Currently being further radicalized against science by economists gushing over this post, and pretending that these graphs convincingly show a strong effect of personal insecurity on black patenting, instead of making it clear that it's probably just noise. 
https://twitter.com/paulmromer/status/1488938625358307331
I don't even get what the story is supposed to be here: the graph shows that black patenting was highest when lynching against blacks was highest and stayed low after lynching violence receded, so what the fuck do people think it shows?
Sure, you can tell a story in which once anti-black violence has peaked around 1900, it had a large effect on black patenting that persisted long after lynching became much rarer, but this hardly seems to be the most obvious interpretation of the data.
In my opinion, to the extent these graphs show a real effect (which I doubt they do), a far more plausible interpretation would be that, for instance, Jim Crow laws (that only became the norm around 1900) made it harder for blacks to go to decent schools or to apply for patents.
But I don't even think it shows anything other than noise. I mean just look at the scale for black patenting on the right: given black population around 1900, aren't we just talking about a deficit of something like 10-20 patents/year?
Imagine how many things beside the alleged lack of personal security -> less patenting mechanism alleged by the paper could have produced these data!
Looking at white patenting, it's clear that there is a lot of variability in patenting rate even at much higher rate, so whatever causes this variability could easily have produced this pattern for black patenting independently of the personal security channel.
Hell, with such a low number of patents, given that some people are probably responsible for more than one, it's possible that even just 2-3 black guys who suddenly decided to pass as white or whatever could have produced this pattern 🙃
Looking at the paper's abstract, Cook refers to segregation laws, so perhaps her story is closer to the one I suggest above than to what the graphs in Romer's post show. But she also talks about the effect of race riots and it's hard to see any on the second graph.
I haven't read the paper yet, so maybe Cook makes a better case than Romer does on her behalf, but looking at the graphs in Romer's post I doubt it can be a good case and economists are reacting to Romer's post anyway and often screenshot those graphs as if they were compelling!
These graphs are like a Rorschach test except there is no objectively correct answer to a Rorschach test, whereas in this case there is and it's not the one economists — those who comment on this anyway, I'm sure many think the same thing as me but stay quiet — have picked 🤷♂️
This episode may be even more embarrassing than the "give money to mothers -> change babies's brainwaves" fiasco to be honest. At least in the latter case, that it didn't show anything was quickly acknowledged by most, whereas in this case I doubt it's ever going to happen.
Yet I'm sure that a lot of economists have thought the same thing as me when they read Romer's post, but they're never going to say it publicly, because they don't want to appear to criticize a black woman. I for one am an equal opportunity anti-science guy though 😎
Lol, if you use the data in the first graph, which actually seem more relevant (
https://twitter.com/BarneyFlames/status/1489277899748298759), to compute the "missing black patents" after 1900, it's closer to 5/year than 20/year. But we're supposed to believe this has massive significance? What a fucking joke 😂
I have said it before and I'll say it again, because it cannot be said often enough: science is fake and we must wage the war against it until total victory has been achieved.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
