The referendum on the Alternative Vote voting system is coming soon, it’s about a week away, so I thought I’d write a bit about it. Amazingly there has only ever been one other referendum which has covered the whole country before, the referendum on whether Britain should remain part of the European Community in 1975, so this coming vote is a rare opportunity to directly influence the future of the country. I’ll be voting yes (for the change to AV) on the day. Why? Well, I think voting reform is long overdue in this country. You just need to look at the results from every relatively modern parliamentary election to see that what we have doesn’t work; The number of seats allocated to the various parties bears very little relation to the total percentage of the popular vote they actually received. Whatever way you look at it, that isn’t right. Now, of course, AV isn’t a form of proportional representation, which is where the percentage of seats would match the percentage of votes, instead it is a bit of a compromise, falling somewhere in-between PR and the huge discrepancies of the current First-Past-the-Post system. A fact the “No to AV” campaign keep highlighting is that the Electoral Reform Society and the Liberal Democrats would much prefer the Single Transferable Vote system, which is a form of PR, over AV. And that’s true, but it’s also true that the main backers of the No campaign are the Conservative Party, and if they wanted to they could easily have made this referendum about STV instead. For them to say: “Why should we use this system when it isn’t even the one they really want?” as a negative against AV is like a bully saying “Stop hitting yourself” as he grabs your arm and forces you to punch yourself.
Everybody knows how FPTP works, people vote for one candidate out of several and the votes are counted, whoever has the most votes wins. AV isn’t quite so simple, but it’s not as impossibly complicated as some make out. Each voter ranks the candidates in order of preference, “1” for their first choice, “2” for their second choice and so on, until they have either ranked all the candidates or they decide they no longer wish to express an opinion. The first choice votes for each candidate are then counted. If a candidate has over 50% of the vote then they win, otherwise the candidate with the least votes is eliminated. If the votes for this candidate had any second choices then they are added to the total of the remaining candidates, if not they are discarded. If any candidate now has over 50% of the vote then they win, otherwise the remaining candidate with the least votes is eliminated, and their votes are redistributed as before, except, while we still look at the second choices of the first choice votes, we will need to look at the third choices for any second choice votes the candidate received. This process of eliminating the last placed candidate and redistributing their votes, looking at the numbered preferences as required, continues until somebody has over 50% of the vote. I don’t find this system especially difficult to understand, but I did a degree in Mathematics so I suppose I’m used to complicated things involving numbers. I can definitely see that it’s more difficult to understand than FPTP, but most things in life get more complicated as they get better. For example General Relativity and Quantum Physics are much more complicated than Newtonian physics, and the computer you’re reading this on is much more complicated than the ancient abacus from which it’s descended. It’s the same with voting systems, AV is more complicated than FPTP and STV is more complicated than AV, things get more complicated as they get better. The thing is you don’t need to understand how AV works in detail in order to vote using it and to understand some of it’s benefits, just like you don’t need a degree in Computer Science to use your computer or to know the benefits email has over a written letter, all you need to be able to do is number the candidates in order of preference.
Perhaps the main difference between FPTP and AV is that AV almost completely removes the potential for tactical voting. I’m only 22, but already I’ve seen loads of election leaflets along the lines of: “If you don’t want X to win, you should vote for Y, because Z doesn’t have a chance here”. If you support Z but decide to vote for Y to stop X winning then you’ve become a tactical voter. The problem with this of course is that it means Z has lost a vote, and, if enough people do this, they really do have no chance of winning, even if they did originally have enough supporters to do so. After all, if Y sent you that leaflet do you think they were truly being honest about the chances of Z? With AV tactical voting is almost completely eliminated, you vote for the candidates in order of preference, until you no longer want to rank them. So, in my example, if you supported Z but didn’t want X to win, you would place Z as your first choice and Y as you second choice, then stop. In the first round of voting, if Z has the lowest share, then they would be eliminated and your vote would then count towards Y in the second round of voting. But, of course, it’s also possible that Y could have the lowest share of the vote in the first round and be eliminated, and that with their second choice votes Z could have enough in total to win the second round of voting. With AV you don’t have to sacrifice your vote to another party to stop those who you don’t like from winning, which is a good thing, as it means less seats will turn into two horse races between those who hold it and those who came second at the last election. As a bit of a disclaimer, I should add that AV does not completely eliminate tactical voting, as there is a chance that you could help the party you want to win by not voting for them as your first choice, but it’s complicated, and to do it effectively you need to know how others have voted, which is almost impossible to do accurately, legitimately at least, so is unlikely to occur in practice.
It is easiest to see how this may occur by looking at a system somewhat similar to AV called run-off voting (in fact another name for AV is instant run-off voting), and a particular form of it called the Exhaustive Ballot system. In EB each voter casts a single vote for a single candidate, and after this first round of voting if no one has more than 50% of the vote then the candidate with the least votes is eliminated. Everybody then votes again, and as before if no one has more than 50% they are eliminated, then everybody votes again and so on, until someone has a share of the vote greater than 50%. This is almost exactly the same as AV, except here you vote multiple times instead of ranking the candidates and voting once. Now consider a three-way EB election between A, B, and C. A has the support of about 40% of the voters while B and C have about 30% each. But, importantly, C voters, if their candidate were eliminated, would always prefer B over A, while B voters, if their candidate were eliminated, would split quite evenly between supporting A or C. Hence it is in the best interests of A to have B eliminated in this round of voting and not C. As there are already more A voters, some of them can vote for C instead without jeopardising their chances of progressing to the next round. If, say, 5% of A voters vote for C instead then A and C would both have 35% while B would only have 30%, so B would be eliminated. If the B voters split 50:50 between A and C in the final round then A would win with 55% to C’s 45%. Note that C loses the extra 5% from the A voters who were voting tactically, because under EB you can vote differently in each round. Tactical voting isn’t easy under EB because you only want a certain percentage of your voters to vote for someone else, if too many do so you’re in danger of being eliminated yourself. Under AV it’s made even harder because you must predict everything in advance, and your vote only changes between rounds if your candidate is eliminated. If the situation I just explained were done under AV then both A and C would have 50% of the vote in the second round, because the tactical votes for C by A voters would still count as being for C. This could be remedied by having a lower percentage of A voters voting for C originally, say 3%, which would still be enough for both to beat B in the first round and result in a 52% to 48% victory for A in the second round. As you can see, if a candidate were trying to manipulate the results of a vote under AV like this they would have to get their guesses spot on, to the point that it is nearly impossible to do in practice. I think I should note at this point that one of the uses of EB, the voting system most like AV which isn’t AV, is for internal Conservative Party elections which decide who prospective parliamentary candidates will be. I find in interesting that they can be so against a system which is so similar to one which they themselves use internally.
AV should also lead to less “safe seats”, because voters who oppose the current holders of the seat can put as their various choices candidates from all the opposing parties, which means they effectively vote for anyone but the current holder. For example if one year party X has 49% of the vote, party Y has 26%, and party Z has 25%, then under FPTP it would take a significant amount of X voters to switch to one of the other two for the seat to change hands in the future, as it’s unlikely support for Y or Z will completely shift to the other, but under AV the Y voters can put Z as their second choice and vice-versa, so even if X maintains 49% of the vote at the next election they may be unable to keep the seat. In my example, if the percentages are kept the same and all the Z voters put Y as their second choice, then Y would win the seat with 51% of the vote in the second round. This is an extreme example, but in practice there are a lot of seats which are considered safe under FPTP which may not be under AV for the reason I’ve illustrated: It allows the opposition to group together to defeat the current holder. Safety under FPTP relies on a split of the opposing votes just as much as having a large number of your own supporters, few seats are won by a candidate getting over 50% of the vote, but a lot are won by candidates getting around about 40% of the vote, and in that case the seat is won by a candidate who the majority of voters (60%) didn’t vote for. Under AV the eventual winner should have at least some sort of backing from over 50% of the voters, by which I mean at least 50% will have put the winner as a numbered preference, so they preferred them over somebody else. This won’t always be the case, because there is no minimum for the number of preferences to put, theoretically everybody could just put only one preference and the result would be exactly the same as FPTP, but in practice people will put second and third choices which means that most of the time the statement will be true.
I’m of the view that it’s always worth taking a step in the right direction, however small it may be, rather than sticking with the status quo and waiting for a bus that may never come to take you the whole way. There may not be another referendum on the Voting System for a generation or more if the result of this one is a no vote, “The people have already decided they want First-Past-the-Post” will be the go to comeback for any suggestion of future change. I think this country would be better off with proportional representation instead of FPTP or AV, as it is the most democratic of all the systems, but I don’t think we’ll get there in a single step. For the reasons I’ve given, and some more I haven’t, AV is better than FPTP, but what is perhaps more important is for people to show that they’re fed up with the current system and that they want change, and that can only come by a “Yes” vote next week.