Network games have been commonly used as a formalism to study the provisioning of public goods, and mechanism design as a tool to induce socially desirable effort from agents. When the type of public goods is non-excludable, i.e., an agent can opt out of the mechanism and yet continue to benefit from actions of those who remain in the mechanism, existing literature has shown the difficulty of designing transfer mechanisms that simultaneously achieve social optimality, weak budget balance, and voluntary participation. This challenge stems from the fact that the participation of some agents creates excessive positive externality, thereby reducing the incentive for others to voluntarily opt in. To get around this difficulty, we consider instead a sequential model whereby agents are invited in some order \(\sigma \) to commit to the mechanism, with each agent’s commitment contingent on the decisions of those preceding it and in anticipation of those following it. This gives rise to the notion of sequential participation referred to as \(\sigma \) -voluntary participation ( \(\sigma \) -VP), which is a full-participation outcome under the given order. We show that satisfying the conventional voluntary participation condition is equivalent to satisfying \(\sigma \) -VP for all orderings \(\sigma \) . Furthermore, we demonstrate via examples that it is possible to simultaneously achieve social optimality, weak budget balance, and full participation in the sense of \(\sigma \) -VP, especially in large systems.