How Do Production Networks Affect the Resilience of Firms to Economic and Natural Disasters: A Methodological Approach and Assessment in Japan, Taiwan and Thailand

Abstract

This chapter examines the economic impact of the disaster in the supply chain of electronic products, transportation equipment and machinery sectors in Asia . A Multi-region static CGE model based on GTAP v.9.0 is used to capture full-fledged propagation effects stemming from natural disaster to seven economies in Asia. Three hypothetical disaster cases are simulated based upon the collapse rates estimated: Tokyo earthquake, Taipei earthquake and Thailand floods. The simulation results confirmed that the economic impacts of the natural disaster are propagated to other countries in the region through changes in the production volumes and prices response. Economic impacts stemming from neighbouring economies can be positive or negative depending on the production and trade structure defined by the supply chain network. Increased resilience to the disaster will have an implication for increased macroeconomic stability by reducing propagation of the disaster impacts.

This research was conducted as a part of the project of Economic Research Institute for ASEAN and East Asia (ERIA) ‘Reducing the Vulnerability of Supply Chains and Production Networks.’ The authors are deeply indebted to the members of this project for their invaluable suggestions. The opinions expressed in this paper are the sole responsibility of the authors and do not reflect the views of ERIA.

Appendices

Annex 1. Region Categories

See Table 4.5.

Table 4.5 Region categories from GTAP v.9

Annex 2. Sector Categories

See Table 4.6.

Table 4.6 Sector categories from GTAP v.9

Annex 3 Sector Impact Estimates

In terms of capital loss assumption, following Huang and Hosoe (2016), the geographic building collapse rate is used to estimate the capital loss. In the r-th region, the study uses information of geographic concentration rates of the i-th industry in the r-th region from the IO table in prefecture level. Combining these two datasets, the regional and industrial building collapse rates can be calculated. As the proportion (ratio) of fully- and partially-collapsed buildings was approximately 1:1, the estimated capital damage could be assumed to be twice as large as the original damage. Finally, the capital losses of the i-th industry in the r-th region CL_i,r as $CL_{i,r} = X_{i,r} \times C_{r} \times 2$ could be computed. The total sectoral capital loss rate $\mathop \sum \limits_{r} CL_{i,r}$ is shown in Tables 4.7 and 4.8 We estimated regional labour loss rates in the total labour force (endowment) LL_r by combining the regional employment share SL_r with the national labour force affected by the building damage (C_r × 2) as $SL_{r} \times C_{r} \times 2$.

Table 4.7 Tokyo earthquake (Tokyo Bay North)

Table 4.8 Taipei earthquake

For the sectoral capital stock damage assumption of Thai Flood, we mainly used the results estimated by Isono and Kumagai (2014). For energy sector of coal, crude oil, petroleum and natural gas, we used the statistics from mineral USGS (2014) and multiplied by the share of capital in the value-add part. For labour endowment loss, we grounded the article Koser (2014) and multiplied the affected labour force by labour share in the value-add part.

Generally speaking, the flood impact usually lasts for months, and the factory operation may as well as be disrupted. The flood recovery and aftermath may not take much time of years as it may comparing with earthquake. However, in the static analysis in this research, a one year disruption of supply chain may have similar impact as earthquakes. And thus we run the flood simulation as we do for earthquake simulation (Table 4.9).

Table 4.9 Thai flood