>You could reply that mathematical economics shows that economic intuition is often wrong.
The math happens to match your intuition, many people's intuition is different. Correct math is the same among all people. Intuition varies. Math is the evidence that *your* libertarian intuition is correct and people should believe you instead of the intuition of others.
>You could reply that mathematical economics shows that economic intuition is often wrong.
The math happens to match your intuition, many people's intuition is different. Correct math is the same among all people. Intuition varies. Math is the evidence that *your* libertarian intuition is correct and people should believe you instead of the intuition of others.
The problem there is that one can sneakily use math to create evidence of just about whatever you want. As Caplan pointed out, throw in some odd assumptions, for simplification and making the problem tractable of course, and voila it works out the way you want. If it doesn't throw in some different assumptions until it does. Then maybe forget to mention all those assumptions.
Not that everyone doing econ math is doing that, but it is a very easy thing to do, and frankly I think that if your model is halfway tractable you are probably missing out on way too much reality to have a useful model. Useful in the sense that it won't lead you wildly astray when you need it most, that is.
Using math to create "evidence" is harder than just saying "my intuition says so".
At least without the assumptions looking glaringly stupid.
Using maths constrains the space of hypothesis, as it throws out all the things that are inconsistent nonsense. (It still keeps plenty of self consistent nonsense, you need evidence to get rid of that.)
It's harder, but only in that you have to be a bit better at math to do it. There are some amazingly bad assumptions that are just accepted, and many that are just obscured or glossed over. Likewise, while math does narrow the space, often it narrows it to lead to the desired answer, not to lead to reality, the problem being that people mistake "the math works out" with "this matches reality pretty well".
You aren't wrong that applying math does help sort out some "is this even possible?" kind of issues that our intuition might just gloss over, but it doesn't help so much as one might hope, and often clever people will create math models that get the answer they want, intentionally or not.
>The problem there is that one can sneakily use math to create evidence of just about whatever you want.
You can also sneakily use *any method of argumentation* to create evidence of just about whatever you want. In my opinion, math is the hardest method of argumentation to be sneaky with. Any flaws with economath are even bigger flaws in arguing via intuition.
Maybe, but consider how much more credence and deference we pay to math based arguments. Presenting "data driven" arguments is always considered higher status than merely logical argumentation, yet rarely do people ever ask to see the source of the data, or even check that the math makes sense. (Not just in econ, by the way; I see this all the time in industry.) Indeed, people seem just as willing to accept math they don't understand, perhaps even more willing, than a logical or intuitive argument they do. As a result, people are far more likely to accept a silly argument that is counter intuitive but "the numbers don't lie" than the reverse.
Now, I agree that the math has to be fairly internally consistent, but rarely 100% consistency is needed. For example, many people make arguments about wealth inequality then elide into income inequality policies, despite the fact that the one is a stock variable and the other flow. People get caught up checking that the process is consistent and fail to notice that the inputs don't match the output, that one changed topics partway through. Sure, you can do that with just intuition, but it is much more convincing to lie using math.
>You could reply that mathematical economics shows that economic intuition is often wrong.
The math happens to match your intuition, many people's intuition is different. Correct math is the same among all people. Intuition varies. Math is the evidence that *your* libertarian intuition is correct and people should believe you instead of the intuition of others.
Or the other way around.
The problem there is that one can sneakily use math to create evidence of just about whatever you want. As Caplan pointed out, throw in some odd assumptions, for simplification and making the problem tractable of course, and voila it works out the way you want. If it doesn't throw in some different assumptions until it does. Then maybe forget to mention all those assumptions.
Not that everyone doing econ math is doing that, but it is a very easy thing to do, and frankly I think that if your model is halfway tractable you are probably missing out on way too much reality to have a useful model. Useful in the sense that it won't lead you wildly astray when you need it most, that is.
Using math to create "evidence" is harder than just saying "my intuition says so".
At least without the assumptions looking glaringly stupid.
Using maths constrains the space of hypothesis, as it throws out all the things that are inconsistent nonsense. (It still keeps plenty of self consistent nonsense, you need evidence to get rid of that.)
It's harder, but only in that you have to be a bit better at math to do it. There are some amazingly bad assumptions that are just accepted, and many that are just obscured or glossed over. Likewise, while math does narrow the space, often it narrows it to lead to the desired answer, not to lead to reality, the problem being that people mistake "the math works out" with "this matches reality pretty well".
You aren't wrong that applying math does help sort out some "is this even possible?" kind of issues that our intuition might just gloss over, but it doesn't help so much as one might hope, and often clever people will create math models that get the answer they want, intentionally or not.
>The problem there is that one can sneakily use math to create evidence of just about whatever you want.
You can also sneakily use *any method of argumentation* to create evidence of just about whatever you want. In my opinion, math is the hardest method of argumentation to be sneaky with. Any flaws with economath are even bigger flaws in arguing via intuition.
Maybe, but consider how much more credence and deference we pay to math based arguments. Presenting "data driven" arguments is always considered higher status than merely logical argumentation, yet rarely do people ever ask to see the source of the data, or even check that the math makes sense. (Not just in econ, by the way; I see this all the time in industry.) Indeed, people seem just as willing to accept math they don't understand, perhaps even more willing, than a logical or intuitive argument they do. As a result, people are far more likely to accept a silly argument that is counter intuitive but "the numbers don't lie" than the reverse.
Now, I agree that the math has to be fairly internally consistent, but rarely 100% consistency is needed. For example, many people make arguments about wealth inequality then elide into income inequality policies, despite the fact that the one is a stock variable and the other flow. People get caught up checking that the process is consistent and fail to notice that the inputs don't match the output, that one changed topics partway through. Sure, you can do that with just intuition, but it is much more convincing to lie using math.