This far out, then, the narrow margins shown by my math are anything but determinative. They serve not as predictors, but as mere frameworks for understanding where things stand right now.

Tweet

Conservatives should not be overconfident: The Romney-Ryan ticket still faces a daunting task.

There are several ways to analyze and predict election results, with the best ways combining almost equal degrees of art and arithmetic. Before Mitt Romney chose Paul Ryan as his running mate, I had run my more artistic/impressionistic model in two ways and, amazingly, both came out with a 269-269 tie in electoral votes. Without re-running my mental model state by state, I nonetheless would be inclined now to give the Republicans a slight edge. Unfortunately for them, though, when I run the arithmetic, the tiny edge moves the other way: Barack Obama wins, 275-263.

Before I explain how I arrived there, let’s be clear: Any election this close, with polling so stable, is likely to be particularly subject to having its outcome determined by national conventions, debates, and news events. Those events may move the needle only slightly, but nonetheless could prove amazingly decisive.  This far out, then, the narrow margins shown by my math are anything but determinative. They serve not as predictors, but as mere frameworks for understanding where things stand right now.

Here are the broad numbers to think about (I won’t weigh this down, or reveal my exact indices, by providing exact numbers): Obama won the last race, 69.4 million votes to about 60 million. Four years earlier, the Bush re-election campaign did a superb job of voter turnout and maxed out at about 63 million. Figuring a GOP baseline of halfway between the effective Bush organization and the disorganized McCain mess, Obama still starts with a lead of nearly eight million votes.

Now, boost both sides by some population growth, but boost Obama more because more of the growth has been in demographic groups that lean Democratic. Then, give another boost to Obama because his team has had four years in power to build its ground game. Then, give him another boost through vote fraud, deliberately enabled by Eric Holder’s Justice Department.

The factors hurting Obama are still bigger, though. Enthusiasm for him even among those who lean his way is down. So, subtract a few percentage points for decreased turnout among this base. Then, directly shift another couple of percentage points from Obama to Romney: These are the people who have completely changed their minds. Throw in a tiny erosion at the margins because of Joe Biden’s drag on the ticket, along with some other, very small adjustments of my own devising, and the final, expected, national popular vote for him is between 66 and 67 million – down from 2008, to be sure, but by no more than three million votes.

Now, add to Romney’s base the mind-changers already subtracted from Obama, plus a reasonably significant boost in “base” enthusiasm rooted entirely in severe anti-Obama feelings. Give him a few other adjustments, and he still only reaches a bit above 65 million votes – two million better than Bush-2004, but still a million votes beneath Obama.

So Obama wins the popular vote. The key thing is, of course, the electoral vote. Studies show that shifts in voter sentiment tend to be fairly uniform across the country, rather than varying widely geographically. The key is to apportion the overall percentages described above, among the 11 states most likely to “swing.” (The only exception to proportional apportionment is Wisconsin, which went freakishly strong for Obama four years ago, but which has dual pro-Romney tailwinds in, first, the successes with regard to Scott Walker and, second, in favorite son Paul Ryan.) What happens is that Nevada, Michigan, and Minnesota still look rather safe for Obama, while North Carolina really should be in the bag for Romney.

The two remaining biggies, Ohio and Florida, are very near toss-ups, but the arithmetic puts them slightly in Romney’s camp. Colorado, Iowa, and New Hampshire lean a bit more strongly to Obama, but still within range. And Wisconsin and Virginia are so close as to be virtual coin flips. Assign those states accordingly, with Wisconsin going Romney due to Ryan’s presence while Virginia falls to Obama because so many new government workers in the DC suburbs fear the Ryan budget. Bingo: Even giving Romney both Florida and Ohio (a tall order, despite my model slightly favoring it), he still loses 263-275.

In this scenario, then, the two key battlegrounds become Iowa and Colorado: Without snagging them, Romney can’t get to 269 (presumably winning in the GOP House) or a 270 majority. (If Virginia and Wisconsin switch sides from the above scenario, and Romney also wins New Hampshire, that’s another Romney route to 270.) With one of the two battlegrounds, Romney can win.

The more pro-Romney news is, as I noted in the beginning, my best predictive results usually meld my state-by-state, partly anecdotal, impressions (the art) with the straight numerical analysis. My impressions give both Iowa and Colorado to Romney, with New Hampshire and even Nevada still in play -- while Ohio hangs on a razor’s edge.

Midway between those impressions and the arithmetic… we arrive right back around an exact tie. Many long nights of worry remain.