在各國職業運動的發展中，自由球員制度的出現醞釀出了大革命。本文以一賽局模型，利用Nash 議價及 Shapley 值等數學模型的輔助，嘗試探討自由球員制度出現後球員及球團效用的增減。最後，我們得到談判力量大小左右了勞資雙方談薪優勢或劣勢這樣的結果。當球團擁有高於球員的談判力量時，即使在自由球員制度下球團仍然獲益。

本研究旨在運用經濟的觀點與計量的方法來探討中華職棒球員的表現與薪資間的關係。究竟中華職棒勞動市場中，對於球員是存在超額給付(overpayment)？還是過低給付(underpayment)？我們將運動產業勞動市場之三大決定要素結合在一起：球團收益、球員表現與球員薪資，並蒐集中華職棒1997~2005年間相關的資料，以此探討球員薪資之決定。 本文首先對台灣的職棒運動作一鳥瞰式的歷史回顧，並將中華職棒球員薪資、年資的基本統計量提出與日本職棒做一比較。接著以Scully(1974)、MacDonald and Reynolds(1994)和Krautmann(1998)所提出的薪資估計方法，將球團收益、球員表現和薪資決定變數結合在一起，估計出中華職棒球員的邊際報酬貢獻(marginal revenue product)，以此進行球員薪資的檢視。合理的情形下，球員邊際報酬貢獻應該等於球員薪資。本研究並將中華職棒的球員群組分為本土球員與洋將部分，觀察是否因為勞動市場條件的不同，而使兩群組球員薪資與表現間的關連性存在異同。在本土球員中，我們進一步以美國大聯盟的年資分類方式，將球員分為新人(apprentice)、中堅球員(journeymen)與老將(experienced players)三大群組，以觀察同樣的薪資決定變數在不同年資間造成的影響是否有異。我們並檢視年薪超過240萬的球員，其技術是否存在規模報酬遞增的情況，若存在，則中華職棒與國外成熟的職業運動一樣都存在所謂的「超級明星現象」。 實證結果發現，整體而言，中華職棒聯盟當中存在超額給付(overpayment)，而洋將因為勞動市場條件較為接近完全競爭市場，因此表現更能反映到薪資的改變上。超額給付的存在應是由於中華職棒的賽程密集度不足，且缺乏二軍與完善的球員交易制度。另外，由於球團經營與行銷手法上仍未臻成熟，中華職棒球員尚未存在「超級明星現象」。我們並以在1997~2005間250位至少完成六個球季的選手，以這些球員的薪資檢視是否存在明顯結構性的轉變，以作為將來制訂自由球員年限的參考。實證結果發現這樣的年限並不存在，我們因此推論，中華職棒應要加強基本面的基礎與改善球員交易制度，當這些基礎獲得良好的改善後，自由球員制度就會是一個水到渠成的結果。 / In this thesis we mainly discuss the relationship between players’ performance and their salaries in the CPBL (Chinese Professional Baseball League) through the point of economics view. Are the players overpaid or underpaid? Are the whole environ -ment in the CPBL mature enough to adopt the FA (Free-Agent) system? To analyze these issues, we combine three main factor of determinants in the field of sports economics together, which are team revenue, players’ performance, and players’ salaries, and use the skills in econometrics with data from 1997~2005, we illustrate an overview of these issues. First of all we make a brief description of the history of CPBL, and we also present some comparison of basic statistics between CPBL and NPB (Nippon Professional Baseball) players. In the following section we introduce the most popular model in the similar research, such as Scully’s model (1974) and the salary determination equation (MacDonald and Reynolds, 1994). Following the methodology of these research, we divide players in the CPBL into some different groups by players’ experience and nationalities. With the ZSUR (Zellner’s Seemingly Uncorrelated Regresion) model, and traditional OLS (Ordinary Least Square) regressions, we could have some empirical results which imply that players are actually overpaid. However, players are overpaid due to the degree of tension and intensiveness are neither sufficient in the CPBL rather than their skills are amateurish. We also find that unlike the English Soccer, MLB, and NBA, there does not exist superstar phenomena in the CPBL. Moreover, the conditions of environment in the CPBL are not mature enough for CPBL to apply the FA system: It would be very possible for the CPBL to face the same dilemma which the NPB are facing now if we prematurely implement the FA system.