Purpose: In this paper, the author develops a game theoretical model to understand why Union Government of India, as a third party, has used different schemes at different times in history to assist the State Governments in fighting the Naxalite insurgency. Comparing across schemes, it was found that though Matching Security Grants scheme was preferred in general, during asymmetric information scenario it led to an emergency situation wherein the Union Government had to provide the less preferred Bulk Security Grants. Later, it became difficult to withdraw these grants as the State Governments free rode by reducing own security contribution. The author finds that instead, in this scenario, Matching Development Grants are more suitable, as they incentivize the State Governments to reveal private information and help the Union Government exit its third-party role. For a practitioner involved in conflict resolution, these conclusions imply that as the desirability of policies can change diametrically overtime, Union Government must spend resources only on those heads of expenditure that provide both security and development benefits provided they aid in preventing flow of resources to Naxalites. Further, to end its assistance, the Union Government's expenditures should also complement the capabilities of the State Government rather than substituting them. These results can also guide policy in other protracted civil wars with substantial third-party intervention, which are common these days. Design/methodology/approach: The paper is an historical analysis of strategies used by Union and State Governments and Naxalites. The analysis is based on game theoretic tools supported with examples. Findings: The Union Government must provide matching grants instead of bulk grants such as Central Armed Police Forces, and the grants should be aimed at building complementarities with the state governments' security contributions. Under asymmetric information scenario, the Union and State Governments reduce their expenses incurred to fight the Naxalites. A Matching Development Grants scheme would have done better. Union Government must spend resources on heads of expenditure that provides both Development and Security benefits, to curb flow of resources to Naxalites, besides complementing the Security Contributions of the State Government. Research limitations/implications: The research is limited by disaggregated data to test the hypotheses. It is also limited by the data on hidden variables like the contribution of the Naxalites to fighting. The research is also limited to the extent that individual groups in the war like police commanders, politicians and Naxalite commanders are not incorporated. Multiple asymmetric parties are also not considered; that may generalize the model to other theaters of insurgency. Practical implications: Certain heads of expenditure such as roads, mobile communication, improving quality of investigation, preventing human rights violations by the security forces, etc. are both security and development enhancing. The Union Government's expenditures must be directed toward this end. Therefore, from a practitioner's perspective, the debate between greed and grievances exists not as a limitation but as a guide. The relevant articles of Constitution of India must be redrafted on these principles. Third-party interventions in other insurgencies may be revisited under these conclusions. Social implications: Security and Development policies are tools for controlling Naxalite insurgency, which can also be used to prevent flow of resources to Naxalites. Security and development policies to resolving insurgencies are useful at different information scenarios. Therefore, information neutral policies should be preferred. Originality/value: This paper has contributed theoretically in modeling continuing conflicts like Naxalite insurgency, explicitly. The author also shows that though the field of civil wars may have evolved along the Greed vs Grievance debate (Collier and Hoeffler, 2004), for a practitioner, the lines blur when it comes to solutions, as many heads of expenditures have features of both security and development. This paper also shows that when the Union Government faced asymmetric information scenario, the policy of matching development grants would be beneficial in long run though of limited value in short run. This is an important conclusion as the most intense period of violence was preceded by the asymmetric information scenario. Besides, it has relevance for the other civil wars with third-party intervention, such as NATO in Afghanistan. [ABSTRACT FROM AUTHOR]