Agricultural Growth, Employment and Poverty: Theoretical and Empirical Explorations with Indian data (1970-1993)

There is a rapidly growing literature on the dual concern of promoting agricultural growth and reducing the incidence of rural poverty. However the analysis of the interaction of growth and poverty is an under researched area of economic policy. This paper attempts to further analyse these dual concerns in an integrated manner. A basic endogenous growth model is developed which explicitly includes poor households and a government that has to decide how to allocate resources to the provision of infrastructure and to the public distribution of food grains. The intertemporal maximisation clearly shows the trade-off the government is facing and the indeterminate outcome. The model derives five key relationships: an agricultural metaproduction function (which allows differing temporal and spatial technical progress), rural employment and wage functions, and relationships for the public distribution of food grains and for rural poverty. These structural equations are estimated in a simultaneous setting for fifteen Indian states using eleven years of data for the period 1970 to 1993. Care is taken in the treatment of missing values, the non-stationarity of many of the state variables, the high level of dependencies between the variables (in the form of extreme multicollinearity and endogeneity) and the presence of structural change. We believe that insufficient care has been taken with these important complications in some studies. Robust structural form, net average elasticities and reduced form impact elasticity multipliers are derived. These estimates give valuable insights into the complicated interdependencies of the policy and endogenous variables. Whilst our broad conclusions tend to reinforce the findings of recent studies there are major differences in our estimates and methodology, which includes the conceptualisation, analytic specification and application of appropriate estimation techniques.