席哈克總統時期的權力中心與內閣穩定

Abstract

近年來，半總統制的學術成果日漸豐碩，有關半總統制內閣穩定的研究也累積相當成果。一般認為共治時期容易因為府會對立與政策僵局，而使得內閣較不穩定，但本文認為未必如此。根據李鳳玉與藍夢荷 (2011) 的研究，本文認為，和一般文獻的看法不同，法國共治時期的內閣穩定可能優於非共治時期，其中原因是因為權力中心的不同所致。 根據既有文獻可知，法國共治時期的權力中心為總理；非共治時期的權力中心則為總統。本文認為，當權力中心為總理時，相較於權力中心為總統時，內閣將較為不穩定。權力中心與內閣穩定之間之所以呈現如此的關係，則是因為幾個因果機制造成。首先，權力中心向總統傾斜時，總統可以透過撤換閣員等方式來規避政治責任；第二，權力中心偏向總理時，總理會迴避改組以避免受到領導失敗的指責；第三，權力中心偏向總理時，若總統大選將近，總理即便想改組內閣以提高滿意度，也可能受制於總統不願配合。本研究透過比較席哈克總統任內的共治時期與非共治時期，證實了權力中心與內閣穩定之間的關係確實如同本研究所預期。此外，本研究也根據法國民意調查的結果，觀察總統或總理以更換總理或閣員來回應政策失敗的時機點與回應速度，結果發現，非共治時期權力中心偏向總統，總統可以在短時間內回應政策失敗與民意指責；共治時期權力中心偏向總理，總理為了避免改組而受到領導失敗的指責，回應政策失敗與民意的時間應該較非共治時期的總統來的長。而這些發現也印證了共治時期的權力中心（也就是總理），相對於非共治時期的權力中心（也就是總統），回應民怨的速度是比較慢的，其中的原因應當與前所述及的因果機制相關。

Based on our survey of recent research on semi-presidentialism, we see a growing literature on this topic. It has been widely claimed that cohabitation leads to political gridlock and conflict between a president and a legislature which in turn makes cabinet instable. However, our research on Chirac’s presidency provides a different view and argument. Based on the theories of Lee & Lan (2011), we argue that French cabinet stability is better in cohabitation periods than that in non-cohabitation periods. The reason is because of important differences in who wields actual political power. It is known that the French power centers are the prime ministers and the presidents during cohabitation and non-cohabitation periods, respectively. We indicate that cabinets should be more stable when prime ministers hold power, compared to the non-cohabitation periods when presidents hold power. Three causal mechanisms explain why. First, when a president controls real power, he can easily shift off political responsibility through reshuffle. Second, when prime ministers have real power, he will avoid reshuffle to keep off the accusation of policy and leadership failure. Third, when the power center is a prime minister, even if he is willing to improve his personal reputation by changing cabinet composition, a president may not cooperate. The findings of this thesis provide strong support for our argument. Moreover, we also observe how fast the different power centers react to negative public opinion, based on French surveys and cabinet data. We find that the reaction time is relatively shorter during non-cohabitation periods than during cohabitation periods, which is again in accordance with our expectation.