The most serious drawback of the current building assessment method is that it does not take account for a ‘building grade’. It is widely recognized that the building grade or workmanship of construction plays an important role in estimating build...
The most serious drawback of the current building assessment method is that it does not take account for a ‘building grade’. It is widely recognized that the building grade or workmanship of construction plays an important role in estimating building value. However, the main reason why having not incorporated it in the assessment model is surveyor’s (public officials in assessment office) arbitrariness in building grading. This study identifies important factors affecting the building grade, constructs a proportional odds logit model, and predicts grades of individual buildings. The identified factors to help to predict building grades include physical variables such as the building use, the structure and the supplementary facilities. Land price (assessed land value) was also found to affect building grades in the study. In particular, high land price was found to force high-class buildings to be constructed for maximizing scarce land resources in neighborhood. In addition, a spatially lagged variable was constructed and added into the model in order to reflect the tendency of buildings with similar grades being clustered together. The inclusion of the spatially lagged variable resulted in remarkable increase in classification accuracy of building grades. We expect that the current building assessment method can be improved by providing surveyors with classification results estimated from our proportional odds logit model, and therefore minimizing arbitrariness in building grading. It is anticipated that the result of this study could help to reduce the discrepancy between the assessed and the market value, and to enhance the property tax equity for buildings.