7,613
edits
Changes
Jump to navigation
Jump to search
Urban Start-up Agglomeration and Venture Capital Investment (view source)
Revision as of 23:22, 8 November 2019
, 23:22, 8 November 2019→Analysis and Exploration
**Chain places together and count them?
The Between Estimator just isn't appropriate. We do want fixed effects (we'll test to see if random effects are appropriate - they aren't). See E:\projects\agglomeration\microeconometrics-using-stata.pdf p254 to p262, as well as [https://www.statalist.org/forums/forum/general-stata-discussion/general/395948-in-which-cases-should-one-prefer-xtreg-be-over-xtreg-re this www.statalist.org thread].
Also the maximum r2 approach picked a number of hulls, which should have been constant (it is now - there were errors before) within a place, making it redundant. Note that STATA will run the variable if you include the fixed effect second but not first - try:
reg something else i.control
reg something i.control else
This problem doesn't occur when using xtreg, which also appears to correctly cluster the standard errors.
===Highest1hull===
