Hello! I'm fairly new to stata so there are a few things that I am not sure how to do. I am currently working on my thesis and I'm having a problem conducting a propensity score matching. I will try my best to explain in full detail what I am dealing with.
The topic I'm studying is the effect on apartment prices because of a height regulation on a specific neighborhood, My database consists of 104,403 apartments and it includes prices, square feet, quality among other characteristics from 2008 to 2019. The data was obtained through a website similar to zillow and is not a panel data set because it is rare that a single apartment is being sold multiple times in 10 years.
I have 6,282 apartments located in said regulated neighborhood, of those, about 50 percent are apartments that were built before the regulation took place (the law was implemented around 2011). The rest of the observations correspond to apartments located in a radius of 3 miles of my "treated" neighborhood.
Since there is no perfect control to test the effect of the regulation, I'm planning to weigh the other observations into creating a synthetic control. To find this weight I thought of doing a propensity score matching but I ended up matching the whole database (before and after) so I can't see any parallel trends.
I was doing it by using this command:
gen tratamiento=1 if neighbourhood=="Santa isabel"
replace treatment=0 if treatment==.
global treatment tratamiento
global ylist price_sqft
global xlist sqft_const sqft_balcony year_construction number_parking_units number_torage_units construction_cost year_of_sale
global zlist
global breps 5
pscore $treatment $xlist, pscore(myscore) blockid(myblock) comsup
After doing this my graph looks like two almost identical lines. I think my mistake is that in the treatment I am not differentiating between the years previous and after the regulation law was passed.
As an alternative, I thought of using The Synth_Runner Package but I am not familiar with it and I am a little be lost, to be honest.
Any help would be appreciated.
Thank you in advance
Related Posts with Troubles with Propensity score matching and synth_runner
estate binary indicatorsHello, I am exporting results with a binary variable as the main IV, and the output shows "var=1" a…
how to merge the spatial data when there is no common variables in two dataset?Hi all, My data contains households’ GPS info (latitude, longitude, and altitude) and don’t have da…
How to get the error termDear Statalists, Greetings, I will be so grateful to have an answer for the following issue, pleas…
Twoway graphs: yaxis only on the right-hand side (like in The Economist)Dear Forum, I wonder how to create graphs that have the yaxis only on the right-hand side, like man…
Need Help with Stata AssignmentHi, Not sure if I can use this forum to get academic help so if this isn't right please delete my po…
Subscribe to:
Post Comments (Atom)
0 Response to Troubles with Propensity score matching and synth_runner
Post a Comment