Is Mitt Romney likely to suffer as a presidential candidate because he’s so, so wealthy and because his policies (at least as the Democrats tell it) are targeted to helping other rich folk? That’s what a bunch of people have been arguing this week. Scott McConnell put it bluntly:

[T]here really is no precedent for a fabulously wealthy Wall Street operator running on a  distribute-wealth-from-the-middle-to-the-rich  platform being a plausible contender for the American presidency.

See also Josh Marshall here and Scott Galupo here.

I’m going to tell you it probably doesn’t matter … but it’s complicated, and instead of focusing on Romney, I’ll focus more on why it’s so complicated, and what that tells us about the limits of what we know and can know about presidential elections.

Here’s the problem with predicting presidential elections: There just haven’t been all that many. It’s sort of a basic fact of statistical analysis that when there are more variables than there are cases, you’re not going to be able to do much. In fact, conceptually that’s true for any kind of analysis, statistical or not. And that’s the case with presidential elections. There have only been so many to begin with, and if you want to narrow it down to elections with a stable two-party system, you narrow it down further, and if you want elections during the age of modern communications, you’ll have to narrow it even further, starting in the TV age or perhaps as far back as the radio era.

And then think of all the variables that might be affecting the outcome. In fact, just think of two types of variables: candidate and issues. For candidate, we could think of attributes such as ethnicity, gender, marital status, height, beauty (subsets: bald or not? Facial hair or not?), religion, speaking ability, charisma (if such a thing exists), age … I’m sure you can think of more. Issues? The economy, health care, abortion, Afghanistan, defense policy, gun control …you get the point; there are dozens, easily.

Now, the people who study this can find ways to collapse all this so that they can at least try to draw some regularities out of it, but the bottom line is that there is still going to be a whole lot of “as long as things are basically the same” that will include quite a few things that won’t be the same — but we don’t know, going in, how much they’ll really count. All we can do is to take a best guess about whether something that might be new is really different enough from what we’ve seen before that it might have an effect. Well, that’s not all; we can massage what information we have as carefully as possible, and we can use other kinds of elections (there haven’t been many presidential elections, but there have been thousands of House elections) to help educate our guesses.

Got all that? Now, back to Romney. Is what Romney is different from very wealthy candidates such as John Kerry, John McCain, the Bushes, JFK or FDR? I don’t know! Does there seem to be any super-rich penalty (or advantage?) in the general election? Nope. In fact, I’d argue that from what we know, the personal attributes of a challenger to an incumbent president are almost completely irrelevant, apart from a reputation for ideological extremism hurting. All the rest? Doesn’t seem to matter.

Except that there’s just no way to know whether this particular type of wealth might be different. Now, I don’t think it is. I don’t see other very wealthy candidates seeming to suffer, and it seems hard to believe that the way Romney made his money would hurt him in some special way not available to those who made theirs in other types of business, or by inheriting or marrying it. But that’s just an educated guess.

And the point here is that most of our conclusions about elections are educated guesses based on assumptions that all of the new and different things happening in this cycle — and there are always quite a few of them — aren’t new and different enough to change how the basic relationships seem to work. But we don’t know that. Presidential election campaign spending hasn’t been that important because of diminishing returns, but we don’t know for sure that the fabulous amounts that will be spent this cycle work that way.

Shifting the media away from the old TV-based neutral media and to today’s menu of party-aligned media plus some traditional neutral media, based on a variety of platforms, the media will be different than it was in 2008 and a lot different than it was in 2000, and I don’t think that those differences matter, but I can’t say I know that for sure. Nor can we say, for sure, that this particular context of economic disaster will affect this election just as it would if you extrapolated from how the economy seemed to affect other elections. Probably, and the best working theory is that it will, but we don’t know for sure.

So this is a very roundabout way of saying that while I don’t think that Romney will suffer for his wealth in November, there’s always a chance that he will.