Elections

At the GNU Public Dictatorship we are nothing if not dedicated to serving you, our world citizens, which is why we have "elections" every couple of years to choose potential new members of the Board of Dictators.  While there has been a very low historical rate of acceptance (one new board member ever) there has been a great deal of dialogue about the significant issues facing us, and that, we believe, is the primary benefit from elections.

Unfortunately, in most democratic societies, including the United States of America, which is, technically, not a democracy as business is conducted on behalf of citizens by representatives they choose, but for the purposes of this argument the label will suffice, elections have degenerated into the basest form of competition, where name recognition and slogans trump substance and where money plays all too great a role.  At the GPD we would like you to know that we would never put up with such foolish nonsense, and were our "elections" to tend that way we would curtail them immediately, while at the same time we do believe that the "voice of the people" is generally correct.  The problem we encounter is that the "voice of the people" is impossible to assess analytically to any degree of certainty, as people's opinions are too easily swayed by factors as trivial as that they didn't have a full breakfast.  What's more, there are mathematical proofs that no group decision method can be guaranteed to always be "fair," where fair is defined as taking into account everyone's preferences and creating an group decision that is true as much as possible to all members in the group.  To illustrate this, we must simply create three fictitious voters and three candidates for some office.  Let's call the voters Vlad, Victor, and Vera, and the candidates Camille, Catherine, and Christopher.  Let us say that Vlad prefers Camille over Catherine, Catherine over Christopher.  Let us say that Victor prefers Catherine over Christopher, but Christopher over Camille, and that Vera prefers Christopher over Camille and Camille over Catherine.  If we treat each voter as equivalent and combine their preferences (assuming they are of equal strength) then we end up with a cyclical preference graph, where Camille is preferred to Catherine who is preferred to Christopher, who is preferred to Camille, clearly illustrating the difficulty of combining even the preferences of three voters.

At the GPD we are nothing if not innovative, which is why we do all of our real work in Board Meetings.  The three of us discuss the issues, and more often than not a decision becomes apparent immediately.  In some circumstances we disagree, and then we keep discussing it until it becomes apparent which option is the best for society.  You might question how this works, but it is quite simple: members of the Board of Dictators are willing to increase or decrease the strength of their own preferences relative to those of other Board members, allowing an optimal decision to be reached.  Because we solve the problem of preference combination, we are able to provably make socially optimal choices without encountering a contradiction.