Tuesday, April 17, 2018

Hip Hop and Econometrics

Yesterday I gave a lecture on an econometric technique known as instrumental variables regression. I drew upon Professor Scott Cunningham's excellent new book, Causal Inference: The Mixtape.

Scott has generously made this book available as a free PDF download. He has a refreshing and unique perspective on modern econometric methods as well as some inspiring personal stories and maybe best of all, poetry of hip hop lyrics. Matching song lyrics to econometric concepts is fun and pedagogically effective.  In learning foreign languages, I often found song lyrics easier to remember, and more enjoyable to read than other mediums, and I think the same could be true for many econometrics students.
At the moment I am working on a new econometric study of commuting behavior using American Community Survey microdata.  The instrument is an indicator variable denoting that the children in a household have the same gender. (See Grazi et al. for an example of the use of this instrument to study travel behavior.) This made me think of the following lines from "Get By":

Mi abuela raised three daughters all by herself, with no help.
I think about her struggle and I find the strength in myself. 

                                                                    - Talib Kweli

This lyric is a powerful reminder that the people in our data sets have real lives and aren't "just statistics". But at the same time, in the context of instrumental variables regression, the fact that his abuela's children were all female may predict choices she made. For example, in my preliminary results I am finding a small but statistically significant effect where households with different gender siblings have more bedrooms on average than observationally identical households with same gender siblings.


  1. At the beginning of your blog you mentioned instrumental variable regression and in my eyes the meaning of this is finding instruments that are correlated with endogenous regressors but uncorrelated with the error. For example, if the wage of education jointly depends on ability which is not directly observable. This view can relate to a lot economic ways that are brought out in hip hop like you mentioned. A aspect of this blog I found very agreeable is finding song lyrics to be easier to remember and enjoyable and bringing the relevance of the hip hop culture into the real world because a lot of it does have relevance when the lyrics are truly listened to. In my opinion, hip hop is something in society that is often overlooked or looked at in a negative connotation when in reality it can come with a lot of insights and reality. Rappers will speak there truth and incorporate reality, like police brutality etc into there work.
    People in society are often viewed as just a statistic. Another way people wil ltreat there statistic is trying to beat it and not become one. The detail about Kweli's quote and how it was three females was interesting to me and something that I had not thought of. The economic status of his grandmother and the way she carried herself with what was given is how Kweli pushed himself in his own economic/life ways.

  2. You first started off the blog with the term " instrumental variable regression" which is something I am unfamiliar with. However, from the term itself, I can infer that it has to do something with an experiment based on the word "variable". After looking it up, it is "a third variable, Z, used in regression analysis when you have endogenous variables — variables that are influenced by other variables in the model."In other words, this method relates to economics. For example, in this blog, the 3 daughters were raised by their grandma and one of the possible variables to this might had to do with gender. Kweli uses the struggle of his grandma to find motivation.