With early voting now underway in Virginia and other states, the United States is once again embarked on an exercise that is unusual among other democracies: Midterm elections.
These have been an American ritual since 1790 when George Washington was two years into his presidency and the nation did not yet have formal political parties. (Some call these “the good old days.”) There were, though, clearly two factions in Congress, identified generally as a pro-administration faction and an anti-administration faction, the former eventually becoming the Federalists and the latter the Democratic-Republicans. The 1790 elections didn’t change much – the pro-administration faction gained a few seats to retain a majority in both houses but not enough to change the nation’s overall political dynamics.
Subsequent presidents, particularly in more recent times, haven’t been quite so lucky as Washington was. In 1994, the midterms of Bill Clinton’s first term, Republicans won back both the Senate and the House, and held them for the rest of his presidency. In 2006, the midterms of George W. Bush’s second term, Democrats won back the House of Representatives. In 2010, the midterms of Barack Obama’s first term, Republicans returned the favor and regained the House of Representatives. In 2018, Donald Trump’s midterms, Democrats again won back the House of Representatives. In 2020, Joe Biden’s midterms, well, we’ll see, won’t we?
What we’ve seen is that midterms often – not always, but often – produce a voter backlash against the party in power. Since voters can’t change the party in the White House at that point, they do the next best thing and change the party in control of one or both chambers in Congress – the House being the most volatile since all 435 members are up for re-election at once. Sometimes this presages a subsequent change in power in the next election – the 2006 Democratic takeover of the House foreshadowed a Democratic victory in the 2008 presidential election. Likewise, the Democrats’ 2018 House majority set the stage for a Democratic victory in the 2020 presidential election. On the other hand, the 1994 Republican takeover in the midterms did not lead to Clinton being defeated. We won’t know what the 2022 midterms mean until 2024, although plenty of analysts on the morning after – including perhaps yours truly – will try to tell you what they mean anyway.
In any case, the United States is unusual among major democracies in that we sometimes have a chief executive of one party and a Congress of another. In fact, most of the time we don’t have the presidency and the Congress controlled by the same party. Since 1980, we’ve had 21 elections. Only five times did those elections produce a clear trifceta for one party or another – seven times if you count tie Senates such as the one we do now. Voters say they don’t like gridlock in Washington, but that’s not what they say where it counts – in the voting booth. Most of the time they vote in favor of gridlock. And when presented with a Congress controlled by the same party as the president – which, in theory, ought to unlock that gridlock – voters have been quick to restore said gridlock. Our founders set up a system of checks and balances and voters seem pretty enthusiastic about that.
Historically speaking then, Democrats ought to expect bad news on election night, simply because their margin in the House is small and their margin in the Senate is non-existent, and history suggests they’ll lose seats. If they retain either chamber, that would be something of a surprise. That’s not making a prediction; that’s just looking at the past four decades of election returns.
The French have a cheeky term for this kind of split government. When they have a president of one party and a National Assembly of another, they call it “co-habitation.” Eventually, though, even the French tired of such arrangements. In 2000, the French approved a referendum to effectively eliminate midterm elections. Now, their president and their National Assembly are elected at almost the same time – almost being the key word here. In April of this year, the French held their presidential election. In June, they held their elections for the National Assembly. In theory, those two elections could produce different results, although that’s a pretty short time in which to have buyer’s remorse. This year’s elections saw Emmanuel Macron re-elected as president in April, although his party lost seats (and its majority) in June. Still, it retained a plurality in a legislature where 11 different parties are represented, so technically there is no “co-habitation.”
Most other democracies avoid that kind of awkwardness altogether. They don’t hold midterms because they’re parliamentary systems where the nation’s chief executive is the leader of whichever party holds a majority in parliament (or can at least put together a majority coalition in places with a bazillion parties). Thus, Prime Minister Liz Truss in Great Britain won’t have to face a midterm election; she’ll just face a regular election at some point. If her Conservative Party wins, she’ll stay in No. 10 Downing Street. If it doesn’t, then Labor Party leader Keir Starmer will move in.
We now engage in some philosophical thinking: Are midterm elections helpful to us? If you’re the party winning them, you certainly think so. Democrats in 2006 and 2018 were thrilled to be able to win back the House and be able to thwart at least some things that the Republican presidents of the day wanted to do. Likewise, for Republicans in 1994 and 2010 with Democratic presidents.
If we summon up the ghost of James Madison with a ouija board, he’d probably tell us that this is exactly what he had in mind. The general idea was that a House that’s up for election every two years would reflect the most current mood of the people while a Senate whose members serve for six years – and with staggered terms so only one-third are up for election at any one time – would proceed with “more coolness, with more system, and more wisdom.” (You can decide whether the “wisdom” part of that has worked out.) To complicate things further, our founders created a president with an entirely different term of service – four years. Our founders wanted gridlock. They feared both a government that was too energetic, but also one that was unable to do much of anything. So, all these years later, things are pretty much working according to plan.
Or are they?
There’s increasing attention on how much we’ve become polarized geographically into red states and blue states and how that leads to a situation where a minority of the country, living in small, rural states, can frustrate the will of the majority, which lives in larger, urban states. The framers intentionally created a Senate to represent all states equally, regardless of size, but they also did not foresee our current union. In 1790, the largest state (Virginia) was 12.6 times bigger than the smallest state (Delaware). By 2020, the largest state (California) is 66 times bigger than the smallest (Wyoming). It’s not healthy for a democracy to have a majority run roughshod over a majority, but, likewise, is it healthy for a democracy to have a system where a minority can so often block the will of the majority? We see this most clearly in the Electoral College, where, as we’ve twice now in recent times, it’s quite possible for the candidate with the most votes to still lose.
Some believe the solution is to abolish (or somehow reform) the Electoral College, but there’s another way that I’ve never heard anyone propose: Should we trash the whole separation of powers business and adopt a parliamentary system? If we did, we’d effectively have a system where the leader of whichever party controls the House becomes president – or prime minister.
That would change our political dynamics in lots of ways, some easy to describe, some no doubt unforeseen. For one thing, we wouldn’t have “outsiders” get elected president, of which Trump was the most obvious but by no means the only example. We’d have presidents (or prime ministers) who were the ultimate party insiders. Jimmy Carter, Ronald Reagan, George H.W. Bush, Bill Clinton, George W. Bush, Barack Obama, Donald Trump – none of those would have been president because none had been party leaders in Congress. Most of the names on that list hadn’t served in Congress at all. Joe Biden’s not on the list, because he had been a senior senator before becoming vice president, but otherwise you have to go back to Gerald Ford to find a president with a long legislative history. Still, they weren’t majority leaders, which is what a prime minister is (or, more technically, a Speaker of the House). We can’t say that with a parliamentary system we’d have Prime Minister Nancy Pelosi (although we might) because if we had a parliamentary system, so many of our political dynamics would be different — candidates who now run for the Senate would instead be in the House, trying to rise through the ranks there. Feel free to debate among yourselves whether we’d have been better off or worse off without those presidents listed above; just keep in mind it’s a package deal, all or nothing.
The same principle applies on the state level: If we had a parliamentary system, we wouldn’t have had George Allen, Jim Gilmore, Mark Warner, Tim Kaine, Terry McAuliffe or Glenn Youngkin as governor. We might eventually have had a Bob McDonnell or Ralph Northam – both were legislators of some stature before they ran. But neither were party leaders, either. In Virginia, we tend not to see senior legislators – such as a House or Senate Majority or Minority Leader – run for governor. Why should they? They feel they have more power where they sit – a different kind of power but a more long-lasting power. In Virginia, governors come and go every four years but legislators often remain for decades.
A parliamentary system offers some clarity that we don’t have now: A party with a majority really can enact its program, for better or for worse.
One thing a parliamentary system would not do is guarantee that the party with the most votes gets to govern. In 2019 and 2021, Canada held national elections. Both times Justin Trudeau’s Liberal Party won – won in the sense that it won the most seats in parliament (an outright majority in 2019, a plurality in 2021). Both times, though, the Conservative Party actually polled more votes. In 2019, the Conservatives polled 34.34% of the vote nationally to the Liberals’ 33.12%. In 2021, the Conservatives took 33.74% of the vote while the Liberals won 32.62%. How did the Conservatives win the most votes but still lose? It’s because Canada has some of the same political cartography that we do. The Conservatives ran up big margins in some rural, western provinces such as Alberta and Saskatchewan. Meanwhile, the Liberals won by narrower margins in more places, so they won more seats with fewer overall votes. Takeaway: It’s not just our Electoral College that distorts things. The only way to guarantee that the candidate with the most votes wins is to hold a direct election for president, as France does.
Don’t get distracted by that, though. That’s not what I’m here to explore, although it’s an interesting byproduct of our system. Instead, I’m curious about this: Why didn’t we set up a parliamentary system? After all, that’s what we were most familiar with in 1776. Of course, that’s also why the founders didn’t do it. They weren’t impressed with the British Parliament – they were even less impressed with King George III, though – and had no interest in emulating either. Instead, they were philosophers and intentionally created an entirely new model.
Still, it’s notable that all 50 states have copied the federal government with a chief executive here and a state legislature there. There’s nothing to stop any of those states from adopting a parliamentary system; the Constitution merely guarantees “a republican form of government.” It says nothing else about how state governments should be structured. If states really are the laboratories of democracy, then some aren’t showing much imagination in that laboratory. It would be curious to see what a state government would look like if it were a parliamentary system.
We don’t, though, so we have midterms. We have them this year nationally, and we have them next year in Virginia. In terms of number of offices on the ballot, next year is the big election cycle in Virginia. That’s when all 100 seats in the House of Delegates and all 40 seats in the state Senate will be on the ballot – no staggered Senate terms here. Politically, there’s always something at stake, but there’s even more at stake next year. If Republicans retain the House (currently 52-48 Republican) and win the Senate (currently 21-19 Democratic), then they will be able to pass any restrictions on abortion they want. (And lots of other things.) If Republicans pull off that sweep, then that’s a powerful talking point for Glenn Youngkin, particularly if he decides to run for president. If the parties split and things stay as they are now, well, things stay as they are now. Youngkin doesn’t get a national talking point, but he won’t be embarrassed, either. On the other hand, if Democrats win the House and retain the Senate, then Youngkin might well be embarrassed nationally – that certainly wouldn’t be seen as a vote of confidence in his program. He’d also be forced into two years of “co-habitation” that I feel certain he would not enjoy.
Here’s an irony: Youngkin hopes this year’s midterms go against the nation’s chief executive but hopes that next year’s Virginia midterms go in favor of our chief executive. Is that too much for him to hope for?