Forecasting State-Level Fiscal Imbalances in India

Although forecasting models have been built for the Indian economy as a whole or its fiscal sector, there are few forecasting exercises for state economies and government finances. This paper develops a forecasting framework for projecting state-level fiscal imbalances as measured by fiscal deficit and government debt relative to respective GSDPs in the medium term. For this purpose, a panel modeling approach within a dynamic multi-equation model has been used. Since state-wise economic aggregates are not available in adequate detail, we take advantage of the linkages between state-level fiscal and economic variables and the corresponding variables at the central level or at the consolidated level of center and states. The model consists of 10 equations where five are stochastic and five are identities. The model is estimated for India’s 17 medium and large (ML) states. The sample period covers 2003–04 to 2019–20, that is, 17 years while the forecast period covers 2020–21 to 2025–26. The model reliably predicts variables in the COVID-affected year of 2020–21. Model properties are studied with respect to the quality of in-sample estimation. Using this framework, we identify Punjab, Rajasthan, Kerala, West Bengal, and Haryana as states that are headed toward larger fiscal imbalances in the medium-term requiring policy corrections.