Hi, everyone
I'm running a fixed effect regression with an instrumental variable. The dependent variable is the log of house price and the variable I'm interested in is the unemployment rate. The dataset I'm using is a panel data set by the UK local authority from May 2020 to Sep 2021.
My command is:
xi: xtivreg lprice (UnemploymentbyLocal = EmploymentonFurlough) MortgageSalesVolume ConsumerPricesIndex RegionalMothlyGDP lprice_1(lagged price) i.monthlydate [aw=population2020], first fe vce(cluster ID)
My first question is why 'weights not allowed' showed up when I try to use population as weights [aw=population2020]?
The second question is: Can I add controls to my regression that are not vary by my ID(local authority) but only vary with time? Since I noticed some time dummies (i.monthlydate) are deleted due to collinearity, such as CPI which can be only obtained at national level.
Many thanks in advance!
I'm running a fixed effect regression with an instrumental variable. The dependent variable is the log of house price and the variable I'm interested in is the unemployment rate. The dataset I'm using is a panel data set by the UK local authority from May 2020 to Sep 2021.
My command is:
xi: xtivreg lprice (UnemploymentbyLocal = EmploymentonFurlough) MortgageSalesVolume ConsumerPricesIndex RegionalMothlyGDP lprice_1(lagged price) i.monthlydate [aw=population2020], first fe vce(cluster ID)
My first question is why 'weights not allowed' showed up when I try to use population as weights [aw=population2020]?
The second question is: Can I add controls to my regression that are not vary by my ID(local authority) but only vary with time? Since I noticed some time dummies (i.monthlydate) are deleted due to collinearity, such as CPI which can be only obtained at national level.
Many thanks in advance!

Comment