A STUDY OF MICRODATA SETS FOR HOUSEHOLD SAVINGS

Abstract

In this study, we draw attention to an untapped area of research in Singapore, that is, the use of longitudinal microdata in the study of household savings. There are several advantages in using microdata rather than macrodata in the analysis of household saving behaviour. Among others, microdata gives information on future expectations, income uncertainty and other variables pertinent to the study of household savings. A survey of five major microdata sets in the United States revealed that these microdata sets have the advantage of tracking individuals and their changing economic circumstances along with behaviour over time which is particularly useful in studying the intrinsically dynamic phenomenon of savings. Consequently, studies from the microdata sets have disproved previous findings based on macro aggregates and contributed much to the existing knowledge base. These new evidences will shape future policy needs differently. More importantly, a part of these studies examine the asset and wealth profile of the elderly households in the context of household savings. It follows that we would be able to study the adequacy of old age support in Singapore. Especially valuable, future empirical works on household savings using microdata would serve as useful inputs for policy formulation regarding the rapid ageing population. In addition, the use of microdata allows us to decompose the aggregate national savings into its various components and determine the relative contributions of the public, corporate and household savings to Singapore's impressive gross national savings. On a similar note, microdata allows the analysis of the effects of the mandatory Central Provident Fund (CPF) on private voluntary savings at the micro level. Unfortunately, we have found no studies on household savings in Singapore due to the absence of longitudinal microdata sets. Among all the microdata sets at the household level, only the Household Expenditure Survey (HES) contains detailed information on household income and expenditure. However, we found that albeit HES is highly comparable to the Consumer Expenditure Survey (CEX) microdata set that has been widely used in the study of household savings in the United States, HES data has hardly been utilised for the study of household savings. The HES is unquestionably a rich source of data for more in-depth studies on households' assets and wealth profile as well as consumption pattern. Hence, we propose that Singapore builds her first longitudinal microdata set by piggybacking on the HES and expanding it. Several recommendations are made such that the HES data set can be used for the studying of intergenerational transfers, asset and wealth profile of the elderly and the saving behaviour of each cohort in the context of household savings, while retaining its original purpose for the computation and revision of the Consumer Price Index (CPI).