Analysis of a Time Series of Household Expenditure Surveys for India

Abstract

ECONOMETRIC studies of household expenditure occupy an important place in government policy formulation with estimates of expenditure and price elasticity proving useful in several planning models. Although the empirical literature is large, relatively few studies have considered the simultaneous impact of total expenditure, price and family size on household demand using time series of budget data and within a framework that is consistent with economic theory. Weisskoff (1971) and, more recently, Sener (1977) have tested and rejected the hypothesis of negligible price effects and established the need to incorporate price and size variables in any meaningful study of expenditure patterns in developing countries. The limited literature on estimation of demand systems from budget data includes Tsujimura and Sato (1964) on Japan, Bhattacharya (1967) and Joseph (1968) on India and, recently, Muellbauer (1977) and Pollak and Wales (1978) on British data. The present exercise differs from the above in investigating the impact on budget share, rather than quantities, and in using a variant of a recent model (AIDS)' that is consistent with economic theory without requiring additive separability of the utility function. The model is introduced by Deaton and Muellbauer (1978). The principal objectives of our exercise include (a) extension of the AIDS model by including family size and applying it to Indian budget data to estimate expenditure, price and size elasticities, (b) testing the hypotheses of (i) no price effects, (ii) no economies of household size, (iii) no money illusion, and (c) comparison of the Indian expenditure pattern with those of other developing economies. The AIDS in its micro version and its adaptation for use on the published data is discussed in section II, and the data are described in section III. Section IV presents the results and we end with the concluding note of section V.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 80 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.