The word “spin” can mean many things. One definition is to present information in such a fashion that it makes people see something that isn’t really there. A classic example would be a memo from the Democratic firm Democracy Corps on a recent poll they conducted in “competitive House districts” for the 2014 midterm elections.
The memo’s authors want readers to believe that the Democrats have a chance to win back House of Representatives in the midterms based on Democracy Corps data. History and their own polling data, in fact, suggest the very opposite.
The president’s party rarely picks up seats during midterm elections. It has occurred only three times since the American civil war: 1934, 1998, and 2002. All three featured presidents who were very popular. President Clinton in 1998 and President Bush in 2002 had approval ratings into the 60s in most surveys. Despite that high approval, their parties picked up only five and eight seats respectively. The Democrats need to pick up 17 to gain control of the House in 2014. The president’s party has not picked up more than nine seats in a midterm since 1865.
In order for that to occur, we would almost certainly need to see an extremely popular president. We don’t.
Among registered voters, President Obama’s approval rating is in the mid 40s. No poll since the middle of May has had President Obama’s approval rating above his disapproval rating among registered voters. The best estimate I have is that President Obama has somewhere in the neighborhood of a -4pt approval among registered voters. It’s probably slightly worse among those who turn out to vote in midterm elections.
Indeed, the Democracy Corps survey shows that President Obama’s approval rating in the swing districts is a measly 44%. His net approval among these 2014 likely voters is -8pt. This is despite the respondents saying that they voted for President Obama by a 3 pt margin in 2012. It’s very difficult to imagine that Democrats can win back many seats when Obama is this disliked in these districts. In the last two midterms, the percentage of the vote won by the president’s party was pretty much equal to the percentage who approved of the president’s job performance.
You might say that the Republican brand is so toxic that House Democrats can overcome a relatively unpopular president. The Democracy Corp poll demonstrates the opposite. The tested Republican candidate in the poll has a 2pt advantage over the Democratic candidate. That’s little changed over the 3pt margin by which respondents said they voted for Republicans in 2012. Such a difference is worth a few seats at most, but certainly not 17.
A closer look illustrates more problems for a possible Democratic takeover. In the seats that Democracy Corps identifies as the most vulnerable, Republican candidates are 1pt ahead. In this same category at this point in the 2012 cycle, Republican candidates were actually down 1pt. A few months before the 2012 election, Republicans were down 6pt in this category.
So, the most vulnerable Republican candidates are actually in a stronger position now than they were for the 2012 election. When Republicans were far more at risk in 2012, they lost only 11 seats in this category and eight overall.
The reason Republicans lost fewer seats overall than just the Republican vulnerable category is because it isn’t just Republicans who are vulnerable. The poll also asked 500 respondents in Democratic districts how they planned to vote. Democrats lead in these districts by 2pt. This certainly does not spell a Republican wave, but it’s worse than the 4pt edge these same respondents said they gave to Democrats in the 2012 elections. This could lead to a few Democratic seats actually falling to the Republicans.
The overall picture the ballot test points to, at this point, is a status quo election. That matches the Washington expert ratings of the Cook Political Report and Rothenberg Political Report – both of which have a near equal number of Democratic and Republican seats up for grabs, with, in fact, a few more Democratic-held seats in play.
Could the political environment change to favor Democrats? It can, but I doubt that would be enough. Joseph Bafumi, Bob Erikson, and Chris Wlezien have shown that the president’s party position in the ballot test deteriorates as you move closer to the actual date of the midterm election. It’s why the Democrats lost all of their Democracy Corp-designated most vulnerable seats in 2010, even though they had a 4pt lead in them at this point in the cycle.
Given the president’s approval rating at this point, it’s more likely for the Democrats to lose ground than gain it. Only an unlikely 15pt improvement in Obama’s approval might conceivably reverse it.
The truth is that Democrats face a very uphill battle to take over in the House of Representatives. The actual data from Democracy Corps, whose polling I trust, proves Democrats are quite unlikely to take back the House. No amount of spin will change that fact.