So I know we’ve beaten this issue to death but I was just wondering, what about just increasing the value of abstention from 0 to 1 or 2. Then being unknown hurts you less, but you don’t have to deal with added zeroes or anything (perceived to be) complicated.
That’s what net approval voting does, except it puts it right in the middle which might be too high.
OK, so net approval involves an explicit approve and disapprove option then, so you lose the simplicity. Hmm… worth it?
“Post must be at least 20 characters”. Well it is now.
Do you know of anyone who calls it that? I’ve only heard “combined approval”, “evaluative”, or “dis&approval”. (Unless you’re talking about something different.)
If you change only the ballot design and refrain from doing anything unnecessarily complicated in the tabulation procedures – that is, if you just sum the vote totals – this “dis-and-approval” method ((-1), (0), (+1) voting) should give exactly the same results as a simple score (+1), (+2), (+3) voting method.
It would initially be perceived differently but the electors would quickly adjust to the superficial difference. Due to the low number of voting options, it would not be very responsive.
I think that 0…10 and -5…5 “make sense”, but something like -2…8 feels like a “magic number”. Why should we prefer -2…8 over -1…9, -3…7, or -4…6? (Not to mention that an 11-value range cannot be divided exactly into thirds or fourths.)
-5…5 could be interesting. Although I wonder if even more people would skip the extremes with that than with 0…10 (I know that -5…5 gives me a stronger urge to rate in the center than 0…10, so I wonder if other voters would behave similarly.)
Obviously any of these could be combined with the STAR rules (for which I always thought 10 or 11 values was superior to 6 anyway… and STAR on -5…5 should IMO treat blanks as less than an explicit zero for the runoff).
There’s also the risk of the abstention value itself feeling like a magic number.
But “minimum support” and “middle of the scale” would be easier to market and actually make sense.