Grokking Republicans: The Non-Cooperator's Dilemma

To create More and Better Democrats means to increase cooperation. Punishing cooperation is the declared Republican mission. The Evolution of Cooperation, by Robert Axelrod, proposes a theory that says they lose, and recommends particular political strategies to make it happen faster. What we want to know, of course, is when and how they lose, and how those strategies speed things up.

The theory has been proved, in a very specific scientific sense. It fulfills the two essential requirements of a successful scientific theory:

• It has a mathematical content, provable in itself from its definitions and axioms apart from the observable world of human interactions, implying important real-world consequences.
• It has been verified by observation and experiment that the definitions and axioms apply under a wide range of conditions, and in many variations, so that we reliably get the results originally predicted, and more.

Even better than that, the theory can be extended in many other useful ways, with provable properties verified in the real world. We also know some limits of the theory, that is situations in which it cannot be applied usefully.

The content of the theory includes specific advantages that cooperators using certain strategies have over non-cooperators (defectors in the language of the book) in general. I will explain some of the theory, and then show how to apply it to Republicans.

The Prisoner's Dilemma Game

The very simple Prisoner's Dilemma game sets up a tension between cooperation and defection (not just neutral non-cooperation, but punitive action) between two people at a time. It is modeled on the familiar police tactic of putting two prisoners, criminal suspects in this case, in separate rooms and offering them each a plea deal to rat out the other. We suppose that the police have probable cause for the arrests, but not enough evidence for a sure conviction in court. So if the prisoners both cooperate with each other in denying everything, they both have a good chance of getting off. Let us, somewhat arbitrarily, award them three points each in that case. If they both rat each other out, they both lose, and we will award them one point each. If one confesses and the other does not, the defector may get off entirely (score of 5) while the cooperator gets done completely (score of 0). We can represent the game in a table.

Prisoners' Dilemma game values table

The first number in each cell is the score for the Row Player, and the second number for the Column Player. Thus the top right cell is for the Row Player cooperating and getting 0 points, and the Column Player defecting and getting 5.

Axelrod refers to these values as

• R=3 Reward for mutual cooperation
• T=5 Temptation to defect
• S=0 Sucker's payoff
• P=1 Punishment for mutual defection

The numeric values are not legally realistic in terms of the plea deals actually offered and the sentences that may result, but that is no matter, because the legal situation is meant here only to be suggestive and mnemonic. The exact numbers do not matter, but certain relationships among them do.

It is essential to the game that the reward for cooperation is less than the temptation to defect (R<T), and the sucker's payoff is less than the punishment for mutual defection (S<P). This means that in a single round of the game, defecting always results in a higher score for the defector. It is also essential that the rewards for mutual cooperation, here 6 for both players together, add up to more than the scores for defection, here 5+0=5. This means that in a group contest with many rounds, consistent cooperators can beat out those who defect even occasionally, because they score higher when playing each other, and can punish defectors and keep them from getting consistently high T scores by retaliating in the next round, or even more than once if they choose.

The values chosen for the standard Prisoner's Dilemma game give cooperation nearly the minimum possible advantage. It is therefore all the more remarkable to find how robust cooperation can be in both computer simulations and the real world even against the most entrenched opposition. But it is also important to be aware of how entrenched opposition can be, so that the victory of cooperation in a given social contest can take a very long time, commonly decades to centuries.

An example is the survival of Jews and Roma (Gypsies) in Europe against massive discrimination and even murder since the Roman Diaspora of the Jews in the 1st century, and the arrival of the Roma in Europe from India about a millennium ago. But the discrimination has not ended, even in the era of Human Rights. Other religions have not only survived, but prospered, using different strategies. Buddhism, Jainism, and Daoism are quite cooperative, even with outsiders. Islam veers between cooperation and astonishing intolerance and murderous defection in various historical periods and different communities, but then so does Christianity of various kinds.

Successful examples of overcoming massive defection include the end of the Troubles between Protestants and Catholics in Northern Ireland, and the end of Apartheid between Whites, Blacks, and the three other categories in the system in South Africa. In each of these cases, many on each side incorrectly considered their opponents to be subhuman and incapable of cooperation for decades. We have the same problem in Israel and the Palestinian territories.

I do not have room here to present the details of any but the simplest of contests. It turns out that in a variety of contests, the rule called TIT FOR TAT has consistently though not universally beaten out all comers, winning in many early human trials and in all but one of the computer tournaments reported in the book, and also winning an ecological contest in which the number of copies of a program in the tournament pool increases or decreases in proportion to its previous success.

It is one of the simplest strategies possible, after RANDOM (cooperate or defect with equal probability), always cooperate (ALL C) and always defect (ALL D). The rule of TIT FOR TAT is to cooperate in the first round, and after that do what the opponent did the time before, holding up a mirror to the opponent. Thus it punishes an opponent for defecting exactly once each, and rewards the opponent for each and every round of cooperation.

ALL D scores consistently high against programs that cooperate freely and rarely or never defect, and consistently low against those that punish defections. No strategy can beat ALL D one on one by attempting to cooperate with it or punish it. It is impervious to any action by the opponent, which it completely ignores. The best you can do is drive its score down nearly to the limit (1 point per round for mutual defection, only 1 point above the sucker's payoff).

ALL C scores consistently high with other cooperative programs, but gives the maximum scope for defectors such as ALL D to take advantage of it.

TIT FOR TAT scores as well as ALL C with programs that are willing to cooperate with other cooperators, and successfully fends off programs that try to take advantage of cooperators, but back off when punished. It does almost as well as possible against ALL D, losing only the first exchange.

Dozens of other programs have attempted to beat TIT FOR TAT, but all have failed, mostly by trying too hard in ways that we will look at below.

We can categorize the behaviors of rules in various ways related to their design and their performance against other classes of opponents. Some of the following definitions are mathematically exact, while the rest depend on somewhat fuzzy concepts.

• A nice rule is one that is never the first to defect, so it cooperates with any strategy that starts out cooperating, for as long as that lasts. ALL C and TIT FOR TAT are both nice. ALL D obviously is the opposite. Nice rules can respond to defection in any manner that the designer chooses. Nice rules always cooperate perfectly with each other, and almost all rules that did well in the tournaments were nice.
• A forgiving rule defects no more than once for each defection of its opponent. TIT FOR TAT is forgiving. We could say that ALL C is too forgiving. One could have rules that punish defections twice (TWO TITS FOR TAT), or always thereafter (FRIEDMAN), or by some analytical or probabilistic rule.
• A retaliatory rule defects immediately in response to an "uncalled-for" defection, which is not precisely defined, but definitely includes a defection after a period of mutual cooperation. Not retaliating immediately to defections allows exploitation by the opponent.
• A maximally discriminating rule will cooperate at least once even if its opponent has never cooperated, and once it has done so will never cooperate with ALL D again, but will cooperate with rules using its own strategy. TIT FOR TAT cooperates on the first round, and thereafter defects against ALL D in all succeeding rounds, and cooperates with itself in all succeeding rounds. A nice and retaliatory rule is maximally discriminating. It is not required to be forgiving.
• A kingmaker is a rule that gives different results with different nice rules depending on how they respond to defections. Kingmakers may or may not score high, but can determine which nice rules win. An example called DOWNING (after its author) tests rules and then responds to them by estimating the probability that they will cooperate or defect. It mostly cooperates with TIT FOR TAT and some other rules, but punishes less forgiving rules drastically.
• A sneaky rule cooperates a lot of the time, but tries to see how much defection it can get away with. An example is JOSS, which adds a small probability of random defection to TIT FOR TAT. After it tries to get away with a defection, TIT FOR TAT punishes it, and JOSS punishes TIT FOR TAT, in a cycle that continues to the end of their confrontation.
• An exploitative rule tries to analyze the behavior of opponents, looking for forgiving but not retaliatory rules, to see what it can get away with against them.
• An exploitable rule cooperates too much with some rules that defect fairly often, particularly if they find that the exploitable rule does not punish them enough to deter defection.
• A robust rule does well in a variety of environments, although it can be proved that no rule can win in all environments. TIT FOR TAT won all but one of the tournaments Axelrod conducted. Attempts to improve on it and attempts to beat it both failed, except in the presence of many copies of the rule REVISED STATE TRANSITION, because that rule was vulnerable to exploitation by another rule that otherwise did fairly badly, coming in 49th out of 63 in the second tournament.
• A clear rule makes it easy for other rules to recognize that they can cooperate with it and also that they will be punished if they do not.
• The shadow of the future is a measure of the importance of repeated play. In a single round, there is no way to retaliate for a defection, but as the number of rounds increases, the possibility of future punishment increases with it. In real-world situations, many factors can influence the shadow of the future.

The shadow of the future we are looking for on dKos is a pattern of Democrats always voting to help each other, and conditionally offering to help Republicans with what is really important, that is, everything but feeding their fears and hatreds. The condition is of course giving up the attempted punishments flowing from the relevant fear, and actually cooperating with everybody else on some issue. Congress, the Administration, and the Courts are all designed in the Constitution to provide cooperation where possible in support of the common good, and defection against the rascals who try to exploit government against everybody else. This is known as the Balance of Powers. It has in many ways gotten better over time, but still could do with some significant improvements to make it nicer, more forgiving, more appropriately retaliatory, and less exploitable.

The most important observed facts from the tournaments are

• The field in Axelrod's tournaments divided neatly and almost exactly into higher-scoring nice programs of various kinds that cooperate completely with each other, and lower-scoring programs that defect without provocation and were frequently punished.
• Nice, forgiving, retaliatory, and clear rules like TIT FOR TAT win except in the presence of too many exploitable rules, which do not win themselves but allow some exploitative program to get a lot of T scores against them.
• Analysis in designing rules has to go at least three levels deep. At level 1 the game table gives the advantage to defecting, if nothing else, especially any future rounds, matters. But at level 2 one must consider the possibility of punishment for defections in the next round, as in TIT FOR TAT. That much was obvious to all designers. But the more subtle level 3 is that responding to defections risks a pingpong chain of mutual retaliation, or even permanent defection, so successful rules must be forgiving, with no more than one retaliatory defection per provocation.
• Trying to be sneaky or exploitative gains points against rules that are too easygoing, but loses many more to retaliatory programs.
• If the number of future interactions expected is sufficiently small, defection pays off better than any other strategy. The con man expects never to see the mark again. Mobile populations such as Roma (Gypsies) or migrant workers have much worse relations with surrounding communities than permanent residents. Companies often refuse to pay bills to other companies nearing bankruptcy. Walling off Palestinians guarantees that there will be no repeated social or business interactions between them and Israelis, while interactions with Muslim citizens inside Israel continue, though not at the same levels as between Jews.
• In the ecological simulation where rules reproduce according to their success in the previous round a small group of cooperators can survive and will routinely grow to be dominant whether in a friendly or hostile environment, even 95% ALL D. However, even sizable groups of defectors among non-exploitable cooperators will be wiped out in short order. Highly exploitable rules vanish if there are exploiters about, and exploiters vanish if their prey does. Other strategies may rise and fall for an extended period depending on their relations with each other.
• Many designers consistently underestimated the power of cooperation, and assumed that the primary problem was how to exploit other rules effectively, not how to cooperate effectively without being exploitable. Many political scientists and nearly all politicians seem to fall into this error, routinely predicting that things will stay the same or get worse, not better. But over decades and centuries many things do get better, even when faced with retrogressive forces, except when fearful events like terrorist attacks, the rise of powerful enemies, plague, or economic collapse occur, or at least threaten a society where the values of cooperation are not strong enough.
• One of the most important facts about the dominant performance of TIT FOR TAT in Axelrod's tournaments is that it never won a single match. It came out even with other cooperators, and lost slightly to non-cooperators. It won the tournaments because it could not be exploited, and because tying with even modest numbers of other cooperators provided much higher scores than others could get trying to beat out their opponents, and more often than not being punished for it.

Examples

One of the most important observations Axelrod made about the first computer tournament is

Even expert strategists from political science, sociology, economics, psychology, and mathematics made the systematic errors of being too competitive for their own good, not being forgiving enough, and being too pessimistic about the responsiveness of the other side.

We see such lacks constantly in human interactions such as union negotiations, criminal law, confrontations between nations, and the multitude of struggles for every kind of basic human right everywhere. The other side does not consist of subhuman monsters, unless it has been carefully and completely taught that it is superhuman. Nazis, torturers, divine right of kings, the one true religion, all that sort of thing. But even among the worst of those, their children are curable, as witness post-war Germany, after the Marshall Plan applied cooperation on a massive scale.

Contrary to pre-war Nazi CT, which claimed that Jewish bankers were out to destroy the German government and economy, it was evident that Jewish bankers and politicians were actually cooperating in building Germany back up again in the Marshall Plan, greatly reducing though not eliminating German anti-Semitism. Compare that with the French policy of permanent defection after World War I that provoked the growth of the Nazis as ALL D defectors against democracy, Jews, Communists, and many others.

The retaliatory Chinese Legalist political philosophy of Han Feizi, adopted by the first Emperor, Qin Shih Huang Di of the Qin dynasty, recommended that the ratio of punishments to rewards should be 10 to 1. Many of the rules submitted for the tournaments were not very forgiving, but few other than FRIEDMAN went as far as orthodox Legalism. The Emperor took it much further than that, for example ordering 460 scholars of all kinds buried alive after being deceived by two alchemists. Disproportionate collective punishment has been a theory of government in several cultures through the ages. It always leads to rebellion and overthrow eventually, or to conquest by other powers due to its internal weakening effects. North Korea is presently the only surviving example, although there are a few lesser tyrannies about.

One of the most important historical examples of several of the principles discussed above is the Dutch rebellion against the Spanish Empire. It was provoked by the Inquisition attempting to root out Protestants, and the antique economic theory which led Spain to forbid spending any of the wealth of the Americas, gold, silver, or jewels, outside the Empire, thus creating runaway inflation. The Dutch were driven to declare the highly cooperative and unheard-of principles of Freedom of Thought (including religion) and Freedom of Trade in almost any goods with any country that was willing. The result was the Eighty Years War, which suggests how strongly both sides felt about it.

The atmosphere in the Netherlands, notably their hosting of the Jewish philosopher Spinoza, produced the seeds of the Enlightenment that in turn produced the American Revolution, the Bill of Rights, and much more. English Separatists fleeing oppression in England settled for a time in the Netherlands before building their theocracy in Massachusetts, from which Roger Williams left to establish Rhode Island, also with freedom of thought.

The other colony that began with freedom of thought and freedom of religion was Quaker Pennsylvania. Quakers radically changed business practices in Britain and British North America by being the only honest, that is, cooperating businessmen. They had collectively and cooperatively concluded that God required them to be honest if it killed them, or more precisely if they lost so much money by not cheating customers and other businessmen that they all went out of business. Nobody was more surprised than they when they prospered mightily. We, of course, can see in these results that being honest in business is cooperation, trying to exploit the customer as much as possible is defection, and nothing is more valuable to a business than satisfied customers who tell their friends and relations. This does not work for an individual in a culture of cheating. In order to spread, it requires some modest fraction of the relevant population to start by being honest among themselves.

I have to leave out much of what Axelrod discussed in the book, including the math, details of the tournaments, and important biological and historical examples. He devoted an entire chapter to the Live and Let Live system of World War I trench warfare, in which both sides mostly stopped shooting at each other except when explicitly ordered to. He also drew on Richard Dawkins's analysis of biological cooperation and apparent altruism in The Selfish Gene, among other evolutionary biology, as part of an entire chapter on the subject. Examples include symbiotic bacteria living in plant and animals cells as chloroplasts and mitochondria, or the fungus and photosynthetic species that combine to form a lichen.

I also have to leave out all of the variations on the Prisoner's Dilemma, of which the most important is turning it into a multiplayer game with the ability to focus punishment on defectors. This is discussed in Axelrod's followup book, The Complexity of Cooperation. The application of the multiplayer version to international relations, particularly to the application of sanctions, should be obvious. A key lesson is not to go all in on sanctions (comparable to the rules not to defect first, and then only once per defection of the opponent), but to have plenty of levels in reserve in order to be able to ratchet them up.

Thinking about reasons why abolishing the death penalty for more than 200 crimes in England in 1808 greatly reduced the murder rate is left as an exercise for the reader. Hint: In for a penny, in for a pound, or As well be hanged for a sheep as a goat.

Personal Strategies

Axelrod's advice for those who find themselves in an iterated Prisoner's Dilemma, say with a Republican or a boss or the police, is as follows.

1. Don't be envious.
2. Don't be the first to defect.
3. Reciprocate both cooperation and defection.
4. Don't be too clever.

These are difficult rules to follow. They require thought and practice, and preferably a community working on them together, like the Quakers or some Buddhist groups or the Civil Rights Movement or the NAACP or Planned Parenthood or the League of Women Voters or Gandhi organizing against the British Empire. This is just like learning non-violence from Rev. Dr. King back in the day. Cooperate and don't fight back at the lunch counter, but bring legal actions and call on Congress to pass laws to punish the defectors appropriately, or to enhance the power of cooperation. Voting rights laws and labor rights laws, for example.

Axelrod has valuable explanations of these suggestions that I suggest you read for yourself. I am adding a few explanations of my own here.

Not being envious means maximizing your own score, not trying to minimize your opponent's. The Prisoner's Dilemma is not a zero-sum game, in which you win only by harming opponents. The most effective strategy in Prisoner's Dilemma is mutual cooperation so that all cooperators do well, and both retaliatory cooperators and defectors harm defectors, preventing them from exploiting cooperators.

To be effective we need to establish cooperation as widely as possible so that we all do well. It simply cannot work well in the political realm to set out to punish the opposition party, particularly those who are already paranoid about us, and those who lie about our hostility to them. They will respond like Redeemer Democrats after the Civil War creating Jim Crow, or like Germans punished by France after WW I. What if there could have been a Marshall Plan for the South instead of Radical Reconstruction?

We may need to call for legislation and prosecution here and there in order to break entrenched social power, but in the end we have to convince people that they can get along with us. Or if not them, then their children or grandchildren. Which means cooperating with them as often as possible on anything of real value. Which in turn means interacting with them as often as possible, even while they try to refuse any real interaction, and limit contact to trolling. This is the advice über-troll Ann Coulter gave in How to Talk to a Liberal, If You Must.

Don't feed the trolls. Cooperating with some of them is impossible and meaningless. Here on dKos we have a simple and effective means of defecting when they turn up to make trouble: Hide-Rate, or HR, them whenever they say something really offensive, but cooperate if they drop the shtick and discuss facts. A few have been taught manners, and sometimes even some sense. In extreme cases, we can HR on sight and bring down the banhammer. (Thanks, elfling.)

The detailed application of this theory to the evolving dKos Hide Rating system is left as an exercise for the reader. It is worth understanding, because it is one of the most important factors distinguishing dKos from all other Internet discussion sites, most of which provide unmoderated and starkly trollish comment threads, or use staff to moderate, sharply cutting down on the volume of discussion. Start with the Troll Rating FAQ.

Hints: Recs show cooperation, HRs are defections, and you can also be neutral, not even reading most Diaries and passing over most comments where you do read. The community is encouraged to HR spam, insults, deliberate outing of private information, and some other general nastiness. The stock rule for avoiding being HRed is DBAD (Don't Be A Dick). HRs should be used for some purposes and not for others that people try out from time to time, but are considered HR Abuse. There are various rewards and punishments provided for getting certain levels of Recs and HRs. Are the system parameters to your liking? Could we improve the system? Do you thank elfling frequently enough? Who are your favorite trolls? Slashdot has a very different community moderation system. I don't like it that much. Do you have an opinion on it?

Anyway, we need to find ways to cooperate with some of those who do not even know that they are looking for cooperation, especially the marks of the professional Right, the ones they extract cash and votes from with the Noise Machine in the Echo Chamber by feeding their fears and hatreds, and their Cognitive Dissonance and sense of entitlement, their propensities to defect against others. Every once in a while you will encounter one whom you can find common ground with on some genuine human issue, as singer Bono did with Sen. Jesse Helms, otherwise one of the worst professional defectors, on childhood AIDS in Africa. Each one will lead to others. We can even cooperate with stopped clock Rand Paul on the issues where he supports useful measures for the wrong reasons, and we can do it while defecting on all of his Ayn Rand nonsense.

TIT FOR TAT, always and everywhere. Obama needs us to defect on Too Big or Too Bad to Jail and his dreaded All of the Above energy policy. Republicans finally defected enough to undo his previous mindless bipartisanship. Certainly we will have to cooperate some and defect some with HRC, too. I think we almost all agree that we don't like her warmongering and corporatism, even if we disagree wildly on what to do about it. Nobody in politics roxors, and relatively few completely suxor, although the competition at that end of the scale is fierce.

Not being envious overlaps with not defecting first. Offer cooperation to the oppressors in everything except their essential acts of oppression. Jesus went further. He said to pay Roman taxes and do what soldiers require you to do.

If a man makes you go a mile with him (a soldier requiring a civilian to carry his pack in accordance with Roman law) go with him two.

Jews are told to do good to those who injure them, and thus pour coals of fire on their heads. Buddhists traditionally said

How happy we are, not hating anybody.

Even if you can't believe that they are really cooperating on anything, take them at their word if they make the offer. You can always defect later if it turns out not to be real. Defect in the relevant communities and in the media if you can, and at least here. But don't be so open-minded that your brains fall out. Don't accept a pretended offer of cooperation that requires you to join in some other oppression.

Being too clever in a Prisoner's Dilemma in a given environment means, to begin with, trying to beat reality, like a gambler with a system. Usually it means trying to find a way to exploit the opponent, to do better than TIT FOR TAT. Forget it. You can't even make that work in a computer tournament, much less as a long-term political strategy with people. We are trying to establish cooperation, remember? Being clever and having unclear motives mostly sets off the paranoid alarms on the other side, which commonly responds with defections.

However, there are situations where we want to change the environment to reduce cooperation among those who will not cooperate with us, no matter how clever they may be. This is a particular problem when writing financial laws and making regulations that might be exploited by clever financial advisors or lawyers. Other cases are members of a close-knit criminal gang such as the Mafia when the code of omertà was absolute, or economic price-fixers and monopolists, or Conspiracy Theorists, or a particular faction of Republicans out to injure certain segments of society. In that case cleverness may be essential.

The Defector's Dilemma

The essence of government is cooperation even where there is conflict. The problem of politics is do make decisions concerning questions that do not have answers, or at least do not have agreed-on answers. It requires that enough of us agree to cooperate to hold things together even against the forces of those who do not agree. This condition has been rare in human history. Almost all societies have fallen under rule by individuals or relatively small groups who have it all their own way domestically. And yet there have been Republics off and on since Greek times.

In slaveowning societies, of which Haiti was the worst, and major tyrannies, of which Nazi Germany is the archetype and North Korea is currently the worst, the penalties for non-cooperation can be raised to spectacular heights.

The theory presented here gives us reasons for much of this. In the past, cooperation within a population has been extremely difficult. At one time it would have required a great deal of face-to-face contact, but cooperation became easier to extend with writing, then paper, then the printing press, then the telegraph, then the telephone, then the fax machine, then radio and TV, and now electronic communication and the Internet. Also various social arrangements, from scientific societies to coffee houses to public libraries to online media. The printing press was enough to support the Enlightenment. We have only just begun to see the power of the Internet to aid in organizing the public and bringing down tyrannies, although we must recognize the limits to its power in a place like Egypt. It is obviously not enough in comparison with organizations such as the Muslim Brotherhood and the Egyptian military, which have their own channels of communication and cooperation that exclude almost everybody else. So far.

North Korea can only maintain itself by limiting cooperation among the populace as much as possible to only what has been ordered from above, and punishing attempts at any other kind of cooperation severely. Collective punishment is the norm. Similarly it has to enforce cooperation within the Party and the Military against the people by similarly draconian punishments. All of this requires maximum control of access to information. North Korea is the only country with no public Internet. Private publication of anything is strictly forbidden.

I lived for a time with a North Korean refugee family as a Peace Corps Volunteer in South Korea. I remember when South Korea for the first time realized that it had enough TV sets to put people on the air who had been separated from families and friends during the Korean War, and expect that a significant fraction could be recognized and reunited. Many were brought together in tears on-air with people from decades before whom they had assumed were lost to them forever. It was an amazing national catharsis. Reunification and national reconciliation is the great national desire in the South, but how it will work out is largely unimaginable at the levels of defection that North Korea practices against its own population and the entire rest of the world. The Northern regime can only imagine reunification as conquest of or surrender by the South. Kind of like Klingons on Star Trek, or Sontarans and any number of other villainous aliens on Doctor Who.

The Republican's Dilemma

The application of the theory to our current politics, as to any real-world situation, requires that we establish certain facts by observation, and so we shall. But in accordance with the theory of Cognitive Dissonance that we examined a few weeks ago, Republicans mired in fear and hatred are commonly incapable of recognizing facts as a problem for themselves. For various reasons that we can trace back to the Civil War and beyond, they are even less capable of doing anything about it.

In US politics the Democratic and especially the Progressive platform is cooperation. We want every sector of society to support every other, or as you might have heard from time to time

to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common Defence, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity…

or perhaps, as we used to say,

One Nation, indivisible, with Liberty and Justice for All

or even

to secure these rights, governments are instituted among men, deriving their just powers from the consent of the governed

However, the avowed platform of the Party of No is defection at every possible point on every possible issue. Whatever the question, the answer is to punish someone. For a substantial minority of Republicans this includes shutting down the Federal government permanently, a fever dream that goes back to the Civil War. Right now defection takes the well-known form of the War on Everybody Including Each Other. I documented the various Republican factions that loathe and despise each other in A Republican Bestiary. I don't need to recite the list of issues that Congress is unable to act on, starting with more than 200 bills passed in the House under Nancy Pelosi and killed by the far-too-exploitable filibuster in the Senate. This will continue at least until we take the House again and put through a new anti-defection Voting Rights Act, and reform (all right, nuke) the defector's filibuster on legislation in the Senate. Nor do I need to list the segments of US society who are taken as enemies to be oppressed with an ALL D strategy at every turn in the supposed name of Liberty according to the RWNJs.

Historically, since the beginnings of the Republican Southern Strategy in the Goldwater campaign of 1964, the defectors had an advantage in recruitment, because of various advantages accruing to those who cooperated with them in defecting against others. Goldwater misunderstood the situation, thinking he could run as an extremist, and was crushed, but Nixon's campaign strategist Kevin Philips (who came over to our side much later) made it work with Southern racists and sufficient Dog Whistles about crime, among other things. The Republican plan was to add to its Big Business/Chamber of Commerce/Country Club/Wall Street base (the 1% today) all of the Southern racists who would vote for tax cuts and deregulation in exchange for keeping Blacks down. Then they added the Religious Right and working class Reagan Democrats, and many more targets of defection.

In 2000 there was nobody to add, but election dirty tricks and the Supreme Court were enough for W to be installed. Now even that cannot be made to work any more. Whoever moderate Republicans proposed even the least cooperation with gets a veto from the Tea Parties defecting from those they call RINOs. The Right has nothing left but fear, hatred, and Cognitive Dissonance, as we saw in the unskewing of the polls, leading to the claim of massive but undetectable voter fraud when Romney lost.

What do we do?

For a start, bipartisanship in the relentless ALL C manner in which President Obama used to pursue it, is not the answer. It is far too exploitable. You cannot cooperate with the ALL D rule in any form. Compromise is not the answer when compromisers are RINOs to the base. Reasoning together is not the answer with the unreasoningly unreasonable.

The answer is, of course, More and Better Democrats, that is, retaliative but forgiving cooperators, in the senses defined here. You don't forgive ongoing defection. You forgive them when they stop, and stay ready to retaliate when necessary. The third tournament that Axelrod ran shows us how. This is the ecological tournament where the number of programs following each rule in each round after the first was proportional to that rule's success in prior rounds. Each rule started with the same number of instances, and TIT FOR TAT had only modest advantages in play. Many rules grew or shrank in importance depending on what rules they could exploit or cooperate with in successive rounds, but soon enough the exploitable fell away, followed by the exploiters. All the while TIT FOR TAT and other similar nice, forgiving, retaliatory programs kept on growing.

Now we are obviously not a computer tournament, particularly not one as simple as this. But the idea of increasing the number of instances of effective rules translates to GOTV and electing more effective Democrats, which means those who understand how to cooperate among themselves, and how to defect and effectively punish the ALL Ders using election law and Congressional rules, House or Senate to the full. We have to make the overt and undeniable Republican ALL D strategy the major issue, preferably all at once, and not one issue at a time. But we must at least fight on the particular issues that elicit cooperation in a particular state or district.

Recruiting more cooperators to our cause is most directly done by simple GOTV among those who are currently not playing the game, mostly because Republicans have made it seem not worth playing. We will come back to this topic of Learned Helplessness in a few weeks. The short version is that it can be a lot of work to get people out of it, but once you do they cannot be put back in again. We are getting help from demographics, including immigration and the raising of ever more young cooperators. The other side is helpfully chasing out their own young in droves.

The evidence-based world is now available to the Right, particularly to their children, as never before, and it is eroding their base at a slow but fairly steady pace, enough to throw a serious scare into them. See, for example, The Incredible Shrinking Church, by Frank Page, previously President of the Southern Baptist Convention. The SBC was shrinking so much that in desperation it renounced racism and set out to recruit minorities. This would be to the good, except that they doubled down on bigotry, misogyny, and science denial at the same time.

In sum, the remedy lies in large part in demographics, the new Generation Gap on the Right, and the Democratic ground game of voter registration and GOTV. We need more and better cooperators, and we have to recruit them and teach them how to do it.

There is a chapter in the book on How to Promote Cooperation. It suggests the following practices:

• Enlarge the shadow of the future
• Change the payoffs
• Teach people to care about each other
• Teach reciprocity
• Improve recognition abilities

How do we do that?

The best way to enlarge the shadow of the future in general is to increase interactions where cooperation is an option. This cannot be done with those practicing ALL D, and at the same time minimizing interactions, as with private schools as a reaction to school desegregation, or creating the Right wing radio Noise Machine for those wishing to live in the bubble.

The Internet is our friend here, because we can engage the victims of the Noise Machine at any point where they show any openness, and they can search out real evidence themselves any time they want to. Little Green Footballs is a prime example, a RWNJ Web site that came over to the light in 2007, although it obviously has had trouble shedding its former reputation. Media Matters is the work of former Republican hatchet man David Brock. Kos was once a Reagan Republican. Nixon strategist Kevin Philips and Nixon's White House Counsel John Dean came over. There are others of note. But the young are much more productive targets. We hear from the Barna Group, an Evangelical polling company, that 38% of Christian Millennials fact check sermons on their smart phones.

Organizing is one of the most effective ways of increasing interactions and thus strengthening cooperation. Voter registration and GOTV, currently rapidly increasing in swing states, is a more effective investment of time and money than campaign ads and events, but suffer from the opprobrium given to productive work as opposed to political Conspicuous Consumption on the TV machine.

Axelrod points out that changing the payoffs is one of the most important functions of government, quoting Rousseau, who said that the role of government is to create conditions in which citizens

will be forced to be free.

That's what the Right always calls Tyranny, in opposition to the Liberty they crave to be able to oppress everybody else.

When my mother taught high-school civics, she liked to begin with the statement

Everybody agrees that aristocracy is the best system of government, as long as they and their friends get to be the aristocrats.

Certainly the self-proclaimed Wall Street Masters of the Universe and Southern Aristocracy believe it, as do White Supremacists, Dominionists, and other factions in the Republican Bestiary. Many of those who do not think that way have to work hard on legislation for human rights, financial regulation, and industrial regulation, each a case of unequal power. There are many other such problems that would otherwise be worse than Prisoner's Dilemmas, with no effective incentive for the Malefactors of Great Wealth and the Economic Royalists to cooperate with the rest of us.

Teaching people to care about each other and to cooperate in helping each other is the essence of the Progressive program. Letting people know that you do care is equally important. I once mentioned my intention to see the Malcolm X movie while on a bus, and the driver was astonished. He had evidently never heard a White man express even that much interest in the Black condition.

Teaching reciprocity is a familiar idea. Cooperate with "good" people who cooperate with you, and defect with "bad" people who defect with you. But it is not so simple. What happens when you decide that "bad" people are permanently bad? Why, you get prejudice, of course, and the intention to oppress such people. This leads those people to fear oppression and to react badly in their turn. Thus some Black groups that have rejected any possibility of working with Whites. In real life, as in game design, it is easy to be too pessimistic about the power of cooperation and too cynical about the force of defection.

So how do you deal with ALL D, or with those who form groups to cooperate among themselves and punish some other group regardless of how members of that group behave? This is particularly difficult in the presence of prejudice, the presumption that some group always behaves badly. The Religious Right often goes so far as to call atheists, adherents of evidence-based knowledge, feminists, LGBTs and some others disciples of the Antichrist, and we have heard of many theories about which races are genetically morally inferior.

Well, basically, you form a coalition of cooperators big enough to sustain itself in adversity even as a minority, and then you look for the opportunity to join several such coalitions into a majority that can require the would-be oppressors to respect basic human rights. Thus More and Better Democrats.

On the opposite side, some say that we should always cooperate, on very different religious grounds. Quaker pacifists may be taken as an example. But Quakers do not preach cooperation with evildoers on everything. They also withdraw cooperation, but they are not willing to be violent or vengeful about it. Their strategy worked well within British and Colonial American societies, overcoming intense and sometimes vicious hostility. Quakers were at the heart of the Abolition movement and Women's Suffrage. The American Friends Service Committee is one of the world's great treasures.

An essential point of reciprocity is not trying to beat out other cooperators by taking advantage of their cooperation. Excelling at whatever you choose to do is fine, but not knocking others down in the pursuit of rewards for what you do.

Improving recognition abilities first requires us to have something worth recognizing. Here we are talking about establishing Progressive cooperation and non-exploitability. President Obama presented himself as endlessly exploitable in his first term. Republicans and Blue Dogs set out to take full advantage. He gradually stopped doing that, and began defecting forcefully right after Republicans defected against him, the global economy, major segments of the US population, and foreign enemies whose power and malice they systematically exaggerate. OK, you can't readily exaggerate the malice of ISIS, but you easily can, and many do, with Putin, Iran, Hamas, even North Korea.

Conclusions

Defecting against the enemy is nothing new. Building up cooperative political alliances is nothing new. What is new in Axelrod's analysis is how to cooperate with rivals and sometimes with enemies. The extreme cases in our times that happened after his book came out are the end of Apartheid, the fall of the Soviet Union, and the end of the Troubles in Northern Ireland. The fall of the Soviet Union in turn led to the fall of a multitude of tyrannies backed by the Soviets and the US both, starting immediately and still continuing, as in Ukraine.

Offer to cooperate with your enemies at every turn, but do it in a way that leaves them no doubt of the consequences for defecting. Ending global poverty would be the best form of cooperation I can think of at the moment, although ending Global Warming is up there as a priority. But we know what to do about Global Warming, and in spite of the political posturing from fossil carbon interests and the Right, we are doing it. Right now, we can't even have serious public discussions about ending poverty, the worst mass defection in the world.

We can look forward to the collapses of several more tyrannies and the Republican Party. A Marshall Plan for poor Muslim countries, with a strong focus on Internet availability to support widespread contacts and the formation of friendships and alliances, would eventually undercut and destroy the various insurgencies and terrorist groups. Now that's cooperation.

Further Reading

Axelrod gave an extensive bibliography in his book, and in its successor The Complexity of Cooperation, which I also recommend highly. I only want to pick out a few titles of particular interest. Much more has been published since.

Axelrod has applied his theory to some particular questions in politics in The Complexity of Cooperation, mentioned above. He had taken up similar questions earlier in Conflict of Interest: A Theory of Divergent Goals with Applications to Politics.

Anatol Rapaport was the only contributor to submit TIT FOR TAT to the computer tournaments. That makes him worth looking into as the only contributor not to fall into the traps that Axelrod talked about, of being insufficiently cooperative and forgiving, and too exploitative. His book Prisoner's Dilemma came out long before Axelrod's, in 1960, and greatly influenced Axelrod's approach. Rapaport also wrote Fights, Games, and Debates in 1965.

In biology, I mentioned The Selfish Gene by Richard Dawkins, the best starting point for examples of cooperation and altruism in evolution.

Coming Soon

On Doctor Who the new Doctor (played by Peter Capaldi) and Clara (Jenna Coleman) set off into darkness this next week, and so will we with Stanley Milgram's book Obedience to Authority: An Experimental View, the classic Yale study on getting people to harm others. We will follow that with John Dean's application of Milgram's theory to Republicans in Conservatives Without Conscience, and then come back to the light to consider what psychologists tell us about overcoming Learned Helplessness.

The reading list for this series can be found in the Grokking Republicans Book List Diary. The updates in that Diary contain links to the Diaries so far.