The original Bucklin ballot allowed only the expression of two preferences, which could be considered a problem if there were a large number of candidates, since the voter is not allowed to express a preference between the candidates below his second rank. And additionally it suffered from one feature that certainly could be considered a flaw: second preferences were considered equal to first. Certainly, when a voter cast a vote with rankings in this manner, he might expect that his ballot would count more for his first-preferred candidate than for his second. The system adopted in Oklahoma was clearly an attempt to remedy these flaws. In the first place, three preferences could be listed (still not, in my belief, sufficient if there are many candidates); but more importantly, lower preferences were weighted inversely by preference level. That is to say, a second-preference vote counted half as much as a first-preference vote at the stage where second preferences were added, and a third-preference vote counted 1/3 as much as a first-preference vote at the stage where third preferences were added.
So a ballot would look like this:
Counting proceeds as follows:
For a simulation of an Oklahoma-type Bucklin count, see the sample provided here.
|Sign up for your own site|
Last modified April 2, 2009.
|Click here to send any comments, suggestions, or other help.|