sandbox10042024

Separate regressions

Estimating pass-through separately

Regular

==============================================================================================
                       Model 1        Model 2        Model 3        Model 4       Model 5     
----------------------------------------------------------------------------------------------
logcost                     0.78 ***       0.78 ***       0.68 **        0.96 *        0.96 * 
                           (0.01)         (0.01)         (0.22)         (0.34)        (0.34)  
v_integrated                               0.36 *         0.43 **                             
                                          (0.14)         (0.13)                               
logtax                      0.16 ***       0.16 ***       0.39 ***       0.46 **       0.46 **
                           (0.01)         (0.01)         (0.08)         (0.13)        (0.13)  
logcost:v_integrated       -0.02          -0.02          -0.02          -0.02         -0.02   
                           (0.01)         (0.01)         (0.01)         (0.01)        (0.01)  
v_integrated:logtax        -0.05 *        -0.05 *        -0.05 **       -0.05 *       -0.05 * 
                           (0.02)         (0.02)         (0.02)         (0.02)        (0.02)  
----------------------------------------------------------------------------------------------
Num. obs.              755985         755985         755985         755985        755985      
R^2 (full model)            0.77           0.66           0.64           0.78          0.78   
R^2 (proj model)            0.71           0.64           0.10           0.01          0.01   
Adj. R^2 (full model)       0.77           0.66           0.64           0.78          0.78   
Adj. R^2 (proj model)       0.71           0.64           0.10           0.01          0.01   
Num. groups: station     1244                                         1244          1244      
Num. groups: town                         17                            17            17      
Num. groups: month                                       59             59            59      
==============================================================================================
Spec:log~log. Sample: regular gasoline. Cluster:municipality.

Using the estimates on column 5, the %increase in the price of REGULAR gasoline after a 1% increase in the tax for a vertically integrated stations is 0.41 (.46-0.05). The mean tax is COP 139 and the mean price is COP 9050. So an increase of COP 1.39 in tax leads to an increase of .43% increase in price or COP 37.1.

Premium

================================================================================================
                       Model 1        Model 2        Model 3        Model 4        Model 5      
------------------------------------------------------------------------------------------------
logcost                     0.84 ***       0.85 ***       0.84 *         0.43           0.41    
                           (0.03)         (0.03)         (0.34)         (0.54)         (0.54)   
v_integrated                              -2.44 ***      -2.29 ***                              
                                          (0.49)         (0.40)                                 
logtax                      1.46 ***       1.46 ***       0.59 **        0.47           0.46    
                           (0.03)         (0.03)         (0.18)         (0.23)         (0.23)   
logcost:v_integrated        0.20 ***       0.20 ***       0.19 ***       0.19 ***       0.19 ***
                           (0.04)         (0.04)         (0.04)         (0.03)         (0.03)   
v_integrated:logtax         0.13 **        0.11 *         0.11 **        0.12 ***       0.12 ***
                           (0.04)         (0.05)         (0.04)         (0.03)         (0.03)   
lshare                                                                                 -0.15 ** 
                                                                                       (0.05)   
------------------------------------------------------------------------------------------------
Num. obs.              100804         100804         100804         100804         100804       
R^2 (full model)            0.71           0.65           0.79           0.87           0.87    
R^2 (proj model)            0.67           0.64           0.11           0.02           0.02    
Adj. R^2 (full model)       0.71           0.65           0.79           0.86           0.86    
Adj. R^2 (proj model)       0.67           0.64           0.11           0.01           0.01    
Num. groups: station      769                                          769            769       
Num. groups: town                         17                            17             17       
Num. groups: month                                       53             53             53       
================================================================================================
Spec:log~log. Sample: premium gasoline. Cluster:municipality.

Here we have to use the estimates estimates on column 3 (having all the FE and fewer observations does not leave us with enough variation to identify several parameters. The %increase in the price of PREMIUM gasoline after a 1% increase in the tax for a vertically integrated stations is 0.70 (.59-.11).The mean tax is COP 138.8 and the mean price is COP 12528 So an increase of COP 1.38 in tax leads to an increase of .70% increase in price or COP 87.7.

Plot tax and price

Plot detrended tax and prices in log and levels

Price and tax changes ($)

Price and tax changes (logs)

Similar plots to Bajo and Borrellas

Average prices by type of fuel

Average prices by vertical integration

Average prices by type of fuel (linear)

Average prices by vertical integration (linear)