South Carolina may put an end to Mitt Romney’s claim to inevitability. Indeed, if the polls that show Newt “Open Marriage” Gingrich surging into a competitive—perhaps even a top—position in the Palmetto state primary are right, the premise that Republicans favor anyone-but-Romney will be well established.

But Romney was never as inevitable as he seemed for a few weeks there.

Yes, he had the most money, Yes, he had some good endorsements. But Republican despise the guy. Seventy-five percent of Iowa Republican Caucus goers went out to the trouble of leaving their homes, driving to school auditoriums and community centers and spending an hour or so caucusing for the sole purpose of opposing Romney. Sixty percent of New Hampshire Republican Primary voters sent the same signal a week later.

But Romney secured the “inevitable” tag when he became—in the language of the breathless horse-race coverage that passes for political reporting—“the first Republican challenger in the modern primary system’s 35-year history to win both New Hampshire and the Iowa caucus, establishing a formidable lead in the contest to pick a candidate to face Barack Obama in November’s presidential election.”

Even before the New Hampshire win—which was always pretty much a given for the New England candidate who essentially lives in the Granite State—no less a GOP commentator than former White House political czar Karl Rove was suggesting that the Iowa caucus results represented a huge coup for the unloved front-runner. Rove wrote in the Wall Street Journal on January 5 about “A Big Win for Romney in Iowa.”

Some of us questioned at the time whether an eight-vote advantage (out of 123,503 cast) qualified for the “big win” title. But because of the way that the Iowa results produce a headline that tends to define the next stages of the nominating process, the fact that Romney came out on top—even by the narrowest of margins—was a big deal.

Or not.

Because, of course, Romney did not “win”Iowa.

He lost.

Or, at least, it looks like Romney lost.

The Iowa Republican Party has released the “final, certified totals of the January 3 Iowa Caucus presidential preference vote,” and they show Rick Santorum with 29,839 votes to 29,805 for Romney.

Now, some will say that a thirty-four-vote lead is not exactly a landslide.

But Santorum’s “certified”lead is four times as substantial as Romney’s preliminary “big win.”

Unfortunately, Santorum’s claim to “victory” is muddied by the fact that the Iowa Republican Party’s back-room counting operation could not find and certify the results from eight precincts, where hundreds of votes were cast in the preference poll.

So, after candidates campaigned for years, after campaigns and Super PACs spent tens of millions of dollars, after the media turned all its attention to the state for the better part of two weeks, and after the Republican race was shaped by their “results,” the Iowa Caucuses have done nothing but confuse things.

The problem is not corruption, as the able Iowa blogger John Deeth explains. Rather, the Iowa caucuses are set up in such a way that they are prone to being defined by “plain human error.”

This is the dirty secret of what happens in Iowa. The caucuses are not elections. They are media events, organized by the political parties, that in the case of the Republicans hang on a glorified straw poll. And they can’t even get the straw poll right.

This is not the first time the caucuses have made a mess of the political process. In 1976, according to the Des Moines Register, “Iowa Democratic officials ignored a rule in calculating delegates, resulting in exaggerated projections of delegates for front-runner Jimmy Carter.” In 1980, Republicans declared that George H.W. Bush had beaten Ronald Reagan but two days later a CBS New review suggested that Reagan had actually won. In the 1984 Democratic caucuses, again according to the Register, “the party’s own counts proved unreliable: Party leaders were sure Walter Mondale had come in first, but they weren’t certain where other candidates had finished. Many votes were never turned in.” In 1988 and 1992, “Democrats refused to share caucus vote totals…” In 1996, “Republicans decided to do no official count…” In 2008, Republicans “experienced some data entry problems.”

Now it is 2012. This year’s caucuses were a bigger deal than ever, more closely watched than ever, But one constant remained: They produced a screwed-up, inconclusive result from based on a ridiculously small turnout dominated by extremists.

This is not what democracy looks like.

Neither party should begin a nominating process with caucuses run by political parties that have neither the capacity nor, it seems, the will to produce reliable results.

The process has to be reformed.

Some Iowans, such as Deeth, suggest that the state should make the caucuses a more official operation, using “the existing infrastructure” of “our Secretary of State, county auditors, and poll workers.” That reform might produce clearer results. But it would not address the fact that the caucuses are ridiculously unrepresentative. They attract less than 5 percent of Iowans. They do not even do a good job of attracting partisans; this year’s GOP caucuses drew less than one in five registered Republicans.

Yet, as a CBS News analysis notes, “Every four years, this relatively tiny group of people—a subset of a subset of a subset—holds extraordinary power. Their whims, and those of their Democratic counterparts, are breathlessly followed by a narrative-hungry media laser-focused on the first-in-the-nation voting event, and candidates that do poorly in the caucuses often drop out of the presidential race.”

Caucuses are a remnant of the old-fashioned insider politics that progressives fought a century ago to end with the development of open primaries. There is no democratically appropriate or acceptable reason for Iowa, or any other state, to maintain a caucus system.

Primaries, especially open ones, are far more representative.

So why doesn’t Iowa just switch to a primary? Because of the unofficial “deal” that makes New Hampshire the first primary state but allows Iowa to start the process with caucuses. Both parties accept this charade. And, as such, they are guilty of warping nominating processes that are already too closed, too costly and too prone toward manipulation.

There is nothing wrong with Iowa going first.

There is something very wrong with Iowa going first—and warping the process of picking presidents—with unrepresentative and dysfunctional caucuses.