June 14 , 2019
Trump’s Reelection
Donald Trump became President in 2016 after winning only 46% of the popular vote. He lost to Hillary Clinton by 3 million votes, but due to a very favorable spread of his voters, he won narrow victories in Wisconsin, Michigan, Pennsylvania, North Carolina, and Florida. With the 2020 race already in full swing for Democratic candidates, it is not too early to start to think about what Trump’s chances are of winning reelection.
Traditionally, an incumbent President gets reelected. Since World War II, eight incumbent Presidents have been reelected, while three have lost. One of those was Gerald Ford, who was an accidental President due to Nixon’s resignation stemming from the Watergate scandal and imminent impeachment. Only Jimmy Carter and George HW Bush have been elected Presidents that lost. Several political scientists have constructed prediction models to show who will win an election. These models are heavily influenced by three factors. The state of the economy, the involvement of the country in an unpopular war, and whether a major scandal has involved the President. None of them rely on the President’s approval rating at the time of the election, but assume that the approval rating will reflect those three key factors. Based on several of these models, Trump should easily be reelected.
But, these models are flawed. Most importantly, the models are based on a very few data points. There have only been 18 elections since 1948, and 18 is a very small number of observations from which to draw a firm conclusion. Secondly, the models don’t take into account the dramatic change in the nature of the political parties and of partisanship in voting that has occurred in the last thirty years. Before 1980, there was real overlap in the ideology of the two parties. There were prominent liberal Republicans and there were racist conservative Democrats (mostly from the South). Voters also were smeared across the two parties. In 1960, the iconic Jackie Robinson was a proud Republican, even though many African-Americans had already begun to vote Democratic due to FDR and the New Deal. The result of this overlap was that voters could swing dramatically between the two parties. The Republican Dwight Eisenhower won 57% of the vote in 1956, while the Democrat LBJ won 61% in 1964. Nixon won 61% in 1972, while Carter won back a large share of those voter to win 50% in 1976. Reagan won big again in 1984 with 59%. After that Reagan landslide, the political parties underwent a very sharp and clear ideological sorting. Conservatives moved entirely to the GOP, while liberals moved to the Democrats. The fraction of voters who were truly in-between shrank. As a result, no party has been able to score a big win in over 30 years. The biggest popular vote shares have been George HW Bush in 1988 and Barack Obama in 2008, both with only 53% of the vote. What this means is that the number of voters who will swing based on the state of the economy or war or scandal has shrunk, and so the old models are not valid.
What model is valid? For Presidents running for reelection a direct measurement of their personal popularity is the most reliable indicator. This is reflected in polling of “approval rating”. If we look at the last seven Presidents running for reelection, it is obvious that their approval rating (I use the Gallup poll numbers for consistency in all cases) just before the election predicts the vote share they will win. In 1976, Ford approval was 45%, he won 48% of the vote. In 1980 Carter approval was 37% and he won 41%. In 1984 Reagan was 58% and won 59%. The elder Bush in 1992 was at 34% and won 37%. Clinton in 1996 was at 54% and won 49%. The younger Bush in 2004 was at 51% and won 51%. Obama in 2012 was at 51% and won 51%. The numbers are quite clear, it is very hard for a President to significantly outperform his approval rating, and has a difficult time getting voters that do not approve of him to vote for him.
Some observers claim that this does not apply to Trump because he still won despite being disliked by most voters in 2016. But that was a different metric. In that case, voters were not rating his job performance as President, but whether they viewed him as a person “favorably” or “unfavorably”. Trump’s favorable rating was abysmal at about 35%. However, Hillary Clinton was also abysmal at about 45%. In most elections, the public in general has positive favorable ratings about both candidates. In 2016, the opposite was the case. On election day, both candidates got the votes of those who liked them, but there were 20% of voters who disliked Trump and Clinton, and Trump won because the vast majority of those voters went for Trump.
Trump currently is in deep trouble. His approval ratings in the last two years have been stuck around 41%. In the latest Gallup poll he was at 40% approve and 55% disapprove. Morning Consult, another polling firm, also has approval ratings in May by state. Trump is in trouble in a host of states that he needs to win. His net approval (approval minus disapproval) is negative by 7 points in Arizona and Pennsylvania, by 8 points in Iowa, by 10 points in Michigan and by 13 points in Wisconsin. He is even negative by 2 points in Florida and North Carolina, and is only positive by 3 points in Texas, Georgia, and Indiana. The map looks very grim for him, and the latest polls nationally have him losing to Biden by over 10 points. He even trails Biden by 4 points in Texas, the second biggest state and one that has not voted for a Democrat since 1976.
Can Trump turn this around? Nothing is impossible but it would take a second miracle. He would need to somehow convince enraged liberals not to vote or to vote for a third party. He would need to win a number of very close states and not lose any. And he would need to develop a message that appeals to more than just his base. But Trump is probably incapable of doing that. To a large extent, he knows only one thing about politics, and he will go back to that over and over again. He won in 2016 on the issues of immigration, Muslims, trade, demonizing Hillary Clinton, and “America First”. Not to mention a little help from Russia. These themes are played out and will not work in 2020. He tried them in 2018 during the midterms, but the Democrats had their biggest midterm wins since Watergate. He has no specific policy agenda, and no good explanation for why he never built his border wall with Mexican money. Trump may be so toxic in 2020 that he will bury the Republicans in such a loss that the Democrats not only get the White House but also the Senate, giving them full control of the government and the chance to enact their full agenda.
