A number of liberal and conservative commentators make the case that Mitt Romney is not “closing the sale” with voters. For liberals, this is a relief. (This is why the Obama surrogates repeat it so often on cable TV and in print.) For conservatives, it’s a worry. But neither side should take this seriously for now.

The “undecided” voters are that same 7 to 10 percent of the electorate we’ve been looking at since the end of the primary. They are the least-engaged voters. This group includes a lot of Obama voters from 2008 who don’t know whether they will vote for him again or even vote at all. Don’t be surprised to see a lot more Romney ads pushing the “buyer’s remorse” notion or telling Obama’s 2008 supporters that the president didn’t keep his promises. (In other words, no one expects you to say you were conned, but there’s no excuse to be snowed twice.)

The most critical voters will clue in at discrete points in the next 96 days: for the VP announcement, for Romney’s convention speech (and Obama’s, I suspect to a lesser extent, since they know him already) and for the debates. Overwhelmingly they will be influenced by the economy, which they think stinks.

I wish it were otherwise, but such voters don’t care about foreign policy. They generally tune out TV ads, and they think most politicians lie. They don’t care about 16 point-plans; they care about the “sense” they get from each candidate. Candidates need to have plans, but the details of those plans are not as critical as impressions like “he seems competent” or “he’s a decent guy” or “he ‘gets’ what is wrong with the country.”

Welcome to the much-sought-after undecided voters — probably no one who is reading a political blog in August — who will decide the next president. Oh, and they only matter if they live in Florida, Ohio, Virginia, New Hampshire, Michigan, Wisconsin, Pennsylvania, Colorado or Iowa. (Democrats can stop trying to make the case that North Carolina, Missouri or Arizona is in play. None is.) The rest of the electorate is fairly set, although each side has to be concerned about motivating the base.

Given this reality, we can conclude a number of things. First, national polls are useless. I’d be shocked if either campaign is paying money to do their own. Second, most of the summer has been background noise, polling for which will be irrelevant after the three events identified above. And third, if the economy continues skidding downward, Romney needs only to clear the “acceptable” bar, not sweep the undecideds off their feet. If August and September’s economic numbers are bad, it will become increasingly difficult for Obama to win, absent a significant Romney error (not a phony media-created “gaffe”) or revelation.

It may be mind-boggling to the politically aware, but just about nothing that happened before this month will affect the election outcome. With this in mind, you may understand what the camps are up to and why the most critical factors (the economy, a sudden foreign policy emergency) are largely out of the hands of the politicians.