Let me ask a few simple questions to try to understand better.
Let’s say we defeat our enemies in battle and offer them death or slavery and they choose slavery. And they do all the work and we get 90% of the cooperative rewards (10% feeds the slaves). Is this just as long as we all agree to the terms of the social contract? What would make it more just?
Let’s say we have a system where some people are soldiers and some are laborers. The laborers suggest they get 90% because they did all the work. The soldiers suggest they get 90% “or else”? Which is just? What is the more efficient social contract? Is it the point(s) along the spectrum where laborers work and soldiers don’t take everything?
You point to a key question that I have left for a later post. In the present and previous posts, I have considered situations where bargaining power is equal. What about when it is not? The answer is a bit unpleasant to our modern ears: in the long run, what is considered “fair” will reflect the balance of power in society. So, in the examples you give, over time, people’s views about fair practices will reflect differences in bargaining power. This conclusion is unavoidable when you understand fairness norms as a coordination device to reach sustainable agreements. Fairness norms that are not in line with the balance of power would be trampled on by the powerful.
Step back to the Middle Ages in Europe, and people would not dare to think of liberal democratic ideas as achievable or workable. The "right" social order was one in which kings and aristocrats were good Christians who respected their subjects, protected them, and did not impose extortionary taxes.
Saying this does not mean we have to agree with it. By our modern standards, in societies with largely equal bargaining power, these social contracts are unappealing. Understanding that fairness norms track the balance of power helps explain why and how fairness norms have changed over time towards a greater focus towards equal rights and equal consideration.
> You point to a key question that I have left for a later post. In the present and previous posts, I have considered situations where bargaining power is equal.
What does it even mean for bargaining power to be "equal"?
I think that the "fair" think will be negociated. I mean, none of the other parts can afford losing the other parts job. So, they Will discuss until they find something that the two parts consider just.
Appreciate your posts as always, Lionel. What I have still struggled with when it comes to this game theory and bargaining account of morality is the question of scope. We are embedded in a nested hierarchy of groups that each claim our loyalty (i.e., our commitment to cooperate). Much conflict (or need for cooperation) is not between "self" and "group" nor between "my team" and "your team" but between "my team" and the larger organization that contains my team and is trying to coordinate its efforts with other teams.
To borrow a classic (if extreme) example: In joining your nation-state's resistance against the conquering empire (which, after, all, just wants to bring "peace, freedom, justice and security" to the Galaxy), you are cooperating at the nation-state level, "defecting" at the empire level, and "defecting" from your family by putting them in danger. Entire mafia-themed epics are built on these tensions.
Each level has a narrative telling you why that level should command your highest loyalty. What's the calculus that tells you what level to optimize for? Any choice you make constitutes a defection at some level, cooperation at another.
Hi Nathan, indeed there are many groups in society: firms, associations, sporting clubs, cities, political parties, and so on. The logic I describe applies within each group. Each group has social contracts: a modus operandi that makes the group work and attracts sanctions when violated. "Cooperation" sounds positive, but cooperation at one group level can conflict with cooperation in the larger group. Your point about the mafia is exactly that.
So what you describe is not a problem for that framework, it is just an extension. Binmore does not talk about these vertical layers because he thinks coalitional game theory has not progressed enough to say something very rigorous. But others have discussed what you describe, like Nichola Raihani in "The Social Instinct".
Thanks, Lionel! Will check that book out. I understand how I can model the decision space from the perspective of a sub-group vis-à-vis the super-group using the same logic. But how I, as an individual, choose from which perspective in the nested set to make the decision still eludes me.
Thanks for posting this essay. It clarified my thinking on the meaning of social justice significantly. If social justice depends on the social contract that determines who gets what, what happens in a democracy when the people’s representatives in government no longer seek the compromises that will best benefit everyone in the society? And what happens when people in the society no longer understand that achieving social justice requires compromises? What happens when good governance no longer is the objective, and, instead, destroying one’s political enemies becomes the goal? When this happens, Carl Schmitt appears to a better guide than Mr. Binmore.
But Schmitt’s excessive focus on conflict is misguided. Cooperation is everywhere, even in conflicts. There are rules of engagement, conventions to limit the magnitude and scope of the conflict, and so on. So we should not understand the observation that cooperation is everywhere as meaning that everybody is being nice to everybody.
Now, about your precise question. It is one of the topics of my next post. When a society becomes divided and polarised, the outcome is a reduction in the range of social contracts that can be reached. If Alice and Bob stop trusting each other, the size of the grey area of possible social contracts will shrink. Why would Alice agree to compromise with Bob today if she does not trust that he will reciprocate tomorrow?
So all that you describe is nicely explained by the framework I present here.
Thanks for your response. I’m not sure you have the time to read responses to responses, but here’s mine anyway. (And it’s rather lengthy.)
I think that the main goal of politicians is not compromise, but control. In other words, they want to dominate the political process and the structure of government at all levels (city, county, state, and national) so that their constituents always win and the opposition always loses. In the United States, this happens when, for example, the Alices outnumber the Toms by a wide margin. How does this work in practice? The Red-Blue split in the United States is largely a rural versus urban divide. The Blues dominate in the cities and the Reds dominate the rural areas. I live in a Deep Blue City. Only the voices of the Alices matter…the Toms have been marginalized and are mostly irrelevant to what happens politically in the city. In fact, there are no elected Toms on the city council or in the government at the city and county levels. What does this mean in practice? There's no need for Alice to compromise with Tom. Tom either has to put up with the situation or move to another state. So, the debate about policy is between groups of Alices. All Alices believe alike, but some Alices are more extreme in their views than others. So, all Alices want to tax the rich heavily, but some Alices want to tax them at higher rates than others. Alices believe that the way to alleviate homelessness, drug abuse, poverty, and poor student performances in schools is to spend lots of money on government programs designed to address these problems. And when the programs fail to solve the problems, the solution is to spend even more money on the government programs (and on the people who administer the programs). So, income taxes go up, property taxes increase, sales taxes rise, and fees and licenses become more expensive. And deficits balloon. But because there are no Toms in the government to oppose the Alices, the Alices always get their way.
What about the electorate? Why do the voters put up with a situation that costs them more and more money each year without much to show for it? Why indeed? For one, many of the voters benefit from the money being spent on the programs. Government employees, union members (especially public union members), NGO employees, teachers and school administrators, the police and firefighters, social workers, etc. all receive excellent pay and benefits, and these people tend to vote in high numbers.
And then there’s the moral question. (“Although politics is largely about the allocation of material resources and social recognition in the Game of Life, it is often conducted in the Game of Morals. Social groups and their representatives do not merely haggle over money and status directly. They argue about merit, need, dignity, and respect, because these are the moral categories through which claims to a larger share of social resources are made.”) The Alices are convinced that the Toms are not only wrong, but evil. They’re selfish people who only care about making money. They lack empathy for the poor and the downtrodden. And they got where they are in life, not because they earned it, but because of their privilege.
So, the neat system of cooperation and compromise that you describe in your essay breaks down entirely when one political party totally dominates. And that is precisely what has happened in the United States. Heavily urbanized states are totally dominated by Democrats at all levels, and they don’t need to compromise. And because they are entrenched in the government structurally, the system and the policies it generates won’t and can’t change until the well runs dry. (“What can’t go on, won’t.)
P.S. And there’s also gerrymandering and the politicization of the judicial system, which further undermines the rational system of cooperation and compromise that you describe in your essay. It’s a zero-sum game, and to the victors go the spoils.
To put the point differently, the reason humans experience the feeling of love is pretty obviously to help us reproduce and raise children. Yet it is pretty obvious that I can't infer anything about whether I should keep loving my wife or leave her if I find out she is barren from that analysis of the source of the concept.
Similarly, finding out the concept of justice arose to solve social bargaining problems in no way entails anything about what society should be like. You "can't infer ought from is".
At which point why should we accept this social bargaining model in the first place as relevant to the kind of things we think are desirable for society to have?
The problem with saying something like: ohh justice is just a solution to a bargaining problem is that -- even if it is a correct analysis of why we have the concept -- creates no reason why we should want to have it.
Just replace the word justice with "heuristics that have been useful in solving bargaining problems" in any kind of discussion about justice and the right response is: why should I care about that.
It's perfectly philosophically respectable to deny the existence of a concept like social justice. But the *key* property of what people mean when they invoke it is that it is something that is intrinsically good in some sense. When you offer a reductive definition like this you give up the reason anyone bothers to talk about what is just.
Let me ask a few simple questions to try to understand better.
Let’s say we defeat our enemies in battle and offer them death or slavery and they choose slavery. And they do all the work and we get 90% of the cooperative rewards (10% feeds the slaves). Is this just as long as we all agree to the terms of the social contract? What would make it more just?
Let’s say we have a system where some people are soldiers and some are laborers. The laborers suggest they get 90% because they did all the work. The soldiers suggest they get 90% “or else”? Which is just? What is the more efficient social contract? Is it the point(s) along the spectrum where laborers work and soldiers don’t take everything?
Hi Swami,
You point to a key question that I have left for a later post. In the present and previous posts, I have considered situations where bargaining power is equal. What about when it is not? The answer is a bit unpleasant to our modern ears: in the long run, what is considered “fair” will reflect the balance of power in society. So, in the examples you give, over time, people’s views about fair practices will reflect differences in bargaining power. This conclusion is unavoidable when you understand fairness norms as a coordination device to reach sustainable agreements. Fairness norms that are not in line with the balance of power would be trampled on by the powerful.
Step back to the Middle Ages in Europe, and people would not dare to think of liberal democratic ideas as achievable or workable. The "right" social order was one in which kings and aristocrats were good Christians who respected their subjects, protected them, and did not impose extortionary taxes.
Saying this does not mean we have to agree with it. By our modern standards, in societies with largely equal bargaining power, these social contracts are unappealing. Understanding that fairness norms track the balance of power helps explain why and how fairness norms have changed over time towards a greater focus towards equal rights and equal consideration.
> You point to a key question that I have left for a later post. In the present and previous posts, I have considered situations where bargaining power is equal.
What does it even mean for bargaining power to be "equal"?
I think that the "fair" think will be negociated. I mean, none of the other parts can afford losing the other parts job. So, they Will discuss until they find something that the two parts consider just.
Appreciate your posts as always, Lionel. What I have still struggled with when it comes to this game theory and bargaining account of morality is the question of scope. We are embedded in a nested hierarchy of groups that each claim our loyalty (i.e., our commitment to cooperate). Much conflict (or need for cooperation) is not between "self" and "group" nor between "my team" and "your team" but between "my team" and the larger organization that contains my team and is trying to coordinate its efforts with other teams.
To borrow a classic (if extreme) example: In joining your nation-state's resistance against the conquering empire (which, after, all, just wants to bring "peace, freedom, justice and security" to the Galaxy), you are cooperating at the nation-state level, "defecting" at the empire level, and "defecting" from your family by putting them in danger. Entire mafia-themed epics are built on these tensions.
Each level has a narrative telling you why that level should command your highest loyalty. What's the calculus that tells you what level to optimize for? Any choice you make constitutes a defection at some level, cooperation at another.
Hi Nathan, indeed there are many groups in society: firms, associations, sporting clubs, cities, political parties, and so on. The logic I describe applies within each group. Each group has social contracts: a modus operandi that makes the group work and attracts sanctions when violated. "Cooperation" sounds positive, but cooperation at one group level can conflict with cooperation in the larger group. Your point about the mafia is exactly that.
So what you describe is not a problem for that framework, it is just an extension. Binmore does not talk about these vertical layers because he thinks coalitional game theory has not progressed enough to say something very rigorous. But others have discussed what you describe, like Nichola Raihani in "The Social Instinct".
Thanks, Lionel! Will check that book out. I understand how I can model the decision space from the perspective of a sub-group vis-à-vis the super-group using the same logic. But how I, as an individual, choose from which perspective in the nested set to make the decision still eludes me.
Thanks for posting this essay. It clarified my thinking on the meaning of social justice significantly. If social justice depends on the social contract that determines who gets what, what happens in a democracy when the people’s representatives in government no longer seek the compromises that will best benefit everyone in the society? And what happens when people in the society no longer understand that achieving social justice requires compromises? What happens when good governance no longer is the objective, and, instead, destroying one’s political enemies becomes the goal? When this happens, Carl Schmitt appears to a better guide than Mr. Binmore.
I would agree to some extent with Schmitt in a realist critique of a naive view of liberal democracy where people debate to find the collective good. See my post on this: https://www.optimallyirrational.com/p/democracy-without-illusions-a-realist
But Schmitt’s excessive focus on conflict is misguided. Cooperation is everywhere, even in conflicts. There are rules of engagement, conventions to limit the magnitude and scope of the conflict, and so on. So we should not understand the observation that cooperation is everywhere as meaning that everybody is being nice to everybody.
Now, about your precise question. It is one of the topics of my next post. When a society becomes divided and polarised, the outcome is a reduction in the range of social contracts that can be reached. If Alice and Bob stop trusting each other, the size of the grey area of possible social contracts will shrink. Why would Alice agree to compromise with Bob today if she does not trust that he will reciprocate tomorrow?
So all that you describe is nicely explained by the framework I present here.
Thanks for your response. I’m not sure you have the time to read responses to responses, but here’s mine anyway. (And it’s rather lengthy.)
I think that the main goal of politicians is not compromise, but control. In other words, they want to dominate the political process and the structure of government at all levels (city, county, state, and national) so that their constituents always win and the opposition always loses. In the United States, this happens when, for example, the Alices outnumber the Toms by a wide margin. How does this work in practice? The Red-Blue split in the United States is largely a rural versus urban divide. The Blues dominate in the cities and the Reds dominate the rural areas. I live in a Deep Blue City. Only the voices of the Alices matter…the Toms have been marginalized and are mostly irrelevant to what happens politically in the city. In fact, there are no elected Toms on the city council or in the government at the city and county levels. What does this mean in practice? There's no need for Alice to compromise with Tom. Tom either has to put up with the situation or move to another state. So, the debate about policy is between groups of Alices. All Alices believe alike, but some Alices are more extreme in their views than others. So, all Alices want to tax the rich heavily, but some Alices want to tax them at higher rates than others. Alices believe that the way to alleviate homelessness, drug abuse, poverty, and poor student performances in schools is to spend lots of money on government programs designed to address these problems. And when the programs fail to solve the problems, the solution is to spend even more money on the government programs (and on the people who administer the programs). So, income taxes go up, property taxes increase, sales taxes rise, and fees and licenses become more expensive. And deficits balloon. But because there are no Toms in the government to oppose the Alices, the Alices always get their way.
What about the electorate? Why do the voters put up with a situation that costs them more and more money each year without much to show for it? Why indeed? For one, many of the voters benefit from the money being spent on the programs. Government employees, union members (especially public union members), NGO employees, teachers and school administrators, the police and firefighters, social workers, etc. all receive excellent pay and benefits, and these people tend to vote in high numbers.
And then there’s the moral question. (“Although politics is largely about the allocation of material resources and social recognition in the Game of Life, it is often conducted in the Game of Morals. Social groups and their representatives do not merely haggle over money and status directly. They argue about merit, need, dignity, and respect, because these are the moral categories through which claims to a larger share of social resources are made.”) The Alices are convinced that the Toms are not only wrong, but evil. They’re selfish people who only care about making money. They lack empathy for the poor and the downtrodden. And they got where they are in life, not because they earned it, but because of their privilege.
So, the neat system of cooperation and compromise that you describe in your essay breaks down entirely when one political party totally dominates. And that is precisely what has happened in the United States. Heavily urbanized states are totally dominated by Democrats at all levels, and they don’t need to compromise. And because they are entrenched in the government structurally, the system and the policies it generates won’t and can’t change until the well runs dry. (“What can’t go on, won’t.)
P.S. And there’s also gerrymandering and the politicization of the judicial system, which further undermines the rational system of cooperation and compromise that you describe in your essay. It’s a zero-sum game, and to the victors go the spoils.
To put the point differently, the reason humans experience the feeling of love is pretty obviously to help us reproduce and raise children. Yet it is pretty obvious that I can't infer anything about whether I should keep loving my wife or leave her if I find out she is barren from that analysis of the source of the concept.
Similarly, finding out the concept of justice arose to solve social bargaining problems in no way entails anything about what society should be like. You "can't infer ought from is".
At which point why should we accept this social bargaining model in the first place as relevant to the kind of things we think are desirable for society to have?
The problem with saying something like: ohh justice is just a solution to a bargaining problem is that -- even if it is a correct analysis of why we have the concept -- creates no reason why we should want to have it.
Just replace the word justice with "heuristics that have been useful in solving bargaining problems" in any kind of discussion about justice and the right response is: why should I care about that.
It's perfectly philosophically respectable to deny the existence of a concept like social justice. But the *key* property of what people mean when they invoke it is that it is something that is intrinsically good in some sense. When you offer a reductive definition like this you give up the reason anyone bothers to talk about what is just.