民主失靈與會議議程操縱－對董事會股東會的啟示
此文編譯自Harvey S. Rosen和Ted Gayer合著的「Public Finance」一書，第9版，Chapter 6 Political Economy，第110頁。系合法出版物及廣泛使用之教材，應無違法違規情由。切望明察！！！！！！！！！！
民主有多種實現程式，一致可通過多種途徑求達。民主實現的程式有很多種，如過半決「MajorityVoting Rules」，超多決「Supermajority Voting Rules，如3/5，2/3，90%等」等。此處討論其中兩方式實現之民主，可能出現的失靈情況，即過半決和超多決「以2/3為例」。
以下討論，有助於思考民主或一致的實質內涵，及其實現程式。有時，民主的實現程式定義了民主。真理有時掌握在少數人手中，而民主成為多數人偏見的暴政。這也啟示我們思考董事會、股東會的投票表決程式。
需要注意的是，儘管過半決民主和超多決民主，可能有某些不足，但並不表示獨裁是正義的，或者民主是不正義的。
Because unanimity is difficult toattain, voting systems not requiring it may be desirable. With a majorityvoting rule, one more than half of the voters must favor a measure to gain approval. 鑒於一致難殝，強求一致的公決或非所需。在過半決，惟過半選民投票某案，方得通過。在超多決，更須超過約定多數的選民傾向某專案，才謂公允。
Although the mechanics of majorityvoting are familiar, it is useful to review them carefully. 雖過半決及超多決之機關為眾所熟，然仔細檢視仍有禆益。
導向均衡之選民喜惡
Consider a community with three voters,Brad, Jen, and George, who have to choose among three levels of missileprovision, A, B, and C. Level A is small, level B is moderate, and level C islarge. The voters' preferences are depicted in Table 1. Each column shows howthe voter ranks the choices. For example, Jen most prefers level C, but given achoice between B and A, would prefer B. 設某村有選民者3，子，丑，寅，擬就3導彈「國防服務供給」水準公決。水準A規制小，B中等，C規制大。選民喜惡如表甲所示。每列列示了選民對選項的偏好。比如，丑傾向於水準C，如若在B、A間抉擇，選B。 [td] Table 1 Voter preferences that lead toan equilibrium
[td] 表甲導向均衡之選民喜惡
設一公投，決於A、B、C3者。則，子投A，丑投C，寅投B。A、B、C各得一票。依過半決，或依2/3超多決，公決失敗。
Suppose an election were held onwhether to adopt A or B. Brad would vote for A while Jen and George would votefor B. Hence, B would win by a vote of 2 to 1. 設一公投，決於A、B2者。則，子投A，丑寅投B。依過半決，或依2/3超多決，水準B專案以2：1勝。
Similarly, if an election were heldbetween B and C, B would win by a vote of 2 to 1. 類似地，設一公投，決於B、C2者。水準B專案以2：1勝。
Level B winsany election against its opposition, and thus is the option selected bymajority rule. Note thatthe selection of B is independent of the order in which the votes are taken. 水準B專案在二公決中勝出，成公允之優選專案。注意，B之優勝與公決次序無涉。
依選民喜惡表甲，將公決安排及結局總匯於下表乙。專案後括號內的數字示意所得選票數。箭頭後表示結局。 [td] | | | | | | | A、B、C | A(1)=B(1)=C(1) -> 0 |
|
| 0 | | A、B | A(1)<B(2) -> B | B、C | B(2)>C(1) -> B | B | | A、C | A(1)<C(2) -> C | B、C | B(2)>C(1) -> B | B | | B、C | B(2)>C(1) -> B | A、B | A(1)<B(2) -> B | B |
表乙導向均衡之選民喜惡公決結局表
可見，依過半決，或依2/3超多決，若3者同決，爭生異存，同求而不得。若兩兩配對而決，水準B專案優勝，同求異存，且與公決次序無涉。
Majority decision rules andsupermajority decision rules do not always yield such clear-cut results. 過半決和超多決，終非利器，結局亦非清歷如是。
導向迴圈之選民喜惡
Consider the preferences depicted inTable 2. 設選民喜惡如表丙。其中，僅丑位序2、3之喜惡與表甲反。 [td] Table 2 Voter preferences that lead tocycling
[td] 表丙導向迴圈之選民喜惡
設一公投，決於A、B、C3者。則，子投A，丑投C，寅投B。A、B、C各得一票。依過半決，或依2/3超多決，公決失敗。
Again, imagine a series of pairedelections to determine the most preferred level. In an election between A andB, A would win by a vote of 2 to 1. If an election were held between B and C, Bwould win by a vote of 2 to 1. Finally, in an election between A and C, C wouldwin by the same margin. 同樣，設一系列兩兩配對公投，以決優選水準專案。A、B決，A以2：1勝。B、C決，B以2：1勝。最後，A、C決，C亦以2：1勝。
This result is disconcerting. The firstelection suggests that A is preferred to B, the second that B is preferred toC. Conventional notions of consistency suggest that A should thereforepreferred to C. But in the third election, just the opposite occurs. 此結局令人訝異不安。AB決A優於B，BC決B優於C。依常理，A優於C。然AC決A劣於C，悖於常理。
Although each individual voter'spreferences are consistent, the community's are not. This phenomenon isreferred to as the voting paradox. 諸選民喜惡一以貫之，而全村喜惡莫衷一是，此稱公決悖論。
Moreover,with the preferences in Table 2, the ultimate outcome depends crucially on theorder in which the votes are taken. Ifthe first election is between propositions A and B and the winner (A) runsagainst C, then C is the ultimate choice. On the other hand, if the firstelection is B versus C, and the winner (B) runs against A, then A is chosen. 且，依表丙喜惡，結局關涉公決次序。若AB決在先，勝出之A與C決，C優勝。若BC先決，勝出之B與C決，A優勝。
依選民喜惡表丙，將公決安排及結局總匯於下表丁。專案後括號內的數字示意所得選票數。箭頭後表示結局。 [td] | | | | | | | A、B、C | A(1)=B(1)=C(1) -> 0 |
|
| 0 | | A、B | A(2)<B(1) -> A | A、C | A(1)>C(2) -> C | C | | A、C | A(1)<C(2) -> C | B、C | B(2)>C(1) -> B | B | | B、C | B(2)>C(1) -> B | A、B | A(2)<B(1) -> A | A |
表丁導向迴圈之選民喜惡公決結局表
可見，依過半決，或依2/3超多決，若3者同決，爭生異存，同求而不得。若兩兩配對而決，結局關涉公決次序。
Under such circumstances, the abilityto control the order of voting—the agenda— confers great power. Agendamanipulation is the process of organizing the order of votes to assure afavorable outcome. 本此，掌控公決次序之力－議程－巨大。議程操縱，乃經由公決次序安排以保障有利結局之過程。
A related problem is that paired votingcan go on forever without reaching a decision. After the election between A andB, A wins. If C challenges A, then C wins. If B then challenges C, B wins. Theprocess can continue indefinitely, a phenomenon called cycling. 不止於此，配對公決綿綿續續，求同無期。AB決，A勝出。勝出之A與C決，C勝出。若勝出之C再與B決，B勝出。此過程綿延不絕，謂迴圈。
此文編譯自Harvey S.Rosen和Ted Gayer合著的「PublicFinance」一書，第9版，Chapter 6Political Economy，第110頁。
| 
雷达卡
发表于 2014-4-30 10:26:35
提升卡
沉默卡
喧嚣卡
变色卡
千斤顶