Page 233 - 360.revista de Alta Velocidad - Nº 6

Impacts of station accessibility and regional heterogeneity on HSR ridership

unobserved effects, presented by month and city as random variables. Baltagi (2013) noted
that a panel data model differs from a regular time-series or cross-sectional regression in that
it has a double subscript on its variables, i.e.

where i represents location and t represents time. The i subscript therefore denotes the cross-
section dimension, whereas t denotes the time-series dimension. Here is a scalar, β is a Kx1
matrix, and X is the i observation of the K explanatory variables. In this study, panel data
th
it
applications utilize a two-way error component disturbance as follows:

where u denotes the unobserved regional effects, μ denotes the unobservable time effect, and
t
i
ν is the remainder stochastic disturbance term. Here, u ~ IID(0,σ ) and/or μ ~ IID(0,σ ) and ν
2
2
i
it
u
it
t
u
~ IID(0,σ ) independent of each other, it is possible to obtain the two-way random effect error
2
u
component model. In addition, X is independent of u , μ and ν for all i and t . Inference, in
i
it
t
it
this case, pertains to the large population, which the sample was randomly drawn from.
In this study, we estimate two models. First model only includes socioeconomic variables
and then transport accessibility variables would be considered as an additional factor in the
second model. We are going to identify whether accessibility variables are important as much
as socioeconomic factors to the model which explains HSR ridership. Our hypothesized model
could be expressed as follows;

Here, taking a log to equation (4)

Where, i=city (region) and t=month, X =socioeconomic factors(population fuel price, business
k
scale, road supply, car ownership), X =year, Ds=dummy for summer, Then, our hypothesized
y
second model with transport accessibility variables is presented as follows:

Where, A =station location (distance from CBD to (HSR station(km)) and A =Rail and bus transit
n
n *
(dummy variable) are added as transport accessibility variables.

International Congress on High-speed Rail: Technologies and Long Term Impacts - Ciudad Real (Spain) - 25th anniversary Madrid-Sevilla corridor 231