Guest blog by David Gould, Chair of Bristol for Democracy
After a month and a half of US primaries, Donald Trump has by far the most delegates at 673. Ted Cruz is quite far behind on 411 and John Kasich is barely in the picture at 143.
So this means that voters in Republican primaries much prefer Trump, right? Wrong.
What’s happening is roughly 35% of voters in Republican primaries want Trump. The other 65% however aren’t very keen on him at all.
Indeed, Fairvote commissioned a Alternative Vote poll a few weeks ago and showed that Cruz would beat Trump in a 1 v 1. They provide further evidence that Trump isn’t the most preferred candidate in a HuffPost article.
Buying pizza provides a great analogy for single-seat elections eg who the Republican Presidential candidate should be.
You’ve probably been in the situation where you’re ordering pizza for a group of roughly 5 people. John and Alex want inferno meat pizza, Rachel wants ham and pineapple and Rebecca wants four cheese. Tom is a vegetarian and wants spinach & sun-dried tomato.
How do we resolve this in real life? We might find everyone fancies a spinach & sun-dried tomato or four cheese pizza. We might get two small pizzas. Or Tom might go raid the snack machine. What we definitely wouldn’t do is force Tom to chip in for an inferno meat pizza.
But that’s exactly how FPTP works. Here, Trump is the inferno meat pizza being forced on everyone else, simply because the GOP didn’t bother to update to a proper electoral system.
This problem is compounded due to the delegates being disproportionally allocated in many states eg Mississippi gave Trump all 20 delegates on Tuesday even though he got less than half the votes.
So if Trump being President scares you, remember FPTP is to blame.
FPTP for a General Election is Even Dumber
FPTP is failing Republican primaries because the anti-Trump vote is split between up to 16 candidates. Likewise in the UK, the Green Party and UKIP only won one out of 650 seats partly because their vote is split with other parties in all the other seats. The SNP won 56 out of 59 seats in Scotland in spite of Labour receiving a quarter of the vote. Only a few of those losses for Labour were because of vote-splitting though. Most were because votes that don’t elect an MP are actually ignored in the final tally. Green Party and UKIP votes were completely ignored in 99.85% of seats.
The other reason that the Green Party do badly is because a lot of their supporters don’t want their vote to be ignored — and so might vote Labour instead. The Liberal Democrats and many smaller national parties have the same problem.
The UK needs proportional representation in both General and Local election. This would be good for the US general election but primaries are a different matter.
Should States use Alternative Vote?
No. Alternative Vote, or Instant Run-off as the Americans call it, doesn’t actually tell you relative candidate popularity, which states need in order to allocate delegates proportionally. The only preferential system I know which does this is Score Voting, also called Range Voting.
Score Voting is a relatively new voting system which is so simple that you wonder why no-one ever thought of it before. It asks voters to score each candidate from 0 to 9. The final result produces an average score for each candidate.
If you’re looking for one winner, that’s the candidate with the highest average. Obviously. If you’re looking for proportional distribution of delegates, you simply make allocate them proportional to the candidates’ scores.
Note: whilst Score Voting would be great for a separate ballot on picking the President, it isn’t suitable for multiseat elections eg Congress and Senate.