In this talk, I propose a new approach to studying elite politics by focusing on changes in network structures that emerge from elites’ informal conversations. I apply this method to solve two interrelated puzzles from China’s early reform era (1977-1992): on one hand, why did intense political conflicts frequently occur (often among reformers) despite their common reformist goal? On the other hand, why did economic reform persist despite these intense power conflicts? To tackle these issues, we employ network analysis and narrative to analyze an original dataset of elite conversations and primary sources that have only been made available recently. We find that elite conflicts were contingent on the relational structure of their conversational interactions. Meanwhile, the persistence of reform was the result of a dynamic process of elite clashes and the rise and fall of mass politics, leading to compromises on economic policy while ultimately restricting the possibility of political pluralization. By integrating micro-sociological theories, historical sociology, and network analysis, this project offers significant methodological and theoretical implications for unpacking the black box of elite politics and its role in macro-historical change.