A Bayesian Equal Part Regression Analysis of the Influencing Factors of Taiwanese People’s Regime Acceptance of Mainland China and U.S. Governments

Abstract

This study utilized the newly-designed Bayesian equal part regression (BEPR) model to analyze the Taiwan National Security surveyed data from 2015 in order to construct a model of Taiwanese people’s regime acceptance of Mainland China and U.S governments. The study also used the Bayesian Regression model to make a comparison with the BEPR model results and attempted to explore the fluctuations of post mean and post probability of non-zero coefficients for each independent variable in the BEPR model. The major findings are as follows: First, the first equal part of respondents who believe that China would coerce Taiwan to make concessions have the lesser negative level of the regime acceptance of Mainland China, while the second equal part of the respondents who have the same attitude have the higher negative regime acceptance level. The second equal part of respondents who deem the higher possibility of unification have the lesser positive view on the regime acceptance level. Additionally, the first equal part of respondents who have higher evaluation of cross-strait relations have lesser positive impact on the regime acceptance of Mainland China. Second, we obtain the results that the second-third of Taiwanese respondents who have the optimistic household economic outlook or agree to reduce the purchase of U.S. military weapons if Mainland China withdraws its missiles have a higher negative impact on the regime acceptance of the U.S. However, the third equal part of Taiwanese respondents who agree with the current “R.O.C.” country name have a higher negative regime acceptance level of the U.S.