Voting is the ultimate popularity contest. In the United States, voters have the opportunity at least every two years to cast ballots in federal elections, either filling seats in the U.S. House of Representatives, the U.S. Senate or the White House.
But operating non-stop between those elections is a political prognostication machine designed to predict the next batch of winners. There are at least a dozen major polling organizations who quiz Americans on a daily basis about their approval or disapproval of current officials, their opinion on the latest hot-button political issues, and whether or not the nation is on track or off the rails.
There's a lot at stake in these polls. Political donors use them to decide which candidates and political issues to bankroll, and savvy congressional representatives consult them to calculate the political risk of backing or splitting with the White House on its legislative agenda.
But the polling number that grabs the most headlines is the president's job-approval rating. Recently, President Donald Trump tweet-bragged about a job-approval rating of "around 50%" according to Rasmussen Reports, even though Rasmussen's approval numbers on that particular day were only 44 percent.
The reason the president is tweeting in March about his rounded-up approval ratings — and the reason Democratic political pundits rushed to correct him — is that job-approval ratings are about more than just stoking or bashing the president's ego. They are perhaps the single best predictor of who will win the 2018 midterm elections in November.
Approval Ratings and Midterms
Midterm elections happen exactly two years after each presidential election and almost always pose a serious challenge for the party in the White House. The president's political party has lost seats in the House of Representatives in all but two midterm elections since World War II (exceptions were Bill Clinton in 1998 and George W. Bush in 2002). What's interesting is that the severity of the midterm losses appears to be directly related to the popularity of the president.
Gallup reported in 2010 that when postwar presidents had approval ratings of 50 percent or above at the time of a midterm election, their party only lost an average of 14 seats in the House. But if they had an approval rating of 49 percent or lower, the average loss was 36 seats. (Using more recent data from 1970-2014, Bloomberg put the average loss at 33 seats for presidents with approval ratings under 50 percent.)
There have been some exceptions, however — Ronald Reagan had an approval rating of 63 percent in 1986 and his party still lost five seats in the midterms. On the other side of the aisle, John F. Kennedy had a 61 percent approval rating in 1962 and lost four seats.
The connection between presidential job-approval and midterm elections is particularly strong during the first midterm election after a new president takes the White House. Pundits see those first midterms largely as a referendum on the president's job performance.
That's why Democrats are making such a big deal about the March 13 special election in Pennsylvania, where their candidate won by the slimmest of margins in a region that went heavily Republican in 2016. They point to it as a sign that Trump's poor job performance and low approval ratings will lead to a "blue wave" of Democratic victories in 2018.
Approval Ratings and Re-election
What about presidential elections? Does an incumbent's job-approval rating make or break his chances of winning re-election?
The short answer: yes. Since World War II, not a single incumbent presidential candidate has won re-election with a job-approval rating below 50 percent. Barack Obama almost became the exception in 2012, with an approval rating in the mid-40s a few months before election day. But by November, it had cleared the hurdle at 51 percent.
Does that mean that if Trump's approval rating continues to languish below 50 percent over the next two years that he will definitely lose re-election in 2020? Absolutely not. If there's one thing that Trump has proven in both his candidacy and short time in office is that precedent means nothing to this president.
Just look at this one startling fact, Trump was elected in 2016 despite being the least-liked major party presidential candidate of all time. His unfavorability rating on election day was a whopping 61 percent. It helped that his opponent, Hillary Clinton, had the second highest unfavorability rating in history at 51 percent.
You might notice that not every job-approval poll comes up with the same numbers. That's because each polling organization uses a slightly different methodology. Some pollsters, like Gallup and Quinnipiac, call up a representative sample of Americans over 18 and simply ask them if they approve or disapprove of the president's job performance. In Trump's case, those polls result in lower approval and higher disapproval numbers.
But what about Rasmussen Reports, Trump's favorite pollster? In that case, Rasmussen doesn't simply poll American adults or even American registered voters, a tighter demographic. Rasmussen only counts answers from "likely voters," people who say they are likely to vote in the next election. Rasmussen is also the only company polling people on a daily basis.
On top of that, Rasmussen Reports give respondents four options: "strongly approve," "somewhat approve," "somewhat disapprove," and "strongly disapprove." In Trump's case, the inclusion of "somewhat approve" seems to be enough to lift his Rasmussen approval numbers as many as 10 points higher than other major national pollsters.