An Evolutionary Theory of Resource Distribution (Part 2)

A 25% chance. That’s the likelihood that when I tell someone I’m searching for a job, they’ll say:

Remember, Blair … to land a job, it’s not what you know that matters. It’s who you know.

OK, maybe I’m exaggerating this chance. Still, it’s an open secret that when it comes to landing a job, it matters who you know. Many people, it seems, like to remind me of this fact.

A 0% chance. That’s the likelihood that when I tell someone I research income distribution, they’ll say:

Remember, Blair … when it comes to income, it’s not what you know that matters. It’s who you know.

Does this discrepancy strike you as weird? It should. It highlights a blind spot in how we think about the distribution of resources. We all know that our social network matters for landing a job. But once we’ve got the gig, we don’t think about how our relationships determine our income.

But what if we did think about the social nature of income? What would the resulting theory of resource distribution look like? I explore this question here using the evolutionary framework developed in Part 1 of this series.

To recap, my thesis is that humans have a dual nature. We are both selfish and selfless. This duality, I argue, stems from a tension between two levels of natural selection. At the group level, selfless behavior is advantageous. But at the individual level, selfish behavior is advantageous. This tension, I believe, is key to understanding how we distribute resources.

In An Evolutionary Theory of Resource Distribution Part 1, I explored this tension by looking at how groups compete with each other and suppress competition internally. In this post, I look at the same tension from the opposite angle. I discuss how individuals cooperate to build groups, and how this cooperation gets used by individuals for selfish gain.

We’ll journey first to the fantasy world of neoclassical economics to see what’s wrong with mainstream theory. Then we’ll journey to the real world and look at how humans use social relations to distribute resources.

The neoclassical bartender

When we study resource distribution, neoclassical economics is always the elephant in the room. It’s the lumbering theory that, despite many bullet wounds, refuses to die. I’ve previously called the neoclassical theory of distribution a thought virus and an ideology. Here, I’ll treat it as the punchline to a joke.

A janitor and a CEO walk into a neoclassical bar. Envious of the CEO’s exorbitant income, the janitor hits the CEO. A brawl ensues. What does the neoclassical bartender say to stop the fight?

“Stop fighting. You both get paid what you produce.

OK, this punchline isn’t very funny. But it’s the line delivered by neoclassical economist John Bates Clark. At the end of the 19th century, social ferment was in the air. In response to this ferment, Clark developed a theory of income distribution that essentially said to society: “Stop fighting. Everyone gets paid what they produce”. Here’s how Clark put his bartender punchline:

It is the purpose of this work to show that the distribution of the income of society is controlled by a natural law, and that this law, if it worked without friction, would give to every agent of production the amount of wealth which that agent creates.

(John Bates Clark in The Distribution of Wealth)

In one of the great ironies of history, Clark’s punchline managed to become a respectable scientific theory (at least among economists). The punchline goes by the name of ‘marginal productivity theory’. It proposes that in a competitive market, every person receives exactly what they produce.

When he delivered his punchline, Clark’s main interest was the income split between workers and capitalists. Clark wanted to show that capitalists got what their property produced, and hence deserved their income. Only later did neoclassical economists focus on income differences between workers. In the 1960s, theorists like Gary Becker and Jacob Mincer proposed that workers’ income was proportional to their ‘human capital’. This human capital was a stock of skills that made workers more productive.

Soon after it was proposed, however, human capital theory ran into trouble. Ironically, it was human capital pioneer Jacob Mincer who revealed the problem. In his initial work, Mincer defined human capital restrictively as an individual’s years of formal schooling. But Mincer soon found that formal schooling explained very little about individual income. Here’s Mincer admitting the problem:

Simple correlations between earnings and years of schooling are quite weak. Moreover, in multiple regressions when variables correlated with schooling are added, the regression coefficient of schooling is very small.

In response to this empirical failure, many economists doubled down. Instead of abandoning their theory, they broadened their definition of human capital so that it could explain everything and anything about income.

Take, for example, Gregory Mankiw’s bestselling economics textbook, Principles of Microeconomics. In it, Mankiw defines human capital as “the accumulation of investments in people”. With vague definitions like this, human capital theory became immune to evidence. Sadly, neoclassical economists didn’t see this as a problem.

Neoclassical Robinson Crusoe

When it comes to explaining resource distribution, neoclassical theory is missing something obvious. To see what’s missing, we’ll tell another joke.

A neoclassical version of Robinson Crusoe gets stranded on a desert island. How much does his standard of living decrease from before he was stranded?

None. Crusoe took his human capital with him!

Again, this punchline isn’t very funny. But it’s a true representation of human capital theory, which assumes that people carry their income-earning potential around with them. All that matters for workers’ income is their stock of human capital.

Maybe it’s just me, but human capital theory seems to be missing something big. Hmmm .. what is it? Oh, just the rest of society!

Neoclassical theory assumes, quite literally, that the rest of society is irrelevant for determining one’s income. In a competitive market, the theory says that we all get what we produce [1]. If some people produce more than others, it’s because they have more human capital, or own more physical capital. The social context, in other words, is irrelevant to one’s income. Put Robinson Crusoe in London or strand him on an island … it doesn’t matter. His skills stay the same, so his income stays the same.

The message of neoclassical economics is that we’re all self-sufficient Robinson Crusoes — islands unto ourselves. This sounds like an April Fool’s day joke, but it’s not. It’s what passes for science in the discipline of economics. Neoclassical economists have built a towering theoretical edifice on the idea that there’s no such thing as society.

Evolutionary microfoundations

To build a more realistic theory of resource distribution, we need a new ‘microfoundation’ for economics. This is the term economists use to describe their assumptions about human behavior. Most economists assume that humans are purely selfish. But this idea has outlived its usefulness.

A better approach, I believe, is to assume that humans are both selfish and selfless. And we should take a hint from biologists and ground this duality in an evolutionary framework. Yes, I’m arguing that the principles of evolutionary biology should form the microfoundation of economics.

I’ll base my approach on a theory called group selection (sometimes called multilevel selection). According to this theory, the duality of human nature stems from an evolutionary conflict between two ‘levels’ of natural selection. Selfishness stems from selection at the individual level. Altruism stems from selection at the group level.

E.O. Wilson and David Sloan Wilson summarize this tension between individuals and groups in a succinct motto:

Selfishness beats altruism within groups. Altruistic groups beat selfish groups. Everything else is commentary.

I propose that we use this principle as the ‘microfoundation’ of economics. Out with the old assumption that individuals are selfish utility maximizers. In with the evolutionary hypothesis that humans are both selfish and selfless — a duality shaped by the tension between individual versus group benefit.

Relations: the building block of groups

The building block of our evolutionary theory should be the human relation. By forming networks of relations, humans are able to form groups. These relations then determine how resources are distributed within groups.

To develop our ideas, we’ll look at a group of two people. We’ll call them Alice (A) and Bob (B).

group of two people

Figure 1: A group of two people, Alice and Bob

Now we ask ourselves — what kind of relation do Alice and Bob have?

One possibility is that they have a purely altruistic relation. This means that Alice and Bob respect each other’s will. They don’t do anything as a group unless they can both agree on it. We’ll represent this purely altruistic relation using a double-headed arrow. The two heads indicate that influence is symmetrical. Alice influences Bob as much as Bob influences Alice.

Symmetrical Relation

Figure 2: A Purely Altruistic Relation

In the real-world, the closest we get to a purely altruistic relation is probably the bond between two people who are ‘in love’. While we should celebrate love, we should also realize that most human relations are not purely altruistic. Instead, pure altruism is an ideal. It’s one end of the spectrum of human relations.

So what’s the opposite end of the spectrum?

It’s tempting to say that the opposite of the purely altruistic relation would be the purely selfish relation. But the problem with this response is that a purely selfish relation is actually no relation at all. If two people pursue only their own selfish ends, we can hardly call this a relationship. It’s just an aggregate of selfish individuals.

No, the opposite of a purely altruistic relation isn’t a purely selfish relation. It’s a pure power relation. In such a relation, one person acts selfishly while the other person acts altruistically. But in power relations, ‘altruism’ takes a special form. We call it submission. The altruistic person submits to the will of the dominant person.

We’ll represent a pure power relation using a single-headed arrow. The direction of the arrow indicates the direction of influence — the flow of power. In the relation below, Alice has power over Bob.


Figure 3: A Pure Power Relation

In a pure power relation, power is absolute. If Alice says “jump off a cliff”, Bob jumps off a cliff. Like pure altruism, pure power is an ideal. In the real world, the closest thing to a pure power relation is probably the bond between a master and slave. A slave must obey their master, even to their own detriment.

Pure altruism and pure power, then, are idealized relations that define the spectrum of human bonds. We’ll use this spectrum to think about how groups distribute resources.

Dividing the pie

To explore how groups distribute resources, we’ll return to Alice and Bob. Imagine that Alice and Bob are a group that together exploits resources. One day, the two of them find an apple pie. How do they divide it up?

The answer depends on Alice and Bob’s relation.

Let’s first imagine that Alice and Bob treat each other as equals. They have a purely altruistic relation. In this case, the two of them will likely divide the pie equally. Why? Because every decision depends on consensus. If Alice wants to take more resources than Bob, she must convince Bob to give up his share. That’s a hard sell if Bob thinks himself equal to Alice.

The only convincing reason for Alice to take more resources than Bob is if she needs more. Suppose Alice is an endurance athlete and Bob is a couch potato. In this case, Alice needs more food, and Bob is likely to let her have it. So in purely altruistic relations, resource distribution will follow the communist ethos: to each according to their need.

Now let’s imagine that Alice and Bob have a pure power relation. Alice has absolute power over Bob. Now how do they divide the pie?

It’s tempting to say that Alice and Bob would follow what I’ve called the red-claw ethos: to each according to their ability to take (see Part 1 for details). But there’s a problem here. The red-claw ethos is about individual competition — a war of all against all. It’s how resources are distributed in purely selfish relations. But this mutual competition is not how power relations work. Instead, in power relations only one person acts selfishly. The other person acts altruistically by submitting to the will of their dominant partner.

In our example, Bob submits to Alice’s will. This submission is key to understanding how resources get distributed. When Bob submits to Alice, he gives her complete control over the resource pie. Alice could be a despot and hoard everything. Or she could be a benevolent dictator and give Bob his fair share. The choice is hers.

In pure power relations, then, resource distribution is determined by the whim of the dominant individual. Still, there is a regularity to how those with power distribute resources. Even the most selfless individuals inevitably use their influence for personal gain. Power, as they say, corrupts.

Here’s how it happens. Suppose that Alice has absolute power over Bob. Suppose also that Alice is a fervent Marxist, and believes in the communist ethos. So she initially shares resources equally with Bob. So far so good.

But as time goes by, Alice’s power goes to her head. She starts to feel that she’s not like Bob. Because of her power, Alice starts to believe that she’s special. She has innate abilities that Bob doesn’t have — abilities that Bob could never have. And because she has these abilities, Alice thinks to herself, “I deserve more resources than Bob”. And so she takes more resources — a little at first but more over time. Slowly Alice turns from a benevolent dictator to a gluttonous despot. It’s a story as old as time.

The power ethos

The moral of our Alice and Bob story is that people inevitably use their power to enrich themselves. In power relations, then, resource distribution has its own ethos. We’ll call it the power ethos:

Power ethos: To each according to their social influence.

When I’ve discussed the power ethos with mainstream economists, they’ve reacted with bewilderment. The problem, I’ve realized, is that economists reject the idea of a social cause. Instead, they insist that resource distribution must be tied to characteristics of individuals or their property. Philosophers have a name for this thinking. They call it methodological individualism. Geoffrey Brennan and Gordon Tullock summarize how the philosophy works in economics:

[I]n modern economics … the ultimate unit of analysis is always the individual; more aggregative analysis must be regarded as only provisionally legitimate.

Economists are bewildered by the power ethos because power is not a property of the individual. Instead, power is a social relation between people. And trying to understand social relations, it seems, produces error messages in the brains of economists. So to them, the power ethos is incoherent.

But while incoherent to mainstream economists, the power ethos makes perfect sense in our evolutionary theory. In fact, the essence of our theory is that there is a conflict between social (group) goals and individual goals. Let’s look at the power ethos through this evolutionary lens.

Power is a social relation shared by both groups and individuals. At the group level, power is a mode of organization — a way to coordinate human activity. At the individual level, power is a tool for selfish gain — something to be accumulated for personal enrichment.

This duality of power creates a tension between levels of natural selection. Concentrating power may be good for the group but not good for (some) individuals within the group. Likewise, when individuals use their power to enrich themselves, this is good for (some) individuals within the group, but not good for the group as a whole.

Power is a double-edged sword that cuts to the core of our dual nature as humans.

How does centralized power benefit groups?

So why might centralized power benefit groups? Peter Turchin thinks it’s because centralized power allows groups to get bigger. When power has a nested structure (a hierarchical chain of command), it limits the need for social interaction. In a hierarchy, you need to interact only with your direct superior and direct subordinates. This structure, Turchin argues, allows humans to sidestep biological limits in our ability to organize. Centralized power allows group size to grow without increasing the need for social interaction.

If Turchin is correct, it still begs a question. Why are bigger groups better? Turchin’s answer is that big groups have a military advantage over small groups. “Providence”, the saying goes, “is always on the side of the big battalions”. Turchin argues that over the last 10,000 years, large hierarchical groups tended to defeat small egalitarian groups. With each defeat, concentrated power spread as an organizing principle.

Turchin’s argument is compelling, and I agree with most of it. But I think it’s missing one big ingredient — energy. The availability of energy, I believe, is a key constraint on the growth of large groups, and thus a constraint on the concentration of power. I’ll explore this idea in a future post.

Centralized power as convergent evolution?

Humans’ use of centralized power as a coordination tool is not unique in nature. In fact, it may be an example of convergent evolution. Think of the evolution of multicellular animals. As they’ve gotten bigger, animals have all evolved centralized control as an organizing principle inside their bodies.

The human body, for instance, isn’t an aggregate of autonomous cells. Instead, it’s a network of cooperating cells that are controlled by the central nervous system. The cells of the brain, in effect, have power over other cells. When brain cells say “jump”, muscle cells say “how high”.

There is, however, a fundamental difference between cells in the body and individual humans in groups. Body cells don’t use their power for selfish gain. We never catch brain cells using their control of the nervous system to take resources from muscle cells. (If we do, it signals that something has gone very wrong in the body). We don’t see this because body cells are altruistic. So even though the body is centrally controlled, the communist ethos prevails. Each cell gets exactly the resources it needs.

In contrast to the cells in our body, individual humans are not purely altruistic. We may be social animals, but we still have a strong selfish streak. (On the ladder of sociality, humans rank far below the cells in our own bodies). So when human groups centralize power, individuals predictably use their power for personal gain. Instead of the communist ethos, then, we get the power ethos. Each person gets what their social influence allows them to take.

This is why power is a double-edged sword.

The double-edged sword

Concentrated power is no panacea for groups. If it were, we’d all be living in totalitarian regimes. Yes, power is a tool for coordination. But it’s also a tool for despotism. And this despotism can easily undermine the coordinating benefits of power.

Here’s an example. Imagine two large armies meet to do battle. Both armies are the same size and have the same weapons. And both armies are organized using concentrated power. On the surface, these armies appear equally matched. But below the surface, there’s a gaping difference. One army is commanded by a gluttonous despot who keeps his subordinates in rags. We’ll call this the ‘slave army’. The other army is commanded by a benevolent dictator who shares resources equally with his soldiers. We’ll call this the ‘professional army’.

Which army wins the battle?

I’d wager on the professional army. The problem for the slave army is that the leader’s despotism undermines his chain of command. Think about it. Would you put your life on the line for a commander who kept you in rags? I wouldn’t. But I might put my life on the line for a commander who shared resources with me.

The professional army probably has better morale than the slave army, and thus a stronger chain of command. The professional army fights as a unit, keeping the group advantage of centralized power. In contrast, the slave army has a tenuous chain of command. At the first sign of misfortune, the slaves will abandon their despotic leader to whom their allegiance is thin. As the battle rages, the slave army collapses and gets slaughtered.

This is a hypothetical example. But there is real-world evidence that inequality undermines groups’ ability to compete. The evidence comes, not surprisingly, from sports — the modern surrogate for violent conflict. In his book Ultrasociety, Peter Turchin notes that sports teams with more equal pay tend to win more games. Here’s Turchin describing work by Frederick Wiseman and Sangit Chatterjee:

Frederick Wiseman and Sangit Chatterjee sorted the Major League Baseball teams into four payroll classes, ranging from those with the biggest disparities to those with the smallest. Between 1992 and 2001, teams in the most equal class won an average of eight more games per season than those in the most unequal class. The corrosive effect of inequality on cooperation is not limited to baseball. The same effect was observed when researchers analyzed the performance records of soccer teams in Italy and Japan.

(Peter Turchin in Ultrasociety)

If inequality undermines sports teams’ performance, we expect it to do the same among warring groups. The lesson is that when groups concentrate power, they must walk a fine line. They must reap the coordinative benefits of power while avoiding the perils of despotism.

Pathways to power

While power is a double-edged sword for groups, it’s a panacea for the individuals who accumulate it (until their group collapses because of their despotism). Amassing power is a proven way to increase reproductive success [2]. So it’s no wonder that humans (some of us, anyway) have an urge to seek power. If a behavior leads to reproductive success, organisms will develop an urge to do it.

But what exactly do we mean when we say that an individual ‘accumulates power’? Ultimately, my goal here is to develop a quantitative theory of resource distribution. To do this, we need to measure the accumulation of power.

With this measurement in mind, I’m going to discuss three ‘pathways to power’ [3]. These are ways that individuals can increase their social influence within a group.

Pathway to Power 1: Make your subordinates more submissive

One way to increase your power is to make your existing subordinates more submissive. By doing so, you make your power more absolute.

An obvious way to do this is to coerce your subordinates. If I hold a gun to your head, you’ll immediately become more submissive. As totalitarian regimes have discovered, coercion is a good way to make people more obedient.

But while potent, coercion is an expensive way to increase your power. The more you coerce someone, the more they’ll dream of killing you. This is the fear of every despot — that their subordinates will turn on them. So while coercion can exact obedience, it requires constant vigilance. Ignore your coerced subordinates for a moment and you may find a dagger in your back.

A less expensive way to make your subordinates more submissive is to turn to the power of ideas. Convince your subordinates that you have the legitimate right to command them and you immediately increase your power.

What kind of ideas work? Convincing your subordinates that you speak for God seems to do the trick. Convincing your subordinates that you are a God is even better. Regardless of the content of your ideology, what matters is its virulence. To work, your ideology must infect the minds of your subordinates. It must convince them that your power is legitimate.

Whether you use coercion or ideology (or both), making your subordinates more submissive can increase your power. That being said, this approach is not an effective way to accumulate power. Why? Because having absolute power over a few subordinates hardly makes you Napoleon. To achieve great power, you need to become the master of many. You need to accumulate subordinates.

Pathway to Power 2: Accumulate direct subordinates

With Napoleon as your inspiration, you set out to accumulate subordinates. How do you do it? One way is to accumulate direct subordinates.

A direct subordinate is someone who is directly under your control. They listen to you and no one else. Figure 4 shows an example of this pathway to power. Here, our budding despot Alice has accumulated 3 subordinates — Bob (B), Charlotte (C) and David (D). In idealized form Bob, Charlotte and David obey Alice and ignore each other.


Figure 4: Accumulating direct subordinates

While simple, there are obvious limits to this pathway to power. Even the most charismatic person will find it hard to maintain direct power relations with hundreds of people. Yet to be powerful like Napoleon, you need hundreds of thousands of subordinates.

Pathway to Power 3: Accumulate indirect subordinates

If you’re not satisfied with the number of subordinates you can control directly, the next step is to encourage your subordinates to find their own subordinates. By doing so, you accumulate indirect subordinates. Figure 5 shows an example. Here Alice controls Bob, who controls Charlotte, who controls David.

indirect path to power

Figure 5: Accumulating indirect subordinates

We assume here that power is ‘transitive’. So if Bob controls Charlotte and Alice controls Bob, then Alice also controls Charlotte. When power is transitive, it forms a chain of command. By virtue of this chain of command, Alice has 3 subordinates (1 direct and 2 indirect).

The advantage of accumulating indirect subordinates is that you don’t need to manage relations with many people. In our example, Alice directly commands only one person — Bob. To wield power, she gives orders to Bob, who then passes these orders down the chain of command. This is far easier than giving orders to hundreds of people.

The disadvantage of having indirect subordinates, however, is that power isn’t perfectly transitive. Power gets diluted as it’s passed down the chain of command. So when Alice says “jump”, there’s no guarantee that David will get the message. The intermediaries in the chain of command (Bob and Charlotte) have their own agendas which may be different from Alice’s. So having indirect subordinates still requires work. You have to maintain the chain of command so that power flows smoothly.

The road to hierarchy

If you’re a budding Napoleon, you’ve likely realized that the best way to accumulate power is to use all three pathways: make your subordinates extremely submissive, accumulate direct subordinates, and encourage your subordinates to accumulate subordinates.

When you pursue all three pathways to power, what do you get? In a word, hierarchy.

A hierarchy is a network in which every relation is a power relation. To maintain the hierarchy, you must make sure your subordinates are submissive (pathway 1). Next, you encourage all of your subordinates to command multiple subordinates of their own. This combines pathways 2 and 3, and leads to the quintessential feature of a hierarchy — the branching chain of command.


Figure 6: The branching chain of command in a hierarchy

It’s by commanding a growing hierarchy that you can accumulate power that rivals Napoleon’s. Think about the structure of hierarchy. As each new layer of hierarchy is added, the number of subordinates you command grows exponentially.

Suppose you start out as a bit player who controls 4 subordinates. But as a budding Napoleon, you soon attract more people to your cause. You convince each of your subordinates to get 4 subordinates of their own. Now you have 20 subordinates (4 direct + 16 indirect). Repeat this process again and you have 84 subordinates (4 + 16 + 64). Repeat again and you have 340 subordinates (4 + 16 + 64 + 256). It doesn’t take long (about 10 levels of hierarchy) before you have millions of subordinates.

Now you have power that rivals Napoleon’s. And our evolutionary theory predicts that you’ll use it to your advantage. You’ll use your immense power to take an exceptional share of the resource pie.

It’s not what you know that matters …

I began this post by noting that when I discuss my research with friends, they don’t comment that one’s income is about ‘who you know’. This, I argued, is a sign that most people don’t think about the social nature of income. Most of us prefer to believe that our income is about our skill, not our place in society.

But with a bit of evolutionary reasoning, it becomes clear that our relations must be what drives the distribution of income inside groups. Moreover, when groups organize using power relations, we have a clear prediction for how income should be distributed. I’ve called this prediction the power ethos: to each according to their social influence.

Hierarchies, I’ve argued, are the quintessential tool for concentrating power. In them, the power ethos should dominate. In other words, in a hierarchy it’s not what you know that shapes your income. It’s who you control.

Now, I realize that I haven’t given you a shred of evidence that the power ethos is true — that income in hierarchies correlates with social influence. There is evidence, I assure you. But to see it, you’ll have to wait for the next instalment of this series. Stay tuned.

Continue to Part 3


[1] ‘Getting what you produce’. You’ve probably noticed that in the real world, very few people (subsistence farmers aside) literally consume what they produce. Musicians don’t exclusively consume their music. And bakers don’t exclusively consume their bread. In the real world, we exchange many different commodities with one another. Not so in neoclassical theory. Clark’s theory of marginal productivity only works in a world with one commodity. Think of it as the widget world. Everybody makes widgets. And everybody consumes widgets. In this world, everyone literally consumes what they produce, because they all produce the same thing.

[2] The anthropologist Laura Betzig has spent much of her career studying how hierarchical power leads to reproductive success. Check out her book Despotism and Differential Reproduction: A Darwinian View of History.

[3] I’m borrowing the term ‘pathways to power’ from Brian Hayden, who used it as the title of a paper about the origin of inequality. Douglas Price and Gary Feinman later used the phrase as the title of their book on the same topic.

Support this blog

Economics from the Top Down is where I share my ideas for how to create a better economics. If you liked this post, please consider becoming a patron. You’ll help me continue my research, and continue to share it with readers like you.


Stay updated

Sign up to get email updates from this blog.

This work is licensed under a Creative Commons Attribution 4.0 License. You can use/share it anyway you want, provided you attribute it to me (Blair Fix) and link to Economics from the Top Down.


  1. Thanks for these links, Kate. I am very interested in using Odum’s work to study the biophysical properties of human societies. But I disagree with his ideas about monetary value. I think it’s a mistake to connect prices to anything biophysical.

    Yes, ‘human nature’ is definitely controversial. I subscribe to the view that anything humans do is in our nature. I know this is tautological, but at least it’s not wrong!

    But I think it’s useful to at least try to distinguish instinct from learned behaviour (both are ‘natural’). This is, however, notoriously difficult to do.

    Looking forward to reading these papers.

  2. I just read this and Part I on Evonomics. Great work. Very encouraging. I’m an outsider watching the changes in economics, rooting for the science-based and evolution-based approaches.

    I wish you great success in your future. The world needs this work. Thanks!

  3. What you are doing is so important and I’m not sure where this is going to end up but I wish I could convince you that group selection is both wrong and not necessary for your criticism of classical economic theory. Homo economicus is wrong even looking at standard genetic evolution. Our prosocial behaviours evolved the normal way. Group selection is neither necessary to explain them, nor does it explain the type of prosocial behaviour we actually exhibit. We do not act “selflessly” or “altruistically,” if we did we’d be dead. We help and care for others using our *spare* time and energy which as humans, we have plenty. We do not need to use fitness costing energy to help and indeed we don’t. If we did we’d be dead or an incel.

    You can cite the most compassionate prosocial behaviour humans display and you still don’t need group selection to explain them, in fact, considering group selection as their source will cause you to misinterpret them. Our prosocial behaviours are indeed at bottom self interested, but this in no way makes classical capitalism the right plan for us. Quite the opposite. I would be happy to go into more detail but for now please know that your very important work dismantling classical capitalism does not require a belief in group selection to continue and thank goodness for that because group selection is a false theory of human prosocial behaviour.

    Looking forward to following the rest of your theory.

  4. Hi Timothy,

    After writing this post, I realized that I need to define selfishness and altruism better.

    In their book Unto Others, Elliot Sober and David Sloan Wilson define two ways to think about selfishness/altruism: evolutionary vs psychological.

    In evolutionary terms selfishness and altruism have nothing to do with intent. You are selfish if your behavior increases your fitness (i.e. reproduction). You’re altrustic if your behavior increases the fitness of others. Intent never enters into the equation.

    But the fact that humans have psychology makes things more complicated. Psychological selfishness means that you intend to benefit yourself. Psychological altruism means you intend to benefit others.

    Sober and Wilson note that there can be any combination of evolutionary and psychological selfishness/altruism. Economists are concerned only with psychological selfishness. In this post (thinking about it after the fact) I’m mostly interested in evolutionary selfishness/altruism.

    Regarding group selection. My impression is that you’re falling into the trap that many critics of group selection fall into. To see this trap, think about how you define an individual. Biologists accept that evolution acts on individuals. But where do individuals come from? And how do we define them?

    Individuals are a group of cells that have managed to suppress within-group competition enough that we can think of them as a cohesive group of cells that live or die together. How else can you understand the evolution of individuals other than through group selection?

    I’d recommend you read Unto Others to understand the problems with most critiques of group selection.

    To say that group selection is a ‘false’ theory is to say that only genes exist — that no other higher units of organization exist. Evolution acts on cells, and cells are groups. Evolution acts on individuals, and individuals are groups. All biologists accept these facts … they just don’t use the language of group selection. They call cells and individuals ‘vehicles’ of evolution.

    The important question is the relative strength of between-group vs within-group selection. In individuals, within-group competition is almost completely suppressed. So we forget that we’re even talking about a group, and treat the individual as a cohesive whole.

    With human groups, within-group competition is still very strong. But that doesn’t mean that groups don’t exist, or that they’re not an important unit of natural selection.

  5. “You are selfish if your behavior increases your fitness (i.e. reproduction). You’re altrustic if your behavior increases the fitness of others.”

    There is the error right there! Human prosocial behaviour actually increases both your fitness and the fitness of the group. Above you present a false dichotomy which you, no doubt, picked up from Sober and Wilson. The prosocial behaviours that we evolved increase *both* your fitness and the fitness of others. This is precisely why humans are so bloody fit. We have developed this win win scenario whereby helping others helps us individually as well. There is no dichotomy between helping yourself and helping others. It”s the same thing. No altruism involved. Prosocial behaviour = self-interest.

    “Psychological selfishness means that you intend to benefit yourself. Psychological altruism means you intend to benefit others.”

    Again, false dichotomy. Helping others is intended to help yourself. Evolution made sure of that. The suffering of others causes you distress. You feel the urge to help end their suffering to relieve you of your own suffering and also because helping behaviour can get you laid. What you are calling “psychological selfishness” is unwise self-interest, and what you are calling altruistic behaviour is wise self-interest. But it’s all self-interest even if that self is a “group of cells.”

    You can call that group selection if you like because we are a group of cells, but that doesn’t change the fact that our prosocial behaviours are not self-sacrificing, not fitness losing, but rather fitness gaining behaviour.

    I have read Unto Others. It makes the errors I point out above. Human groups do not replicate, so group selection at the human group level does not occur. If it did we would expect to see truly altruistic/sacrificial behaviour in humans but we don’t.

    The good news is that your criticism of classical economic theory does not require group selection theory. Good old fashioned natural selection at the individual human level has created a prosocial creature that does not resemble homo-economicus in any way.

    Truly altruistic sacrificial behaviours do not exist in humans except in extremely anomalous instances. I invite you to give an example of regularly occurring altruistic/sacrificial behaviour in humans if you can. You will likely point to seemingly self-sacrificial behaviour that, upon further examination, is not self sacrificial at all but rather fitness increasing. All human behaviour is self-interested. Again, what you are calling “selfish behaviour” is unwise self-interest and what you are calling “altruistic behaviour” is actually wise self-interest. That’s why it evolved. It increase personal fitness.

    Sober and Wilson’s entire error is based on the mistaken belief that helping others needs to cost personal fitness. The opposite is true. It increases fitness of the individual.

    • Hi Timothy,

      Lots to respond to here. Let me start with what we agree on. We don’t need group selection to critique neoclassical economics. As many people have pointed out, neoclassical theory fails on its own terms.

      Where we disagree. It seems to me that you are not critiquing altruism and group selection so much as defining them out of existence. In particular, you’re committing what Sober and Wilson call the “averaging fallacy”.

      Let’s start with your comment: “Human prosocial behaviour actually increases both your fitness and the fitness of the group.”

      This is perfectly true, but it’s not a critique of group selection. What is good for the group is, on average, good for individuals within the group. This doesn’t mean that group selection doesn’t exist, or that altruism doesn’t exist.

      We can use the same reasoning to define away selection at any level. Take individuals. What is good for individuals is, on average, good for cells within individuals. Therefore there is no selection at the individual level. Only cell selection.

      Take genes. What is good for genes is, on average, good for molecules within genes. Therefore there is no selection at the gene level.

      What is good at a higher level of selection is always, on average, good at every lower level of selection. That’s the whole point. This is a truism, not a critique of group selection.

      Now onto a specific example that illustrates your point. If altruism is beneficial, it will always be a net benefit to both the recipients of altruism and those who perform it. But this benefit happens within the group. The net benefit occurs for individuals within the group relative to individuals in other groups.

      Let’s use war to illustrate this point. When two armies meet, some individuals must be on the front line. Charging on the front line is altruistic because, relative to individuals in your group who don’t charge, you’re increasing your chance of getting killed. Therefore, within the group, altruistic front-line soldiers are less fit than cowards who hide at the back of the army. To be clear, this is relative fitness within the group.

      Now suppose that an army with many altruistic individuals meets an army with mostly selfish individuals. The altruistic army charges as a whole and crushes the selfish army. Now, within the altruistic army, the selfish individuals are more fit than altruistic individuals. Their army wins the battle, yet these selfish individuals don’t risk death on the front lines.

      Still, altruism benefits the altruistic individuals — relative to the selfish individuals in the other army. Again, the altruistic army wins the battle, so all individuals within it benefit.

      To commit the averaging fallacy, we average fitness across all individuals. We find that altruistic individuals are (on average) more fit than selfish individuals. This is because the army composed of selfish individuals was decimated. Looking at average fitness tells us only what evolves. It says nothing about why it evolves.

      Now, you say that human group selection cannot occur because human groups don’t reproduce. This is another fallacy. Groups don’t need to reproduce for selection to occur. What’s needed is for individuals within groups to periodically mix.

      This is just like genetic selection. Genes must periodically mix (within different individuals) for selection to occur. The same with groups.

      Take our altruistic army. If groups never mix, the altruistic army will eventually be composed entirely of selfish individuals. Why? because in every battle, more altruistic individuals are killed than selfish individuals. If the group lasts forever, all the altruistic individuals will get killed.

      To stop this from happening, what’s needed is for groups to mix. Some groups will end up with more altruists than selfish individuals. These groups will win out. Altruism increases in the general population. Mix groups again and repeat. Altruism survives and thrives.

      Regarding psychological altruism, this too you are defining out of existence. You’re adopting the hedonist argument that any form of pleasure seeking is self interest. Helping someone makes me feel good, so it’s masked self interest. This is a valid argument, but it succeeds by defining altruism out of existence. To be altruistic, you need to help people and this needs to make you feel bad. How odd!

      We need to think about why it is pleasurable to help people. I address this here ( so I won’t repeat myself.

      To recap, let’s be clear about our definitions. Evolutionary altruism means decreasing your fitness relative to selfish individuals within your group. If altruism benefits the group, it will benefit all individuals within the group (relative to other groups). In this case, altruism gets selected by group selection. If altruism doesn’t benefit the group, it won’t provide a net benefit to individuals within the group (relative to other groups). In this case, altruism doesn’t get selected because within group competition prevails.

      Regardless of what gets selected, altruistic individuals have lower fitness relative to selfish individuals within their group. Because of this, altruism cannot evolve without group selection.

      • Thanks for the detailed reply, Blair. Let’s start with your final line which I am in complete agreement with.

        “altruism cannot evolve without group selection.”

        Agreed. My argument is that we see no evidence of altruism and therefore no evidence of group selection. If group selection occurred we’d see an abundance of altruistic behaviour but in fact we see virtually none. The only example you gave of the phenomenon is frontline soldiers in war and I’m afraid that example fails on so many levels.

        For starters, soldiers are *sent* to the front lines by authority command. Perhaps you are familiar with the term “cannon fodder.” This is what we (under our breath) call front line soldiers because we know the likely end they will meet. But we don’t call them that to their face! We don’t put that on the sign up sheet! If we did we’d have no front line volunteers. You and I sure as hell wouldn’t sign up to knowingly be cannon fodder. We would need to be tricked by propaganda about glory for us if we survive (prestige) or for our family is we die (kin selection) or we would need to have a gun pointed at our head (force) or we would need to be suicidal or deluded by the belief that we are invincible warriors. Frontline soldiers are far from examples of altruism. They are examples of coercion by force , or victims of propaganda that appeal to the self-interest of glory and prestige, or dumb as fuck warriors with an invincible god complex. I honestly can not think of a worse example of altruism than front line soldiers. I hope you have better examples than that up your sleeve.

        “Now onto a specific example that illustrates your point. If altruism is beneficial, it will always be a net benefit to both the recipients of altruism and those who perform it.”

        This could not more directly contradict the definition of “altruism.” It is only altruism if it costs fitness to the “performer.” If there is a net benefit to the performer it is not altruism by definition. It is prosocial behaviour, but not altruism.

        “But this benefit happens within the group. The net benefit occurs for individuals within the group relative to individuals in other groups.”

        Not true. I help my neighbour build his garage and his sister takes a shining to me and my manly helpfulness and she invites me to put my seed in her. My instinct to help evolved because it was a good strategy for passing on my personal genes. And there is no guarantee that me helping my neighbour will help our group. What if me and my neighbour are badd actors who knowingly or unwitingly harm the group. Even if it does help my group by extension, that has no effect on why I did it. I did it for me. If it helped the group that’s a side bonus. It’s a win win situation. “Altruism” is the antitheses of win win. By definition it requires a loss for the performer.

        “This is a valid argument, but it succeeds by defining altruism out of existence.”

        Agreed, my valid argument means that altruism does not exist. It’s okay for us to have a word for something that doesn’t exist. It doesn’t render the word useless. The word “supernatural” is still useful for example. The words “knowledge” and “free will” refer to something many philosophers don’t think we have. I’m putting “Altruism” in that category. It’s okay that we once mistook self-interested behaviour as self sacrificing fitness costing behaviour but we know better now. We should stop using the word “altruism” in a way that contradicts its own definition. We should use the term “prosocial” in its place because the definition of “prosocial” does not require a fitness loss for the performer and thus is not an incoherent way to describe helpful but self beneficial behaviour.

        “To be altruistic, you need to help people and this needs to make you feel bad. How odd!”

        Right, altruism sounds odd (makes no sense) when you examine it closely. Prosociality on the other hand is win win. You help and it feels good because you are gaining, not losing fitness. Nothing odd about that at all.

        “We need to think about why it is pleasurable to help people.”

        I will read your article on it but it seems quite clear it is pleasurable to help people for the same reason it is pleasurable to eat, drink and procreate. Pleasure is nature’s way of letting you know you are doing something that will help pass on your genes. Good old fashioned evolution by natural selection selecting for self-interested individuals who’s self-interest is best achieved by prosocial behaviour. No group selection required.

        For true altruism, group selection would be necessary to explain it. But we see no evidence of true altruism and from this we can conclude that group selection is not the way in which our prosocial behaviours evolved.

  6. Timothy,

    Sorry for the delayed response. It seems we are mostly splitting hairs about definitions of altruism. You’re criticizing my example of evolutionary altruism using a different definition of altruism than I am.

    Let’s go through the definitions.

    1. Evolutionary altruism. This definition of altruism is only about actions and has nothing to do with intent. You act altruistically if your action lowers your fitness relative to others in your group. Again, it’s your actions that matter, not your intent.

    2. Evolutionary altruism with psychological selfishness. Here, you do something that lowers your fitness relative to others in your group (evolutionary altruism) but your actions are actually self serving (psychological selfishness). Your action gives you pleasure (hedonism) or you think it will benefit you in the future.

    3. Evolutionary altruism with psychological altruism. Here, you do something that lowers your fitness relative to others in your group (evolutionary altruism), and your intent is explicitly to benefit others (psychological altruism).

    Now that we have these definitions, let’s look at my example of frontline soldiers in an army. These soldiers are clearly lowering their fitness relative to the generals at the back of the army (frontline soldiers are far more likely to get killed). So these soldiers are evolutionary altruists. Since frontline soldiers exist in all armies in history, it seems that this is an unassailable example of evolutionary altruism.

    Your response is that this is not an example of altruism. Frontline soldiers, you note, usually don’t want to be there. They know they’re likely to get killed and so they’d rather be somewhere else. They charge on the front lines because they are forced to by the military hierarchy. (On a side note, this is a good point. I think hierarchies excel at making people do things that they would not otherwise do outside the hierarchy).

    Your point is well taken. Most frontline soldiers don’t explicitly want to lower their fitness. But this doesn’t mean they’re not evolutionary altruists because, under this definition of altruism, intent is irrelevant. Only actions matter. So as long as the soldiers follow orders, they’re being evolutionary altruists.

    You criticism is that most (if not all) of the soldiers are psychologically selfish. They are either forced to act in a way that is not in their self interest, or they think that charging on the front lines will benefit them in the long run. This, you say, makes soldiers non-altruists.

    The problem is that this is not a fair criticism of my example, which is explicitly about evolutionary altruism. You’re saying that to be altruistic, you must sacrifice your fitness (evolutionary altruism) and do so with the intent to benefit others (psychological altruism). It’s fine if you want to define altruism this way, but you can’t use this definition to criticize evolutionary altruism.

    The problem of definitions, I think, is a big reason why group selection is controversial. In common language, altruism is all about intent. You have to want to benefit others to be a true altruist. But in group selection theory (indeed, all of evolutionary theory) intent doesn’t matter. Only actions.

    This is a good example of where scientists have taken a term and used it in a different way than in common language. It’s bound to lead to confusion.

    As I understand it, you want to use different terms than Sober and Wilson. You want to reserve ‘altruism’ for fitness-lower actions done with the intent to benefit others. If you lower your fitness with the intent to benefit yourself, you call this ‘pro-social’.

    It’s probably a good idea to give these different concepts different names. But when we criticize theories, we need to be sure we’re using the same definitions. You can’t criticize group selection using your definition of altruism because it’s different than the definition used by theory.

    • Thanks, Blair,

      What you are saying means that if I and some group members are running away from a lion, and I accidentally trip and fall causing the lion to stop chasing others in my group and it kills me instead of them, I have just committed an act of altruism according to your definition of evolutionary altruism. I really don’t think that’s right. Either you are making a mistake with this definition or evolutionary biology is. Honestly I think it’s you. I have never heard of this description of altruism in evolutionary biology. Altruism by force or by accident. It seems antithetical to the word altruism. I think for an action to be altruistic it needs to be intentional. Front line soldiers are not intentionally giving their lives for others. They are forced and also hoping to survive and be heroes. There is no intention to be altruistic, evolutionarily speaking or psychologically speaking.

      Also, psychological altruism does not exist in humans as far as I can tell, other than at an anomalous level. What we call “extreme altruists” are the only examples of actual altruism IMO and this phenomenon is as anomalous as psychopathy. Psychological altruism is precisely the kind of behaviour we would expect to see if group selection occurred as theorized, but we don’t see that kind of behaviour which is why I and others are extremely skeptical of GS. Can you give a real world example of psychological altruism?

      Take your time. Delayed responses are understandable and don’t bother me. I appreciate the engagement.

  7. Tim,

    Your example is interesting, but can be used to make selfishness too seem ill defined. Continuing your example, suppose you and your friends are running from a lion. You accidentally fall in a deep hole. The lion runs past you and catches your friends, eating them all. You increased your fitness (relative to your friends) so this act is “selfish”. It’s “selfish” in the same way that tripping and getting eaten is “altruistic” … defined purely in terms of outcomes, not intent.

    This is obviously ridiculous, as your example is intended to show. But not, I think, for the reasons you intend. It is not ridiculous because conscious intention matters when defining evolutionary altruism (or selfishness). It is ridiculous because it can’t be part of any evolutionary theory.

    Theories of evolution are not concerned with freak accidents. They are concerned with heritable traits. I can’t see how “accidentally” falling (into a hole, or into the lion’s mouth) when running from a lion could be a heritable trait.

  8. Agreed. Can you explain what heritable trait is selected for when a person is forced by warring uncaring authoritarians to charge into certain death?

    And just a reminder, I also asked you for an example of psychological altruism in humans. Did you provide one and I missed it?

  9. “What heritable trait is selected for when a person is forced by warring uncaring authoritarians to charge into certain death?”

    Answer: submissiveness, deference to authority, willingness to believe in an ideology

    The problem here is the word “force”. Rarely is pure force used to motivate humans. Think about how hard it would be to make someone do something that they are absolutely and completely determined not to do. It’s almost impossible.

    It’s like trying to make a wild animal do want you want. You can’t do it … they simply don’t have submissive traits.

    Humans, however, do have submissive traits. With just a little bit of coercion, you can get humans to submit in a hierarchy — charge into battle to their certain death. With a dose of ideology, this is even easier.

    There’s good evidence that since the dawn of agriculture, humans have “domesticated ourselves”.

    This fits in with selection for submissiveness.

    “And just a reminder, I also asked you for an example of psychological altruism in humans. Did you provide one and I missed it?”

    I did not provide one in the last response. As I have said before, if you define as selfish any act that gives you pleasure, then there is no such thing as psychological altruism. We clearly derive pleasure from helping others.

    However, I don’t subscribe to hedonistic notions of selfishness. As a teacher, I personally like helping other people. I do it because I find it enjoyable. If you want to call this selfishness, fine. But I think of it as altruism, because my intent is to benefit other people. I could be making much more money working in the banking sector. But I don’t do this because I think the banking sector is parasitic.

    What you are trying to do is get me to agree with your definition of altruism. As you define it, altruism doesn’t exist. But I don’t agree, with your definition. As a result, I see altruism everywhere — anytime one person helps another person.

    That being said, I find the whole intend side of this uninteresting. The utilitarianism pursued by neoclassical economics has got to be one of the biggest dead ends in science.

  10. Hi Blair,

    “Answer: submissiveness, deference to authority, willingness to believe in an ideology”

    Weird. How would these qualities be selected for? Those people all died and did not pass on their genes, so submissiveness, deference to authority, or willingness to believe in ideology is not selected for. Quite the opposite.The warring authoritarians who sent them to war lived on and procreated. So it is aggression, Authoritarianism, crafting ideologies of control, these are the qualities that would be selected for as they are the traits who’s possessors survived and procreated. I’m guessing politicians and generals have more offspring than cannon fodder.

    “The problem here is the word “force”. Rarely is pure force used to motivate humans.’

    We are talking about a specific situation. Conscription is force. Fight or go to jail is force. The common quality shared by front line soldiers is not submissiveness but rather low wealth. Front line soldiers could be anything from psychopathic to extreme altruists and everything in between. These qualities would be equally selected for because draft boards do not discern by personality trait and neither do the enemy’s bombs and bullets. This is such a weird argument it’s hard to believe anyone with intelligence would attempt it. Sending random poor people to die on the front lines does not select for any personality traits except maybe those that make you rich.

    “As a teacher, I personally like helping other people. I do it because I find it enjoyable. If you want to call this selfishness, fine. But I think of it as altruism, because my intent is to benefit other people.”

    I don’t call it selfishness. I call it self-interest because that’s what it is. You only intend to help others because it brings you joy. If it didn’t you would not help others. That’s how evolution designed us. That is why Homo sapiens are the fittest. Because evolution designed us to get personal individual satisfaction by sharing and cooperating. If it didn’t design us to gain personal satisfaction and pleasure from those things we would not do them.

    “What you are trying to do is get me to agree with your definition of altruism.”

    No I am using the dictionary and biology definitions of the term. Yours is the private language definition.

    “As you define it, altruism doesn’t exist.”

    As the dictionary and biology textbooks define altruism it does not exist in humans. You are the one changing the definition to save the term.

    “As a result, I see altruism everywhere — anytime one person helps another person.”

    There is no such definition of altruism. That is your own private definition. (You’re not alone of course others make the same mistake but it’s not in the dictionary or in the textbooks.)

    “That being said, I find the whole intend side of this uninteresting.”

    I find nothing interesting but the intention. The intention is what gets selected for not the accident or forced actions.

    “The utilitarianism pursued by neoclassical economics has got to be one of the biggest dead ends in science.”

    On this we agree! But group selection offers us nothing to combat this because there is no evidence it occurred in humans. In eusocial insects yes. In Humans no.

  11. […] Although power relations are one sided, they can (under the right circumstances) benefit both the master and servant. This benefit happens when groups compete. By concentrating power in a single leader, a large group can act cohesively in a way that would otherwise be impossible. If this cohesive group beats its competitors, the altruism of subordinates is rewarded. For an in-depth discussion of this principle, see Peter Turchin’s book Ultrasociety. (The caveat here is that a despotic leader can use their power to hoard resources, wiping out any benefits to the rank and file. That’s something I discuss here and here.) […]

  12. […] Although power relations are one sided, they can (under the right circumstances) benefit both the master and servant. This benefit happens when groups compete. By concentrating power in a single leader, a large group can act cohesively in a way that would otherwise be impossible. If this cohesive group beats its competitors, the altruism of subordinates is rewarded. For an in-depth discussion of this principle, see Peter Turchin’s book Ultrasociety. (The caveat here is that a despotic leader can use their power to hoard resources, wiping out any benefits to the rank and file. That’s something I discuss here and here.) […]

Leave a Reply