Do Presidential Approval Ratings Predict Reelection Results?
Follow me on Facebook and LinkedIn!
Check out my website!
Want me as a podcast guest? Contact me on my Matchmaker FM profile!
Check out my book!
When a presidential election is on the horizon, election forecasters and pundits in the media regularly scrutinize the president’s job approval rating. Even when a presidential election isn’t about to happen, media outlets, depending on their partisan and ideological leanings, either love to highlight the poll or disregard it if it suits their viewpoints. When Donald Trump’s approval rating was low, Democrats were eager to point it out. Now when Joe Biden’s approval rating is low, Republicans are keen to make note of it.
At face value, it makes sense to use the poll as a gauge of how presidential elections will play out. In theory, people that think the current president is doing a good job are going to vote for him, and people that think that he is doing a bad job are going to vote for a different candidate. Sounds simple enough.
But is it really that straightforward?
Caveats Concerning Presidential Job Approval
While useful, job approval ratings for president aren’t necessarily as predictive of who will win the election as one may think. One of the biggest reasons is that job approval is a national poll for an election decided through the states. If presidential elections were direct elections — i.e., decided simply by whichever candidate gets the most popular votes — then job approval ratings might be much more indicative of who wins.
But presidential elections are decided through the states by way of the Electoral College. Candidates get electoral votes by winning the most votes in a given state. So, the job approval rating may reflect the popular vote, but the way those votes shake out might lead to the president getting less than enough Electoral College votes. For example, 48% of the poll’s respondents may approve of the president’s performance and 48% of the popular vote goes to him in his reelection bid, but many of those popular votes might come from states that the challenger wins instead. Usually, the Electoral College and the popular vote both go to the winner anyway, but as was seen in 2016, for example, that isn’t always the case.
Here are some other nuances to consider:
- It’s also important to gauge not only the overall population but segments of the population, such as by race or gender. A poll with 2,000 total respondents is a good-sized sample, but even if the number of Black respondents is representative (i.e., the percentage of Black respondents in the poll is the same or a similar percentage as Black voters in the overall electorate), that sub-sample is still going to be fairly small. The vast majority of Black voters also vote Democratic, so depending on how they are sampled a poll might over- or underestimate a Democrat’s support.
- There is more than one job approval poll, and some of them are made for a partisan or ideological purpose. Campaigns sometimes do their own approval poll and will highlight it in press releases if it suits their message. Media outlets with politically-slanted views will also point to some approval polls if they fit their messages, too. The most reliable job approval rating comes from Gallup, a well-established and respectable polling organization (more on them later).
- People who respond to job approval polls aren’t necessarily going to vote. They may be willing to answer the survey, but they might not be compelled enough to actually go to a polling place to cast a ballot if they don’t feel particularly strongly about the president’s performance one way or another.
- Job approval and favorability are two different things. Respondents to a survey may disapprove of the president’s performance and view him unfavorably, but they may also view the main challenger unfavorably. We’ve regularly seen and heard lots of voters be unhappy with their own candidate, but the disgust, anger, etc. they feel toward the other party is sometimes enough for them to vote against the other candidate. How to word polls is a critical component to their efficacy.
A Brief Overview of the Job Approval Rating
All of what I just said isn’t to say that the job approval poll is worthless. It can be quite helpful, in fact . . . as long as you don’t rely on it too heavily to predict an election and you consider the caveats that come with it.
I briefly mentioned the polling organization Gallup earlier. They are going to be your most reliable source for the president’s job approval rating. They are highly reputable, respected, and neutral, and they have done a lot of public opinion analysis since 1935.
In the case of the job approval survey, Gallup has been administering that since Harry Truman’s presidency, refining it over time. They also have a nifty feature where you can view job approval over time for each president since Truman and even compare multiple presidents’ job approval at the same point in their tenures in the Oval Office. There are also segments of the poll sample you can break down job approval by for a given president, such as by party.
The Data: Presidential Approval Ratings and Reelection Bids
That same feature is where I gathered the following data. Note that since this article is about predicting reelection contests for president based on job approval, and this is a reelection year for Joe Biden, the data only accounts for presidents’ first terms. Below is a table showing each of these data points for the presidents in Gallup’s survey:
- The highest and lowest ratings, and roughly when they occurred (approximating by month and year instead of using the exact days)
- The first term rating average (conveniently provided by Gallup already)
- The last rating before the date of Election Day, which will be abbreviated “RBE” from here on (not accounting for early voting), for whether that president would win a second term, and when that rating was gathered by Gallup (which has been closer to election time for more recent presidencies)
- The year the president ran for reelection
- The percentage of the Electoral College and popular votes from that election that the incumbent won
- The difference between the last rating before the election and the Electoral College and popular vote percentages — a negative number means that the RBE was higher than the popular vote/Electoral College vote, while a positive number means the RBE was lower
- Whether that president was reelected
The rows are color-coded based on the president’s party — red for Republican and blue for Democrat.
In all but one case where the last rating before Election Day (RBE) was 45% or below, the incumbent president lost. The exception to this is Harry Truman, but the RBE was almost five months before the election, which is famous for the “Dewey Defeats Truman” blunder by the polls and media. That debacle led George Gallup to make serious refinements to his polling methodology. (Note that the poll that Gallup overhauled was not the job approval rating that we are talking about here, and there were polls besides his that predicted Thomas Dewey would win.)
And if you view the first Gallup tracker’s job approval rating that was done after the election — for January 7–12, 1949 — it came out at 69 percent. If the RBE had been done much closer to Election Day instead of June, the rating would have been a lot closer to this. Instead, there was a whopping 24 percentage point difference between the June 1948 and January 1949 polls.
Nevertheless, a 45% or lower RBE has usually indicated impending defeat for the incumbent president, including Trump. Each president that has had a higher RBE than that has been reelected, the lowest being George W. Bush at 48 percent.
Now, let’s compare the RBE to the popular vote. Early on, there was a considerable difference between the two, though that is partly because the RBE was well before Election Day. Job approval ratings can substantially change over the course of four or five months. For Truman, Eisenhower, and Johnson, the gap between the RBE and popular vote was 9.5 percentage points or higher in one direction or the other.
From Nixon to Trump, however, that gap has been much smaller — between 0.8 and 4.8 points. Whether a difference such as 3.0 points (Ford) or -4.8 points (Clinton) is close is perhaps a bit subjective, but job approval ratings have been more reflective of the popular vote for more recent presidencies, in part because polling methods have been much improved since their early days. What is also interesting is that in every case where that difference was negative — that is, the RBE was higher than the popular vote — the incumbent still won the election.
The difference between the Electoral Vote and the RBE is included in this chart just to demonstrate how starkly different it can be from a job approval rating, but the Electoral College is ultimately what decides presidential elections. An interesting data point here, though, is that in the three instances where that difference was negative but close — that is, the RBE was only a few percentage points lower than the Electoral Vote percentage — the incumbent lost (Ford at -0.4 points, H.W. Bush at -2.8 points, and Trump at -1.9 points).
How Joe Biden’s Approval Rating May Affect His Reelection Chances
At present, the current data mostly suggests that Biden may be in trouble. If an RBE of 48 percent or more is what we consider to be necessary to get reelected (which is what every president after Truman who was reelected had when they won a second term), then Biden has an uphill climb. The most recent job approval poll that Gallup has administered for him was for January 2–22, 2024, and it came out to 41 percent.
Now, let’s compare that to the two presidents who served full first terms but lost reelection at similar times in their tenures — Republican Donald Trump and Democrat Jimmy Carter.
At about the same point in their presidencies, Trump (46%) and Carter (58%) both had higher job approval ratings than Biden has right now. And as seen in the chart below, Trump and Carter’s approval ratings went up-and-down a couple times before dropping lower toward election time. Trump’s RBE was 45% and Carter’s cratered to 37 percent.
But there are still critical unknowns at this point for Biden. Eight or so months before Election Day is a lot of time for Biden’s approval rating to either jump or dip further. Current issues — the economy, healthcare, the Israel-Hamas War, etc.—and issues yet-to-come will have an impact that we aren’t entirely certain of yet.
There’s also the issue of who the Republican challenger to Joe Biden will be. As things stand, Trump is likely to win the Republican nomination again in a bid to win a second, nonconsecutive term in office. The only other time we’ve had a rematch between two candidates who both served as president was when Grover Cleveland defeated Benjamin Harrison in 1892 after losing in 1888 (though Cleveland won the popular vote both times). That was well before the “modern” era of polling and Gallup’s job approval survey. And the favorability ratings for both Biden and Trump are very close at the moment.
But what if the efforts to keep Trump off the ballot are successful, and Nikki Haley wins the Republican nomination? Will Biden’s low approval rating be enough of a boon to her campaign?
Or what if those legal efforts falter and Trump does win the Republican Primary again? Biden already defeated Trump once back in 2020 by about 7,000,000 popular votes, and Trump’s approval rating was also relatively low when he lost to Biden. Some voters may refer back to their disapproval of Trump’s performance, even if they don’t approve of Biden’s performance either, and will “suck it up” and vote for Biden anyway. Or they could pick a third choice such as Robert F. Kennedy, Jr.
On the other hand, many voters don’t necessarily think about what happened in a previous presidency. Rather, they think, for lack of a better phrase, “What have you done for me lately?” Since Biden’s performance is more recent than Trump’s, such voters who disapproved of both of their jobs may vote accordingly by “sucking it up” and picking Trump. Or they could, again, pick a third choice.
All of these questions, speculations, and data points demonstrate this: Job approval ratings are a valuable supplement for election forecasts. But take heed not to lean too heavily on them. There is never a “silver bullet” that is a perfect predictor of how an election will shake out.
