Oh, c'mon! Hahaha. True enough about the studies... But 1) the discount rate of System 1 and System 2 are likely very different. And 2) the advice we give others we love and our revealed preferences differ. It's not obvious revealed preference should dominate.
3) We know that gambling follows something of a power law of users and quantities. Supposing that you can't gamble more than you have liquid assets, we shouldn't worry about individual quantities, since wealthy people gambling isn't the issue of concern from a social welfare perspective. But number of bets / period is probably a better indicator of self described bad behavior.
4) Some gambling is totally opaque: slot machines, scratchers. There should be some required disclosures (maybe there are!).
I just don’t think you can infer harm from someone choosing to save less. The indifference curve they are on, if their budget is the same, is necessarily one which dominates the prior curve. And if one seriously follows this logic, you come to absurd conclusions — is a kid who spends all their birthday money on pokemon cards harmed by this? You wouldn’t say that if they had simply received the trading cards and no money, they would be harmed — so why say they’re harmed when they now *demonstrably* prefer it to all other options?
I DID spend all my birthday money on Pokemon cards and regretted it! I should have been forced to ration spending a bit more by my overly permissive parents. So maybe I'm typical minding, and maybe I'm a horrible parternalist - due to low conscientiousness personality traits. Nonetheless, I'm pretty sure being totally illiquid is bad!
Nicholas is correct this hypothetical kid isn't realistically harmed by impulsively spending birthday money. That is because he is under a paternalistic authority providing for his quality of life which has to be *far more coercive* to his autonomy than any government regulating gambling a bit more.
Sebastian's strongest point is about the difference between discount rates for "System 1 vs System 2."
Setting aside whether or not System 1/System 2 is the right way to think about human preferences, the broader point is that people's preferences in reality are much more complex than the simple preferences we write down when we write down indifference curves. Our standard assumptions are probably fine for most things, but I suspect they dont match some important wrinkles when applied to addictive goods ("which goods" is likely an individual-specific property).
For example I don't know if SARP/WARP hold for goods that are addictive. If they don't hold then it is much harder to find utility functions to represent such preferences, and presumably therefore much harder to trace out indifference curves as a practical matter.
Preference theory isn't my specific area of work, but it is extremely interesting in its own right.
If you want to dive into this area further, here's an LLM prompt to get you started, this should provide some reasonable context:
> Can you help me find papers on modeling the preference orderings of addictive versus non-addictive goods? I specifically want academic economics papers on preference orderings, think SARP/WARP.
Whatever list of papers it gives you, plug each one into scholar.google.com and see if each one is real. If one isn't real, you can point this out to the LLM and ask for an alternative. Once you have a list, here is an example follow-up Q to the LLM to help it explain the papers to you:
> I'd love for you to walk me through each paper, and "tell the story" of how the field (as a disaggregated, decentralized entity that doesn't fully agree internally) thinks about addiction. (Essentially this is what an academic economist does when introducing a colleague to a new a literature.)
After the LLM "walks you through" the papers, here is one more prompt to get at empirical exploration of the results:
> Do any of these papers propose or execute any sort of experiment to see if different models of preferences could be distinguished from one another?
Bankruptchy is not just spending more, it's spending so much more you're out of proportion with your income _and cannot deal with that without a bailout_.
Oh, c'mon! Hahaha. True enough about the studies... But 1) the discount rate of System 1 and System 2 are likely very different. And 2) the advice we give others we love and our revealed preferences differ. It's not obvious revealed preference should dominate.
3) We know that gambling follows something of a power law of users and quantities. Supposing that you can't gamble more than you have liquid assets, we shouldn't worry about individual quantities, since wealthy people gambling isn't the issue of concern from a social welfare perspective. But number of bets / period is probably a better indicator of self described bad behavior.
4) Some gambling is totally opaque: slot machines, scratchers. There should be some required disclosures (maybe there are!).
I just don’t think you can infer harm from someone choosing to save less. The indifference curve they are on, if their budget is the same, is necessarily one which dominates the prior curve. And if one seriously follows this logic, you come to absurd conclusions — is a kid who spends all their birthday money on pokemon cards harmed by this? You wouldn’t say that if they had simply received the trading cards and no money, they would be harmed — so why say they’re harmed when they now *demonstrably* prefer it to all other options?
I DID spend all my birthday money on Pokemon cards and regretted it! I should have been forced to ration spending a bit more by my overly permissive parents. So maybe I'm typical minding, and maybe I'm a horrible parternalist - due to low conscientiousness personality traits. Nonetheless, I'm pretty sure being totally illiquid is bad!
Nicholas is correct this hypothetical kid isn't realistically harmed by impulsively spending birthday money. That is because he is under a paternalistic authority providing for his quality of life which has to be *far more coercive* to his autonomy than any government regulating gambling a bit more.
Sebastian's strongest point is about the difference between discount rates for "System 1 vs System 2."
Setting aside whether or not System 1/System 2 is the right way to think about human preferences, the broader point is that people's preferences in reality are much more complex than the simple preferences we write down when we write down indifference curves. Our standard assumptions are probably fine for most things, but I suspect they dont match some important wrinkles when applied to addictive goods ("which goods" is likely an individual-specific property).
For example I don't know if SARP/WARP hold for goods that are addictive. If they don't hold then it is much harder to find utility functions to represent such preferences, and presumably therefore much harder to trace out indifference curves as a practical matter.
Preference theory isn't my specific area of work, but it is extremely interesting in its own right.
If you want to dive into this area further, here's an LLM prompt to get you started, this should provide some reasonable context:
> Can you help me find papers on modeling the preference orderings of addictive versus non-addictive goods? I specifically want academic economics papers on preference orderings, think SARP/WARP.
Whatever list of papers it gives you, plug each one into scholar.google.com and see if each one is real. If one isn't real, you can point this out to the LLM and ask for an alternative. Once you have a list, here is an example follow-up Q to the LLM to help it explain the papers to you:
> I'd love for you to walk me through each paper, and "tell the story" of how the field (as a disaggregated, decentralized entity that doesn't fully agree internally) thinks about addiction. (Essentially this is what an academic economist does when introducing a colleague to a new a literature.)
After the LLM "walks you through" the papers, here is one more prompt to get at empirical exploration of the results:
> Do any of these papers propose or execute any sort of experiment to see if different models of preferences could be distinguished from one another?
...as above, check that papers exist using scholar.google.com
I had good results from these prompts with Claude 3.5 -- "your results may vary" of course!
Bankruptchy is not just spending more, it's spending so much more you're out of proportion with your income _and cannot deal with that without a bailout_.