DYNAMIC PROGRAMMING WITH APPLICATION TO HDB HOUSING DEMAND

Abstract

Singapore's housing market is characterized by a high percentage of house ownership. Because of the durability of housing, conventional neoclassical theoretical frameworks built on static optimization analysis do not provide relevant insights to housing policies for the policy-makers in Singapore. There is a need for a customized and dynamic model to explain the consumer behavior in the housing market, which is under the active intervention by the Singapore Government. The objective of this work is to analyze a household's demand for public housing in Singapore. To achieve this, a dynamic model has been built based on the dynamic utility optimization behavior of consumers. The household is assumed to have perfect knowledge of the future economic conditions. He makes decisions on acquisition of public housing such that the total utility over his working lifespan will be maximized. In this dynamic model, durability of housing and some features of Singapore's public housing system are captured. Dynamic programming, being a versatile and comprehensible approach to the solution of dynamic optimization problems, has been used to solve the model. A numerical example of a household's utility maximization problem is set up for demonstration purpose. The household's working lifespan is divided into eight periods and it is assumed that he will make decisions on the size of public housing to acquire once a period. The optimal decision path of the household is obtained by applying dynamic programming. With reliable estimations of parameters and appropriate sensitivity tests, numerical solutions will provide significant policy implications to the policy-makers. This study has successfully explored the possibility of using dynamic models to analyze households' demand for public housing. Suggestions on the uses of the dynamic model in the other aspects of housing are made. The model may be modified to explain a household's decision on his residential location. Alternatively, one may use this model to forecast the housing demand from the theoretical viewpoint. This may help to reduce the risks associated with central planning in the supply of public housing.