Hello all,
I am currently doing the econometrics for my dissertation about the extent to which the quality of public transportation access defines an area's unemployment rate.
The data is cross sectional, with 982 observations. Each observation is a London MSOA (geographical region). I argue that an MSOAs characteristics may plausibly be, to some extent, a function of the neighboring MSOAs characteristics. I would therefore like to implement some sort of Spatial HAC errors to correct for spatial dependence.
What would be the best practice here?
Here is an example regression that would require such standard errors:
Many thanks,
Alex
I am currently doing the econometrics for my dissertation about the extent to which the quality of public transportation access defines an area's unemployment rate.
The data is cross sectional, with 982 observations. Each observation is a London MSOA (geographical region). I argue that an MSOAs characteristics may plausibly be, to some extent, a function of the neighboring MSOAs characteristics. I would therefore like to implement some sort of Spatial HAC errors to correct for spatial dependence.
What would be the best practice here?
Here is an example regression that would require such standard errors:
Code:
reg UnemploymentRate MEANPTAL_Clean percent_no_car lWDist lmean_income low_edu mid_edu Workingage lHousePrices ovr60_on_credit, robust ivreg2 Unemployment (MEANPTAL_Clean=iv_dist10) percent_no_car ovr60_on_credit lmean_income lWDist Workingage low_edu mid_edu lHousePrices, robust first
Alex
Comment