Table 5: Estimated Causal Models for TC and JTC

Table 5: Estimated Causal Models for TC and JTC

In logs

Excel | CSV

  12TC ∇∇12JTC
Constant 0.0544* -0.0368*
(0.0073) (0.0076)
DAYS 0.0018 0.0026*
(0.0024) (0.0010)
EASTER -0.0535* -0.0710*
(0.0106) (0.0041)
FEB93   0.0579*
  (0.0136)
APR93   -0.0870*
  (0.0137)
MAR94 0.0368*  
(0.0165)  
JUL95   0.0492*
  (0.0085)
ΔPTCi 0.1576 -0.0134
(0.3032) (0.1387)
MRLi 0.2517*B4 - 0.4175*B5 0.1800*B4 - 0.2087*B5
(0.1344) (0.1665) (0.0861) (0.0977)
(1 + 0.9897*B)  
(0.0672)  
θ1 0.4557*  
(0.1316)  
θ2 -0.5399* 0.4671*
(0.0997) (0.1641)
θ3   -0.5261*
  (0.1628)
Θ12 -0.1285 -0.4048*
(0.0939) (0.0949)
φ1 0.3240* 0.9753*
(0.1042) (0.1044)
φ2   -0.6448*
  (0.1997)
φ3 0.2694* 0.3357*
(0.0964) (0.1314)
φ4   -0.3400*
  (0.0942)
lowercase sigma caret subscript {lowercase a} 0.0256 0.03035
R2 0.7 0.86
LBQ (12, 24, 36) 11.2, 22.8, 45.6 10.6, 14.6, 21.9
Jarque-Bera (p-value) 2.75 (0.252) 2.69 (0.260)
White's heterosk (p-value) 6.68 (0.946) 5.86 (0.97)

Notes: Standard errors in parenthesis. * represents significant at 5%; ∇ = (1 - B) and ∇12 = (1 - B12): regular and seasonal differences; B: lag operator; ΔPTCi (i = TC, JTC): rates of growth of deflated price for TC tickets; MRL: Metro route length; LBQ: Ljung-Box Q statistics; lowercase sigma caret subscript {lowercase a} = residual standard error.