Jiraphan Suntornchost.. Modelling the household debts in Thailand using Fay-Herriot models. (). King Mongkut's University of Technology North Bangkok. Central Library. : , 2565.
Modelling the household debts in Thailand using Fay-Herriot models
Abstract:
In this study, we investigate the performance of the Fay-Herriot area level model to
estimate the Household Debts of Thailand by using the 2017 household socio-economic survey
from the National Statistics Office (NSO) of Thailand. Two versions of the model are considered
which are the classical Fay-Herriot model and the Bayesian approach to the model, called as
the Hierarchical Bayesian Fay-Herriot model. Results from our study show that the two models
can reduce mean square etrrors of the direct estimates, particularly for the areas with large
sampling variances. Moreover, we apply results from the two models to discuss behavior of
household debts in Thailand by comparing household debts among different regions of
Thailand. Moreover, we study the difference in household debts between municipal area and
non-municipal area. From our analysis, household debts in municipal areas are higher than
those of non-municipal areas. Comparing among different regions, we find that Bangkok has
much higher average household debt than other regions. The northern and the northern east
regions have higher household debts than the country average while the central, the east, and
the southern regions have average household debts lower than the country average."
King Mongkut's University of Technology North Bangkok. Central Library