The data analysis in this research are devided into analyses by LPI component, analyses by different commodity and the overall spatial correlation. The software which used in this analysis is STATA 12.1.
4.1 Analyses by LPI Components
Table 3 presents the result of spatial regression model where the spatial weighted matrix was constructed based on total trading value.
Table 3. The result of analyses by LPI components
variable
CUS: custom, INF: infrastructure, INT: international shipment, LOG: logistics competence, TRA: tracking and tracing, TIM: timeliness
*) significant at =10%, **) significant at =5%, ***) significant at =1%
Table 3 shows that the spatial correlation exists and significant. The spatial correlation for the model with LPI as the dependent variable is 0.110 meaning that the LPI of a country is affected 0.110 point by the neighbor country’s LPI. For the independent variable values, such as the liner shipping connectivity, political stability, total tax rate, and road density, those independent variables influence the LPI score of a country 0.250, 0.843, -0.003, and 0.116 respectively. Several variables have been included into model but those variables causing the spatial correlation not significant (see appendix 6). Furthermore, the table 3 also presents the result of spatial correlation where the other six LPI components were used as dependent variable. The spatial regression model with LPI infrastructure as dependent variable has the highest spatial correlation value, while the lowest is LPI timeliness.
16
4.2 Analyses by Different Commodity
This analysis was conducted by changing the basis in constructing the spatial weighted matrix. In addition to total trading value, other commodities are selected as the basis in constructing the spatial weighted matrix. The commodities which have been selected are electronics, automotive, and agriculture, forestry and fishing.
The electronic comodity was chosen as a basis of constructing the spatial weighted matrix since this commodity have complex nature of supply chain. Electronic Industry Citizenship Coalition (EEIC) in their report stated that the supply chain for any given electronics product can include hundreds of companies. This is primarily due to the complex nature of electronic products. Unlike a garment (e.g. shirt), an average laptop computer consists of hundreds of individual parts that must be sourced and assembled according to precise specifications. These parts come from all over the world. It means that the electronic commodity is highly depend on global supply chain. For this reason, the electronic commodity is choosen.
The automotive commodity was chosen as a basis in constructing the spatial weighted matrix because this commodity have similar properties with electronic commodity. Automotive product consist of hundreds parts (components) for which most of them must be sourced from outside supplier. Like electronics commodity, these supplier come all over the world and highly depend on global supply chain.
The agriculture, forestry and fishing commodities were chosen in sector analysis because these commodities were used by some manufacturers as a raw materials of their product, but on the other hand most of countries can fulfill their need for these commodities by themself. It is based on this reason that these commodities are assumed to have low dependencies on global supply chain.
Table 4. presents the comparison of spatial regression analysis result with LPI overall as the dependent variable. There are 4 models shown in table 4, where every model use different basis in constructing spatial weighted matrix.
17
Table 4. The result of analyses by different commodity
Variable
Basis in constructing spatial weighted matrix Total
Liner shipping connect 0.250*** 0.205** 0.268*** 0.226**
Political stability 0.843*** 0.911*** 0.984*** 0.943***
Total tax rate -0.003* -0.005** -0.002* -0.004*
Road density 0.116* 0.123** 0.094* 0.130**
Spatial correlation () 0.110*** 0.242*** 0.151*** 0.008**
*) significant at =10%, **) significant at =5%, ***) significant at =1%
Table 4 shows the model where the spatial weighted matrix was constructed based on electronic trading value has the spatial correlation () 0.242 which means that in term of electronic trading value the LPI of a country is affected 0.242 point by the neighbor country’s LPI. The independent variables value for liner shipping connectivity, political stability, total tax rate, road density are 0.250, 0.843, -0.003, and 0.116 respectively.
The model where the spatial weighted matrix was constructed based on automotive trading value has the spatial correlation () value 0.151. It means that in term of automotive trading value the LPI of a country affected 0.151 point by neighbor country’s LPI. The independent variables value for liner shipping connectivity, political stability, total tax rate, road density are 0.205, 0.911, -0.005, and 0.123 respectively.
Meanwhile, the model which is the spatial weighted matrix was constructed based on agriculture, forestry, and fishing trading value has the spatial correlation () value 0.008, which means that in term of agriculture, forestry, and fishing trading value the LPI of a country affected 0.008 point by neighbor country’s LPI. The independent variables value for liner shipping connectivity, political stability, total tax rate, road density are 0.226, 0.943, -0.004, and 0.130 respectively.
Table 4. also shows that the spatial regresssion model where the spatial weighted matrix was constructed based on electronic trading value has the highest spatial correlation () and the model where the spatial weighted matrix constructed based on
18
agriculture, forestry, and fishing trading value has lowest spatial correlation (). It means that electronics commodity is highly depend on global supply chain. Meanwhile, the agriculture, forestry, and fishing commodity have low dependency on global supply chain, this may be due to most of the countries been able to fulfill their own needs in term of agriculture, forestry, and fishing.
4.3 The Overall Spatial Correlation
Figure 2. shows the comparison of the spatial correlation value () among three commodities in the analyses by different commodity.
Figure 4. The comparison of spatial correlation value of sector analysis.
Figure 2 shows that electronics commodity has the highest spatial correlation followed by automotive commodity, then comes the total trading value, and agriculture, forestry and fishing commodity as the lowest. This indicates that electronics commodity have the highest dependency on international trading (global supply chain) than the other two commodities which were analyzed in the section 4.2.
For the electronic commodity, the spatial correlation for LPI is 0.242, meaning that in term of electronics trading, the LPI of a country is affected 0.242 point by the neighbor country’s LPI. Meanwhile, for the other six LPI dimensions, the highest spatial correlation is LPI infrastucture and the lowest spatial correlation is LPI timeliness. The
19
automotive commodity has the LPI spatial correlation value 0.127 which means that in term of automotive trading, the LPI of a country is affected 0.151 point by the neighbor country’s LPI. The agriculture, forestry and fishing commodities have the LPI spatial correlation value 0.008. it means that in term of agriculture, forestry and fishing trading, the LPI of a country is affected 0.008 point by the neighbor country’s LPI.
Based on the spatial regression result that electronic commodity has the highest dependency on global supply chain followed by automotive commodity, the Indonesian government should make a policy to improve the electronics and automotive industry by strenghten either the large scale industry and the small and medium industries. Indonesia have a lot of potential electronic and automotive companies which could become a supplier for world class electronic and automotive companies. Indonesian electronic and automotive industries have to participate in global supply chain, because by participating in global supply chain, Indonesia not only become a market for but also become a player in global supply chain. The Indonesian government should enhance the research facility for electronic and automotive industries so that their product able to compete in a global market. Moreover, government should giving the easiness for export import procedure.
Furthermore, in regard to the factors that influencing logistics performance in this research, the Indonesian government have to improve the road and shipping infrastructure and also maintain political stability, because political stability will affect to the security and the business is needs stability for business continuity.
20