Wednesday, May 13, 2009

Range Voting

My Post article on the Indonesian electoral system provoked a number of email responses with further ideas on electoral systems.  Amongst these were emails from two campaigners who advocate in support of range voting (also known as score voting).  In their opinion (and I have some sympathy with their position) range voting is the ultimate democratic preferential voting system.  In range voting each voter gives each candidate/party on the ballot list a score out of ten or out of one hundred.  The winner is determined by the candidate/party with the highest average score at the end of counting.  It is very similar to the system of determining a winner in a diving competition.  

I maintain that this system is to complex and ungainly for most electorates.  However, the more I have thought about it, the more I find the system attractive at least in an academic sense.  If you would like to know more I would strongly recommend that you check out the range voting website to understand the beauty and complexity of this novel electoral system.       

1 comment:

brokenladder said...

Well, score voting is much simpler than pretty much all ranked methods, such as Instant Runoff Voting. It's simpler in the sense that voters statistically spoil fewer ballots with score voting (whereas they tend to spoil more ballots with IRV). It is also simpler to count, and is better for election integrity, because it is additive (unlike IRV) and therefore preserves the notion of a precinct sub-total (extremely beneficial for auditing purposes). Finally, just ask any computer programmer you know to write the shortest possible program to implement an election with a variety of election methods. The score voting program will be shorter.

Finally, the simplest form of score voting is approval voting, which is just like the ordinary voting system we call plurality voting (or "first past the post") -- except that you are not limited to voting for just one candidate. You can vote for 1 or more. (This is actually algorithmically simpler than plurality voting, because there is one less rule.)

This should all be cut-and-dry, but unfortunately you have election reform "leaders" (like FairVote's Rob Richie) trying to cloud the issue. In his case this is because his ultimate goal is to achieve proportional representation via the Single Transferable Vote method (in fact, the original name for FairVote was Citizens for Proportional Representation). Since IRV is the single-winner form of STV, but is simpler, Richie and his associates see the adoption of IRV as a "stepping stone" to STV, and do not particularly care about the important qualities I'm discussing here.