We have daily prices for regular gasoline sold by all stations operating in the markets not subject to price ceilings. We control for the unit-cost of gasoline. Importantly we have variation in the carbon tax levied at each station. The variation in the carbon tax comes from two sources. First, the gross carbon tax (an amount of pesos per ton of carbon) is updated yearly in February with the previous year inflation, this gives us some time variation. The carbon tax is levied only on gasoline and the ethanol component of the fuel is exempt. The fuel sold at each station varies in the proportion of ethanol (it ranges from 2% to 10%). So the actual tax paid on each gallon of fuel varies depending on the proportion of ethanol. This gives us variation across stations during the same period. Importantly, as far as we know, the percent of ethanol mix is exogenous to the stations and it probably depends on the international price of sugar. But so far this is just a conjecture.
m1 = lm(logp~logimpocarb, data=subset.reg)
m2 = lm(logp~logimpocarb+logcosto, data=subset.reg)
m3 = lm(logp~logimpocarb+integrada, data=subset.reg)
m4 = lm(logp~logimpocarb+logcosto+integrada, data=subset.reg)
m5 = lm(logp~logimpocarb+logcosto+integrada+I(logimpocarb*integrada), data=subset.reg)
screenreg(list(m1,m2,m3,m4,m5),
custom.model.names = c("(1)","(2)","(3)","(4)","(5)"),
custom.coef.names = c("Intercept","log(Tax)", "log(Cost)","Integrated","log(Tax) x Integrated"),
digits = 5,stars = c(0.01,0.05,0.1), omit.coef = c("as.factor"))##
## ===============================================================================================================
## (1) (2) (3) (4) (5)
## ---------------------------------------------------------------------------------------------------------------
## Intercept 10.00054 *** 2.71611 *** 10.00304 *** 2.73077 *** 2.69072 ***
## (0.00539) (0.01042) (0.00535) (0.01030) (0.01050)
## log(Tax) -0.18320 *** 0.09823 *** -0.18291 *** 0.09803 *** 0.10610 ***
## (0.00109) (0.00090) (0.00108) (0.00089) (0.00098)
## log(Cost) 0.66323 *** 0.66211 *** 0.66214 ***
## (0.00087) (0.00086) (0.00086)
## Integrated -0.01973 *** -0.01867 *** 0.18220 ***
## (0.00019) (0.00014) (0.01010)
## log(Tax) x Integrated -0.04073 ***
## (0.00205)
## ---------------------------------------------------------------------------------------------------------------
## R^2 0.03546 0.45011 0.04921 0.46242 0.46270
## Adj. R^2 0.03546 0.45011 0.04921 0.46242 0.46270
## Num. obs. 764716 764716 764716 764716 764716
## ===============================================================================================================
## *** p < 0.01; ** p < 0.05; * p < 0.1m1 = felm(logp~logimpocarb| municipio, data=subset.reg)
m2 = felm(logp~logimpocarb+logcosto| municipio, data=subset.reg)
m3 = felm(logp~logimpocarb+integrada| municipio, data=subset.reg)
m4 = felm(logp~logimpocarb+integrada+logcosto| municipio, data=subset.reg)
m5 = felm(logp~logimpocarb+logcosto+integrada+I(logimpocarb*integrada)| municipio, data=subset.reg)
screenreg(list(m1,m2,m3,m4,m5),
custom.model.names = c("(1)","(2)","(3)","(4)","(5)"),
custom.coef.names = c("log(Tax)", "log(Cost)","Integrated","log(Tax) x Integrated"),
custom.note = "Market fixed effects",
digits = 5,stars = c(0.01,0.05,0.1), omit.coef = c("as.factor"))##
## ================================================================================================================
## (1) (2) (3) (4) (5)
## ----------------------------------------------------------------------------------------------------------------
## log(Tax) -0.18421 *** 0.09851 *** -0.18375 *** 0.09883 *** 0.10680 ***
## (0.00107) (0.00089) (0.00106) (0.00088) (0.00096)
## log(Cost) 0.66824 *** 0.66792 *** 0.66794 ***
## (0.00088) (0.00087) (0.00087)
## Integrated -0.01848 *** -0.01818 *** 0.18040 ***
## (0.00019) (0.00014) (0.00987)
## log(Tax) x Integrated -0.04027 ***
## (0.00200)
## ----------------------------------------------------------------------------------------------------------------
## Num. obs. 764716 764716 764716 764716 764716
## R^2 (full model) 0.07538 0.47475 0.08724 0.48622 0.48649
## R^2 (proj model) 0.03730 0.45312 0.04965 0.46506 0.46534
## Adj. R^2 (full model) 0.07536 0.47473 0.08722 0.48620 0.48647
## Adj. R^2 (proj model) 0.03728 0.45310 0.04963 0.46504 0.46533
## Num. groups: municipio 17 17 17 17 17
## ================================================================================================================
## Market fixed effects# m1<-felm(logp~logimpocarb | date+municipio, data = subset.reg)
m1 = felm(logp~logimpocarb| date+municipio, data=subset.reg)
m2 = felm(logp~logimpocarb+logcosto| date+municipio, data=subset.reg)
m3 = felm(logp~logimpocarb+integrada| date+municipio, data=subset.reg)
m4 = felm(logp~logimpocarb+integrada+logcosto| date+municipio, data=subset.reg)
m5 = felm(logp~logimpocarb+logcosto+integrada+I(logimpocarb*integrada)| date+municipio, data=subset.reg)
screenreg(list(m1,m2,m3,m4,m5),
custom.model.names = c("(1)","(2)","(3)","(4)","(5)"),
custom.coef.names = c("log(Tax)", "log(Cost)","Integrated","log(Tax) x Integrated"),
custom.note = "Market and time fixed effects",
digits = 5,stars = c(0.01,0.05,0.1), omit.coef = c("as.factor"))##
## ================================================================================================================
## (1) (2) (3) (4) (5)
## ----------------------------------------------------------------------------------------------------------------
## log(Tax) 0.31978 *** 0.26744 *** 0.29393 *** 0.24157 *** 0.24191 ***
## (0.02720) (0.02722) (0.02690) (0.02692) (0.02691)
## log(Cost) 0.85669 *** 0.85710 *** 0.85244 ***
## (0.02491) (0.02464) (0.02463)
## Integrated -0.01801 *** -0.01801 *** 0.17758 ***
## (0.00014) (0.00014) (0.00982)
## log(Tax) x Integrated -0.03966 ***
## (0.00199)
## ----------------------------------------------------------------------------------------------------------------
## Num. obs. 764716 764716 764716 764716 764716
## R^2 (full model) 0.48639 0.48719 0.49755 0.49835 0.49861
## R^2 (proj model) 0.00018 0.00173 0.02190 0.02345 0.02396
## Adj. R^2 (full model) 0.48534 0.48613 0.49652 0.49731 0.49757
## Adj. R^2 (proj model) -0.00188 -0.00033 0.01989 0.02144 0.02195
## Num. groups: date 1555 1555 1555 1555 1555
## Num. groups: municipio 17 17 17 17 17
## ================================================================================================================
## Market and time fixed effectsm1 = felm(logp~logimpocarb| date+codigo_sicom, data=subset.reg)
m2 = felm(logp~logimpocarb+logcosto| date+codigo_sicom, data=subset.reg)
m3 = felm(logp~logimpocarb+integrada| date+codigo_sicom, data=subset.reg)## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinitem4 = felm(logp~logimpocarb+integrada+logcosto| date+codigo_sicom, data=subset.reg)## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinitem5 = felm(logp~logimpocarb+logcosto+integrada+I(logimpocarb*integrada)| date+codigo_sicom, data=subset.reg)## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinitescreenreg(list(m1,m2,m3,m4,m5),
custom.model.names = c("(1)","(2)","(3)","(4)","(5)"),
custom.coef.names = c("log(Tax)", "log(Cost)","Integrated","log(Tax) x Integrated"),
custom.note = "Station and time fixed effects",
digits = 5,stars = c(0.01,0.05,0.1), omit.coef = c("as.factor"))## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinite
## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinite
## Warning in chol.default(mat, pivot = TRUE, tol = tol): the matrix is either
## rank-deficient or indefinite##
## ===================================================================================================================
## (1) (2) (3) (4) (5)
## -------------------------------------------------------------------------------------------------------------------
## log(Tax) 0.27945 *** 0.23197 *** 0.27945 *** 0.23197 *** 0.23201 ***
## (0.01778) (0.01778) (0.01778) (0.01778) (0.01777)
## log(Cost) 0.78567 *** 0.78567 *** 0.78244 ***
## (0.01634) (0.01634) (0.01633)
## Integrated
##
## log(Tax) x Integrated -0.03843 ***
## (0.00133)
## -------------------------------------------------------------------------------------------------------------------
## Num. obs. 764716 764716 764716 764716 764716
## R^2 (full model) 0.78170 0.78236 0.78170 0.78236 0.78259
## R^2 (proj model) 0.00032 0.00335 0.00032 0.00335 0.00444
## Adj. R^2 (full model) 0.78089 0.78156 0.78089 0.78156 0.78179
## Adj. R^2 (proj model) -0.00335 -0.00031 -0.00335 -0.00031 0.00078
## Num. groups: date 1555 1555 1555 1555 1555
## Num. groups: codigo_sicom 1245 1245 1245 1245 1245
## ===================================================================================================================
## Station and time fixed effects