本研究以政治學的分配理論為基礎，利用行政院「擴大內在需求方案」當中屬於內政部營建署所主管的「創造城鄉新風貌」補助款政策利益於八十八至八十九年度分配在各縣市（含臺北市與高雄市）的情形來觀察該項政策利益分配的結果，藉以判斷獲得利益的縣市在政策利益分配的型態上究竟是屬於最小獲勝聯盟？還是通通有獎的型態？ 本研究發現，「創造城鄉新風貌」政策利益的分配既是最小獲勝聯盟又是通通有獎的型態。最小獲勝聯盟的理由是因為當我們將政策利益定義為現金價值的補助金額時，我們發現國民黨籍立法委員比例越高的縣市或是立法委員越資淺的縣市獲得越多的政策利益。而當我們將政策利益定義為補助計畫案數時，我們卻發現國民黨籍立法委員比例越低的縣市或是立法委員越資深的縣市獲得越多的政策利益。政策利益分配型態是通通有獎型態的理由在於當我們合併觀察上述兩個最小獲勝聯盟時發現，這兩個最小獲勝聯盟的成員並沒有交集（概念上來說）。也就是說，國民黨籍立法委員比例越高的縣市或是立法委員越資淺的縣市獲得越多的補助現金，而國民黨籍立法委員比例越低的縣市或是立法委員越資深的縣市獲得越多的補助案數，大家各取所需不同類型的政策利益，彼此井水不犯河水。雖然同時從兩種不同形式的政策利益定義來觀察「創造城鄉新風貌計畫」，其補助利益的分配是屬於通通有獎的型態。但是行政機關所分配的多數實質利益（現金價值的補助金額）絕大部分還是由國民黨籍立委或資淺的立委所掌握，只是分配較多的補助計畫案給非國民黨籍或資深的立法委員，使其可以向選區的選民交代。獲得多數的計畫案數並不代表獲得多數的實質經濟利益，因此「創造城鄉新風貌計畫」補助利益的分配型態實質上仍屬於「最小獲勝聯盟」。

　　　　　The present study explores whether the pattern of geographic distribution of policy benefit among cities and counties in Taiwan is minimum winning coalition or universalism. By looking at the grant allocation of "Creating New Features of City and Rural Areas" offered by Department of Interior in the Taiwan central government from fiscal year 1999 to 2000, we find that the pattern of the distribution of grant is either minimum winning coalition or universalism. We say it is minimum winning coalition because the minority exploits the majority of constituencies either we define the policy benefit as the total amount of money or define the policy benefit as the number of projects respectively. We find that, on the one hand, the disproportionate amounts of money flow to the cities or counties with higher ratio of KMT legislators or with junior legislators. On the other hand, we find that most of projects go to the cities or counties with higher ratio of non-KMT legislators or with senior legislators. When we put these two things together, we say it is universalism because the members of winning coalition who receive disproportionate amounts of money and members who receive most of the projects are exclusive to each other. In other words, cities or counties receive disproportionate amounts of money tend not to receive disproportionate number of projects, and vice versa, in the manner that each coalition gets whatever it needs and would not be deprived of the other. However, we still allege that cities or counties with higher ratio of KMT legislators or with junior legislators are the real winners. Because we find that the coefficient of correlation between the amount of money and the number of projects, which cities and counties receive, is significantly negative in the sense that cities or counties receive more money tend to receive less projects, and vice versa. It implies that cities or counties with higher ratio of KMT legislators or with junior legislators tend to receive more money despite they got few projects. Based on these results, we conclude that minimum winning coalition is still at work in the substantive sense (monetary value).