IMPACT OF SPECIAL EDUCATION SCHOOLS ON SURROUNDING CONDOMINIUM PRICES

Abstract

This study aims to explore the relationship between proximity of SES and surrounding condominium prices. This study focuses on the preference of people towards the distance of SES from their choice of home and its impact on condominium transaction prices. The primary data analysis focuses on 8 SES schools in Singapore and its pre-post effects on condominium transaction prices within a 5 year period before and after the respective schools are opened. The initial assumption and hypothesis is that SES in closer proximity to condominiums have a negative correlation their market transaction value. On one hand, there is an intrinsic need to locate Special Education Schools in close proximity to Housing and other amenities. However, we must consider the optimal placements of these schools without harming the economic development of the neighbourhood. One hypothesis to consider is if the proximity of SES influences the perception of the buyer towards a specific location’s attractiveness. Although an individual may not harbor any preconceived negative notions towards children with special needs, he has the freedom of choice and personal preference. Given the same amount of money, would he “objectively” purchase a unit located nearer or further from the SES? Primary data study was collected in Singapore with aim to empirically assess the proximity of SES on Condominium transaction values in Singapore. There are over 20 special education (SPED) schools in Singapore catering to young children who experience disorder on the autism spectrum and intellectual disabilities. For the scope of this study, the primary data covers the following 7 Special Education Schools in Singapore, focusing on children aged 7-18 years old with autism. Data was collected on condominium transaction prices 5 years before and after the date when the school was established, within a total of 2km radius from the SES. The treatment group represents condominiums within 1km from the SES which are considered in close proximity. The control group represents condominiums within 1-2km range from the SES and considered far. This housing transaction data was obtained from REALIS portal. A total sample size of 29,267 condominium prices were analysed in this study. This data was analysed in 3 parts; Summary statistics, Hedonic regression and Difference in Difference table to account for treatment and control groups effects. When dealing with transaction price, the logarithmic function is applied to it. This helps assume a log-normal distribution in terms of the statistical distribution of the values. For properties within 200m, the regression value transitioned from 0.199 to -0.00697 to -0.361. This shows a 36% decrease in condominium price for close proximity properties. For other close proximity ranges; 200-400, 400-600m, 600-800m, 800-1000m, the respective regression values are negative as well indicating -34.1%, -27.0%, -30.0%, -16.7%, -22.5% respectively. There is a clear negative causal relationship as properties within 200-600m have a lower average of -35.1% while those within 600-1000m averages -23.4%. Using difference-in-difference method, the opening of SES reduce surrounding condominium housing price by 1.98% on average. I further conduct heterogeneity analysis in Resale and New Sale types. I find that this result is driven by the New Sale market. In order to better understand the effect of treatment groups on price, a Parallel Line Trend graphical visual shows the trend lines for each of the groups.Over the course of the dataset period, we notice that treatment housing prices experienced a dip after 2013 and started diverging from control price further. This is a strong suggestion that in spite of market fluctuations, the treatment group does produce lower prices as a result of the SES located in close proximity.