Wouldn't it be in the beneficiary's interest to behave like a free rider? Rationally, the beneficiary has an interest in maintaining the income of the altruist, but not of the altruistic group (the government for example). In natural selection, the beneficiary has an interest in favouring his own genes but not directly those of the group.
The beneficiary has an interest in the welfare of other beneficiaries of the same altruist because the better off they are the less the altruist will transfer to them, leaving more to transfer to him.
I'm imagining situations where he can do something that produces net gains for the group at some cost to himself. Under Becker altruism he is better off doing it than not doing it.
One thing to note is that the reproductive effects on the benefactor due to his altruism might well be zero on the margin, not always negative. The reason is that success often is a question of “good enough” and the marginal utility of extra resources often goes to zero. So if after a certain threshold more resources don’t have a real impact on reproductive success (say because sickness or violent accidents are random and unpreventable after a certain point) the cost to reproduction is zero from altruism and you don’t expect the genes to be competed out. Depending on the nature of the society, that threshold might be really quite low, if there is say ample food but nothing else to really spend resources on like functional medicine.
Nope. It is very rare for even kings and emperors to have more than a few dozen kids, or even more than a handful. Females have an even harder upper bound. Likewise, how many wealthy people in the last few hundred years had swarms of children?
Typically people have far fewer children than their resources can support.
This is true for other animals too. Octopi for example rarely live for more than one or two years, even in captivity. Their natural environment is so dangerous that despite spending all the resources on intelligence they just die of old age after reproducing once or twice. If one could live another few years they could produce thousands more offspring, but apparently being built to do so only once or twice is good enough, presumably e cause the chances of surviving to do it a third time are so low that being able to live that long isn’t a benefit.
Likewise, human women start having problems with child birth around 40. If they could have kids into their 50’s they could have many more, except that it is very risky to do so for other reasons. The marginal benefit from a few more years of being able to have kids is apparently zero all considered in the eyes of evolution adopting that pattern.
I am not sure I agree with the "evolution is slow for species with a long time between generations" aspect. In many cases, yes, but there is much opportunity for punctuated evolutionary periods where things change quickly. If, for example, a super strain of deadly drug resistant malaria were to spread around the world, we would see very quickly a increase in the prevalence of partial sickle cell genes in the (rapidly decreasing) over all population. Or, to the extent that personality traits are genetic, a particularly bad Maoist or Pol Pot style dictatorship could wipe out entrepreneurial and intelligent in a wide swath of the population in short order, effectively breeding it out of the population. (I think it was Gordon Tullock that made that point about totalitarian nations unnaturally selecting self determination out of the population, but I don't remember.)
The explanation is that the marginal value of one additional child in terms of reproductive success of genes that push for more children goes to zero for reasons other than resources. That's what I was getting at in the previous comment, that's why I brought up octopi.
How many kids one pops out is only one aspect of how likely one's genes are to persist. As the other aspects dominate the returns on more kids decrease.
Imagine a mammal species where the mother raises pups alone for 6 months. She has to go out and get food for herself and then return to feed her babies during this time, going out more frequently the more pups she has at once. Going out is dangerous, not just for her but for her pups because if she dies they all starve.
So even given an unlimited amount of food in the environment, how many pups is optimal depends on how dangerous it is go out and how frequently she will have to go out for a given number of pups. If it is sufficiently dangerous fewer pups makes sense because she can more readily ensure they reach adulthood, but less danger means more pups will reach adult hood. The number of resources available doesn't matter, just the likelihood of coming back and raising the pups at all.
(The function can be U shaped as well; some octopi have thousands of babies at once, but basically starve themselves to death caring for them in one big "get all the reproduction done at once" go.)
In summary, resource availability is only one aspect determining whether your genes get adopted over time. Going all in on resources to the detriment of the other aspects is likely to be a bad strategy past a certain required threshold.
> Imagine a mammal species where the mother raises pups alone for 6 months. She has to go out and get food for herself and then return to feed her babies during this time, going out more frequently the more pups she has at once. Going out is dangerous, not just for her but for her pups because if she dies they all starve.
Except modern people already have way more resources than they use on children.
Sorry, I am trying to explain why there is a point where more resources cannot be spent to make more children in an evolutionarily relevant manner. That's the explanation for why "Typically people have far fewer children than their resources can support."
“If Becker's analysis is correct, genes for altruism should become less frequent over time within groups containing one or more altruists, but the genes of such groups should be becoming more frequent over time; only if the second process at least balances the first will altruism survive.”
I’m not sure what this means. I suppose there is a distribution of genes among the in group, and a different but related distribution among the population in general, and the same phenomenon causes one to tend upward and the other downward. How does that work? If there were two forces at work, I could see how they would come into equilibrium. But it is a single force, so how does that work? The altruist genes become a smaller proportion of a growing sub-population? The local bands with altruists get larger more often/faster than those without, but as they get larger the altruist genes occupy a reduced proportion of the local group, but as they larger proportion of the total population?
If we bring in the factor of reciprocity the calculation becomes more complicated, but probably more favorable toward altruists. In this view, an altruistic act is in part an invitation to scratch my back consisting of scratching yours first. It seems like an empirical question - how long does altruism continue when not reciprocated in any way?
Yep, some people directly value the welfare of other people. But what is value? And what is welfare? Quibbles. Say it like this: the first person is made happy to some degree directly by the other's being happy, and from that we very quickly get to the economist's formulation, that one of the items in the utility function of the first is the utility of the second.
But I have worries: (i) Is it really altruism?—In ordinary parlance, an altruist is prepared to some extent to sacrifice their happiness for the happiness of others, but the first person is not like this, since they always do what makes them most happy; (ii) What happens when two altruists get together?—If my utility function is defined in part by yours, and yours by mine, it seems both are undefined.
Perhaps the solution is to distinguish between all-in utility (= what is revealed by choice, and what we have to this point been talking about), and nontuistic utility (= what your all-in utility would be, ignoring other people's utility). Then we could say (i) the altruist is prepared to sacrifice their /nontuistic/ utility for the /nontuistic/ utility of others, and (ii) one of the items in the /all-in/ utility function of the first is the /nontuistic/ utility of the second. Would that work?
Neither party knows with certainty or accuracy the utility function of the other (not to mention their own). So they use estimates, perhaps after gathering info. Such estimates will not be perfect, but what else could we do? They are usually pretty close to the margin, so perfect accuracy is not required.
The analysis in the post leaves out the repetitional boost the altruist gets by acting generously. This also factors into the equation.
Wouldn't it be in the beneficiary's interest to behave like a free rider? Rationally, the beneficiary has an interest in maintaining the income of the altruist, but not of the altruistic group (the government for example). In natural selection, the beneficiary has an interest in favouring his own genes but not directly those of the group.
The beneficiary has an interest in the welfare of other beneficiaries of the same altruist because the better off they are the less the altruist will transfer to them, leaving more to transfer to him.
Yes, he has an interest in the well-being of the other beneficiaries, but does he have an interest in financing it himself?
I'm imagining situations where he can do something that produces net gains for the group at some cost to himself. Under Becker altruism he is better off doing it than not doing it.
One thing to note is that the reproductive effects on the benefactor due to his altruism might well be zero on the margin, not always negative. The reason is that success often is a question of “good enough” and the marginal utility of extra resources often goes to zero. So if after a certain threshold more resources don’t have a real impact on reproductive success (say because sickness or violent accidents are random and unpreventable after a certain point) the cost to reproduction is zero from altruism and you don’t expect the genes to be competed out. Depending on the nature of the society, that threshold might be really quite low, if there is say ample food but nothing else to really spend resources on like functional medicine.
You can use resources to improve your children's chance of surviving and having children.
Couldn't extra resources always be used to make extra children?
Nope. It is very rare for even kings and emperors to have more than a few dozen kids, or even more than a handful. Females have an even harder upper bound. Likewise, how many wealthy people in the last few hundred years had swarms of children?
Typically people have far fewer children than their resources can support.
This is true for other animals too. Octopi for example rarely live for more than one or two years, even in captivity. Their natural environment is so dangerous that despite spending all the resources on intelligence they just die of old age after reproducing once or twice. If one could live another few years they could produce thousands more offspring, but apparently being built to do so only once or twice is good enough, presumably e cause the chances of surviving to do it a third time are so low that being able to live that long isn’t a benefit.
Likewise, human women start having problems with child birth around 40. If they could have kids into their 50’s they could have many more, except that it is very risky to do so for other reasons. The marginal benefit from a few more years of being able to have kids is apparently zero all considered in the eyes of evolution adopting that pattern.
> Typically people have far fewer children than their resources can support.
And this is something that needs an explanation.
We are not in the environment where we evolved, and evolution is slow for species with a long time between generations.
I am not sure I agree with the "evolution is slow for species with a long time between generations" aspect. In many cases, yes, but there is much opportunity for punctuated evolutionary periods where things change quickly. If, for example, a super strain of deadly drug resistant malaria were to spread around the world, we would see very quickly a increase in the prevalence of partial sickle cell genes in the (rapidly decreasing) over all population. Or, to the extent that personality traits are genetic, a particularly bad Maoist or Pol Pot style dictatorship could wipe out entrepreneurial and intelligent in a wide swath of the population in short order, effectively breeding it out of the population. (I think it was Gordon Tullock that made that point about totalitarian nations unnaturally selecting self determination out of the population, but I don't remember.)
The explanation is that the marginal value of one additional child in terms of reproductive success of genes that push for more children goes to zero for reasons other than resources. That's what I was getting at in the previous comment, that's why I brought up octopi.
How many kids one pops out is only one aspect of how likely one's genes are to persist. As the other aspects dominate the returns on more kids decrease.
Imagine a mammal species where the mother raises pups alone for 6 months. She has to go out and get food for herself and then return to feed her babies during this time, going out more frequently the more pups she has at once. Going out is dangerous, not just for her but for her pups because if she dies they all starve.
So even given an unlimited amount of food in the environment, how many pups is optimal depends on how dangerous it is go out and how frequently she will have to go out for a given number of pups. If it is sufficiently dangerous fewer pups makes sense because she can more readily ensure they reach adulthood, but less danger means more pups will reach adult hood. The number of resources available doesn't matter, just the likelihood of coming back and raising the pups at all.
(The function can be U shaped as well; some octopi have thousands of babies at once, but basically starve themselves to death caring for them in one big "get all the reproduction done at once" go.)
In summary, resource availability is only one aspect determining whether your genes get adopted over time. Going all in on resources to the detriment of the other aspects is likely to be a bad strategy past a certain required threshold.
> Imagine a mammal species where the mother raises pups alone for 6 months. She has to go out and get food for herself and then return to feed her babies during this time, going out more frequently the more pups she has at once. Going out is dangerous, not just for her but for her pups because if she dies they all starve.
Except modern people already have way more resources than they use on children.
Sorry, I am trying to explain why there is a point where more resources cannot be spent to make more children in an evolutionarily relevant manner. That's the explanation for why "Typically people have far fewer children than their resources can support."
“If Becker's analysis is correct, genes for altruism should become less frequent over time within groups containing one or more altruists, but the genes of such groups should be becoming more frequent over time; only if the second process at least balances the first will altruism survive.”
I’m not sure what this means. I suppose there is a distribution of genes among the in group, and a different but related distribution among the population in general, and the same phenomenon causes one to tend upward and the other downward. How does that work? If there were two forces at work, I could see how they would come into equilibrium. But it is a single force, so how does that work? The altruist genes become a smaller proportion of a growing sub-population? The local bands with altruists get larger more often/faster than those without, but as they get larger the altruist genes occupy a reduced proportion of the local group, but as they larger proportion of the total population?
If we bring in the factor of reciprocity the calculation becomes more complicated, but probably more favorable toward altruists. In this view, an altruistic act is in part an invitation to scratch my back consisting of scratching yours first. It seems like an empirical question - how long does altruism continue when not reciprocated in any way?
Very interesting, I wrote a post on 'Empathy Economics' (https://ohmurphy.substack.com/p/empathy-economics?r=bhqc9) a little while ago, but I actually missed Gary Becker's contributions. Seems like a glaring omission now.
Yep, some people directly value the welfare of other people. But what is value? And what is welfare? Quibbles. Say it like this: the first person is made happy to some degree directly by the other's being happy, and from that we very quickly get to the economist's formulation, that one of the items in the utility function of the first is the utility of the second.
But I have worries: (i) Is it really altruism?—In ordinary parlance, an altruist is prepared to some extent to sacrifice their happiness for the happiness of others, but the first person is not like this, since they always do what makes them most happy; (ii) What happens when two altruists get together?—If my utility function is defined in part by yours, and yours by mine, it seems both are undefined.
Perhaps the solution is to distinguish between all-in utility (= what is revealed by choice, and what we have to this point been talking about), and nontuistic utility (= what your all-in utility would be, ignoring other people's utility). Then we could say (i) the altruist is prepared to sacrifice their /nontuistic/ utility for the /nontuistic/ utility of others, and (ii) one of the items in the /all-in/ utility function of the first is the /nontuistic/ utility of the second. Would that work?
Neither party knows with certainty or accuracy the utility function of the other (not to mention their own). So they use estimates, perhaps after gathering info. Such estimates will not be perfect, but what else could we do? They are usually pretty close to the margin, so perfect accuracy is not required.
The analysis in the post leaves out the repetitional boost the altruist gets by acting generously. This also factors into the equation.