidates at the beginning of campaigns and others learn later. Many, Cook said, won’t start paying attention until after the 2008 World Series.

I thought that was smart because it fits a new theory of mine.

I call it the theory of the two electorates and this is it:

I think that because of the internet and cable theory, people like us who live and breathe politics (let’s call us the informed electorate) know much, much more about it than ever before. Years ago, a person who wanted to know about contests for congress in places like Montana had to subscribe to publications like the Cook Political Report. Before the advent of CNN, Fox News & MSNBC, the only time to see reporters give their opinions about who was up & down in various campaigns was on the Sunday morning network interview shows.

Conversely, people who vote but don’t care as much about politics (let’s call them the vast majority) know less than before for the same reasons. Before cable and the internet most Americans watched Walter Cronkite or one of his network rivals to learn what was the most important events that happened in the world that day. Even if they didn’t care about politics, they learned something about it when Walter reported something.

Now these same people don’t have to watch the network news. If they care more about celebrity news or sports or crime, there are plenty of web sites and cable shows to watch so they don't learn who the Des Moines Register endorsed for President.

These changes in the knowledge of the electorate are, I think, a major reason for the volatility of political polls. It’s why a Mike Huckabee can gain so much ground in such a short time in Iowa as social conservatives begin to focus on the Primary. It’s also why he may drop as quickly, as the same voters learn about his record as Governor as Arkansas.

My two electorate theory also helps explain why some many pundits make so many embarrassing predications. Just over a month ago, the consensus was that a Hillary-Rudy contest in November was inevitable. Part of this was because, despite all their disclaimers, they do believe the polls. But it was also because pundits just assume that voters know as much about Rudy and Bernie Kerik or Obama and Oprah as their friends and they do and are shocked when they find out otherwise.

So, my advice to all my fellow members of the informed electorate is to keep this in mind and don’t get ahead of yourself in deciding who’s winning or losing in not just Presidential elections but others too – remember the first publicized poll that had Mike Bloomberg ahead in 2001 appeared the Friday before Election Day!