An econometric exploration of consumer behaviour in India, 1950-51 -- 1972-73.

In this thesis, I attempt, to analyse consumption behaviour in India during, the period 1950-51 -- 1972-3. Before applying to Indian, data, the existing theories of personal consumption, have been analysed. The aggregate time-series tests of the major, theories of consumption reveal that current per-capita absolute, income strongly, influences current per-capita consumption, and that there has been, a declining trend, of consumption over, time, because of, growing urbanisation, rising, inequality in. the distribution of income and spread of education, during the period studied. The effect of past consumption standards is seen to be weak because of high priority attached to saving. For want of Data, no wealth theory, other, than, the, permanent income theory could, be tested for India. Results of testing permanent income theory reveal that both permanent and transitory incomes are equally significant in explaining personal consumption and therefore, the theory is denied by Indian data. Tests of the same theory with the help, of cross-section data for the year 1967-8 do not reveal conclusive results. I also attempt to explain, with family budget data for 1963-4, the existing regional, and rural-urban disparities in consumption of food and clothing in terms of disparities in total consumpion expenditure, price of the commodity and household size between regions and rural-urban areas within a region. Finally, an aggregate consumption function has been developed and tested in the context of an econometric model for India. The former separates income into agricultura1 and non agricultural and considers non-working population as another determinant of aggregate consumption. The time-series data for 1950-1--1970-1 reveal, signifacently higher MPC out of agricultural than, out of non-agricultural. Income and a negative influence of non-working population. The Ordinary Least Squares Estimates of the consumption function in real per-capita variables support the conclusions obtained from the Two-Stage Least Squares estimates of the aggregate specification.