What I inferred from your first comment was that it is not irrational to be averse to risky ventures, even if the probabilities seem beneficial.
That is correct. You are not obligated to value X the same amount as a 50% chance of getting 2X, whether X is a unit of money, lifespan, or whatever. But that’s because your utility function does not have to be linear with respect to X. If you say that X is worth 1 util and 3X is worth 2 utils, that’s just another way of saying that X is just as valuable as a 50% chance of getting 3X. A utility function is just a way of encoding both the order of your preferences and your response to risk.
Or to put it another way, the Endowment Effect is not irrational.
No, the Endowment Effect is status quo bias, which is different from risk aversion, and which changes your relative preferences when your assessment of the status quo changes, potentially making people decline deals that, if all added together, would leave them strictly better off, so that still is irrational. There are models of risk aversion which are completely time-symmetric (not dependent on the status quo), like exponential discounting.
Given where this conversation is going, I should clarify that the Endowment Effect does not strictly speaking violate the expected utility axioms. It’s just that most people have a strong intuition that temporary changes in your ownership of resources that get reversed again before you would even get a chance to use the resources cannot possibly matter, and under that assumption, the Endowment Effect is irrational.
I am starting to think the Endowment Effect might be responsible for a lot of the hesitancy to engage in lifespan gambles.
Only partially. Our risk-aversion with respect to our future lifespan has very little to do the Endowment Effect, and can be modeled by perfectly status quo-ignoring exponential discounting.
However, we also have an intuition that once a person has been created, keeping them alive is more valuable than creating them in the first place. In a sense, this is the Endowment Effect, but unlike in the case of material resources, it does not seem obvious that someone continuing to live a certain amount of time should be just as valuable as someone starting to live the same amount of time. Hence, it is possible to value 60 years of future life less than twice as much as 30 years of future life for someone who already exists, but also value creating one person who will live 60 years more than creating two people who will live 30 years each.
That is correct. You are not obligated to value X the same amount as a 50% chance of getting 2X, whether X is a unit of money, lifespan, or whatever. But that’s because your utility function does not have to be linear with respect to X. If you say that X is worth 1 util and 3X is worth 2 utils, that’s just another way of saying that X is just as valuable as a 50% chance of getting 3X. A utility function is just a way of encoding both the order of your preferences and your response to risk.
No, the Endowment Effect is status quo bias, which is different from risk aversion, and which changes your relative preferences when your assessment of the status quo changes, potentially making people decline deals that, if all added together, would leave them strictly better off, so that still is irrational. There are models of risk aversion which are completely time-symmetric (not dependent on the status quo), like exponential discounting.
Given where this conversation is going, I should clarify that the Endowment Effect does not strictly speaking violate the expected utility axioms. It’s just that most people have a strong intuition that temporary changes in your ownership of resources that get reversed again before you would even get a chance to use the resources cannot possibly matter, and under that assumption, the Endowment Effect is irrational.
Only partially. Our risk-aversion with respect to our future lifespan has very little to do the Endowment Effect, and can be modeled by perfectly status quo-ignoring exponential discounting.
However, we also have an intuition that once a person has been created, keeping them alive is more valuable than creating them in the first place. In a sense, this is the Endowment Effect, but unlike in the case of material resources, it does not seem obvious that someone continuing to live a certain amount of time should be just as valuable as someone starting to live the same amount of time. Hence, it is possible to value 60 years of future life less than twice as much as 30 years of future life for someone who already exists, but also value creating one person who will live 60 years more than creating two people who will live 30 years each.