Quantifying the influence of policy variables on rooftop solar economics with evidence from India

India is one of the world’s largest electricity consumers, with nearly 75% of its power generation still dependent on coal-fired thermal plants. To reduce carbon emissions and energy-import dependence, the Indian government has increasingly promoted residential rooftop solar photovoltaics (PV) through subsidies and net-metering policies. However, the effectiveness of these policy instruments varies across Indian states owing to differences in regional policies and economic conditions. This study develops an interpretable machine-learning framework to quantify the relative importance and sensitivity of policy variables influencing the return on investment (ROI) of residential rooftop solar installations across Indian states and Union Territories. State-wise electricity tariffs, installation costs, subsidies, and net-metering compensation were incorporated into a discounted cash-flow (DCF) model to evaluate Payback Period and Net Benefit post Payback Period. Conditional Permutation Importance (CPI) and Double Machine Learning (DML) quantified feature importance and elasticity while accounting for nonlinear interactions and multicollinearity. Grid electricity tariffs were the dominant determinant of rooftop-solar economics across most regional settings, accounting for up to approximately 89% of the normalized feature importance for net benefit and 52% for payback period. In contrast, net-metering compensation dominated in geographically remote states, contributing up to approximately 94% of the feature importance. Elasticity analysis revealed significant regional variation in policy sensitivity, indicating that a uniform national policy framework may not maximize rooftop-solar adoption or financial returns. The proposed framework provides policymakers with a quantitative tool for identifying region-specific policy priorities and can be extended to other countries with heterogeneous regional energy policies