Evidence of Spatial Dependence in Thailand
1) Introduction
2) Data
In this study, satellite nighttime light data of Thailand published by GeoQuery is used. The data cover 77 provinces of Thailand between 2012 and 2020
.
The geographical data of Thailand (map) published by GeoQuery is also used to merge with night-time light data.
Table 1 shows non-spatial nighttime light data of 77 provinces of Thailand between 2012-2020.
Table 2 shows spatial data of 77 provinces of Thailand.
Table 3 represents the merge data between night-time light and geographical data of Thailand.
3) Methodology
This analysis employ Exploratory Spatial Data Analysis (ESDA) which includes 1) Global Spatial Autocorrelation (Moran's I) and 2) Local Spatial Autocorrelation (LISA, Local Indicators of Spatial Association)
4) Data Exploration
Thailand is strategically located in the center of ASEAN countries.
Most of the ASEAN nations are centered around Thailand, including Myanmar on the west, Cambodia and Lao PDR on the east, Malaysia, Indonesia, and Singapore on the south.
This picture shows the distribution of the average night-time light in Thailand in 2012 and 2020
Geographically, Thailand's economic activity are generally dispersed throughout the country both in 2012 and 2020.
Some provinces have witnessed a rise in economic activity over the past several years, while others have seen a decline in fostering economic activity. It's seen that those provinces tend to locate near to each others.
5) Results
Global and Local Spatial Dependence in 2012 and 2020
The Moran's scatter plot for nighttime lights in 2012 and 2020 provides indications of spatial dependency, with the majority of the districts falling under the HH (high-high), LH (low-high), and LL categories (low-low) cluster in 2012, as well as the LL, HL, and HH (high-high) (low-low) in 2020. The Moran's I statistics for the nighttime light data are 0.18 in the first year, and increase to 0.33 in the final year.
Evidently, the trend of spatial dependence is on the rise, and the results of Moran's I, as previously indicated, show that there is spatial clustering in 2012 and 2020 meaning that the Moran's I test results that are significant and positive show that, on average, province tend to be surrounded by neighbors that have simillar values. For example, refer to 2020, Bangkok, the capital city of Thailand, which is categorized as HH is surrounded by HH provinces which includes such as Rayong and Chachoengsao.
Regarding the LISA analysis which is used to identify the location of spatial relationship, the two pictures below depicts the locations of spatial clusters, hotspots and coldspots, in 2012 and 2020. Provinces in the maps are regarded as clustering members if the local Moran's I's p-value is below 0.10.
The result appears to be different between the years 2012 and 2020, as indicated on the map. It brings up the concern that the geographical clustering, in terms of cluster size and member, is not lasting over time.
The result is partly consistent with the previous study (Regional convergence and spatial dependence in Thailand: global and local assessments by Tipayalai, K., & Mendez, C. (2022)). For the dissimilarity, the study by Tipayalai, K., & Mendez, C. (2022) reveals that an economic disparity is more likely to continue over time which is untrue in this study. However, according to the result in 2020, we agree that the capital city (Bangkok) and its surrounding areas (Nakhonpathom, Nonthaburi, Pathumthani, Samutprakan, and Samutsakhon), the Central (Phranakhonsriayuthaya), and the Eastern parts of the country (i.e., Chachoengsao, Chonburi, and Rayong) are where the majority of the high-income provinces are concentrated spatially. And these regions are likely to contribute to higher economic activities as the industrial activities (manufacturing) are mainly concentrated in these regions as well as the government has various plans to support these particular areas as an investment hub of Thailand.
2012
2020
5) Conclusion and Policy implication
The results make it abundantly evident that Thailand's development is clearly unequal as a result of the government's inconsistent application of its many policies. The government must adopt separate policies for each province in order to take into account their individual differences and avoid concentrating exclusively on a few key sectors. Given that Thailand receives a significant number of tourists each year, the service industry is particularly crucial. Supply-side policy should thus be included in certain other policy domains.
6) Discussion
On the one hand, the fact that the findings of this study and earlier studies differ suggests that future research should compare nighttime light and official GDP per capita to identify any discrepancies. Furthermore, to capture informal activity and lower the likelihood of widespread government involvement, the next research may incorporate nighttime light and GDP per capita.
References
Miranti, R. C., Yan, S. S., & Aginta, H (2021). REGIONAL GROWTH, CONVERGENCE, AND HETEROGENEITY IN SUMATERA: EVIDENCE FROM NEW SATELLITE DATA.
Mveyange, A. (2018). Measuring and Explaining Patterns of Spatial Income Inequality from Outer Space: Evidence from Africa. World Bank Policy Research Working Paper, 8484.
Nordhaus, W., & Chen, X. (2015). A sharper image? Estimates of the precision of nighttime lights as a proxy for economic statistics. Journal of Economic Geography, 15(1), 217– 246.
Lessmann, C., & Seidel, A. (2017). Regional inequality, convergence, and its determinants–A view from outer space. European Economic Review, 92, 110–132.
Tipayalai, K., & Mendez, C. (2022). Regional convergence and spatial dependence in Thailand: global and local assessments. Journal of the Asia Pacific Economy, 1-28.