Hi,
I have a dataset with different stores and I want to analyze the relative prices of that stores.
I run a OLS regression with the price difference as dependent variable.
Therefore, I define local markets: A market is defined by one store and all its competitors within 2 miles.
I cluster the errors by market.
However, it is possible that one store operates in two local markets.
For example: There is store A, 1.5miles north is store B and another 1.5 miles north is store C.
Hence, store B is in market A as well as in market C
If I consider relative prices, relative price AB and CB, then it is likely that the errors are correlated since B is involed in both cases.
Does anyone have an idea how I can adjust the errors in order to control for that?
Thanks!
Christoph
I have a dataset with different stores and I want to analyze the relative prices of that stores.
I run a OLS regression with the price difference as dependent variable.
Therefore, I define local markets: A market is defined by one store and all its competitors within 2 miles.
I cluster the errors by market.
However, it is possible that one store operates in two local markets.
For example: There is store A, 1.5miles north is store B and another 1.5 miles north is store C.
Hence, store B is in market A as well as in market C
If I consider relative prices, relative price AB and CB, then it is likely that the errors are correlated since B is involed in both cases.
Does anyone have an idea how I can adjust the errors in order to control for that?
Thanks!
Christoph
