If there's one thing we in the political media have been certain about in this election, it's that debates matter.

After all, without the Republican primary debates, we would never have had the boomlets of Herman Cain and Newt Gingrich -- badly underfinanced candidates who rose to prominence based on their star turns in the debates.

And, just as Mitt Romney appeared to be on the edge of electoral oblivion at the start of this month, he pulled himself back from the brink with a star turn in the first presidential debate of the general election. Heck, we wrote an entire column this morning on how the third presidential debate tonight in Florida could swing the election.

Right? Well, no -- at least according to voters.

In the brand new NBC-Wall Street Journal poll, 47 percent of those tested said that the debates had made no difference in their candidate preference. We're no math majors -- no duh -- but that's roughly half of the voting public. By comparison, 27 percent said the two debates had made them more likely to support former Massachusetts governor Mitt Romney while 24 percent said the debates had made them more likely to vote for President Obama. Two percent said they weren't sure. (Who are these people?)

Those findings run entirely counter to the "debates matter" narrative that has been in the political ether for decades.  So, what gives?  A few theories:

1. We are living in extraordinarily partisan times. The simple truth is partisans and even those loosely aligned with one party or another have been aligned in their respective camps for months. No external event has shown an ability to move these people because their minds are totally made up.  So, it should be no surprise that 47 percent of people say that the debates haven't changed their opinions; nothing will.

2. People don't like to admit debates sway them. Most people think of themselves as a) smart and b) free thinkers. And so, when asked a question about whether they had their mind changed by a 90-minute debate between the two candidates, the first reaction is to say that it didn't have that effect. (This phenomenon, by the by, is why so many people like to call themselves either independents or undecideds when they are, in fact, neither.)

3. Debates do matter...but not as much as we think. While we tend to focus on the fact that nearly half of all voters said the debates had no impact on how they will vote on Nov. 6, you could also flip the script --boom! -- and focus on the fact that 51 percent of people did say that the debates made them more likely to support one or the other candidate.  We may just be looking for too much from the debate; 75 percent of people are never going to say that the debates deeply changed how they feel about the candidates and their vote -- because of the reasons we outlined above.

Regardless of what the reason is that voters tend to say that debates don't matter, the evidence in this election that they do is hard to ignore.

Take a look at the Real Clear Politics poll of polls of the general election matchup Obama and Romney:


On the day of the first presidential debate -- Oct. 3 -- Obama had a 49 percent to 46 percent edge over Romney. Five day later, his lead was down to .5. Two weeks after the debate, Romney led by .4.

So, something happened during that time. And, aside from Romney's overpowering performance in the debate -- more than 70 percent of voters in a Gallup survey said the challenger had won -- nothing was a big enough event to move the numbers like they moved.

While it's impossible to ascribe the movement toward Romney as solely sue to the debates -- we would need to interview everyone who either got off the fence and went for Romney or who switched from Obama to Romney in that time -- it's a very reasonable conclusion that the debates were the prime mover in re-shaping the race.

The debate over the importance of debates has always been with us -- and always will be. But simply because people say debates don't impact them doesn't mean they are right.