The influence of a political idea rarely relies only on a few scholars and politicians. Instead, it emerges, evolves, and transforms into feasible policies when policy experts of various backgrounds and reputations cooperate and communicate through organizations and their personal networks. Previous research on the roles of policy experts in the policy making process, however, focus primarily on celebrities whose personal connections with other political actors are more publicly visible and politically influential. Two narratives are thus missing: First, what are the roles of the majority of lesser-known experts? Second, how much are experts’ cooperation and communication shaped by organizations such as think tanks? To provide a more complete picture, I construct a Twitter follow graph which covers 609 experts with valid Twitter accounts, which is about 70% of all listed experts on the websites of the Brookings Institution, the American Enterprise Institute, the Center for American Progress, and the Heritage Foundation. Then, I analyze the structure of this Twitter follow graph and explore the implication and limits of Twitter-based connections. With a DeepWalk Algorithm, I argue that the Twitter connections among experts are highly reciprocal and organizationally based. Finally, with edge-based regressions, I show that the organizational based connection between experts is significant even when the similarity of experts’ policy interest is controlled. That is, social media usage seems ineffective in exposing experts to opinions from those beyond the reach defined by their offline constraints. In all, I construct a framework to study the expert network through Twitter data and provide a complementary narrative to the existing qualitative discussions on how opinions and information may circulate among policy expert networks.