A walk through presidential first-term midterms election history of the last 100 years
[NOTE: Whenever I have a math-heavy post like this, I always get nervous that I’ve made an arithmetical error.]
I see it again and again: the assertion that the 2022 midterm results were almost unheard-of, because a president’s first midterm always involves the opposite party gaining a large number of seats in the House and a significant number in the Senate. Just to take one example, we have this:
Historically, every president in history facing his first midterm experiences a tough day with automatically 20 to 30 House seats lost…and 4 or more Senate seats lost…but this terribly unpopular President Biden is brain dead with severe dementia, and can’t put 3 coherent sentences together…Yet Biden defied history?
When I read something like that, I think: is it true? Did every president in history facing his first midterm experience results like that, or pretty much like that? Or nearly every president?
So I decided to actually take a look – not at every single president in history, but at least at presidents for the last 100 years, the group one would imagine would be most relevant to today. Then maybe, if it appeared true for all of them, I’d go back further in time and see what I found.
Let’s just say I didn’t have to go back further in time, because it wasn’t true during the last 100 years.
We’re not talking about midterms in a president’s second term; just the first. Here’s the record for the Senate midterms in presidents’ first terms:
(1) Calvin Coolidge (R) – first half of first term 1923-1925, Senate was R53 D43 and midterm results were R54 D42. Doesn’t follow the rule.
(2) Herbet Hoover (R) – first half of first term 1929-1931, Senate was R56 D39 and midterm results were R48 D47. This follows the rule.
(3) Franklin Roosevelt (D) – first half of first term 1933-1935, Senate was R36 D59 and midterms were R25 D69. This goes dramatically against the rule.
(4) Harry Truman (D) – first half of first term can be calculated either from the time FDR died (which was quite early in the term) or the time of Truman’s first election on his own, when he had already been president for many years. I’ll do both. First half of first non-elected term 1945-1947, Senate was R38 D57 and midterm results were R51 D45. This follows the rule.
First half of first elected term 1949-1951, Senate was R42 D54, and midterm results were R47 D49. This also follows the rule.
(5) Dwight Eisenhower (R) – first half of first term 1955-1957, Senate was R47 D48 (plus one other who caucused with the Democrats, making it 47/49), and midterm results were R47 D49. No change. Doesn’t follow the rule.
(6) John Kennedy (D) – first half of first term 1961-1963, Senate was R36 D64, and midterm results were R34 D66. This does not follow the rule.
(7) Lyndon Johnson (D) – I’ll skip his first term, because it wasn’t long enough for a midterm election. His second term and first elected term was 1965-1967, Senate was R32 D68, and midterm results were R43 D57. This follows the rule; however, it has an asterisk because of the special circumstances that involved LBJ having actually been president for 3 years at the time of those midterms rather than two.
(8) Richard Nixon (R) – first half of first term 1969-1971, Senate was R43 D57, and midterm results were R44 D54 but actually 45/55 because of members of other parties and where they caucused. This election did not follow the rule.
(9) I’m going to skip Gerald Ford because his term was too short, and he was unelected as well.
(10) Jimmy Carter (D) – first half of first term 1977-1979, Senate was R38 D61 (but really 38/62 because of caucusing), and midterm results were R41 D58 (but really 41/59 because of caucusing). This follows the rule.
(11) Ronald Reagan (R) – first half of first term 1981-1983, Senate was R53 D46 (but really 53/47 because of caucusing), and midterm results were R55 D45. This did not follow the rule.
(12) George H. W. Bush (R) – first half of first term 1989-1991, Senate was R45 D55, and midterm results were R44 D56. This follows the general rule, but in a very small way with the loss of only one seat.
(13) Bill Clinton (D) – first half of first term 1993-1995, Senate was R43 D57 (changing later to R44 D56 prior to the midterms), midterms were R52 D48. This follows the rule.
(14) George W. Bush (R) – first half of first term 2001-2003, Senate was R50 D50 (there was some small shifting back and forth afterward, but that was the original election result), midterm results were R51 D48 (but really 51/49 because of caucusing). This did not follow the rule, and in fact is similar to what happened in 2022 – a gain of one seat for the party with the presidency.
(15) Barack Obama (D) – first half of first term 2009-2011, Senate was R41 D57 (but really 41/59 because of caucusing), and midterm results were R47 D51 (but really 47/53 because of caucusing). This followed the rule.
(16) Donald Trump (R) – first half of first term 2017-2019, Senate was R51 D47 (but really 51/49 because of caucusing), and midterm results were R53 D45 (but really 53/47 because of caucusing). This did not follow the rule.
I was going to do the same thing with the House, but doing it for the Senate was already tiring and nitpicky and I think it made my point sufficiently anyway.
Here’s the tally: in the last 100 years, it’s exactly even for the Senate. Half of the first-term midterm results followed the supposed rule and half did not. So it’s not even close to “always.”
Just taking a really quick look at the House – the legislative branch that is more sensitive to political winds that blow because every member is up for re-election every two years, rather than a third every two years as in the Senate – reveals that the following presidencies did not follow the rule in their first term midterms (that is, the president’s party did not lose House seats in the first midterms): Coolidge, FDR, Eisenhower, and George W. Bush. What’s more, two other presidents only lost a small number of House seats in the first midterms, much like what happened to Biden and the Democrats this year: JFK only lost 6 seats, and George H.W Bush only lost 8 seats. So even in the House, results like those in 2022 have not been especially unusual.
What’s more, this year’s Senate breakdown was a difficult one for the Republicans, and yet they only lost one seat (perhaps two in the end, but we don’t yet know). By “difficult” I mean that fourteen seats held by Democrats and 20 held by Republicans were up for grabs this time. Not only that, but more of the seats held by Republicans were considered to be in battleground states (see the map here).
What does it all mean? Obviously, the common assertion that the 2022 midterm results were highly unusual is incorrect. However, I think we see the result this year as highly unusual and counter-intuitive for two reasons. The first is the predicted “red tsunami” set the right up for a huge win, and it wasn’t forthcoming. The second is that the awfulness of the present Congress and president, as well as some of the Democrat candidates, predisposed us to think that of course there would be a huge Republican victory. There had to be.
But there wasn’t. And whether you think the results were fraudulent, or the result of a legal exploitation of the relaxed voting rules, or an actual fair and square election, the results were nowhere near as unusual as so many people have been saying.
Make of it what you will.
Ten days post election & they still can’t call Kiley in CA 03 or Valadao in CA 22. 61% and 75% of the vote counted.
So, an oft repeated lie, or common journalistic ignorance? Maybe a combination of both.
Costanza rule applies.
Posts like this are why I come here. The media tells you all kinds of things that are not only false, but easily shown to be false with a few hours’ work.
We used to hear about the taller candidate winning the Presidential election, but it had no real predictive value, it was true until it wasn’t.
What Frederick said.
Neo; you mean the conventional wisdom is wrong?
Obviously you are a denier !!
The most amazing part is this supposed ‘historical fact’ didn’t even happen the last mid term in 2018.
So many of the journalists reporting on this just skipped right over that.
Political moods are way more complicated than this easy ‘fact’ implies.
Any trend that exists is very weak and not very consistent.
Thank you for making a strong case of likely fraud. Popular presidents who were perceived as doing a good job didn’t get punished. Presidents who were unpopular and doing horrible jobs got hurt. That’s what your breakdown very clearly shows.
Which applies to Biden?
The results are unprecedented on an extraordinary level.
Responding to stan (5:19 pm) . . .
But we need to bear in mind that the visceral venom hurled at Trump was, I think, unprecedented, at least in part due to the sophisticated state of the omnipresent media, who were very, very happy to deliver and reinforce the venom (over and over . . .).
The 2022 vote still featured the odiferous impetus of voting against the hated Trump.
(No, I am in no way suggesting that the above was the *sole* explanation for the 2022 midterm results.)
And many thanks, neo, for that nuts-‘n’-bolts research. A mini tour de force there. Wow.
stan:
I don’t see it that way. What I see is a trend from domination of Congress by one party or another (but mostly domination by Democrats, and often very strong domination by Democrats) to a situation where in recent years the split between the 2 parties is much more even and also relatively stable. After that, you don’t see strong party majorities from either party.
In Coolidge and Hoover’s time it was consistent Republican domination, with ups and downs in the manner you describe tracking in sync with the presidents’ popularity. Then beginning with FDR, there was a lengthy period of tremendous and consistent Democrat domination (tracking in sync with presidents’ popularity). This trend to a strong Democrat majority was only interrupted one short time during Truman’s presidency (with a Republican majority that was not only brief but small) and a somewhat even split during the Eisenhower years. Then with Kennedy’s election it goes back to a tremendous Democrat majority again (with the usual ups and downs tracking presidential popularity) and this Democrat skew lasts till the Reagan years.
With George W. Bush, however, we enter what I consider a new pattern, which is the fairly even split. This pattern has persisted with some ups and downs. And even those smaller ups and downs don’t track presidential popularity quite as well as they did in the past. For example, Trump gained seats during his first-term midterm even though his popularity as measured by polls was low at the time. It is especially relevant to the discussion of 2022 because he was the most recent president prior to Biden.
That’s for the Senate. The House is different because all its members are elected every two years, among other reasons. However, it generally follows a somewhat similar pattern over the last 100 years – that is, strong majorities (mostly Democrat ones) yielding to closer splits on average in more recent years. Joe Biden’s tenure has featured particularly close splits, with an eight-seat Democrat advantage giving way this time to what will probably be something even smaller for Republicans.
I don’t think these trends have as much to do with presidential coattails as they used to. I think they have taken on a life of their own and represent a very fundamental and more or less even split.
Another trend I’ve noticed is that presidents no longer enjoy high approval ratings for as much of their terms. More sustained times of high approval used to be more common, and although there certainly were also times in which a president had low approval, it was usually not for a large portion of their tenure and often came towards the end of it (Truman, LBJ, Nixon). Now, both Trump and Biden have had low approval most of the time they were president, and even Obama – who started out very popular and finished up somewhat popular – had popularity ratings in the 40s most of the time. So that is another change.
Therefore I think it’s a mistake to think that the figures in this post prove fraud. Nor do they disprove it. Because of the change in recent patterns, I don’t think they tell us all that much about fraud.
Fraud vs. no fraud was not the thrust of my post, of course. The point of the post was to disprove the commonly-heard statement about first-term midterms.
The chart here might help:
https://www.thoughtco.com/historical-midterm-election-results-4087704
This is good analysis. I think we’ve had big midterm rebukes recently because we’ve had highly unusual and even radical presidencies. Take away Obama’s two rebukes in 2010 and 2014 and there’s barely a pattern.
But that leaves us with Biden this year. If you’re judging in terms of radicalism, he richly deserved a 2010 or 2014 level defeat.
I think one of the reasons that didn’t happen is people judged “stop the steal” to be more radical that Biden and the Democrats. I think a lot of people on the right who are concerned (rightly) about eletion integrity fail to recognize just how radical and destructive their rhetoric can sound. For example, Doug Mastriano promised that, if elected governor of PA, he would not certify the 2024 presidential election for PA unless the Republican won. FWIW, I doubt that he would have carried through on that promise if the Demcorat’s nominee won a clear victory, but that’s what the man said.
What is more radical? Illegally preventing renters from getting kicked out of their homes, illegally “forgiving” hundreds of billions of dollars of student loans, or promising to ignore the results of future elections if they don’t go your way.
That’s not to apologize for Biden or excuse him in any way, but just put yourself in the shoes of a non-partisan voter and ask which of those things is more radical and more of a threat?
The Democrats did an amazing job of tilting the voting playing field through efforts fair and foul.
Census “reapportionment” mistake was impressive skullduggery. Oh, I’m “sure” it was an honest mistake…
And interestingly no challenges were done to it.
And the GOP dropped the ball in many areas. Deliberate incompetence? The Gop needs to up its game.
One other factor is that this was a bad year for Republican Senate incumbents. Republicans had to defend a lot more seats this year and 2024 is a much friendlier map for the Republicans.
And to further connect that point to the midterms the 2024 Senate races are the same ones that happened in Trump’s midterm year of 2018 when he defied the ‘conventional wisdom’.
What seats are up for election also has a big impact on all of this which further makes this a shaky ‘trend’.
Thanks for doing all the homework. It was not so much the history as the expectation that a president doing as badly as Biden is would have to lose a lot of seats. People filled in the history afterwards based on watershed elections they remember (1994, 2010). As I remember it though, it’s often in the sixth year, rather than the second year, of a party’s tenure in the White House that voter boredom or impatience or disgust sinks in (1938, 1946 (six years after FDR’s 1940 win), 1958, 1966, 1974, 1986, 2006). Call it the Sixth Year Itch. Clinton 1998 was something of an exception. His party had already had a big loss four years before, the country wasn’t in bad shape, and voters were as tired of impeachment as they were of him.
What made this election different is that the Democrats have such support and control in their states that it was impossible to dislodge even incompetent incumbents even when an incompetent Democratic administration is making a mess of the country. Abortion had a lot to do with it. So did Trump — more the mere existence of Trump, rather than anything he said or did this year. But I have to wonder if this is the future for the Blue States. Have they all become Massachusetts?
Democrats were probably also able to hold on to three seats they could have lost due to census overcounting in NY, RI, and MN. Add to that that some states didn’t have a reapportionment plan approved until it was too late to be applied to this election.
some compared to 1970, the dems lost some seats, but some of the ones like lowell weicker, good grief he was terrible, back then buckleys brother actually won in new york,
Abraxas:
I think you are quite correct about how people filter their memories. Thing is, some of these people are pundits/authors who are supposed to be experts in history.
I think that many incumbents have long been difficult to dislodge no matter what they do.
Too many bad candidates for Senate and Trump and his giant ego and enormous mouth frankly should step aside.
I noted this earlier, I believe, but it’s a good commentary on why the “Red Wave” did not happen, yet the Dems do need to very much fear for 2024…
And this year may not have been what was hoped for but was a very bad sign for Dems going forward.
Mark Levin:
https://www.youtube.com/watch?v=5fhkvbocrmk
Take from it what you will.
I think Victor Davis Hanson has correctly identified seven reasons why the midterms were a disappointment. Here are some of them:
“…Five, this time the silent and undercounted voters were not disillusioned MAGA supporters who hung up on pollsters’ calls.
Instead, pollsters missed the 70% of those under 30, along with single women, who voted straight Democratic tickets.
Mannered Republicans may have scoffed at how Biden and the Left demagogued the abortion issue, or slandered Republicans as semi-fascists and un-American insurrectionists. They shrugged at Biden’s hokey efforts at buying off young voters with amnesties for marijuana convictions and student loans or offering slightly cheaper gas by draining the Strategic Petroleum Reserve.
But all those low-minded strategies resulted in high left-wing enthusiasm and turnout.
Six, usually reliable conservative pollsters forecast a huge Republican victory. Apparently, they oversampled conservative voters, reasoning that left-leaning pollsters usually undersampled them.
They were not just wrong, but way off. And the ensuing hubris of certain victory led to nemesis as Republicans let up the last few weeks. Thousands of conservative voters may have passed at the chance to go to the polls deeming their votes superfluous.
Seven, the Left smeared conservatives as democracy destroyers and violent insurrectionists. So, when the Republicans offered nonstop negative appraisals of Biden’s failed policies without commensurate alternative positive agendas, they unknowingly fed into the Democrats’ false narrative of cranky nihilists.
Could not Republicans have offered an upbeat and coherent contract with America that offered uplifting, concrete solutions to each of Biden’s messes?…”