Analyzing Factors Affecting Tax Morale Levels in Turkey Using Alternative Ordered Response Models

Abstract

The concept of tax morale is introduced to measure the reason of paying taxes by subjects. The main objective of this paper is twofold. Firstly, the present paper examines potential factors that may influence the degree of tax morale in Turkey. Secondly, the paper aims to determine the best model to explain the determinants of the degree of tax morale by a comparison of alternative ordered response models. For these purposes, a written-questionnaire was administered to 943 subjects in three cities of northern Turkey, namely, Erzurum, Erzincan, and Bayburt. Standard ordered probit (OPROBIT) and alternative ordered response models including generalized ordered logit model (GOLOGIT), partial proportional odds (PPO) model and heteroskedastic ordered logit (HOLOGIT) model were performed to determine the best model fit. Results reveal that marital status, monthly income, and operation time were positively correlated with higher degrees of tax morale. The empirical findings also indicated that higher educated subjects were more likely to have lower degrees of tax morale. Tax amnesty, higher tax load, tax system, tax policies, previous negative experiences, and efficient use of public expenditures were found to have statistically significant impact on the degree of tax morale among subjects.