Mellman: Any number in a storm — looking to predict elections

We’re desperate to know who’ll win the presidency.  

We become like the sailor who’ll choose “any port in a storm” — a phrase that began life in the erotic 1749 British novel “Fannie Hill” and migrated to American politics in 1841 as Winfield Scott’s slogan in his primary campaign for president.  

Today we aren’t in search of ports, but rather any number that might help predict the general election outcome, as we thrash about in a sea of uncertainty. 

Some suggest winning a state’s primary is a sign of candidate strength there. 

It might sound slightly sensible at first blush.  

But while President Biden did far better than Donald Trump in South Carolina, would you give the Democrat the edge there? Nonsensical.  

Winning a state in a primary or caucus has almost no bearing on the ability to carry it in a general election. 

In 1972, World War II hero and anti-Vietnam War movement leader George McGovern (D) won nomination contests in 21 states. He lost 20 of those states in the general election. 

In the 2012 general election, former Massachusetts Gov. Mitt Romney (R) lost more than half the states he’d won on his way to the nomination, while in 2016, Trump lost 17 of the states in which he’d won primaries. 

Predicting general election outcomes from primary results is a fool’s errand. 

The next redoubt of those scouring the tea leaves provided by primary results for general election insights relies on primary results within demographic subgroups. It’s equally fraught.  

The most important political fact about anyone who turns out for a primary is not their gender, age, education or income. Rather, the central distinction is the very fact that they chose to vote in a partisan primary. 

Young people who vote in a Republican primary are very different from young people who do not. Assuming those who didn’t vote in the primary will behave like those who did when the general rolls around ignores their most salient distinction. 

In 2016, Hillary Clinton lost primary voters under 30 by a massive 44-point margin. Did that signal major problems for her with young voters in the general? No. She won them handily. 

In contrast, Clinton bested Sen. Bernie Sanders (I-Vt.) by 28 points among those whose formal education ended with a high school diploma or less. However, in the general, Trump prevailed in that segment. 

The latest attempt to find a general election port during the primary storm considers exit poll questions asking people how they might vote in a general election if the candidate they did not support in the primary becomes the nominee. 

This kind of question has all the problems we described a couple of weeks ago in asking people to project themselves into an alternate universe and predict their behavior.  

Moreover, they’re being probed as they are casting their vote for another candidate—the high point of their support for the rival. 

Such questions just don’t tell us much. The Nation magazine, which sadly ceased being a font of wisdom long ago, featured a piece in 2016 headlined, “Relax, Donald Trump Can’t Win.”  

In reaching this foolishly wrong-headed conclusion, it employed some of the analysis of primary demographics discussed above, but also cited NBC/Wall Street Journal polling which found “47% of Republican female primary voters said they could not imagine themselves voting for Trump.” 

These respondents’ imaginations were obviously limited. Such defections would have rendered a Trump victory impossible. Yet he won, with 88 percent of Republican women, within one point of his support among GOP men.  

Nonetheless, capable prognosticators are giving serious weight to the fact that 24 percent of South Carolina Republican primary voters claimed they would be so dissatisfied with Donald Trump as the nominee that they would not vote for him in November.  

Even if those voters just stayed home, it would be enough to transform South Carolina into a swing state.  

Does anyone believe that’s possible? Of course not.  

Our desire to read tea leaves is insatiable, so we read too much into each numerical leaf we happen upon. 

Sometimes, though, tea leaves are nothing but shriveled foliage in the bottom of a cup. 

Mellman is president of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House members. Mellman served as pollster to Senate Democratic leaders for over 20 years, as president of the American Association of Political Consultants, a member of the Association’s Hall of Fame, and is president of Democratic Majority for Israel.