Looking at dissent in the U.S. House of Representatives

This all came up when a friend and I were recently chatting politics. At one point he expressed frustration with how one of the two major US parties was, according to him, way more dogmatic than the other. Interestingly, he felt that this manifested in the way members of congress from that party always seemed to vote according to party guidelines and not according to their own values and positions.

I really liked this framing because it is quite simple to measure, so I thought I’d give it a try.

Measuring dissent

If we assume that a party line would usually be represented by firm majoritarian positions in that party (Yea or Nay votes), then we can measure dissent by counting the votes that were cast on other positions (Nay, Yea, Present, NV).

While, I couldn’t locate aggregate party voting stats on any of the chambers, The Office of the Clerk for the US House of Representatives lists row data on all roll calls since 1990. Here’s one example:

I downloaded all 18,281 of those and if you’d like to spare yuorself the hassle, you can get the aggregate data here.

After looking at things a bit more I decided to only focus on the firm positions: Yea and Nay votes. I don’t know the inner workings of the House well enough to fully grasp the strategic significance of the “Present” and “Not Voting” positions, but it felt like strictly opposite votes would be a better indicator for dissent anyway.

The next thing that comes to mind when thinking about dissent is that both parties would have an incentive to manage it. If we have learned one thing from Frank Underwood it is that whipping people into line is hard work.

A utilitarian view of the process would hence imply that more of this work would happen if it is likely to payoff.

Also, as my friend Jonathan Lennox just explained to me: the majority gets to define the agenda, so they get to keep more controversial topics off.

All this to say that, it felt important to check if there is a correlation between levels of dissent in a party at any moment and its current role in congress.

Here is what we get then:

The first thing that jumps out is that, YES!!!, a party’s role seems to be a very reliablе predictor of the comparative dissent across the two parties. For the thirty years available on house.gov, there were only four instances where this wasn’t true:

In year 2006 and then again in 2015, 2016, 2017, the Democratic party was in opposition, but it generated lower levels of dissent than the Republicans.

The final interesting piece of information that came out of this was the trend:

Both parties seem to be agreeing to disagree, more and more. If lowering levels of dissent are indeed indicative of dogmatism (which is a BIG “IF”), then this doesn’t look good. I am also guessing that this is like yet another indication of America’s growing polarization … not that we were in any lack of indications [0], [1], [2], [3].