Digging into the Polls
t’s March, which means Spring has sprung, flowers are starting to bloom, the sun stays out later, and we are now just eight months away from the mid-term elections. While political prognosticating, especially in writing, is always dangerous – perhaps never more so than following the 2016 Presidential election – I thought we might take a look at some key polling metrics to get a sense of what we might expect as we get closer to the mid-term elections, with comparisons between polling data as of this morning to data from this precise point in the 2010 mid-term election cycle, the first mid-term election of President Obama’s tenure, during which Republicans captured six seats in the Senate and swung 63 seats in the House– the largest swing since 1948 – and reclaimed the majority and the Speakership in the chamber. One important caveat before we dig in: eight months is an eternity in politics and quite a lot could change between now and November. With that proviso out of the way, let’s roll up our sleeves.
A good place to start to set the stage is to look at American voters’ feelings with regard to the direction of the country. Voters who believe the country is on the wrong track are significantly more likely to lend their support to the opposition party whereas voters who believe the country is headed in the right direction are more likely to cast their ballot in favor of the status quo. For all of the divisiveness in society and in our politics, the data indicates that meaningfully more Americans – about five percent – believe the country is on the right track today than did at the same point eight years ago:
(As an aside, the last time this poll indicated that more Americans felt the country was headed in the right direction than the wrong one was in June 2009, when the Great Recession officially ended.)
The natural next dataset to examine then is the President’s job approval. Since 1962, Presidents with an approval rating below 50 percent as voters head to the polls in the mid-terms have seen their party lose an average of 40 seats in the House of Representatives. With Democrats needing to flip just more than 20 seats in November to claim a majority in the House, this particular data point is not looking good for the GOP and President Trump, who finds himself 14 points underwater in aggregate job approval polling across the country:
But the GOP need not despair just yet. At the very same point in his presidency, President Obama was viewed by most voters to be doing a good job, with a net job approval rating of +3 on March 1, 2010:
Between March 2010 and November 2010, President Obama’s approval rating fell sharply, to -5; an eight-point swing. The same eight-point swing for President Trump – just in the other direction – would see his job approval rating hovering right around 50% -- the almost magical threshold at which Republicans would be likely to have a better election night this November. Democrats should therefore not be counting their chickens just yet.
There is an old truism in national politics: every voter hates Congress but loves their Congressperson. How else can one explain why, even as Congress’ favorability rating as an institution hovers near single digits across the country, the average rate of re-election of incumbent members of the House of Representatives over the last 60+ years rests at around 90 percent? Though Watergate, the Vietnam War, and enactment of Obamacare saw the rate of incumbent re-election in the House drop during those election cycles, at no point in our history have less than 80 percent of incumbent members of the House been voted back to Washington:
With that caveat, we will assess generic party polling. These polls have historically served as contextual indicators of the mood of the country rather than scientific assessments of the odds of one party to sweep the other in November. As we saw above, the best bet in Vegas is on an incumbent, regardless of party affiliation, to keep their job. But history shows that national polling of a generic Congressional ballot can be predictive, generally, of which party has tailwinds heading into an election. For example: depending on the poll, Republicans led Democrats on a generic Congressional ballot at this point in the 2010 mid-term election cycle by somewhere between seven and 10 points. Today, the Democrats are in almost precisely the same position as the Republicans were at this point in 2010. A generic Democrat leads a generic Republican candidate for Congress in nationwide polling by an average of 8.4 points:
In the aggregate, the polling data as of today suggests that voters are more interested in casting their ballots in favor of change and are largely dissatisfied with the current slate of leaders in Washington. This, of course, indicates that the opposition party – the Democrats – should expect to have a better election night in November than the GOP. Though these trends, if they continue, are likely to result in this outcome, with eight months between now and the mid-terms, and with incredibly sticky historical incumbent reelection rates – to say nothing of gerrymandered Congressional districts – it is both far too early and far too simplistic to declare that the Democratic Party will see significant gains in the House later this year.