本文使用「高雄市選民對二屆立委選舉賄選議題之看法」選後調查資料分析，結果發現： 1. 高雄市二屆立委選舉期間，賄選現象至為普遍，有半數受訪者表示有人向他（她）買票，更有近四分之三受訪者表示他（她）的親友中有人收到買票錢；這些經驗也因而讓四分之三的受訪者覺得高雄市二屆立委選舉期間，花錢買票的情況「很嚴重」或「嚴重」，覺得「不嚴重」的只占一成多。 2. 向選民賄選的以國民黨籍候選人占絕大多數，但民進黨與其他黨派（含無黨籍）也都有候選人涉入其中。候選人競相賄選的結果，有些選民甚至被三個不同政黨的候選人請客、送禮及買票，可能有人因而發了一筆「選舉財」。但被賄選的民眾，有些並不清楚「選舉禮物」與「買票錢」是那一黨籍，甚至那一位候選人送來的。 3. 候選人（及其椿腳）傾向於以「男性、年齡大及中等教育程度」之選民為主要賄選對象，但本研究卻發現「女性、年齡大及教育程度低」的受賄者賄選效果較大，可見候選人的「賄選策略」與「賄選效果」相矛盾，這形成了當前賄選現象的第一個吊詭。 4. 根據本研究分析之結果，賄選的獲票率只有百分之十六，也就是向一百個選民賄選，得十六票。但這十六票不能全算是賄選的效果，其中可能有幾票是本來（是某種原因，而不是賄選）就會投給該候選人的。同時賄選對選票之獲得，也不能全以「正面效果」視之，固然有些選民是因為「賄選」才給票，但也有些選民會因為「賄選」而把原本要給的票投向他人。 5. 因此「賄選效果」極為有限，但「賄選行為」卻又至為普遍，這形成了當前賄選現象的第二個吊詭。

　　　　　 This study uses the post-electoral survey from the "Perception of the Kaohsiung City Citizens towards the Issue of Vote-Buying in the Second Legislator Election". The result is summarized as follows: 1. During the period of the second legislator election in Kaohsiung City, vote-buying is quite common. Half of the correspondents say that someone had given him (her) money for vote-buying. Almost three fourths of the correspondents express tat his (her) relatives had received money for vote-buying. This experience made three fourths of the Kaohsiung city citizens feel that vote-buying was either "very serious" or "serious", only about 10% of the correspondents feel the issue was "not serious". 2. The KMT candidates occupied the majority of vote-buyers, while the DPP candidates and other candidates also somewhat involved with vote-buying. Because of the competition of vote-buying among the candidates, some people were even contacted by more than three different candidates, thus making a little "fortune" during the election. Yet, some people even did not know which candidates who gave them "money" or "presents". 3. Candidates (and their assistants) are likely to select "males, aged persons and middle-level educated persons" as the objects of vote-buying. But, this study finds out that "females, aged persons and low-level educated persons" would have better effect from vote-buying. Therefore, candidates' "vote-buying strategy" is contradictory with "vote-buying effect" which has formed the first paradox in the vote-buying phenomenon. 4. According to the analysis of this study, only 16% of vote-buying is successful. But, is should be noticed that these 16 votes of every 100 voters are not directly the effects of vote-buying. Some of the 16 votes may go to the specific candidate without vote-buying. In addition, the effect of vote-buying should not be judged from the positive side only. Because some people may vote for someone else if the specific candidate gives them money for vote-buying. 5. Therefore, although "vote-buying effect" is very limited, "vote-buying behavior" very common, which has formed the second paradox of vote-buying.