The Democratic math in the Senate is obviously bizarre, relying on a 50-50 split among senators for any votes that evade or survive a filibuster, and then depending on Vice President Harris to tip the scales. But it’s bizarre, too, in less obvious ways. Like that there are actually only 48 Democrats in the Senate plus two independents who caucus with the Democrats and give them the majority. And then, really, we need to shift two Democrats — Sens. Joe Manchin III (D-W.Va.) and Kyrsten Sinema (D-Ariz.) — into an interparty zone from which they fight to keep the party from moving too far from the center.
Manchin and Sinema are currently irritating everyone else in their party by doing precisely that (as they see it): submarining (as everyone else sees it) President Biden’s policy agenda by slowing the passage of spending bills as they push for nebulously defined alterations. Since every vote is needed in the Senate for anything to pass, it all comes down to what the two of them want, an enormous frustration to the rest of their party.
And yet, in another sense, the party is lucky they have two senators to come down to.
You are by now very familiar with the fact that the Senate is not perfectly representative of the will of the public. It was designed to provide equivalence to states, not state populations, and has always exhibited an imbalance of power between more- and less-populated states. In the moment, though, it means that a country that preferred a Democratic president by more than four percentage points last November has a Senate in which power is split. But, again, that’s better than it could be.
Just imagine that senators were apportioned simply on the basis of presidential voting. If that apportionment happened relative to the national tally, the Democrats would have a 52-to-48 senator advantage — and could ignore Manchin and Sinema. If, however, we simply hand both senators in each state to the winner of the state in 2020, the split falls to 50-50: Biden won 25 states and Donald Trump won the other 25. That’s not fair, of course; plenty of purplish states are more likely to split their votes between the parties. If, for example, you assume that any state that had a margin of a half-point or less between Trump and Biden would have one Democrat and one Republican, you’re already looking at a four-senator advantage for the GOP.
Move that up to a one-point margin or a two-point margin, and it’s a seven-senator advantage for the Republicans. If you figure that states with a three-point margin between Biden and Trump would split their senators between the parties, you’ve suddenly got a 55-45 Republican Senate — even with a 52-48 Democratic country.
This isn’t complicated to explain: A lot of Biden’s national margin was a function of massive wins in big states. His presidency is a function of narrow wins in smaller states — states that skew much more Republican than places like California. Places like Sinema’s Arizona.
That said, it’s not surprising that Sinema was elected. Arizona has been trending blue, and she ran in the favorable Democratic environment of 2018, eking out a win. The stars aligned and there she is. Manchin, however, is something akin to a miracle. West Virginia is one of the reddest states in the country, but Democrats nonetheless get a vote out of it — a function of Manchin’s track record there, a function of his efforts to keep space between himself and the party, and a function of himself getting lucky with election cycles.
To put a fine point on it: There is no scenario under which the Democrats should have a senator in West Virginia, but they do. It is mostly rural and heavily White in a way that has proved disadvantageous elsewhere. And, in fact, he’s a reminder that the party needs to figure out a way to get more senators like him, a way to win places that are more rural and more White, or risk permanent disadvantages in both the Senate and the electoral college.
On Wednesday, Nate Silver offered a useful frame for thinking about this. While the United States is pretty evenly distributed in where it lives — big cities, suburbs, rural areas, etc. — on average, states skew more rural. We can visualize that by comparing the national population (using Pew Research Center’s county-level descriptors for density) with the population across the 50 states.
The result looks like this, with the states having a more-rural and more-small-metro population on average than the country overall. And both of those sorts of counties preferred Trump in 2020.
This isn’t intuitive, so I’ll clarify a bit. New York is a split between New York City and the rest of the state. Overall, a bit less than half of the state lives in the five counties that make up the city. So about 42 percent of the state lives in a county with an average population density of about 12,800 people per square mile. Fifty-eight percent of the state, though, lives in a county with an average population density of less than 140 people per square mile. If New York allocated state Senate seats by county, New York City would be at a massive disadvantage in terms of county type.
The same idea applies nationally: States with large urban populations contribute a lot to the national distribution of urban residents, but there are a lot more states where rural residents are the norm. And that’s the disadvantage for Democrats in holding the Senate.
It is also the context for the moment. There are rumblings about running a candidate against Sinema in a primary, though that’s still three years away. But it’s hard to imagine a Democrat who could replace Manchin in West Virginia, particularly one who would vote any more to his left. It’s similarly tricky to figure out how the party holds seats in other heavily rural places with Democratic senators, like Montana, or gains seats in deep-red ones.
That’s the long-term challenge the party faces moving forward. After it figures out the short-term one with Manchin and Sinema this week — a challenge the party may soon come to view as the good old days.
correction
The density data for New York have been corrected after a calculation error.