立法院黨團協商制度在1999年時正式法制化，此舉被視為是增進立法效率的重要變革。先前的研究大多假定此制度使大小黨團權力均等化，甚至不少輿論及政治評論認為是讓少數霸凌多數。這些斷言不但從未被檢驗且常常被視為理所當然，此外其有效性對於多數黨議程設定能力也有深層的意涵。為了處理此重要的議題，我們依照立法院（第三次國會改革後）制度建立黨團協商賽局模型，從模型中我們導出3個假設並蒐集第五至七屆（2002~2011）的法案資料進行經驗分析。我們主要的理論發現是多數聯盟控制黨團協商的事後程序（即程委會及院長），使得多數聯盟在推動自己偏好法案通過黨團協商階段的優勢，高於主要少數黨。而另一發現是當多數聯盟內協調成本越高（低），多數聯盟在黨團協商階段優勢越弱（強）。我們資料分析結果十分支持本文研究假設，這不但推翻了過去認為黨團協商中各黨團權力均等或少數霸凌多數的觀點，並提供多數聯盟在立法中控制議程設定的更有力證據。

The party negotiation mechanism (PNM), which was formally institutionalized in 1999 in Taiwan's legislature, has been viewed as a crucial step for boosting legislative efficiency by granting veto power to every party caucus in exchange for less obstruction. It is commonly assumed that this institution has equalized the influence of party caucuses with varying sizes or even caused minority dominance. More importantly, its validity has profound implications for the agenda power of majority coalitions or parties in the legislature. To tackle this critical issue, we model the legislative process, derive three hypotheses from the model, and test them with newly collected data spanning from the fifth through seventh term. Our principal theoretical finding is that the control of majority coalitions over ex post procedure of the PNM (i.e., the Rules Committee and the Speaker) gives rise to their comparative advantage in party negotiation on bills that they push, with this edge diminishing (increasing) with higher (lower) coordination costs within a majority coalition. Our empirical analysis demonstrates strong support for our hypotheses, rejecting the assertion of equal power among party caucuses and offering further evidence for majority coalitions' agenda setting control in the legislature.