Cook = fraud? Cook = incompetence? Take your pick. But why is she on the FED board? Easy! AA.
Concerns about Lisa Cook’s research
-
OK, when someone refers to Heckman selection as a 'sophisticated statistical method', it becomes abundantly clear that this board has lost all its economists.
Has anyone replicated Lisa Cook’s research? Reading several of her papers raises red flags.
For example, in her paper on PPP loans, she argues that blacks had less access than whites because of discrimination by banks. However, in the PPP loan data, self-identified race is only recorded in 10% of observations. She claims to use a sophisticated statistical method (“a Mills Ratio derived from a first-stage Heckman selection approach”) to account for selection bias. Given that data, her principle evidence is a reg x, y of “Black” on “log loan amount.”
Similarly, in her work on the affect of racism on black patent issuance, I don’t think that patents actually record the inventor’s race! She also claims to use a sophisticated statistical method to infer the patent owner’s race.
I’m concerned that there might be academic fraud. -
OK, when someone refers to Heckman selection as a 'sophisticated statistical method', it becomes abundantly clear that this board has lost all its economists.
Has anyone replicated Lisa Cook’s research? Reading several of her papers raises red flags.
For example, in her paper on PPP loans, she argues that blacks had less access than whites because of discrimination by banks. However, in the PPP loan data, self-identified race is only recorded in 10% of observations. She claims to use a sophisticated statistical method (“a Mills Ratio derived from a first-stage Heckman selection approach”) to account for selection bias. Given that data, her principle evidence is a reg x, y of “Black” on “log loan amount.”
Similarly, in her work on the affect of racism on black patent issuance, I don’t think that patents actually record the inventor’s race! She also claims to use a sophisticated statistical method to infer the patent owner’s race.
I’m concerned that there might be academic fraud.upvote
-
Where's this data analyst blog that looked into the lynchings and found the results are driven by 1-2 lynchings?
If you're too brain damaged to find it via google searching, and if Karl's investigative journalism skills aren't up to the task, then it appears to be lost.
(phuking rtards...)
-
https://michaelwiebe.com/blog/
February 25, 2021"This seems like a serious problem for the riot results. The paper isn’t estimating the effect of riots on patenting; instead, it’s doing the effect of five specific riots. If we could collect data on the remaining 30 riots, I’d expect the estimate to change. In other words, why should we expect this result to be externally valid for other historical riots?"
This feels like a smoking gun.
-
There was a thread with a link to a website that found problems with the patents paper from several months back.
There is no race data in patent applications, but I think she matched it to decennial census, which has. Of course, this requires name matching, which is not trivial, and can be prone to error.This matching stuff is known to be really sketchy. Mostly you cannot match more than 30-35 percent of the people. The folks who do this assume that the probability of matching is orthogonal to anything else, which is silly. Martha Bailey has a nice paper in the JEL showing this problem. Of course, the NBER/DAE mafia don't like it because it means Princess Leah cannot publish another 20 bogus papers.
TLDR: What Cook does is known to be wrong, but really important people do it. If you question it, you are worse than racist: you are challenging the sacred zipcode
-
There was a thread with a link to a website that found problems with the patents paper from several months back.
There is no race data in patent applications, but I think she matched it to decennial census, which has. Of course, this requires name matching, which is not trivial, and can be prone to error.This matching stuff is known to be really sketchy. Mostly you cannot match more than 30-35 percent of the people. The folks who do this assume that the probability of matching is orthogonal to anything else, which is silly. Martha Bailey has a nice paper in the JEL showing this problem. Of course, the NBER/DAE mafia don't like it because it means Princess Leah cannot publish another 20 bogus papers.
TLDR: What Cook does is known to be wrong, but really important people do it. If you question it, you are worse than racist: you are challenging the sacred zipcodeIf you say so!
-
https://michaelwiebe.com/blog/
February 25, 2021"This seems like a serious problem for the riot results. The paper isn’t estimating the effect of riots on patenting; instead, it’s doing the effect of five specific riots. If we could collect data on the remaining 30 riots, I’d expect the estimate to change. In other words, why should we expect this result to be externally valid for other historical riots?"
This feels like a smoking gun.The research design of the paper sounds very shoddy, but what do you expect from a paper in JEG? Never ascribe to malice that which is adequately explained by incompetence.
-
Why you guys are going along with a miserable failure who can hardly string two sentences together just because he's mean to the right colored people is beyond me. Imagine thinking a PhD dropout foreigner is a man worth following.
because monetary policy is more important than partisan sniping. She is very clearly a terrible pick. Compare her to Yellen for instance.
-
Why you guys are going along with a miserable failure who can hardly string two sentences together just because he's mean to the right colored people is beyond me. Imagine thinking a PhD dropout foreigner is a man worth following.
because monetary policy is more important than partisan sniping. She is very clearly a terrible pick. Compare her to Yellen for instance.
its ok man, you don't have to try to impress me. We can be honest here.