This touches on what for me is one of the big open questions on what it means to act rationally. I question the common position that the kinds of ‘irrational’ decisions you describe are actually all that irrational. Many such decisions seem to be rational decisions for an agent with a high time preference at the moment of decision. They may seem irrational from the perspective of a future self who looks back on the decisions when dealing with the consequences but I see the problem as more one of conflicting interests between present selves and past/future selves than one strictly of rationality. As the recent post discussed, rationality doesn’t provide goals, it only offers a system for achieving goals. Many apparently irrational decisions are I suspect rational responses to short term goals that conflict with longer term goals.
If I decide to eat a chocolate bar now to satisfy a current craving, I am not really acting irrationally. I have a powerful short term drive to eat chocolate and there is nothing irrational in my actions to satisfy that short term goal. Later on I may look at the scales and regret eating the chocolate but that reflects either a conflict between short term and long term goals or a conflict between the goals of my present self and my past self (really just alternative ways of looking at the same problem). It is not a failure of rationality in terms of short term decision making, it is a problem of incentives not aligning across time frames or between present and future selves. In order to find solutions to such dilemmas it seems more useful to look to micro-economics and the design of incentive structures that align incentives across time scales than to ways to improve the rationality of decisions. The steps I take to acquire chocolate are perfectly rational, the problem is with the conflicts in my incentive structure.
Many such decisions seem to be rational decisions for an agent with a high time preference at the moment of decision.
Emotions can also have a lower time preference than your conscious self. For example, a surge of anger can make you stand up against a bully and win much more than the present confrontation in long term self-respect and respect of others, even if you eventually “lose” this particular conflict. My subconscious is always tracking the intangible “social” terms of my long range utility function, and over the years I’ve come to appreciate that.
A great description, funny how it applies to other emotional acts such as cheating on your spouse (increase reproductive chances while risking comfort of family life). It might be enlightening to think of some emotions as optimizations for the very long term—for you and all your descendants (makes sense as emotions were created by evolution) - and the rational mind as optimizing for the short to medium term..
The irrational aspect is not that there is conflict between the short-term and long-term or that you act based on short-term consequences, but that when you think rationally while contemplating eating the chocolate bar about the short-term, medium-term, and long-term consequences, weighting them all appropriately, you still end up eating the chocolate bar even though at this very moment your careful consideration of the weighted factors led to the conclusion that not eating it is preferable.
This is something I’ve pondered myself. I think you’re at least partly right, but I’m not entirely certain.
Let’s say that the desire you usually feel in this regard is to not gain weight. What if, while experiencing the craving, your current desire to eat a chocolate bar doesn’t reflect a temporary change in your incentive structure, but instead reflects a temporary distortion of your mental map of reality? For example, your certainty that eating the chocolate bar will make you gain weight might have decreased from 80% to 1%. If the truth is that eating the chocolate bar will in fact make you gain weight, you will therefore be less rational while experiencing this craving than before (or after).
I suspect there’s a bit of both going on but I’m fairly sure it’s not as dramatic a discounting as an 80% to 1% change (I realize your numbers were only illustrative of the idea). My feeling based on introspection of the decision making process when making a choice that favours short term gain over the more ‘rational’ longer term choice is that I am still fully aware of the negative consequences, I just discount them heavily.
If there’s one area where my judgements are distorted it is in my estimate of how likely I am to be able to ‘make up’ for present choices in the future. I think this is a fairly universal phenomenon and is also reflective of conflicts between present and future selves—I may eat the chocolate bar and commit my future self to exercise or a healthier eating regime but I am far too trusting of my future self and consistently underestimate his incentive to renege on any commitments I attempt to bind him to in the present.
In my personal history I have an unusually explicit example of present/future self conflict. When I was at university I made short term decisions which I explicitly justified to myself and others on the basis that I was making choices that my future self would have to pay for but that I anticipated my future self being a person who my present self would have no qualms about taking advantage of. I was aware of the fact that political views tend to move further to the right with age, as best expressed by the line “Show me a young conservative and I’ll show you someone with no heart. Show me an old liberal and I’ll show you someone with no brains.” and as a young liberal anticipated an older and wealthier conservative self who might not believe in wealth transfer. By taking out student loans I could commit my future self to a wealth transfer that suited my purposes at the time but that my future self would likely not approve of.
As it turns out, I was at least partially right about where my political views would move (though of course if I met my younger self now I would attempt to point out the many rational reasons why my views now are in fact more correct, and the many ways in which his understanding was overly simplistic). Overall I don’t begrudge my younger self the choices he made however, though that may only be because the commitments did not prove to be overly burdensome.
This touches on what for me is one of the big open questions on what it means to act rationally. I question the common position that the kinds of ‘irrational’ decisions you describe are actually all that irrational. Many such decisions seem to be rational decisions for an agent with a high time preference at the moment of decision. They may seem irrational from the perspective of a future self who looks back on the decisions when dealing with the consequences but I see the problem as more one of conflicting interests between present selves and past/future selves than one strictly of rationality. As the recent post discussed, rationality doesn’t provide goals, it only offers a system for achieving goals. Many apparently irrational decisions are I suspect rational responses to short term goals that conflict with longer term goals.
If I decide to eat a chocolate bar now to satisfy a current craving, I am not really acting irrationally. I have a powerful short term drive to eat chocolate and there is nothing irrational in my actions to satisfy that short term goal. Later on I may look at the scales and regret eating the chocolate but that reflects either a conflict between short term and long term goals or a conflict between the goals of my present self and my past self (really just alternative ways of looking at the same problem). It is not a failure of rationality in terms of short term decision making, it is a problem of incentives not aligning across time frames or between present and future selves. In order to find solutions to such dilemmas it seems more useful to look to micro-economics and the design of incentive structures that align incentives across time scales than to ways to improve the rationality of decisions. The steps I take to acquire chocolate are perfectly rational, the problem is with the conflicts in my incentive structure.
Emotions can also have a lower time preference than your conscious self. For example, a surge of anger can make you stand up against a bully and win much more than the present confrontation in long term self-respect and respect of others, even if you eventually “lose” this particular conflict. My subconscious is always tracking the intangible “social” terms of my long range utility function, and over the years I’ve come to appreciate that.
I’d describe that as a situation where your long-term interests and your very-short-term interests gang up on your short- to medium-term interests.
A great description, funny how it applies to other emotional acts such as cheating on your spouse (increase reproductive chances while risking comfort of family life). It might be enlightening to think of some emotions as optimizations for the very long term—for you and all your descendants (makes sense as emotions were created by evolution) - and the rational mind as optimizing for the short to medium term..
It doesn’t optimize for “you”, it optimizes for the gene that increases the chance of cheating. The “future” has very little “you”.
The irrational aspect is not that there is conflict between the short-term and long-term or that you act based on short-term consequences, but that when you think rationally while contemplating eating the chocolate bar about the short-term, medium-term, and long-term consequences, weighting them all appropriately, you still end up eating the chocolate bar even though at this very moment your careful consideration of the weighted factors led to the conclusion that not eating it is preferable.
This is something I’ve pondered myself. I think you’re at least partly right, but I’m not entirely certain.
Let’s say that the desire you usually feel in this regard is to not gain weight. What if, while experiencing the craving, your current desire to eat a chocolate bar doesn’t reflect a temporary change in your incentive structure, but instead reflects a temporary distortion of your mental map of reality? For example, your certainty that eating the chocolate bar will make you gain weight might have decreased from 80% to 1%. If the truth is that eating the chocolate bar will in fact make you gain weight, you will therefore be less rational while experiencing this craving than before (or after).
I suspect there’s a bit of both going on but I’m fairly sure it’s not as dramatic a discounting as an 80% to 1% change (I realize your numbers were only illustrative of the idea). My feeling based on introspection of the decision making process when making a choice that favours short term gain over the more ‘rational’ longer term choice is that I am still fully aware of the negative consequences, I just discount them heavily.
If there’s one area where my judgements are distorted it is in my estimate of how likely I am to be able to ‘make up’ for present choices in the future. I think this is a fairly universal phenomenon and is also reflective of conflicts between present and future selves—I may eat the chocolate bar and commit my future self to exercise or a healthier eating regime but I am far too trusting of my future self and consistently underestimate his incentive to renege on any commitments I attempt to bind him to in the present.
In my personal history I have an unusually explicit example of present/future self conflict. When I was at university I made short term decisions which I explicitly justified to myself and others on the basis that I was making choices that my future self would have to pay for but that I anticipated my future self being a person who my present self would have no qualms about taking advantage of. I was aware of the fact that political views tend to move further to the right with age, as best expressed by the line “Show me a young conservative and I’ll show you someone with no heart. Show me an old liberal and I’ll show you someone with no brains.” and as a young liberal anticipated an older and wealthier conservative self who might not believe in wealth transfer. By taking out student loans I could commit my future self to a wealth transfer that suited my purposes at the time but that my future self would likely not approve of.
As it turns out, I was at least partially right about where my political views would move (though of course if I met my younger self now I would attempt to point out the many rational reasons why my views now are in fact more correct, and the many ways in which his understanding was overly simplistic). Overall I don’t begrudge my younger self the choices he made however, though that may only be because the commitments did not prove to be overly burdensome.