Pulapre Balakrishnan The Growth and Structure of Savings in India by B L Pandit; Oxford University Press, Delhi, 1991; pp xi + 139, Rs 190.
IN the grand manner of theorising about economic development savings is a variable that has been given much importance. When it came to studying the developing economy this was particularly emphasised. For instance, in the classical view contained in the model of Lewis, savings is a pre-requisite for investment, which is itself the engine of growth. Classically, the problem of development is seen as one of raising the level of savings to levels considered commensurate with the capital stock necessary to generate the targeted level of output. This is the origin of the notion of the savings constraint, a notion that is believed to have provided the organising principle for India's First Five- Year Plan If it is accepted that raising the level of savings thai an economy generates is crucial to its long run health, then attempting to understand its determination becomes a worthwhile exercise. B L Pandit's Growth and Structure of Savings in India is one such attempt. Arguing, correctly, that the factors governing the different sources of savings in India vary, the author analyses the determination of the savings of households, the private corporate sector and of the government as a whole separately. Starting with households, much attention is paid to testing the theories of household behaviour with respect to savings that were originally propounded for the United States economy after the Second World War. Both cross-section and time-series data are analysed. With respect to aggregate savings over time, Pandit appears concerned to establish his predilection that while the marginal propensity to save rises with income, the income elasticity of saving actually declines. This not entirely intuitive proposition, he points out appropriately, is a property satisfied by only a limited range of specifications; for instance, the semi-logarithmic S/X = a + bJogX, where S = the real value of savings and X real income With this as template a range of specifications is estimated (see pages 31 and 58). It is argued that, since the marginal propensity to save out of rural incomes is lower than that out of non- agricultural incomes, shifts in the income distribution that arc favourable to agriculture arc likely to reduce the ratio of aggregate savings to aggregate income. Thus the specification is widened to include the ratio of agricultural to non-agricultural income and the inflation rate (or in alternate runs the expectation of inflation) as independent variables. The estimate of such an equation is not unsatisfactory in that while the inflation variables are statistically insignificant, the other two variables are. However, the overall explanatory power is not par ticularly high at about 65 per cent.
