납세자의 정서가 정부의 조세정책 결정에 미치는 영향

In this paper, the effect of taxpayer tax resistance on the tax policy-making process was analyzed. Specifically, in order to measure the tax sentiment for tax policy, a tax sentiment dictionary was established using a textmining technique and a tax sentiment index was developed. A total of 12 tax policies with a significant level of Naver comment were selected and used as tax resistance episodes among the tax law amendments, the Ministry of Public Administration and Security, and emotional scores were established by collecting comments on credit card income deduction reduction policies in 2019. In order to calculate the tax sentiment index, Naver’s comments for each policy were text-minimized in units of words, and then the tax policy ontology established was signed. In this paper, logistic regression analysis was conducted as an example to confirm the correlation between the government's tax policy decision and tax sentiment. As a result of the analysis, it was found that the negative emotions of taxpayers are likely to withdraw government policies. Specifically, if the tax sentiment index is lowered by one unit, the likelihood of the government withdrawing the policy increases by about 1.5%, and when the tax sentiment index –69.87 is set as a threshold, the classification accuracy of the model is observed to increase to 75%.