회계이익과 과세소득 및 관련 세액의 비교

This paper empirically shows whether tax researches using book income and tax expenses are valid as compared with tax income and taxes payable, which topic has been hardly dealt with. The approach to comparing the degree or strength of linear relationship between the two income－tax models is used to verify the validity. The correlation coefficient and the regression coefficient(slope) are selected as the measures of the degree or strength. 2004～2007 years’ data of 283 listed non－financial firms are used for analyses. The empirical results using variables deflated by total assets are as follows. The correlation and regression analyses of tax expenses on book income show larger correlation and regression coefficients as compared with those of taxes payable on tax income. These results can be interpreted that the degree or strength of linear relationship between book income and tax expense is higher than that of relationship between tax income and tax payable. These results support the validity of tax researches using book income and tax expenses as compared with those using tax income and taxes payable. Additional regressions using variables deflated by sales show different results from the above. According to the additional results, the tax income－tax payable model has a closer income－tax relationship than the book income－tax expense model. It means the results are sensitive to deflating variables. Nevertheless, asset－deflated results show stronger linear relationship than sales－deflated results. This paper may have a limitation of sample selection bias because of the sample selection criterion of firms with positive income.