Page 403 - 360.revista de Alta Velocidad - Nº 5

Measuring The Long-Term Regional Economic Impacts of High-Speed Rail in China Using a Dynami

(2012). One of the advantages of using TERM to evaluate regional impacts of HSR in China is that
the model contains a detailed breakdown of trade margin by different transportation modes.
As indicated in Horridge et al. (2005), TERM assumes that all users in each region consumes
commodities from other regions according to common proportions .
4
Specifically, the value of follows is represented in three respects:

A. Basic values = Output price (for domestically produced goods), or CIF prices (for imports);
B. Delivered values = Basic values + (transport or retail) Margins;
C. Purchasers’ values = Delivered values + Tax.

Such a detailed structure enables us to simulate the indirect economic impacts of rail
infrastructure development through the shocks on rail transportation cost change. The original
database includes 137 sectors and 31 regions. To facilitate CGE simulations, we adopted a
condensed version of the SinoTERM database for the assessment, which includes 47 sectors and
8 regions. A detailed bridging table of the sectors and regions are shown in appendix I.

In addition, we made the following updates to improve the accuracy of simulation. First,
since the original SinoTERM is a static model, we upgraded the model into a dynamic
model by specifying its investment/capital ratio as 25 based on findings of Bai et al.
(2006). Second, the elasticity of substitution for factor inputs were updated based on
Guo et al. (2014) and Zha and Zhou (2014).

6. Results

The CGE simulations were implemented in five groups in order to measure the regional economic
impacts of five scenarios, including land use effect, capital investment effect, transportation
cost change effect, productivity change effect and a simultaneous effect. The model was
operated using RunDynam, a windows interface developed by the GEMPACK Software team
at the University of Victoria in Melbourne. Given that the direct impact drivers capture the
period 2002 – 2013, the model was solved recursive-dynamically, in other words, the results are
computed one-period-at-a-time. A short-run closure rule was applied for the simulation in order
to allows wage to be fixed while employment to be adjustable endogenously. In addition, an
additional rule was applied to exogenize the investment variable when the simulation involves
a capital investment shock in order to achieve a convergent optimized solution.

The simulation results of regional economic impacts measured by the change of regional gross
product (GRP) are illustrated in Figure 7. Specifically, Figure 7(a) illustrates the impacts of
land use change as a result of rail infrastructure development on GRP. It is clear that the
land use effect has a negative impact on GRP growth due to the constraint of land factor for
agricultural related sectors. The magnitude of impacts varies substantially across different
regions. For instance, the negative impacts of land use from rail infrastructure development
in the northwest and south-coast regions are found to be relatively higher than other regions.
The two spikes of negative impacts occurred in 2006 in the northwest region and 2013 in the
south-coast region were primarily due to the extended effect of urbanization as a result of rail
infrastructure development.

4 Unlike the conventional data structure for a single-region CGE model, inter-regional trade flows are captured in a TRADE
matrix, which serves as a part of TERM’s data structure. TRADE contains a n X n submatrix, where n represents the number of
regions in the model. Each row corresponds to region of origin and each column corresponds to region of use (destination). Locally
consumed commodities are denoted as diagonal elements.

International Congress on High-speed Rail: Technologies and Long Term Impacts - Ciudad Real (Spain) - 25th anniversary Madrid-Sevilla corridor 401