Voting for Senators is a strange activity in the U.S. Even a cursory glance at the data raises more questions than it answers. The ebb and flow of available seats follows no true regular rhythm. Candidates win and lose seats based on vastly different vote numbers. The elect often remain in office for long periods. Little about the numbers reflects reasonableness.
As it turns out, Senators do leave office, generally due to death. When this happens, the governor of their state generally appoints the replacement. I was surprised to discover that governors have no problem appointing someone from their own party if it differs from that of the deceased incumbent. That seems unfair, but that’s politics. Since terms are six years, some states do hold interim elections where they deem necessary. So, the seats that are up for grabs do not follow an orderly rotation by thirds.
For that matter, a quick refresher may be in order. Every two years, a third of seated Senators are up for re-election. That means that the Senators in office right now were elected across the prior three campaigns held in even numbered years.
And that is why I have so many columns for each party in the data:
- Pop Vote: the national popular vote received by the party in this election cycle
- Seats: the actual seats won in the Senate in this election cycle
- Deserved: the number of seats the party should have received based on their share of the popular vote in this election cycle
- Vote ROI: the return in seated Senators that each voter received by voting for this party in this election cycle
- Tot Sts: the total seats held in the Senate after this election
- Tot PV: the national popular vote received by the party in the three most recent election cycles
- Tot ROI: the return in seated Senators that each voter received by voting for this party across the three most recent election cycles
- Tot Des: the total seats that each party deserves to hold in the Senate after this election
Briefly, ROI (Return on Investment) is the share of Senate seats received divided by share of the popular vote. I like it because it gives a sense of both the over-reaching and the cynicism of each party. Both major parties profit from the futility of third parties and independent candidates.
I obtained the popular vote for the parties that achieved significant support, but ended up extrapolating total turnout from 1940 to 1956 based on the percentages that the two major parties received.
As for the numbers
They prove stark and depressing. That’s because the reality has rarely reflected the will of the voting public. Since 1940, the Republican party has never earned enough of the popular vote to merit seating a majority in the Senate. The Democratic party has done so 18 time out of the 37 calculated. That still leaves 19 sessions (and 38 years) when neither party has merited outright control. Since 1992, the Democrats pulled it off only twice- 2008 and 2016. The most recent results look particularly absurd given 18 million more votes for Democratic Senatorial candidates.
Since 2002, Democrats have been building popular vote success not seen previously. Even so, from 1996 to 2002, the Republican party captured the majority of the popular vote, though nothing approaching the same degree.
One of the reasons that Senate representation is so skewed is because of the dramatic difference in the votes required to win election in the various states. In the 2014 election cycle, 84,196 votes was enough to get elected from Wyoming, while the winner from Texas needed a minimum 2,324,180 votes. If you controlled a national party and had to decide where you might best invest your funds, I think you might find you get the biggest return on your dollar in Wyoming.
Thirty-five Senate seats were up for grabs in 2014. Across the nation, 47,493,119 people voted. 8,207,251 votes decided seventeen of the seats . That means that just over a sixth of the voters decided almost half the available seats. Arguably, the people who voted for the winner in their state received the value of three votes in the remaining states.
What to do
As opposed to House elections where I had ideas about how to proceed, I don’t know that a good solution to misrepresentation is available here. We have never really entertained the idea of eliminating our system of separate states. In many ways, the distinct experiments carried on in governance within each state may be a strength of our nation. No doubt, that is a topic for another time.
All right- maybe I do have one idea.
Distribute the seats based on the national popular vote. Assign the seats to the candidates within each party who receive the highest proportion of votes.
This avoids the strange idea proposed in some circles to eliminate the requirement for direct election of Senators (otherwise known as the 17th Amendment). Congress put that into place because some state governments used to appoint their federal Senators as opposed to holding popular elections. Yes, it was a lovely reward for good service- something like an ambassadorship still is.