How presidential approval can make or break a midterm

There’s debate every midterm year over whether a president is “on the ballot,” or whether the House and Senate elections are a referendum on him. Technically, no, President Obama isn’t up but the political reality is different.

The chart below shows the recent pattern of presidential approval at the time of the midterm, and how their party fared in the House. The last two times a president had majority approval at a midterm – in 2002 and 1998 - coincided with times their party did well. In both cases, surprisingly well. When they’ve been underwater, including in 2010 for Mr. Obama, their party has lost.

Today the president’s approval is 46 percent in the most recent CBS News poll. Which is one reason the GOP starts the year with an edge to keep its House majority, maybe even gain.

In the last midterms, six in 10 voters said their congressional vote was partly to either express “support for” or “opposition to” Mr. Obama. And it intuitively makes sense, too, that any president’s standing can affect outcomes: like it or not he’s the “face” of his party and brand. (It's why even party affiliation in polls can move in tandem with his popularity, especially in this era of high partisanship.) Voters might try to reward a popular president with more legislative power, or put a check on an unpopular one.

There’s also the oft-overlooked, behind-the-scenes role of candidate recruitment in the months prior to the election. Your party wants that star fundraiser, up-and-coming local officeholder to step up and run for Congress? Well, it’s a lot easier to recruit them when they don't have to duck questions about cooperating with the party’s embattled national leader, or to raise money from an excited donor base.

That said, there are some important side notes to know.

If you go back further, popular presidents lost midterms, too, which had given rise to the maxim that these elections were always bad for a sixth-year incumbent. President Reagan was very popular in 1986 and heading into the elections that year, which he called his last campaign, but Republicans did not fare well. President George H.W. Bush was popular in 1990 and didn’t, either. And that trend goes back quite far.

Still, those years are increasingly starting to look like a different era, by comparison, a time before money and media helped parties more easily nationalize House elections. Ticket-splitting, for one thing, was more prevalent, with Republican crossover for Democrats at rates two or three times higher than today’s scant levels (94 percent stayed with their party in 2012.) An incumbent’s votes “for” or “against” a party line, or a president, might not as easily undermine what they did for the district locally.

All of which is to say: not everyone agrees on a single theory – or whether there even is one – behind the historical data. In any given year, are factors like the number of open seats and retirements, and district-specific issues that affect each party's results. And sometimes there are overarching national issues, too. In 2006, Iraq hurt the Republicans; Clinton had an economic boom and backlash against impeachment in 1998; in 2002 President Bush had wide support from his handling of 9/11.

Certainly, though, a constant refrain this year will be candidates being described as either “distancing themselves” from the president or “running with” him; the parties might go back and forth over whether the campaign is a “referendum” on his presidency – which the GOP will surely claim if they win - or about broader issues, as the Democrats might want. This is why. Watch Presidential approval as a leading indicator as we go through 2014. (And if it rises, see whether the parties suddenly start to change those interpretations.)