Against Term Limits; qualitative reasoning

Congress, even when it doesn’t feel like it’s moving, is a dynamical system. What this means is that, as time goes by, Congress changes. It (sometimes) passes bills. It (sometimes) endorses Supreme Court Justices. People make alliances, they argue, they jockey for position and status, and things for the folks back home.

Understanding how something like term limits might affect Congress seems like a daunting problem. Congress seems so complicated — and a lot of the politics governing it are hidden or distorted from public view.

What we’re going to do: we’re going to appeal to some math — qualitatively — and then we’re going to try and understand how the math helps us understand Congress by associating bits of math with human behaviors.

My favorite part about math is that it helps you understand how to decompose complicated things into patterns that combine together to elicit the richness of reality. This example is one of my favorite ones because it leads you to a nontrivial conclusion.

Every dynamical system can be broken up into two behaviors: slow mode and fast mode behaviors. Even though vibrating guitar strings don’t seem like they have anything to do with Congress, play along for a bit. Look at the image below from Wikipedia. These are different patterns of slow and fast mode behaviors; on the top left you have the slowest mode, on the top right a slightly faster mode, and so on until we get to the bottom right, which is the fastest mode.

Source: https://en.wikipedia.org/wiki/String_vibration#/media/File:Standing_waves_on_a_string.gif

A violin string plucked at a B above concert A might look like this:

http://www.ccp14.ac.uk/ccp/web-mirrors/isotropy/~stokesh/violin.html

The complicated sound of a violin string is nothing more than the combination of slow and fast modes. The slowest mode here is that of the B note, and the faster modes reflect the unique characteristics of the violin itself. (Imagine adding each of the funky waves above in different amounts.)

In basketball the slowest mode is the back-and-forth after teams score. A faster but still slow-ish mode is how often you draw a foul against a player known to not do so great at the foul shot.

The fastest mode of a game depends on whether King James is mad.

The point here is that you don’t need to be deterministic (like an ideal clock) for a dynamical system to decompose this way. They all do.

OK — Congress is a dynamical system. It has slow and fast modes. But what do they mean?

I argue that the slowest modes of Congress are the processes of generating consensus or majorities based on their personal beliefs. It’s like momentum: left to their own, a person elected to Congress will continue to be pro-life, or pro-small business, and so on. However, a slightly less slow mode is consensus building; it’s what’s necessary to actually enact their personal beliefs. This interacts with their personal beliefs in unusual ways: sometimes people will vote yes or no on a bill that’s suboptimal according to their personal beliefs, but necessary to build consensus on the next set of bills. A slightly less slow mode is to satisfy the electorate.

Complicated dynamical systems also have “phase changes” which are characterized by abrupt changes. An example of this is the frenzied and sudden backroom deals that we see around budgets in election seasons.

The fastest modes in a dynamical system contain more energy than the slowest modes. Maybe that seems counterintuitive, but think about Congress: you see lots of drama and theatre and posturing in the public eye. They expend tons of energy on the opposite of their innate characters — few people survive in Congress being quiet and voting just as their own fancy takes them.

Let’s now imagine the process of imposing term limits. A term limit for a dynamical system amounts to limiting the extent to which slow modes can smooth out the system. This at first seems like a Good Thing: “I don’t want people to stick around with their stupid ideas forever!”

Consider other slow modes that exist right now (without term limits). Tribal knowledge. Passing and enacting actually plausible legislation is a really hard process — you have to fact find, you have to try and set a budget, you have to try and describe how the money should be allocated and executed by the executive branch, you have to consider how it interacts with the Constitution and existing law. The way we’ve found to deal with this is the oldest human strategy for forming societies: we develop a feel, a system, tribal and nonverbalized behaviors that become slow modes to smooth out problems.

The shorter you make term limits, the less smoothing we’d see. I’m arguing that this is a human nature phenomenon, not something endemic to a constitutional government. Can you imagine an only news-pandering Congress?

Everything you hate about Congress would be — paradoxically — magnified by term limits.

Heck, the slowest mode of government is the bureaucracy. Yes, it’s annoying. But it also ensures peaceful transitions of generations through dissipative power.


I have been dreaming about actually modeling this for some time as a sort of “computational government” /sociology problem. I wish I had PhD students…I find it frustrating when historians claim that we lack data, or that these systems are ill-posed. You can do a hell of a lot of qualitative reasoning even with limited data and gain useful insights you might not have as easily otherwise had.

I should also add the following: I am a Director at Sanofi Pasteur. That means that everything written in this blog, current and past posts alike, were neither endorsed nor approved by my employer, and do not reflect Sanofi Pasteur, Sanofi, their affiliates, their ex’s, their pets, or their favorite sports teams in any way. These opinions are mine, so blame me.