THE USE OF MULTIPLE REGRESSON ANALYSIS APPROACH IN PROPERTY TAX ASSESSMENT (CASE STUDY: ELDORET MUNICIPALITY)
warning: Illegal string offset 'url_cutoff' in /home/unonbi/public_html/uon/modules/acquia/link/link.module on line 385.

Abstract


Property taxation is a major source of revenue to the county governments in Kenya as
provided for by the constitution of Kenya 2010. It accounts for about 20% of the total
revenue collected by local (now county) governments (Olima, 2010). Property tax is imposed basing on the value given by the valuer in the valuation roll. Preparation of the valuation rolls thus requires the valuer to use a more accurate and reliable method devoid of subjective and arbitrary adjustments to comparable sales in the market.
This study has looked at the different methods used by valuers in valuing land ranging from allocation to income capitalization, residual method cost method and sales comparable method. Application, strengths and limitations of these methods have been analyzed in detail in this study. Also discussed in detail are the various factors that determine the value of land. This study proposes the use of statistical tool, Multiple Regression Analysis (MRA) to achieve objective and more accurate values of properties in a robust manner. As a scientific method, it enables the valuer to describe accurately the relationship between sale price and the various factors that contribute to the value of land, test the significance of these factors, and apply the resultant econometric indicators to a credible estimate of value for a particular property. The study was carried out within Eldoret Municipality where prices of thirty two pieces of land were used. The selection of the pieces land was based on simple random sampling. Information required for each piece of land were; actual sale price, area of plot, distance from the main road, distance from CBD, distance from recreational facility, distance from nearest shopping centre, distance from nearest police station, flood area on plot, slope, sub soils, frontage, water facilities, sewerage facilities, ground coverage, plattage and tenure. These factors were used to develop a land valuation model after regressing the various independent variables against sale price (dependent variables) using SPSS version 16.0. The results obtained showed the following variables were significant plot size, type of subsoil, ground coverage, tenure and distance to the CBD. Ground coverage, distance to CBD and plot size were the most critical accounting for about 79.6% of the total variation of land value in Eldoret Municipality. The study recommends the use Multiple Regression Analysis in property tax assessment in order to obtain objective values in an efficient and robust manner. It further recommends adoption of digitised and computerised rating system based on MRA techniques.

Documentation :