seats were elected from 73 single-member districts,with a much smaller portion of seats being produced by proportional representation party votes that cross the 5%threshold. Such a winner-takes-all game not only marginalized smaller parties psychologically,but also created mechanically a big gap between vote and seat shares (38.2%vs.23.9%)of the DPP in the 2008 legislative election.3 Such a game continued in 2012 when the KMT obtained 56.5%of seats with 48.2%of district votes and 44.6%of party votes,and the DPP gained only 35.4%of seats despite 43.8%percent of district votes.Meanwhile,the two tiny parties,the PFP and the TSU crossed the 5%threshold of party votes and gained three seats each.However,because the reduced gap between the KMT and the DPP in seat shares,the effective number of parliamentary parties in 2012 (2.23)is closer to the standard value (2.0)for the two party system than the figure in 2008(1.73),should we employ the definition and calculative formula proposed by Markku Laakso and Rein Taagepera (Ns=1/>vs ).4 The electoral outcomes before and since 2008 have demonstrated the effectiveness of institutional variables in shaping Taiwan's two-party system in the legislative system.The KMT and the DPP now have 56.6%and 35.4%of seats respectively,while the PFP or the TSU has less than 3%of seats,which may vanish in the 2016 legislative election due to the decreasing influence of PFP's chairman James Soong and TSU's spiritual leader Lee Teng-hui.Among the 73 district seats,the KMT enjoys 44 while the DPP has 27.Despite the big difference of seat shares of the two parties,the gap of their vote shares is not so big (48.2%vs.43.8%for district votes).Should party votes gained by the PFP and the TSU be converted respectively to the KMT and the TSU during next legislative election,their seat shares ratio might be 9:8.District seat shares may be drawn closer as well if the DPP can get more votes next time. The gap between vote and seat shares of the two parties is the mechanical function of the single-member-district system.Another function of the system is favoring the incumbents.During the 2012 legislative election,60 district members ran for reelections and 45 won the campaigns,accounting for 61.6%of district seats.Among the 73 districts, 19(26%)experienced power turnovers from one party to another.As a result,district seats owned by the DPP increased from 13 in 2008 to 27 in 2012.If similar changes happen again in 2016,the KMT would no longer maintain or even lose its majority position (64 versus 40 seats)in the current legislative organ. Table 3 Vote and Seat Shares in Legislative Elections(1992-2012) Year 1992 1995 1998 2001 2004 2008 2012 Total seats 161 164 225 225 225 113 113 KMT Vote% 52.7% 46.1% 46.4% 28.6% 32.8% 53.5%(51.2%) 48.2%(44.6%) Seat% 58.4% 51.8% 55.1% 30.2% 35.1 71.9% 56.6% DPP Vote% 31.4% 33.2% 29.6% 33.4% 35.7% 38.2%(36.9%) 43.8%(34.6%) Seat% 31.7% 32.9% 31.1% 38.7% 39.6% 23.9% 35.4% 3 Maurice Duverger has made a conceptual distinction between psychological and mechanical effectiveness of single-member-district plurality system on the number of relevant parliamentary parties.See Maurice Duverger,Political Parties (London:Lowe &Brydone Ltd,1964),pp.224-226. 4 Markku Laakso and Rein Taagepera,"Effective Number of Parties:A Measure with Application to West Europe," Comparative Political Studies,Vol.12,No.12(1979),pp.3-27. 66 seats were elected from 73 single-member districts, with a much smaller portion of seats being produced by proportional representation party votes that cross the 5% threshold. Such a winner-takes-all game not only marginalized smaller parties psychologically, but also created mechanically a big gap between vote and seat shares (38.2% vs. 23.9%) of the DPP in the 2008 legislative election.3 Such a game continued in 2012 when the KMT obtained 56.5% of seats with 48.2% of district votes and 44.6% of party votes, and the DPP gained only 35.4% of seats despite 43.8% percent of district votes. Meanwhile, the two tiny parties, the PFP and the TSU crossed the 5% threshold of party votes and gained three seats each. However, because the reduced gap between the KMT and the DPP in seat shares, the effective number of parliamentary parties in 2012 (2.23) is closer to the standard value (2.0) for the two party system than the figure in 2008 (1.73), should we employ the definition and calculative formula proposed by Markku Laakso and Rein Taagepera (Ns = 1/∑ vs 2).4 The electoral outcomes before and since 2008 have demonstrated the effectiveness of institutional variables in shaping Taiwan’s two-party system in the legislative system. The KMT and the DPP now have 56.6% and 35.4% of seats respectively, while the PFP or the TSU has less than 3% of seats, which may vanish in the 2016 legislative election due to the decreasing influence of PFP’s chairman James Soong and TSU’s spiritual leader Lee Teng-hui. Among the 73 district seats, the KMT enjoys 44 while the DPP has 27. Despite the big difference of seat shares of the two parties, the gap of their vote shares is not so big (48.2% vs. 43.8% for district votes). Should party votes gained by the PFP and the TSU be converted respectively to the KMT and the TSU during next legislative election, their seat shares ratio might be 9:8. District seat shares may be drawn closer as well if the DPP can get more votes next time. The gap between vote and seat shares of the two parties is the mechanical function of the single-member-district system. Another function of the system is favoring the incumbents. During the 2012 legislative election, 60 district members ran for reelections and 45 won the campaigns, accounting for 61.6% of district seats. Among the 73 districts, 19 (26%) experienced power turnovers from one party to another. As a result, district seats owned by the DPP increased from 13 in 2008 to 27 in 2012. If similar changes happen again in 2016, the KMT would no longer maintain or even lose its majority position (64 versus 40 seats) in the current legislative organ. Table 3 Vote and Seat Shares in Legislative Elections (1992-2012) Year 1992 1995 1998 2001 2004 2008 2012 Total seats 161 164 225 225 225 113 113 KMT Vote% 52.7% 46.1% 46.4% 28.6% 32.8% 53.5%(51.2%) 48.2%(44.6%) Seat% 58.4% 51.8% 55.1% 30.2% 35.1 71.9% 56.6% DPP Vote% 31.4% 33.2% 29.6% 33.4% 35.7% 38.2%(36.9%) 43.8%(34.6%) Seat% 31.7% 32.9% 31.1% 38.7% 39.6% 23.9% 35.4% 3 Maurice Duverger has made a conceptual distinction between psychological and mechanical effectiveness of single-member-district plurality system on the number of relevant parliamentary parties. See Maurice Duverger, Political Parties (London:Lowe﹠Brydone Ltd,1964), pp.224-226. 4 Markku Laakso and Rein Taagepera, “Effective Number of Parties: A Measure with Application to West Europe,” Comparative Political Studies, Vol. 12, No. 12 (1979), pp. 3-27