Site Overlay

Democracy Is An Evolutionary Game

This essay by a political scientist is interesting in singling out moral integrity of politicians for explaining the “test result” for the Republican leaders that is depressing: “most elected Republicans are failing the test by refusing to stand up to Trump.” “The only dissenting notes are coming from Republicans who are retiring at the end of the year or don’t have to face voters for several years, such as Senators Mitt Romney of Utah and Ben Sasse of Nebraska.” On the other hand, the good news is that “The vast majority of lower-level Republican office-holders are passing the stress test, many with distinction.” The examples he cited include the Georgia Secretary of State, who has been thrown under the bus by Trump, and the director of the Cybersecurity and Infrastructure Security Agency.

How Evolutionary Game Theory Thinks Differently

Instead of moral compass as the only factor or only guidance for whether to stand up against Trump’s delay of presidential transition, the evolutionary game theory brings in the concept of strategy — the blueprint of various moves players make in various scenarios. What evolutionary game theory differs from other game theoretic models is not strategy or strategies — all game theories have those — but the Darwinian competition and selection of optimal strategies over generations of dynamic processes in a population of multiple players.

Evolutionary game theory also dropped the rationality assumption for game players, because “strategies” of animals are genetically inherited behavioral patterns. This aspect of the model does not apply to the game of democracy, which is played by human agents and is better to make rather than break the rationality assumption.

Why Democracy Is An Evolutionary Game

Now, going back to the behavioral patterns of Republican leaders against Trump, evolutionary game theory contains unique features that apply to democracy well. First of all, democracy is always a population game, open to everyone in the society and always a game of multiple players. Secondly, every political candidate has his /her strategy to win the campaign, and those strategies compete with each other for public support to be selected. Third, the same election game is repeated with no fixed and known point of termination, just like in real life. Finally, just like in classic games, we can see strategies that are evolutionary stable, called ESS (evolutionary stable strategies), which is similar to the Nash Equilibrium, except that the latter is broader than the former. In other words, all ESS must be Nash Equilibrium but not all Nash Equilibria are ESS.

Dominant Strategy & Evolutionary Stable Strategy

Enough of the boring stuff linking evolutionary game theory with democracy or politicians’ behavioral patterns, what makes game theory interesting — or what the game theory is mostly about — is to find out what is the dominant strategy, which is about how individual player (like a senator in our case) decides his/her moves without considering how other players (or other Republican senators in our case) move or play the game.

An even more realistic scenario is Nash equilibrium, a combination of strategies by all game players. Nash equilibrium is similar to the dominant strategy, but this time each player will make his/her best move conditional on how other players move. Most politicians are more likely to follow the Nash equilibrium than the dominant strategy. If each Republican senator acted on his/her own, there would be more to come out accusing Trump, because whether other senators do it or not makes no difference for each of them. Nash equilibrium however would tell them that if there is nobody else doing it, the payoff for him/her to do it would be very low and risk very high. Given Trump’s influence among followers, and his strong sense of revenging, whoever coming out first and alone WILL be Trump’s enemy number one, who is more likely to lose his/her re-election or lose the evolutionary game for not being one of “the fittest.”

Nash equilibrium strategy, or in our terminology in evolutionary game theory, the ESS (evolutionary stable strategy), is useful because it allows predictions to be made about what behavioral patterns of senators will be more likely. Like in an evolutionary game, senators or more generally elected politicians want to get elected /re-elected, which is to say they want survive the election /re-election campaigns over time — or using Darwinian term — become the “fittest”.

It goes without saying that any elected politician or candidate would listen to or watch the “trend of the crowds” so they can promise what voters want — more accurately the majority of voters want — whether it is gun control or birth control.

Appointed officials however do not have that obligation or incentive, nor the retiring or retired politicians or officials. The author of the aforementioned essay made the correct observation, and the game theory provides the reason behind.

Moving by the preferences of voters however is a “super-strategy” but it is the more specific strategies that the candidates are competing about. After all, no politician or no candidate can make everyone happy, so they all must find their targeted voters who will likely to vote for them but not their competitors.

The Mutant Players & Strategies

Evolutionary game theory not only predicts dominant strategy and Nash equilibrium, it also allows mutant players and strategies to enter — and to change — the game. After all, what is evolution without mutations? In the classical or traditional game theory the players are just two random individuals. But an evolutionary game is often played between a traditional and a mutant prayers.

Even more interestingly, the currently popular strategies may not be evolutionary stable in future generations. In other words, just because millions or even billions of players all follow the same strategy does not necessarily mean it is evolutionary stable. It all depends on the playoffs or using the evolutionary terminology, the fitness scores. A mutant player may decide to play a novel strategy and if voters like it, he or she may end up receiving higher score of fitness.

To make it more interesting and realistic, let’s just assume AOC is the new player and her new strategy is the “New Green Deal”. Such a progressive agenda may be popular among some voters and as a result AOC received higher fitness score than say Senator Diane Feinstein in the sense that the former gets a chance to promote her proposed legislation, and stay in power longer.

The Ultimate ESS

But what is an ESS anyway? It is the situation where the current strategy or strategies can resist or defeat invasions of mutant players with their novel strategies (here “strategies” can be agendas, goals, proposals or anything of that nature). Put in everyday language, an ESS is a sustainable combinations of strategies or the opposite of vulnerability to changes.

When we apply evolutionary game theory to human society, keep in mind that while it may take millions of years or generations to change an ESS in animals, in human society it may just take one election cycle to do it.

The other difference between human and animal ESS is that human voters are divided into parties to form alliances, and more importantly, people have different preferences and priorities. For animals, the selection pressure is high and the signals are clear: You inherited a wrong strategy, you die sooner or get killed earlier than others. Such a high pressure makes the ESS clearly defined and results are life or death, black or white, day or night, hardly anything in between.

This is not the case with human ESS. Human preferences are messy and sometimes vague, pending on many things simultaneously. If anything, this leads to a bigger variety of mutant strategies that will not be cleared away but stay in the society, sometimes for a very long time. Such a large reservoir of mutant candidates (i.e., players and strategies) is healthy, because it increases the collective survival rate of humans.

This leads to the most important conclusion of this blog: The best or the ultimate ESS of human society is diversity itself. Any single strategy — no matter how many players are following it now — can’t be stable. Every player following the Hawk strategy is just as evolutionary unstable as every player following the Dove strategy. Having at least two strategies and letting them competing with each other, this is the way to survive the Darwinian test. The only thing that is better is to have multiple strategies in reserve, to have tolerance for different ideas, institutions, paths of development, and so that when the time is right, they can come “out of the closet” to replace the current strategies.

What Can We Learn from the Evolutionary Game?

Human players can change their strategies quickly, depending on the scenarios and most importantly their personal costs and benefits from the moves. Whether Republican leaders would come forward to criticize Trump is not the same as whether they would do so when they are subpoenaed to testify to the Congress or to a court. The latter suddenly increases the personal cost of not telling the truth, whether for or against Trump. We should thank Trump for being the most dramatic and most “eventful” president in the US history, offering an unprecedented number of scenarios of political drama. We only need to look back at his impeachment hearing on the Ukraine military aid case, more specifically the July 25 phone call made by Trump with Ukraine President Volodymyr Zelenskiy. As far as I can remember, all government officials, most or all Republicans, have spoken out to provide evidences not in favor of Trump.

Why China Is On A Wrong Path

China and US are following very different strategies of survival. The biggest difference is not ideologies like Mike Pompeo has wrongly claimed, but whether to tolerate diversity both domestically and internationally. The US is on the right path by allowing battle of ideas — more accurately rule based battle of ideas. For the most part, the country lets voters decide what is right or wrong, what survives or not, what fits or not. By letting voters have the final say on ideas, agendas, strategies and roads, it naturally introduces competition for voters’ attention and attraction, which is essentially social selection in an evolutionary game. Tolerant to players and ideas or strategies, the country has a large reservoir and talents and strategies, which come out at the least expected time or place, like this story from Michigan shows.

China’s biggest vulnerability is to move on the wrong path of promoting single strategy, single path or single idea. I will talk about it more in another blog.