To promote ride-hailing platforms' adoption of electric vehicles (EVs), two primary policies (i.e., control policy and subsidy policy) have been implemented by different governments in the real world. Which policy is more appropriate for the government under different market conditions is still a question that needs to be answered. To answer this question, we develop a Stackelberg game model to capture optimal government decisions. The results indicate that in a market with a high environmental governance cost and a large number of ride-hailing drivers who own traditional vehicles, the government should adopt the control policy; in a market with a low environmental governance cost and a small number of ride-hailing drivers who own traditional vehicles, the government should adopt subsidy policy. These findings provide meaningful policy implications for policymakers who wish to promote ride-hailing platforms' EV adoption.