為探討政黨在SNTV選制中的有效提名行為，本次從政黨提名數的最適性（optimality），以及選票動員配置的有效性兩個面向來測量。初步發現，從提名失誤的角度觀察，國民黨進入九零年代後失誤率增加，伴隨的是超額席次利益的降低；相對的，民進黨則呈現相當的學習能力，在選票增長有限的情況下，透過降低提名失誤，大幅擴張席次率。但是在選票有效動員的面向來看，故事略微不同，國民黨配票能力仍強，民進黨一有長進，差別的是速度快慢，造成目前不同的選舉命運。

To identify the optimality of nominating behavior of both KMT and DPP, this paper produces two kinds of indicators, nomination error and effectiveness of vote mobilization, to evaluate party’s nomination decision from 1983 to 2001 legislative elections. This research finds that KMT committed more nomination errors in 1990’s than before, and induced the loss of seat bonuses. By contrast, DPP learned to reach the optimality by reducing both nomination errors and unused votes to produce positive seat bonuses. But the indicator of effective vote mobilization tells a different story that KMT was still strong at allocating vote while DPP made progress on turning voted into seats. The different electoral fortunes of these two parties were a result of learning capacity discrepancy for coping with changing electoral environments.

單記非讓渡制;國民黨;民進黨;席次紅利;政治學習

SNTV; KMT; DPP; Electoral Politics; Seat Bonus; Political Learning