Why Ohio matters in November, but not on Super Tuesday

This is the first column by longtime political journalist and observer Jeff Greenfield, who joins the Yahoo! News team to contribute to our coverage of the 2012 presidential campaign. The host of PBS' "Need to Know," Greenfield won three Emmy awards over the course of a career covering national politics for several networks. You can follow him on Twitter: @greenfield64.

This primary season has been characterized by the absence of familiar friends. Where is the "momentum" of yesteryear? When did voter preferences start shifting so suddenly, so violently, that a double-digit lead is gone within a matter of days?

But there's one reliable old friend that's been with us from the start: a widespread incomprehension about how to count. What's more, there's every reason to think that this cluelessness will shadow the political conversation once the general election begins.

It began at the beginning, in Iowa, when the cable networks struggled late into the night, breathlessly waiting to report whether Mitt Romney or Rick Santorum had won by 8 votes, or 24 votes, or 32 votes. By any rational standard, it didn't matter. The caucuses ended in a tie; neither candidate would gain any measurable advantage in delegates, and--as has happened more often than not in Republican nomination battles, Iowa had no impact on future contests (unless you count the formal end of Michele Bachmann's candidacy).

So how to explain the frenzied faux-excitement of this non-contest? The same way to explain the persistent misreporting of the primary process in general: the inability to distinguish between a primary and an election.

Had the Iowa contest been part of the November election, those handful of votes would have mattered a great deal; in the fall, a win by any margin wins a candidate all of a state's electoral votes. When George W. Bush won Florida, and thus the presidency, by 537 votes out of the 6 million that were cast in 2000, they still played all the notes in "Hail to the Chief" when he walked into a room.

But that's not the way most primaries work. At one level, most folks who follow politics know this. When Romney beat Santorum in Michigan by three percentage points, his net gain was all of two delegates. That same day, his win in Arizona gave him all of that state's 29 delegates. Yet, had he lost Michigan by three points (or half a point), that would have been described not simply as a hit to his campaign, but as a blow that left it faltering, crippled, on life support--even as his campaign would have ended the night with a net gain of some two dozen delegates. Something like that may well happen on Tuesday should Romney lose Ohio, even though he's all but assured of coming out of that day's 10 contests with far more delegates than any of his rivals. "Romney Loses Ohio!" is a more compelling headline than a laborious exercise through the numbers, even if such a headline would convey the highly misleading idea that Romney had lost every delegate up for grabs.

It's as if we can't remember that primaries are not elections; "winning" them is nothing like winning a state in November. The fact that Barack Obama in 2008 netted more delegates out of Idaho and Kansas than Hillary Clinton did by winning Ohio and Pennsylvania is comprehensible only if you remember that under the Democratic rules of the game, it's generally more profitable to win small states by big margins than big states by narrow ones. In the fall, of course, the exact reverse is true.

The most significant lesson in all this, I think, is not to obsess about the national poll numbers, which will be carpet-bombing us by Labor Day on an almost hourly basis. As often as not, they offer a distorted measure of just how narrow the difference between victory and defeat is in many of our presidential elections.

When a margin is as razor-thin as 1960 (one-tenth of 1 percent) or 2000 (half a percent), no one has to be told how close the outcome was. But there are other times when a clear popular vote spread concealed an astonishingly close election.

In 1916, President Woodrow Wilson beat Justice Charles Evans Hughes by 3 percentage points--a clear, if hardly overwhelming, spread. But in the Electoral College, the result was 277 to 254 (266 were needed to win). The difference was California, and its 13 electoral votes, which Wilson won by 3,773 votes. Had 1,900 Californians changed their minds, Hughes would have won the Presidency.

In 1948, Truman won his historic upset with a decent plurality, a margin of 4.5 percentage points in the popular vote. But shift 3,550 votes in Ohio (out of 2.9 million cast), and 9,000 votes in California (out of 3.8 million), and Truman's electoral majority disappears (the segregationist Strom Thurmond carried four Southern states) and the election would have been thrown into the House of Representatives.

And in 1976, Jimmy Carter's two-point national plurality (1.8 million votes) vastly understates how remarkably close his election was. A shift of little more than 5,600 votes in Ohio (out of 4 million) and 3,600 in Hawaii (out of 287,000) would have given the election to Gerald Ford.

Nor was George W. Bush's 3-million-vote victory in 2004 nearly as wide as that number suggests: change 60,000 votes in Ohio (out of 4.5 million cast) and John Kerry would have become the second president in a row to have won while losing the popular vote.

What these numbers tell us goes well beyond the obvious injunction to remember that state votes, not the national vote, determines who wins the White House. Specific twists and turns within a state can affect the outcome. Hughes likely lost California, and the presidency, in 1916 because Hiram Johnson, the state's governor and the leader of the Republican Party's progressive wing, felt snubbed by a missed appointment and sat on his hands on Election Day. I've long believed that Gerald Ford's "no Soviet domination in Eastern Europe" debate flub cost him a week of ill will with Polish and other Eastern European descendants, who are found in Ohio in large numbers.

In 2004, a same-sex marriage ballot measure in Ohio was crucial. As the New York Times noted two days after the election, "political analysts credit the ballot measure with increasing turnout in Republican bastions in the south and west, while also pushing swing voters in the Appalachian region of the southeast toward Mr. Bush."

I don't mean to repeat the old chestnut that "all politics is local"--as overstated a political maxim as any. Clearly the state of the economy, or the fallout from a potential international crisis, have driven plenty of elections. But some politics is local. And more often than we might think, those local matters have affected a relative handful of votes, votes which in turn have determined whose hand is on the button.

Jeff Greenfield is a Yahoo News columnist and the host of PBS' "Need to Know." Follow him on Twitter: @greenfield64.

More popular Yahoo! News stories:

In Ohio and elsewhere, Obama's campaign quietly organizes while Republicans fight

Romney hoping for a 'reset' with voters after a long and heated primary

'He's not a favorite son': In search of Romney Fever in Massachusetts

Want more of our best political stories? Visit The Ticket or connect with us on Facebook, follow us on Twitter, or add us on Tumblr. Handy with a camera? Join our Election 2012 Flickr group to submit your photos of the campaign in action.