Obama can’t win the election without it.

Two weeks ago, I published a piece arguing that the Obama campaign would do better to take Ohio as their campaign template than to focus on Colorado. Critics pointed out the arithmetic truth that you could remove Ohio and several other Midwestern states from Obama’s 2008 victory column and still have enough electoral college votes to prevail.

That’s true but largely irrelevant. My argument rests on the fact that Ohio is close to being a microcosm of the country—closer than any other pivotal state. As such, winning Ohio is a statistical “tipping-point” for any presidential election: If a candidate can carry Ohio, he will have appealed to a large enough slice of the national electorate to have won the states that tilt even further in his preferred direction, and he is odds-on to win the race. Likewise, a candidate who loses Ohio will almost certainly lose nationally. Here are the numbers from the six post-Reagan presidential elections:

            National Democratic vote (%)   Ohio Democratic vote (%)        Difference

1988                45.6                                                     44.2                             (1.4)

1992                43.0                                                     40.2                             (2.8)

1996                49.2                                                     47.4                             (1.8)

2000                48.4                                                     46.5                             (1.9)

2004                48.3                                                     48.7                              0.4

2008                52.9                                                     51.4                             (1.5)

Over these elections, the Democratic candidate’s share of the Ohio vote averaged 1.5 points below his national share. And the winner of Ohio prevailed in the electoral college in all six elections.

Compare these results to those from Pennsylvania, another state often viewed as contested and pivotal.

            National Democratic vote (%)   Penn. Democratic vote (%)       Difference

1988                45.6                                                     48.4                              2.8

1992                43.0                                                     45.2                              2.2

1996                49.2                                                     49.2                              ----

2000                48.4                                                     50.6                              2.2

2004                48.3                                                     50.9                              2.6

2008                52.9                                                     54.5                              1.6    

Democratic candidates averaged about two percentage points above their national performance in Pennsylvania, but winning that proportion of the vote—without winning enough to carry Ohio—is never sufficient. Gore won in Pennsylvania but lost in Ohio; so did Kerry. Likewise, both Gore and Kerry won in Michigan, Wisconsin, Illinois, and Minnesota, because those states are more liberal than Pennsylvania, but lost the nation as a whole.

You might argue that Ohio is less central to the new majority that Obama forged than it was to Democrats whose appeal was narrower, and there’s something to that. But look at the nine states that Obama won but Kerry didn’t. In three of them—Indiana, North Carolina, and Florida—his share of the vote was actually lower than in Ohio. Unless his relative appeal among them were to change, a loss in Ohio would mean losing at least three other swing states as well, putting him on the brink of defeat. That’s not a chance a sensible campaign would take.

As I showed in my first piece on this subject, a Democratic majority coalition in Ohio doesn’t look much like a majority in Colorado. A focus on Ohio means understanding why Democrats experienced ruinous losses there in 2010—and taking seriously the concerns of the coalition that Democrats need to recreate a majority.

Let’s be clear about what’s at issue in this debate. In a post-election analysis published in TNR last November 5, Ruy Teixeira and John Halpin (two thinkers not exactly opposed to the “new majority” thesis) pointed out that Democratic reverses among white working-class voters were a major contributor to the 2010 Republican avalanche:

“The most significant shift against the Democrats occurred among the white working class—defined here as whites without a four-year college degree. Congressional Democrats lost this group by 10 points in both 2006 and 2008. Yet this deficit ballooned to 29 points in 2010—a deficit even larger than the 22 point margin Democrats suffered in 1994.”

How threatening is that margin? According to an analysis by Teixeira and Alan Abramowitz, Al Gore lost the white working class vote by 17 points, and John Kerry lost it by 23. Obama did better than Kerry, holding the gap to 18 points, which helped boost him to victory throughout the Midwest. Bottom line: Absent an even larger turnout from “new majority” voters than we saw in 2008, there’s no way that Obama can prevail in 2012 without doing much better among white working-class voters than Democrats did last year.

What will that take? Consider an analogy from the other side of the aisle. For the foreseeable future, Republicans will not win a majority of the Hispanic vote. But it matters a lot whether they lose Hispanics by less than 20 points (as Bush did in 2004) or 38 points (McCain in 2008). The difference was the widespread perception that Bush had been an inclusive governor and as president was sympathetic to comprehensive immigration reform.

What is the equivalent for the white working class, which gave as much support to Obama as they did to Gore, and more than they had given to either Kerry or Dukakis? While I advanced some tentative proposals in my previous piece, I don’t want to pretend to more certainty than I have. But one thing seems clear: Right now, the white working class feels left out—not just economically, but also culturally. It doesn’t have to be that way. Bill Clinton figured out how to build a coalition that included them alongside the groups that had risen to power in the Democratic Party since 1972, and he was rewarded with a plurality of their vote in both 1992 and 1996.

Autre temps, autre moeurs. What worked 15 years ago may not work today. But it would be political malpractice for the nascent Obama campaign not to invest the analytical resources needed to come up with a better answer than the administration has found so far.

William Galston is a former policy advisor to Bill Clinton and current senior fellow at the Brookings Institution.

For more TNR, become a fan on Facebook and follow us on Twitter.