Regional and spatial dependence of poverty factors in Thailand, and its use into Bayesian hierarchical regression analysis

TL;DR

This study addresses Thailand's multidimensional poverty by identifying geographically coherent regional partitions that better reflect poverty dynamics than administrative boundaries. It combines spatial statistics (Moran's I and Local Moran's I), PCA, and an isoperimetric-guided regionalization to derive six regions, enabling region-aware analysis. A Bayesian hierarchical regression links income to education across regions, complemented by a geographically weighted regression to validate spatial structure, revealing that Bangkok-Pattaya leads in education and income while Northern, Northeastern, and some other regions lag. The approach offers a practical framework for designing targeted poverty alleviation policies and can be extended with spatial models to inform policy at the regional level.

Abstract

Poverty is a serious issue that harms humanity progression. The simplest solution is to use one-shirt-size policy to alleviate it. Nevertheless, each region has its unique issues, which require a unique solution to solve them. In the aspect of spatial analysis, neighbor regions can provide useful information to analyze issues of a given region. In this work, we proposed inferred boundaries of regions of Thailand that can explain better the poverty dynamics, instead of the usual government administrative regions. The proposed regions maximize a trade-off between poverty-related features and geographical coherence. We use a spatial analysis together with Moran's cluster algorithms and Bayesian hierarchical regression models, with the potential of assist the implementation of the right policy to alleviate the poverty phenomenon. We found that all variables considered show a positive spatial autocorrelation. The results of analysis illustrate that 1) Northern, Northeastern Thailand, and in less extend Northcentral Thailand are the regions that require more attention in the aspect of poverty issues, 2) Northcentral, Northeastern, Northern and Southern Thailand present dramatically low levels of education, income and amount of savings contrasted with large cities such as Bangkok-Pattaya and Central Thailand, and 3) Bangkok-Pattaya is the only region whose average years of education is above 12 years, which corresponds (approx.) with a complete senior high school.

Figures (16)

• Figure 1: From left to right, the maps depicts the following information. First: We present the monthly average income reported per household in each province of Thailand. Second: We group the provinces into five different clusters using the Fisher-Jenks algorithm. Third: We present the topology induced by the spatial network, each province is connected with its five nearest neighbors. Fourth: We present the spatial-lag of the monthly average income.
• Figure 2: Left: Reference distribution for Moran's I in the absence of spatial autocorrelation for the monthly average income. Right: Moran's plot, the observations have been colored if their local Moran's I is significantly different from its expected value in the absence of spatial correlation
• Figure 3: Clusters found using local Moran's I statistic. From left to right, the maps depicts the following information. First: We present the local Moran's I statistic for each province. Second: We present the provinces for which we reject the hypothesis of non-spatial autocorrelation. Third: We colored the provinces accordingly to the quadrant where they belong in Moran's plot. Fourth: We present Moran's clusters.
• Figure 4: Moran's clusters for the considered poverty factors.
• Figure 5: Left: Moran's clusters using the first principal component. Right: Proposed regions using agglomerative hierarchical clustering, adding the spatial constraint. It is interesting to notice that Northern Thailand and Northeastern Thailand correspond approximately with two of the Moran’s clusters for the first principal component. While Bagkok-Pattaya and Central Thailand form, approximately, the third cluster detected with the first principal component.