search

ijyliu / ECMA-31330-Project

Econometrics and Machine Learning Group Project
2 stars 1 forks source link

Cool Ideas Without a Method #4

Closed ijyliu closed 3 years ago

ijyliu commented 3 years ago

Relationship between the number of restaurants in airports, etc. and prices

Relationship between number of college administrators and student outcomes

paul-opheim commented 3 years ago

An idea I've found interesting recently is the relationship between population growth and economic growth (Charles Jones had a paper last year on this: https://web.stanford.edu/~chadj/emptyplanet.pdf). Doing a thorough, convincing study of this would probably be too difficult and time-consuming for this class, but maybe we could do a first-pass analysis looking at how working age population growth seems to relate to per capita economic output?

paul-opheim commented 3 years ago

Joseph Henrich talks about "The Collective Brain":

"One way to capture this is to imagine how many other minds a person in a city could—in principle—access in order to learn from or collaborate with. To get at this, we assume each person can access everyone in their own city, and then we weight the populations of other cities by the cost of traveling to them. So, if you live in a populous city that is tightly connected to surrounding cities by cheap transportation, you are part of a larger collective brain—even if you yourself don’t travel." - The WEIRDest People on Earth

Maybe we could do something with the "collective brain" size and patent applications or something? We could maybe focus on the US and take into account population growth and advances in physical transportation and electronic communication over time to represent how the collective brain size changes over time. This sounds sort of hard lol.

ijyliu commented 3 years ago

Joseph Henrich talks about "The Collective Brain":

"One way to capture this is to imagine how many other minds a person in a city could—in principle—access in order to learn from or collaborate with. To get at this, we assume each person can access everyone in their own city, and then we weight the populations of other cities by the cost of traveling to them. So, if you live in a populous city that is tightly connected to surrounding cities by cheap transportation, you are part of a larger collective brain—even if you yourself don’t travel." - The WEIRDest People on Earth

Maybe we could do something with the "collective brain" size and patent applications or something? We could maybe focus on the US and take into account population growth and advances in physical transportation and electronic communication over time to represent how the collective brain size changes over time. This sounds sort of hard lol.

Reminds me of work on patents and spillovers which I think is mentioned in the notes. Though that isn't geographic, think it was between firms in the same industry. But yeah there is definitely work on this out there

ijyliu commented 3 years ago

An idea I've found interesting recently is the relationship between population growth and economic growth (Charles Jones had a paper last year on this: https://web.stanford.edu/~chadj/emptyplanet.pdf). Doing a thorough, convincing study of this would probably be too difficult and time-consuming for this class, but maybe we could do a first-pass analysis looking at how working age population growth seems to relate to per capita economic output?

Lasso with tons of variables in growth regressions with population as a covariate we never exclude?

paul-opheim commented 3 years ago

An idea I've found interesting recently is the relationship between population growth and economic growth (Charles Jones had a paper last year on this: https://web.stanford.edu/~chadj/emptyplanet.pdf). Doing a thorough, convincing study of this would probably be too difficult and time-consuming for this class, but maybe we could do a first-pass analysis looking at how working age population growth seems to relate to per capita economic output?

Lasso with tons of variables in growth regressions with population as a covariate we never exclude?

Makes sense. Maybe some quasi-experimental thing too like the Baby Boom at the end of WWII or the effect of Roe v Wade on birth cohort size or something like that.

ijyliu commented 3 years ago

An idea I've found interesting recently is the relationship between population growth and economic growth (Charles Jones had a paper last year on this: https://web.stanford.edu/~chadj/emptyplanet.pdf). Doing a thorough, convincing study of this would probably be too difficult and time-consuming for this class, but maybe we could do a first-pass analysis looking at how working age population growth seems to relate to per capita economic output?

Lasso with tons of variables in growth regressions with population as a covariate we never exclude?

Makes sense. Maybe some quasi-experimental thing too like the Baby Boom at the end of WWII or the effect of Roe v Wade on birth cohort size or something like that.

Might be tricky to figure out something original

ijyliu commented 3 years ago

For this: https://github.com/ijyliu/ECMA-31330-Project/issues/4#issuecomment-823749581

What are some interesting variables (independent and dependent) to look at city level spillovers for?