Nash Equilibrium in American politics

For the serious discussion of weighty matters and worldly issues. No off-topic posts allowed.

Moderators: Azrael, Moderators General, Prelates

User avatar
Pfhorrest
Posts: 4595
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: Nash Equilibrium in American politics

Postby Pfhorrest » Tue Feb 06, 2018 7:57 am UTC

I've got too much of a (literal) migraine right now to think if this counts as a literal nash equilibrium or not, but it's a pretty well established theorem in voting theory that our first-past-the-post electoral method pretty much mathematically guarantees a two-party system. I would intuitively expect a two-party system to naturally tend toward extremes as each party tries to differentiate itself from the only competition to gain votes, though Hotelling's law seems to counter that; perhaps the two effects are themselves in some kind of equilibrium.

In any case, it's pretty well-known that the structural thing that we have to change to break out of this binary partisanship is our electoral method. There's a lot of further debate over what electoral method to implement in its place and, further than that, the much harder problem of how to get any electoral reform passed when the parties in power benefit from keeping it the way it is. But one way or another, that's the well-known structural cause.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Tue Feb 06, 2018 6:54 pm UTC

Fixing that directly is certainly what I would like, but I was trying to be open to other ideas. For example, this voting system has been with us for quite a while, and we've weathered the storm of partisan politics before, so how is the case stronger today for this fix compared to times in the past? This is where I tried to lay out the idea of partisan sorting that's going on, and I put forth the idea that it's the lever of social media that's making this possible. So maybe another approach might be to make this a less powerful lever. Of course, I don't know what this solution would look like without completely destroying the freedom we have on the internet. But (assuming my idea is correct), understanding the mechanism can lead to other potential solutions.

But aside from how to solve the problem, another piece to this is simply building the case that we need to solve it. I remember harping on the two-party system since the end of the Bush years, and in fact that's what made me seek out a forum to do discuss this and ultimately led to my first political topic here. It's certainly not hard to find other people that share a similar urgency for fixing this, and I dove headlong into learning about other voting systems. But my case of "Hey isn't this other way better?" didn't really appeal to anyone that wasn't already filled with a similar zeal. I couldn't gain traction on this beyond it being an interesting academic concept.

I suppose this reveals a bit of motivated reasoning on my part for coming up with a compelling story of why two-party politics is such a problem. But in the end, that's another piece of what I want to accomplish. Firstly I want the idea of NE tested (if my desire to have this be true is causing errors in my logic, I want it exposed), but beyond that I want to come up with a way to break outside of the circle of people that already agree that our political duopoly should be fixed. I don't believe we'll convince the parties to do this, so it will taking uniting the people against the parties. Therefore, I do want a compelling story for why people should care. I want this idea either blown up as armchair expert garbage, or I want it placed on firmer ground that I can bring to others.
A gentle answer turns away wrath, but a harsh word stirs up anger.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7418
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Nash Equilibrium in American politics

Postby Zamfir » Wed Feb 07, 2018 2:02 am UTC

You keep returning to this Nash equilibrium, You also talk about social media bubbles. Are these separate ideas, or are they somehow related? .

To me, they seem difficult to fit together. Nash equilibria assume perfectly clear goals, a precise and fixed numerical valuation of any outcome. But if people have those, how would Facebook change their behaviour? Or are voters not actors, in this model?

More in general, I think you need to sketch the game-theoretical model that you have in mind, if you want to pursue this nash-equilibrium idea. What are the actors (are voters actors in the model? Are their special interest within parties, or are parties unitary actors), what payoff function do they have (do party just want to win, or do they have policy preferences of their own?), what actions are open to them, how do actions translate to outcomes? Do the actors have perfect knowledge?

Even if you can't dot every I and cross every t, it would still be good to have an outline. Without such a model, is it meaningful to talk about a nash-equilibrium?

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Wed Feb 07, 2018 7:28 am UTC

Zamfir wrote:You keep returning to this Nash equilibrium, You also talk about social media bubbles. Are these separate ideas, or are they somehow related?

NE describes the strategy, and the filter bubble idea is the tactic. The NE strategy is a hardline approach that favors blocking the opposition over partial wins through compromise. The tactic to achieve this is to employ the lever of social media to create isolated echo chambers that builds a sense of team loyalty and inoculates members from opposing arguments.

As you've probably guessed, I don't have enough of a game structure to actually make predictions, like why this strategy is preferred over one with compromise. I'm just observing that both parties are moving that way, and it's pushing us to a place of gridlock and greater tribalism.

Also I'm being a little loose with the idea of Nash Equilibrium. I wrote an earlier post talking about a solution that's near NE. Part of the rules of the game would be implementations of the electorate reward and punishment for certain behaviors (e.g. the appetite for attack ads). These rules would shift around in ways we can't perfectly know, but if the changes aren't too large, then the solutions could perhaps circle around a Nash Equilibrium. I.e. effectively providing gridlock most of the time, but not perfectly blocking all opposing moves.

Zamfir wrote:Even if you can't dot every I and cross every t, it would still be good to have an outline. Without such a model, is it meaningful to talk about a nash-equilibrium?

I'm envisioning the parties are the actors, but the concept of "party" is poorly defined. I suppose the best definition would be whoever is helping shape the party narrative, or the people with the largest public platforms within the party. We'd need multiple players on each team since there are conflicting voices on what the party should do, especially with the sweeping changes with Trump. But maybe a simpler model with fewer players would still be able to capture the NE that I'm proposing.

I guess I'm stuck here on the model. There's certainly no perfect information, but I have no idea how to build a set of actions and outcomes. Ideally we'd want to simplify things down to a level we could analyze, but I bet it's PhD worthy material to figure out what a useful version would look like.

I'm happy to proceed as far down the road of detail with you that you have patience for. :) But I'm also open to the idea that I'm out of my depths in thinking I could quantitatively model this. If that is true, then is there value is building a qualitative model? Or in other words, could I try to sell you on the concept?

In the end I'm just trying to connect dots. We have deep divisions that are more sorted by party than ever, and a politics that seems to thrive on bringing it deeper still. We have filter bubbles that wrap people up with a language and narrative that effectively lets them talk past the opposition without ever accomplishing anything. We have hardline political responses to each other that offer little space for compromise. And we seem to have an increasing appetite for more of it as entrenchment grows and blame flies back and forth. I feel there's a game theory lens that can bring this all together. But I'm open to the idea that I'm shoehorning NE onto it. I latched onto it because of the description of gridlock that blocks your opponent's moves, and that feels pretty similar to what we see happening.
A gentle answer turns away wrath, but a harsh word stirs up anger.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7418
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Nash Equilibrium in American politics

Postby Zamfir » Wed Feb 07, 2018 2:01 pm UTC

Thing is, I have trouble pinning down the core of your argument. As far as I can tell, it's something like this:

- US politicians are less willing compromise than even in the recent past
- this is because voters are nowadays strongly influenced by social media filter bubbles, and people in filter bubbles do not like compromises with the other party
- Politicians are actively at work to create filter bubbles, and the effect would be far weaker if they didn't do this.
- politicians might not want to do this, but they have no choice because otherwise they would lose to politicians who do. That the 'game theoretical' aspect

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Wed Feb 07, 2018 7:54 pm UTC

A few comments:

1. I'm not sure how much of the game is played by politicians versus by salient voices in the party. In reality, both play a role, but can we model the game in a more clean way with just one? I don't know. In particular I'm thinking of the powers that foster the filter bubble.

2. People in a filter bubble that fosters tribalism do not like compromise. The actors are shaping existing bubbles rather than creating new ones. I don't know if that's an important distinction within the game theory model.

3. Can we boil the goal down to simply a position of no-compromise? It seems like that's the case within the halls of Congress, but at election time there could be other benefits directly derived from the filter bubbles and tribalism. For example, building a sense of identity within the team and animosity to the other team might help with motivating voters and pandering to the base. Can the election game be separated out, or do we need one game that combines all of this? I don't know.

Thank you for the help so far!
A gentle answer turns away wrath, but a harsh word stirs up anger.

User avatar
Pfhorrest
Posts: 4595
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: Nash Equilibrium in American politics

Postby Pfhorrest » Wed Feb 07, 2018 9:37 pm UTC

I have a theory:

Nothing about the structure of the game has changed, but the increasing speed and quantity of communication is like adding heat to a reaction. It's still doing the same thing it was doing before (in this case, increasing partisanship), but because there are more interactions going on more quickly, it's happening faster now.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7418
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Nash Equilibrium in American politics

Postby Zamfir » Thu Feb 08, 2018 11:34 am UTC

It would be good to have some data on the effect that you're trying to explain. Numerical time-series, preferably. Voter behaviour, legislative behaviour, data that say 'it used to be like this, now it's different'. Where does the effect show up (and where is it absent where you might have expected it).

And when does it start! If social media form the core of the effect, then the 90s are the last pure period of the old, more bi-partisan ways. In simplest terms: then how about Ken Starr?

Can we boil the goal down to simply a position of no-compromise?

How so? You yourself appear convinced that compromised change is better than a gridlocked status-quo. Why would you assume otherwise for the actors in your model?

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Thu Feb 08, 2018 6:23 pm UTC

Pfhorrest wrote:Nothing about the structure of the game has changed, but the increasing speed and quantity of communication is like adding heat to a reaction. It's still doing the same thing it was doing before (in this case, increasing partisanship), but because there are more interactions going on more quickly, it's happening faster now.

That could be. There's certainly a much faster news cycle now with stories exploding on Twitter almost instantly. I'm sure that plays a role, though I have no idea if the metaphor of heat is helpful. One other impact could be that since there's a lower barrier for anyone jumping online and contributing, we now have a smorgasbord of options available now, thus letting us really be picky with only the options that taste best.

EDIT: I thought about it more and have come around on the heat metaphor. The fast news cycle gives no time to cool down and reflect, but just keeps feeding us more to be upset about. The smorgasbord is full of plates of outrage that we have a hard time passing up.

Zamfir wrote:And when does it start! If social media form the core of the effect, then the 90s are the last pure period of the old, more bi-partisan ways. In simplest terms: then how about Ken Starr?

If we set our sights a little further back, we have a civil war. My claim doesn't depend on there being a golden age of of reasoned bipartisan togetherness. The system is inherently both competitive and high stakes, which means you'll get people doing mean and nasty things.

The piece of data I'm bringing is the partisan sorting. I cited links in my first post, but here's another that lets you look at polling data over a number of years. What it shows seems to be in line with this shift happening with the Tea Party movement.

I cited a link of filter bubbles somewhere along the way that sets up a demonstration of what a conservative versus a liberal Facebook account would see. This is just demonstrating that the effect is real, but I'm really just relying on this being commonly accepted here. If you're looking for more data on it's existence, I could try to find that.

As for linkage between those two things, I don't have data, that's just me supposing as part of my claim.

Zamfir wrote:How so? You yourself appear convinced that compromised change is better than a gridlocked status-quo. Why would you assume otherwise for the actors in your model?

I meant the goal of shaping the filter bubble, as in can we boil down the benefits to simply fostering a position of no-compromise. I was proposing that other benefits might be relevant at election time, and turning up the volume on tribalism might just check all those boxes.

But your question is interesting. If I'm inherently treating compromise as something better, why wouldn't I assume the actors would feel the same? But compromise is the strategy, and I'm saying that I want a system that doesn't reward rejecting it. My case is that the system has built in incentives that are pushing us away from that. So I'm not assuming how the actors feel on compromise, I'm observing it.
A gentle answer turns away wrath, but a harsh word stirs up anger.

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Fri Feb 23, 2018 5:36 pm UTC

I know this topic lost steam, but I came across another piece that I felt was relevant, so I thought I'd dump my thoughts here.

I just learned about the concept of the Overton Window, which describes what ranges of topics are politically acceptable for debate. Things can be radically too far Left or radically too far Right, and areas that are not shunned on sight fall within the window in the middle.

The context where this was brought up was based on the idea that we now have two separate Overton Windows. There are mainstream ideas on the Right that people on the Left think are beyond the pale, and visa versa with ideas on the Left. This is a similar idea to filter bubbles, but it's not explicitly tied to social media.

This idea of two windows was linked to political strategy in the sense that we may see more candidates who have approval ratings with a high floor but a low ceiling. It's hard to imagine what could bring Trump's rating below 35% or what could bring it above 50%. He was polarizing during the election, and since then he's focused on throwing more red meat to his base. The floor comes from people being happy to have someone that speaks their language and sees through the same window. To me it seems like it comes from a major dysfunction in our communication that has brought us separate windows.
A gentle answer turns away wrath, but a harsh word stirs up anger.

User avatar
Pfhorrest
Posts: 4595
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: Nash Equilibrium in American politics

Postby Pfhorrest » Fri Feb 23, 2018 6:41 pm UTC

Really everyone has their own Overton Window, and the effective Overton Window in a debate is wherever the intersection between the participants windows falls. There's no doubt a different common window among the right than there is among the left, and the common social window is whatever the intersection between them is.

I think a general remedy to problems like this could be described as increasing the size of people's Overton Windows, and thinking about that now kinda makes me realize that a chart I made a while back to illustrate my political stance is basically an attempt to get people to think outside of their Overton Windows, which I almost universally fall outside of.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

guenther
Posts: 1836
Joined: Sat May 17, 2008 6:15 am UTC

Re: Nash Equilibrium in American politics

Postby guenther » Sat Feb 24, 2018 1:31 am UTC

As I said, I'm new to this idea, so there's certainly room for me misunderstanding. But my interpretation is that the Window is an aggregate concept and doesn't make a lot of sense at the individual level. That's because a group's window isn't simply overlaying all the individual windows, but rather the individual window is formed out of the consensus of the group. This consensus is how the boundary gets created.

So seeing this in the aggregate consensus sense, the idea that we have two different windows moving further apart means that we have a dysfunction in communication. There's no space for a consensus to be formed. There might be some actual space of common ground if we could get people to think outside their partisan divides, but the language used to wrap them up will either have to conform to the sides' windows or be rejected.

I'm all for promoting this getting fixed at the individual level, but I'm skeptical of how successful that will be. I'm sure you've heard the expression that keeping an open mind is good, but not so open that your brain falls out. I bet the result of people using introspection to broaden one's own window will be that they move it towards what they see as the middle, but will just be reinforced with how much the other side is the problem because they're so far from the reasonable common ground. This point of "reasonable common ground" is a concept that comes from the group's window, and when the windows are moving away from each other on how to acceptably speak, this just makes the other side seem that much more radical.
A gentle answer turns away wrath, but a harsh word stirs up anger.


Return to “Serious Business”

Who is online

Users browsing this forum: No registered users and 7 guests