Page 234 - 360.revista de Alta Velocidad - Nº 6

Kim, Junghwa. Li, Yeun-Touh. Schmöcker, Jan-Dirk.

5. Model Estimation

The estimated time-random/region-random effects two-way error component models’ results
are shown in Table 3. By the model 1, we examine how socioeconomic indicators and time/
seasonal factors which have been verified variously in previous studies, explain THSR ridership
considering regional heterogeneity. Additionally, this study also examines how the factors
of access-tothe station enhances model fits towards THSR ridership. Thus, Model 2 contains
all socioeconomic elements of the Model 1 as well as accessibility factors. Both models are
statistically significant, supported by the Chi2 statistic of 0.000, moreover all coefficients of
explanatory variables are significant at the 5% level; (coefficients significant at the 1% level are
shown in bold)
From our finding, regional specific socio-demographics (i.e. population, car ownership, length
of road network, business scale), fuel price by year as well as time dummy variables are
always being significant generally explain THSR demand. Population and car ownership are
most influential variables to explain THSR ridership by region, since population may represent
the market size of the travel demand and HSR works as an alternative transportation mode
instead of car use. Therefore, this indicates that higher number of HSR users can be shown in
the cities with high population and low car ownership ration. Furthermore, the length of road
network, fuel price, business scale that shows the revenue per company, is found positive to
local THSR ridership. In addition, the seasonal factors, the summer vacation and the element
which shows how long after the opening of THSR, performed well to explain the local demand
as well. The demand is found positive significant in the summer suggests that vacation activities
has induced HSR demand. Moreover, the number of THSR users increases as time passes after
system opening. This result shows the “mass effects” which argued by Schmöcker et al. (2014)
and “adaptation effects” suggested by Lit et al. (2015). According to Li and Schmöcker (2014),
the negative perception of the traveler at the beginning of operation has been observed,
such as safety concerns, unreliable ticketing, or the reservation system that prohibited some
potential users to taking rides (Cheng, 2010). However, this general perception might possibly
have been changed over time, where more travelers recognized the advantages of HSR, travel
time saving, level of services, easier to access than before and other advantages related with
perception. These conversions of perceptions would enlarge HSR travelers from a small number
of population group penetrates into the majority. This was resonated by Schmöcker et al.
(2014) who discussed that “mass effects” can be significant determinants of long term demand
adaptation. One persuades a few to change their behavior initially in order to encourage
a large number of people to follow later. There is then a potential of enduring significant
demand increases as the new service might increase its attractiveness over time if more start
to join it. Li et al. (2015) proposed an econometric time-series model to predict aggregated
THSR ridership on long-term demand. This study takes into account on THSR local demand to
understand the impact of “adaptation effects” as well as access link regarding other general
explanatory variables.
The result of Model 2 is also quite similar results with those of the Model 1 in the case of a
socioeconomic variable by region. However, we considered three of accessibility variables here
additionally. All are significant but has different sign in the coefficients. Especially the station
location has negative value, unlike the other two accessibility variables. It implies that the
number of passengers is smaller as the station of the HSR is located far from the city center. On
the other hand, the ridership demand would be induced if buses or railway public transportation
is connected to the HSR station. In addition, when we compared to the transit system in an
aspect of access-to-the station, bus system (BRT and shuttle bus) has a larger influence than
rail transit (MRT and TR). From the result of the Model 2, we identified that accessibility to HSR
station is also an important factor in explaining the demand for HSR as much as socioeconomic
indicators.

232 360.revista de alta velocidad