I have retail price data for 31 different products observed weekly in 10 stores of the same chain spread across 3 states for 5 years. I want to study if prices vary differently during holidays across states. In some states, there are policies that prohibit pricing below cost. I do not have cost data, hence as a proxy I use weekly gas prices in these states.
My Model: log(price (ijt)) ~ policy_dummy (j) * holiday_dummy (t) * gas price (j) + product FE + state FE + year FE
subscripts : product i, store j, week t
I need help interpreting the coefficients.
- Does the coefficient of the holiday_dummy only term correspond to when policy dummy is 0? Can it be interepreted as product prices fall by 0.7% in state where there in no policy on minimum markup during holidays?
- holiday_dummy:gas_price -> During holidays, an increase in gas price is associated with 0.2% increase in product prices in states where the policy isn't enforced. Is this right?
Comment