In order to estimate the willingness to pay for environmental conservation contributions, which are non-commidities, this study estimated the amount of willingness to pay using Ulleungdo Island as an example. To this end, the payer''s willingness to pay was estimated and compared with the cost-based imposition, and a method of estimating the willingness to pay was sought considering spatial characteristics, and the marginal willingness to pay was estimated for each attribute of environmental conservation contributions.
In Chapter 2, one of the most important issues in introducing new tax charges is tax resistance from taxpayers, so this study aims to provide implications for calculating the appropriate imposition rate when introducing Ulleungdo environmental conservation contributions. To this end, this study first estimates the willingness to pay(WTP) of the payer of environmental conservation contributions by using the contingent valuation method(CVM) based on the survey. Next, the measured WTP for environmental conservation contributions is compared with the cost-based charge for environmental pollution treatment. This study confirmed that the payer’s WTP of environmental conservation contributions (KRW 20,076 per visit per person) was larger than the cost-based environmental conservation contributions (KRW 9,424 per visit per person). Through this, it can be expected that if the imposition rate is calculated based on cost, it is acceptable and no strong tax resistance will occur.
In the 3rd chapter, to find a more sophisticated wtp estimation method, it was reviewed the spatial characteristics of WTP of Ulleungdo Environmental Conservation Contributions. In order to confirm the existence of spatial autocorrelation, they were checked the dependent variables and residuals of the Logit model performed for CVM estimation. As a result, although the global spatial autocorrelation was not significant, but clusters were found for local spatial autocorrelation, raising the need to review spatial factors. A comparison of the Logit model, the Spatial Autoregressive Model (SARLM), and the Geographically Weighted Logit Regression Model (GWLR) confirmed that the GWLR model was the best fit. Comparing the Logit model and the GWLR model, the average WTP amount of the GWLR model is about 95.7% of the average WTP amount of the Logit model. And in the GWLR model, the WTP distribution of individual observations is also concentrated at lower amounts than the Logit model. Considering spatial factors, the lower average amount of WTP will be an implication for the introduction of the Ulleungdo environmental conservation contributions system. In addition, as a result of reviewing the local characteristics of WTP of individual observations identified in the GWLR model, it can be confirmed that there are HH-type high clusters in the metropolitan area and LL-type low clusters in the southern area.
In Chapter 4, this study aims to find ways to increase tax acceptance by reflecting the preference of taxpayers in the expenditure sector of financial resources raised by Ulleungdo environmental conservation charges. In this study, the attributes of environmental pollution that may occur due to the increase in tourists to Ulleungdo Island are classified as waste emission, air pollution, and biodiversity reduction. In order to understand the preference for each attribute, the Marginal Willingness To Pay(MWTP) for each attribute was estimated using the Choice Experiment(CE) among the non-market goods valuation methods. As a result of the analysis, the Marginal WTP for improving the environment from the worst level to the current level was KRW 20,689 for biodiversity conservation, KRW 18,491 for waste emission treatment, and KRW 12,735 for air pollution purification. Therefore, it is possible to establish an expenditure plan that reflects the preferences of payers, such as an expenditure budget that considers the proportion of the MWTP by attribute.
In this study, to consider spatial factors was not reviewed for the method of estimating the marginal willingness to pay (MWTP) by attribute according to the choice experiment method (CE). Therefore, they are left as future research challenges the review of the existence of spatial autocorrelation for each attribute constituting the environmental goods and the estimation of the marginal willingness to pay for each attribute reflecting it.