Determining State Expenditures in India: Relative Role of Own Revenues, Central Transfers and Fiscal Deficit

This article analyses the relative role of state’s own revenues, tax and non-tax, central transfers, and borrowing in determining state expenditures. We have examined the prevailing hypotheses regarding explaining determinants of government expenditures such as the Wagner, Wiseman-Peacock and Clark’s Critical Limit hypotheses in the context of state governments in India. We have also looked at Revenue and Spend, Tax and Spend, and the Fiscal Synchronization hypotheses. We have quantified the relative magnitudes of the differential effects of these sources of financing state-level primary expenditures for two groups of states namely, medium and large states (ML) and small and hilly states (SH) in a panel data framework. In addition to the variables reflecting financing sources, this study also examines other prevailing hypotheses affecting state expenditures such as those relating to cyclical increases emanating from election years and years of economic slowdown. Based on results of the Hausman test, we have estimated a two-way random effects model. Our findings indicate that with respect to own tax revenues, the magnitude of responsiveness of primary expenditure is 0.22 for the ML group and 0.10 for the SH group of states. For fiscal transfers, these magnitudes are 0.29 for ML states and 0.20 for SH states. For borrowing, the relevant coefficients are 0.33 for ML states and 0.13 for SH states. Thus, for the ML group, the maximum responsiveness is with respect to borrowing, followed by fiscal transfers and own tax revenues. In the case of the SH group, central transfers explain the largest variation in state-level primary expenditure, followed by borrowing and own tax revenues. We also find that for two out of three general elections in the sample period namely, 2009 and 2014, there was a positive and significant response of primary expenditures. Further, in years of economic slowdown, there is a positive response except in one case where the income effect was dominant and primary expenditure also fell along with the slowdown.