Lecture 10 Racial Disparities in Economic Opportunity

Lecture materials and videos for a course taught at Harvard University entitled “Using Big Data Solve Economic and Social Problems.” This course provides a simple introduction to the work of Opportunity Insights and related topics in a manner that does not require any prior background in Economics or Statistics. The course was taught by Raj Chetty and Gregory Bruich in Spring 2019, and, with an enrollment of 375 students, was one of the largest classes in the university.

[@kubagrabowypywanie7825]

What about the fact that even though black americans end up richer on average in census tracts above poverty line than below, their income is shrunk by high housing expenses?

[@gregoryconnor3134]

This an interesting set of lectures, but the development of polygenic risk scores is dramatically changing the research frontier regarding income mobility. Chetty's research project would be much more enlightening if the next version took account of polygenic index scores.

[@caseybunge658]

This is a set of 18 lectures consisting of great work from Dr Chetty, which I find to be very enlightening as it is. Also, his work appears to be many-fold more influential in the field than 'polygenic index score' research.
Why did you only comment on the Racial Disparities lecture?

[@gregoryconnor3134]

@@caseybunge658 Plomin shows that, within developed countries, genetic variation is much more powerful than variation in shared environment in explaining income and educational outcomes across individuals. Only nonshared environment (which is effectively just noise, and includes innate differences from random embryonic development) has equal or close to the explanatory power of genetic variation. These are excellent lectures (I have only watched a few) but Chetty is missing most of the story. This one struck me because the graphics are so excellent and thought-provoking. The moving chart where the two groups with different ethnic backgrounds are set "equal" in high-income houses and then diverge in the next generation does have the dynamic structure of an underlying genetic explanation, or perhaps not, who knows. Such speculation is not really appropriate.

[@caseybunge658]

​@@gregoryconnor3134 Thanks for the source. I've learned quite a few things from reading Plomin's work.
To elaborate - I assume that Chetty, being an economist, seeks to use big data to answer economic questions. Not to speak for him, but the goal of these lectures doesn't seem to be to explain all variation in income and educational outcomes. The goal seems to be to explain the portion of variation in outcomes that is apparently related to current economic realities, and at least suggesting realistic paths to change those realities. This is, after all, "Using Big Data to Solve Economic and Social Problems." GPSs might not fit in that discussion. To be sure, the genetics of outcomes does seem like an interesting topic for a seminar series - just not this one per se.

[@samstone1532]

@@caseybunge658 nature.com/articles/d41586-018-06784-5
I would say that Plomin's work is highly controversial and should be read with great critique and knowledge of it's limitations.

[@gregoryconnor3134]

@@caseybunge658 My sincere apologies for not replying soon to your thoughtful email. Your response was "lost" in my email inbox subfolders in a subfolder I never check. Sorry about that. The problem with Chetty only looking at economic variables not genetic ones is that it creates a severe missing variables bias in his conclusions. Suppose (just to make the effect clear by example) that Chetty finds that students who do not take algebra in high school have much higher unemployment rates in their age 20s. He does not allow for genetic variables in his model. The conclusion he makes is that high school Algebra classes increase employment rates. Then, we add a polygenic score and show that it predicts both high school algebra enrollment and predicts higher employment rates. Including both algebra enrollment and the polygenic score in an expanded model of employment rates we find that the coefficient for algebra enrollment is zero. Algebra class enrollment has no true impact on employment rates. The earlier empirical conclusion, a la Chetty, is false and due to missing variable bias. That is why it is so important to use the very informative new genetics data.

[@hudsonview91]

The

[@person-ie1fe]

If we wanted to be serious about evidence, we might compare where blacks stood a hundred years after the end of slavery with where they stood after 30 years of the liberal welfare state. In other words, we could compare hard evidence on “the legacy of slavery” with hard evidence on the legacy of liberals. (Thomas Sowell)

[@superfreiheit1]

Maybe its the American black male culture. Gangs and Rap does not create earnings.

[@emilboehme6813]

Maybe it's your racist generalization putting all black males in the same box?

[@davidjackson2131]

Well you're just ignorant and you don't know that the only things that black males were allowed to profit off of were gangsterism and rap music most other industry was closed off or if you became an industry leader you were lynched.