> More to the point, she said that the statistics used for almost all mainstream economics is flat-out wrong.
As someone who has worked extensively with both PhD statisticians and economists, but practices neither beyond what I've learned via osmosis, I couldn't care less what the statisticians think anymore. They were definitely smart, but completely useless. We had several projects where the economists came up with solutions that measurably and objectively worked to the order of 9 figure changes in the bottom line and reversals of decades-long trends...and they did it using data sets that the statisticians wouldn't even touch due to some religious moral panic from observational data. Statisticians always found some way to object to everything under grounds of being proper in some way or another, and rarely came to the table with anything more useful than "First we need to find some way to create a perfect alternate universe".
Econometrics has very rightly diverged from statistics. Economists do not have the luxury of the theoretical environments that statisticians require but never have to create. And as a social science, they don't have the luxury of measuring outputs of perfectly understood processes, but rather the ever changing output of cultures and personalities guided by 3.5 billion years of evolution. Maybe their ideas offend statisticians, but it's pretty hard to find something that exists in the real world that doesn't offend them, so I'm not concerned.
I don't know about your industry but the abiding memory i have from before the financial crisis of 2009 was of economists fanning the property bubble with the mantra ' the fundamental are sound' and how the worst scanario would be a "soft landing"
The main problem with economics is that it's hopelessly corrupted by political and financial interests, in a way that hard sciences only intermittently have to deal with, except with stuff like global warming.
When people wave around the conclusions of scientific papers as proof that some political action is required which is going to impact the lives of millions of people, you need to seriously take the papers with a grain of salt.
I don't really know any way to get around it. Any time you're studying the impact of human collective activity the stakes are far too high to trust that anyone is taking a disinterested view of reality.
To support your point, the economist I can most clearly remember and who seemed to be on tv every other day, was chief economist with a bank that later needed rescuing.
> I don't know about your industry but the abiding memory i have from before the financial crisis of 2009 was of economists fanning the property bubble with the mantra ' the fundamental are sound' and how the worst scanario would be a "soft landing"
As someone who has worked extensively with both PhD statisticians and economists, but practices neither beyond what I've learned via osmosis, I couldn't care less what the statisticians think anymore. They were definitely smart, but completely useless. We had several projects where the economists came up with solutions that measurably and objectively worked to the order of 9 figure changes in the bottom line and reversals of decades-long trends...and they did it using data sets that the statisticians wouldn't even touch due to some religious moral panic from observational data. Statisticians always found some way to object to everything under grounds of being proper in some way or another, and rarely came to the table with anything more useful than "First we need to find some way to create a perfect alternate universe".
Econometrics has very rightly diverged from statistics. Economists do not have the luxury of the theoretical environments that statisticians require but never have to create. And as a social science, they don't have the luxury of measuring outputs of perfectly understood processes, but rather the ever changing output of cultures and personalities guided by 3.5 billion years of evolution. Maybe their ideas offend statisticians, but it's pretty hard to find something that exists in the real world that doesn't offend them, so I'm not concerned.