Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> Federal Reserve officials hoped wages would begin rising at today’s 5.1 percent, but economists are increasingly saying the rate might need to fall to 4.9 percent or lower to push wages higher

Whenever I read something like this I have to wonder: do these economists have a server farm running models to backup statements like this or do they just pull it out of their asses?



Well I don't know about the Fed but this is how we do it in England https://en.m.wikipedia.org/wiki/MONIAC_Computer


The best part is that economists always use 2 significative algarisms. And then, when reality sets down it's not rare that they are wrong by an order of magnitude or signal. (No, what it rare is them getting the most significative algarism right.)

No, nowadays I'm going for the explanation that those numbers are dishonest at their inception, and are just formated on the way that will decieve the biggest number of people.


They pull it out of a simple linear model in Excel .

just an educated guess based on my occasional economics study.


I wonder if deep learning could be more accurate for this kind of problem


Deep learning works well when you have a ton of data and don't care about interpreting the parameters of your model. My understanding is this is pretty much the opposite of the typical situation in econ research.


From talking to economic grad students, it appears hard to obtain enough data for typical machine learning approaches. Basically all parameters end up being set from "theory" and "intuition"


Probably but the conversion rate between expert recommendations and policy is already so abysmal it's probably not worth it.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: