What History Tells Us About Midterm Elections

May 2018 -

Presidential elections are the red letter days on the political calendar, for the obvious reason that the U.S. presidency is the world’s most powerful office. Yet that power can be supported or undercut by what happens two years later, in national midterm elections. Voters weigh in on all 435 House seats and elect at least one-third of 100 Senators, and the typical result—losses in the House for the president’s party—can have a major impact on policy and politics leading up to the next presidential election.

Now, as Republicans and Democrats gear up for a showdown in November, every poll and pundit seems to offer a different take on what the result will be. But to judge by the long history of this nation’s midterm elections, it would be a mistake to take any prediction as absolute truth, says Dr. Larry J. Sabato, the founder and director of the University of Virginia Center for Politics. Surprises are the rule, not the exception, Sabato says, as everything from international conflicts to economic turbulence can affect which party does a better job in delivering motivated voters to the polls. OUTLOOK spoke with Sabato about what history has to say about midterm elections, the balance of political power and how the second half of President Trump’s first term may play out.

OUTLOOK: Given what you’ve studied about past midterm elections, what factors are you watching most closely leading up to this fall?

Larry Sabato: We have to start with presidential popularity, because we have never had a first term president this unpopular. President Trump’s approval ratings are pretty much stuck in the low 40s, and to the best of our ability to gauge it, Trump has never been over 50 percent on a single day of his presidency, not even on his inaugural weekend, when he was close but didn’t get there. This is remarkable. Looking at polls going back to the mid-1930s, that has never happened before. And, of course, Democrats have been very angry since the election. Starting on his inaugural weekend, you had massive protests. So, you have to start there and say Democrats are certainly favored to gain seats in the House, at least, and probably governorships and state legislatures. We don’t know how high or far that blue wave is going to reach.

OUTLOOK: What about in the Senate?

Sabato: The current balance in the Senate is 51 Republicans and 49 Democrats. Some people might look at that and assume that Democrats have an excellent chance to take back the Senate. But, in fact, it would take some kind of tsunami for Democrats to get the seats they need. It comes down to the fact that the Senate is a completely different animal from the House. All 435 house seats are up for election every two years, so in a sense the vote for the House is truly a national referendum. By contrast, only a third of Senate seats are elected every two years, so a lot depends on which seats happen to be coming open during which election. This year, you’ve got 26 Democratic seats on the ballot and just nine Republican seats. Ten of those Democratic seats are in states that President Trump carried in 2016; some, such as North Dakota and West Virginia, he won by large majorities. Keep in mind that because every state, whether Wyoming or California (which has 70 times the population of Wyoming) gets two senators, small states have disproportionate representation in the Senate—and small states these days tend to be Republican. I’m not saying it’s impossible. Lots of election results stun you, but it would take a stunning result.

OUTLOOK: What impact is the improved economy likely to have?

Sabato: Republicans have been stressing the tax bill they passed at the end of 2017. Over time, that may well help them to a certain degree. And a great economy should help Republicans. How much it helps depends on whether people really feel it and whether they believe that the good times will last. Do they have confidence in the future and believe their children will have the opportunity to do even better than they have? Yes, the economy has strengthened in many ways, but these are “squishy” conditions. A lot depends on what’s happening at the moment, what’s in the headlines, whether someone’s plant is laying off people or hiring people. Again, it’s too early to know, but I would say the economy and the tax bill are the two best things Republicans can use. Of course, you could argue forever about how much presidents actually have to do with the economy—they always claim too much credit for good economies and take too much blame for bad ones. Politically it matters because people think a president sits in the Oval Office pushing buttons and pulling levers and running a 20 trillion dollar economy.

OUTLOOK: If the House of Representatives or the Senate goes to a Democratic majority, what will that mean from a policy and governing standpoint? What would a return to divided government mean for the U.S.?

Sabato: If the House goes Democratic, it is effectively a veto on anything President Trump proposes in legislation— assuming the proposal is ideological and associated with the GOP. More technical bills as well as consensus bills would not be affected. However, if the Senate stays Republican, then Trump could still get many of his nominees for public posts as well as court picks confirmed. My guess is that, if forced to choose one house to control, the White House would pick the Senate.

OUTLOOK: And if the House and Senate both stay Republican?

Sabato: The margin will matter. Most GOP leaders concede that their current margin of House control will be cut in half at best, and that would mean less margin for error. The House GOP is divided into factions that often disagree on fundamentals. Now it’s perfectly possible, because of the seats on the ballot, that the Republicans could net additional Senate seats. That would allow for a few defections when a Supreme Court seat opens up, just to cite one example.

OUTLOOK: How do national midterm election trends play out at the state level—for example, with governorships?

Sabato: Democrats currently have just 16 of the 50 governorships, and before they won New Jersey last year, they had 15—their lowest total since the early 1920s. Over time, history tells us we should regress toward the mean of about equal governorships on both sides. In the governors’ races you have 26 Republican-held positions being contested, compared with nine Democratic governorships. So, overall, Democrats have a major opportunity to return some balance. I could see Democrats picking up only four or five, or as many as eight, governorships in November. There are still too many unknowns to make predictions, because we don’t know for sure who the nominees will be. But I can tell you that the map at least gives Democrats an opportunity.

The divisions of this country do seem to be becoming more extreme and that should worry everybody. For precedents I would point to the late 1960s and early 1970s.
OUTLOOK: What about state legislatures?

Sabato: If the Democrats do carry a substantial number of governorships you’re going to have a change in state legislatures as well. The Obama presidency was a disaster for the Democratic Party at the state legislative level. There are 7,400 state legislative seats nationwide. When President Obama took office, Democrats controlled 4,100. Today they have 3,100 state legislators. The loss of 1,000 seats is a record among presidents of the post-World War II era. There’s no way the Democrats will make up all of that ground this year, but they can start. They’ll probably add 300, 400, 500 state legislators. It’s too early to know for sure, but there will definitely be a gain for Democrats in state legislatures and that will matter in 2021, when redistricting is done.

OUTLOOK: This year’s midterms will take place in a sharply divided political climate. Are there midterm elections from history that occurred during similarly polarized periods?

Sabato: The divisions of this country do seem to be becoming more extreme and that should worry everybody. For precedents I would point to the late 1960s and early 1970s. I’m 65 and was nearly an adult at that time, and I remember it very well. Between Vietnam and Watergate and the very controversial presidencies of Democrat Lyndon Johnson and Republican Richard Nixon, the American system was pushed almost to a breaking point. The 1970 midterm election during Nixon’s first term stands out as particularly contentious. Elected in 1968, he was the first president in a long time to come into office not having won either the House or the Senate. Both had large Democratic majorities, and from the beginning, he had very little support in Congress. He set his sights on trying to turn at least one, the House or the Senate, to the Republicans. Vice President Spiro Agnew went after the “nattering nabobs of negativism” in the news media, and criticized the intellectual elitism of the networks. Some liberal Democrats lost, especially in the Senate but also in the House. In the end, though, Republicans still had a net loss of 12 House seats.

OUTLOOK: Historically speaking, what tends to happen during midterm elections?

Sabato: I avoid numerical averages because they don’t mean much. Every year and every election is so different. That said, it’s true that midterm elections are usually bad news for the incumbent White House party. Governing requires making choices, and choices about important matters make enemies. It’s the old truism: “For every decision, you make one friend and nine enemies.” Over the first two years you’re going to have people who are disappointed with the new administration, unhappy about decisions that were made. These voters look forward to going to the polls to register their unhappiness. Meanwhile, the president’s supporters tend to be not as energized. After all, their candidate was elected, they got what they wanted. It’s not that loads of people change sides. But turnout drops and the out-of-power party shows up in greater numbers.

OUTLOOK: What other factors tend to affect midterms?

Sabato: You look first at the big picture items, such as presidential popularity or unpopularity, and what the state of the economy is. You have to be concerned about scandal, what the international situation looks like and whether the U.S. is involved in an unpopular foreign war. All of these are related.

OUTLOOK: To what extent do national crises influence midterm elections?

Sabato: In the 1962 midterms, during John F. Kennedy’s Democratic presidency, Congress was heavily Democratic and by historical norms the party might have been expected to lose quite a few seats. But in mid-October, the Cuban Missile Crisis erupted. That standoff with Cuba and the Soviet Union was resolved 10 days or so before the election and the resolution was very favorable to the United States. So Democrats were boosted and there was virtually no change in Congress. The same phenomenon held during the midterm election of 2002, during the first term of George W. Bush. It was a rare instance when the president’s party actually gained seats in the House. That was, of course, just over a year after the terrorist attacks of September 11. A crisis of that magnitude can also lead to periods of bipartisanship. Democrats and Republicans in Congress actually achieved a lot, internationally and domestically, for a couple of years.

There are lots of ways to predict how groups will vote. What we have a tough time predicting is the relative turnouts of various groups, which matters a lot.
OUTLOOK: To what extent do social movements or demographic factors enter into the outcome of midterms?

Sabato: Party is absolutely the most important factor, but beyond party, racial and ethnic and gender lines are very important, and probably age and even education. So, there are lots of ways to predict how groups will vote. What we have a tough time predicting is the relative turnouts of various groups, which matters a lot.

OUTLOOK: Demographic changes—waves of immigration, say, or shifts in the racial mix of voters—tend to happen gradually. Does that limit their impact on midterm results?

Sabato: Over time, demographic changes make a big difference. For example, in the second half of the 20th century, whites often constituted about 85 percent of the electorate. Now, that percentage is down to 70 percent, and by the middle of this century, it may be about 55 percent. Because whites vote about 60 percent Republican, and all minorities together vote about 80 percent Democratic, it’s likely that Democrats will benefit— though there’s also the chance that a decade or two from now, the GOP may be able to attract more minorities, which would change this calculus. But such shifts don’t often have an immediate impact, and may not play out as expected. The 26th Amendment became fully effective in 1971, giving the vote to 18- to 20-year-olds. Predictions at the time were that the youth vote would benefit Democrats substantially. But in the 1972 presidential election, the new youth voters split about 50-50 between Richard Nixon and George McGovern. Moreover, demographic shifts are much more likely to show up in presidential years, because newly activated or registered voters usually come out to vote for president but tend not to vote in midterms.

History shows that a loss, even a substantial one, often helps the sitting president”
OUTLOOK: Are there exceptions to that rule—times when demographic factors have affected the outcome of a midterm election?

Sabato: The last time a midterm election was clearly swung by demographic changes was in 1934, during Franklin Roosevelt’s first term. The New Deal energized union workers, first-generation immigrants and many others who had almost given up hope because of the Great Depression. The Democrats actually gained seats in the 1934 midterm in part because of this—one of only three times in the past 100 years in which the president’s party gained House seats in a midterm.

OUTLOOK: We’ve talked about how the president affects midterm elections. But what effect do midterms have on presidents and their reelection prospects?

Sabato: Paradoxically, history shows that a loss, even a substantial one, often helps the sitting president. Bill Clinton’s first-term midterm election, in 1994, was a disaster. He lost 52 Democratic seats and the House went Republican for the first time since the early 1950s. Almost everybody wrote him off. The pundits were saying, Clinton is going to be a one-termer for sure. The Republicans swept into Congress and, of course, promptly began to over-read their mandate to bring about change. Clinton was able to play himself off against the “Republican revolution,” and that became his springboard to reelection. In 2010, thanks to Obamacare and other things, President Obama lost even more than Clinton had—61 House seats—and was also written off for reelection. And, of course, Obama successfully countered Republicans in Congress and wound up being reelected. So, if you’re president, sometimes you can win by losing. By the same token, though, presidents can also lose by winning.

OUTLOOK: When has that happened?

Sabato: After his election in 1988, President George H.W. Bush was very popular. Then, when we won the Gulf War in very quick fashion, his approval shot up still further, to about 90 percent. Thanks in large part to that goodwill, the Republican Party bucked tradition and lost only a handful of House seats in the 1990 midterm. Bush emerged looking so invulnerable for reelection in 1992 that some Democrats, believe it or not, briefly toyed with nominating Bush as a Democrat. The idea was that they could then nominate a running mate other than Vice President Dan Quayle, because many voters had reservations about Quayle. Bush would be guaranteed a second term but there was an outside chance they could defeat Quayle. The fact that Democrats even considered such a strategy tells you what a foregone conclusion a second Bush term appeared to be. Then, we went into a recession, Bush’s “read my lips: no new taxes” pledge came back to haunt him, Clinton emerged as the Democratic nominee and Bush lost the race. In the end, that strong showing in the 1990 midterm meant very little.

OUTLOOK: Has a midterm result ever been such that a president was discouraged from seeking reelection?

Sabato: That’s only happened a couple of times, and usually there have been other major events going on that contributed to the decision. Democrat Lyndon Johnson stood down in 1968 after a substantial loss of Democratic seats in the 1966 midterms. But it wasn’t just that. By 1968 he realized that he had failed to win in Vietnam. Opposition to the war was so intense that Johnson at some level knew he would probably lose if he ran. And, even if he won, it might tear the country apart. Democrat Harry Truman could have run for another term in 1952, because the 22nd amendment, limiting the president to two terms, didn’t apply to him. But I think the combination of disappointing midterm results in 1950, as well as his failures with the Korean War, caused him to step aside. Plus, deep down, he probably knew he would have lost to the Republican who became the next president, Dwight Eisenhower.

Also in this issue 

  • Interest Rate and Economic Indicators
  • CoBank Reports First Quarter Financial Results

Read Outlook