On Fri, Nov 29, 2002 at 08:54:30AM +0100, Matthias Urlichs wrote: > For a trivial example, consider a vote split three-ways between > - cabd > - badc > - dacb > > Run it through Condorcet: A is unbeaten, thus the winner.
Minor nit: A is unbeaten and A beats all the other options, thus is the winner. If A was unbeaten but did not defeat all the other options, the outcome would be different. > > Actually, it is stronger evidence against McGann's thesis, that majority > > rule provides the best protection for the minority. > > Personally, I don't think that thesis is true _at_all_. In a mathematical sense what he's saying is true. For example, if a minority which wants to exclude other minorities is fully represented in an election to change the election process to favor them, they've got a better chance of doing this with majority rule than with any other process which does not explicitly favor them. And, even if they win that election, it still would be true afterwords that a majority rule system would have protected other minority interests better than the new process. -- Raul