It's not just hard, it's also political.

The amount of nuance can be overwhelming. CPI doesn't account for the ability of consumers to substitute inferior goods, but GDP doesn't account for the sacrifice involved in doing so. CPI focuses on consumers, while GDP focuses on all sectors. CPI includes the price of imports while GDP does not. The list goes on. Making all of these choices and keeping in mind the consequences for doing so is exhausting, but you shouldn't just pick someone who looks like the most credible person in the room and trust them because the very nuance that you were hoping to avoid allows people to torture the numbers to say anything they want, and the most credible person in the room might not want the same thing you do. GP argued that OP was torturing the numbers and I argue that GP is torturing the numbers. Who do you believe?

Here's a test to avoid getting lost in the wilderness of nuance: "What would Elysium look like in this model?"

I am referring of course to the movie where a small number of rich people live in luxury on a space station while Earth and its population have devolved into a gigantic poverty-ridden slum. In this case, CPI might show that a basic basket of food, clean water, shelter, and education have become prohibitively expensive for most humans, while the GDP measure would argue that most members of the slum can't afford clean water or proper shelter so those things shouldn't be included in the index. Instead, people live in tents and drink barely-filtered water, which are both still affordable, so what's the problem? More generally, any figure expressed in terms of dollars (GDP, value creation, etc) runs the risk of placing low weight on what happens in the slums and high weight on what happens on the space station. The average (unstated subtext: dollar-weighted) person can afford luxurious accommodation in all facets of life and sees nothing but perpetual economic improvement, so again, what's the problem?

Which metric is the right one? That depends entirely on the point you want to make, but the "Elysium test" is a very basic honesty check that catches the more egregious errors.

That was a very informative explanation. Thanks. There was a sentence that sounded really interesting, but which I didn't understand:

> CPI doesn't account for the ability of consumers to substitute inferior goods, but GDP doesn't account for the sacrifice involved in doing so.

Would you be down to unpack how that works? I'm intrigued.

This page has a specific example: https://www.economicsdiscussion.net/differences-between/diff...

Keep in mind that their example chose a very gentle substitution (apples and oranges) while my example chose a very extreme substitution (housing and clean water for tents and unfiltered water). A nontrivial amount of your opinion-formation should include deciding where typical substitutions (e.g. moving back in with parents) actually lie on that scale. Of course, this choice is peanuts compared to the choice of whether or not you lump high income earners in with low income earners. Hence my focus on the Elysium test.

Great link. So a GDP deflator corrects for what our current equilibrium would have cost back then, while the CPI deflator corrects for what our equilibrium from back then would cost now?