There are too many AV schemes, each more cumbersome than the other. They involve different, rather complicated ways of counting and sometimes of voting. Generally, not unlike Proportional Representation, they have the effect of reinforcing the power of parties. Some will suit a minority of voters fine, but not others. Their weaknesses can best be illustrated by contrasting AV with Fuzzy Democracy. Further below I discuss the weakness of particular AV schemes in certain countries.

FuzzyD holds to the two principles that

* voting must not be onerous,

* but should nonetheless be precise (i.e. in order to be meaningful).

Indeed, these are principles that any electoral system should observe. However, there is a tension between the two principles.

The question is, then, whether any AV scheme resolves the tension between these principles satisfactorily, or rather, better than FuzzyD.

AV enables voters to rank their preferences, whether just two or many. FuzzyD gives a power of political attorney to just a single candidate: that is, the chosen candidate is empowered to act (vote) on the voter’s behalf.

All democracy (as much else) involves filtering. Ill-informed and lightly held opinions must be eliminated; well-informed and deeply held opinions must be taken into consideration. But it is not easy to see how this can be done. Nor is there any exact dividing line between these.

The task of filtering has traditionally been entrusted to the political parties, but these have fallen into disrepute. Good and well-qualified characters (not to speak of those who wish to keep their personal lives private) have increasingly kept their distance from party activism, while others have seized on the opportunities a political career would seem to offer, i.e. irrespective of their talent or conviction. In their way, the latter may be hard-working, but in a mechanistic, managerial fashion. They are good at sound-bites, but abysmal on joined-up thinking and sustained concentration. In the UK at least, the pay is insufficient, such that professionals without other resources and, say, family obligations, must think twice: many will stand to lose financially (“opportunity costs”) while residing part-time in an expensive location and working in a place where they are high-profile and therefore must present themselves smartly.

Under AV the voter is given more choice (precision) than they maybe wish for. If the voter is certain of who they support, then finding a second or third best may be an extra hurdle and result in that extra vote not being made. If the voter vacillates between two or more, then the AV may make things easier. This is partly a matter of the disposition of the individual voter. AV schemes work with a limited number of candidates, although this number may still be considerable. FuzzyD by contrast, in its fully-fledged form, allows for choice among a huge number of candidates, one of which must be close to ideal even for the most demanding voter.

So we have different kinds of choice. FuzzyD involves trusting a single candidate from a huge pool, whereas AV involves focussing on more than one candidate and trusting an algorithm. It is likely a matter of personality and experience which you prefer.

Another approach to the merits or drawbacks of AV schemes is to examine how they count, or weigh, the votes cast. There are various algorithms, some dating back over a century (one even bearing the surname of the founder of FuzzyDemocracy). Any debate on introduction of AV would be never-ending as the competing virtues of these esoteric methods are debated by statisticians.

For instance, there is a world of difference between being able to register a single second choice, and being constrained (as in elections to the Australian House of Representatives) to list umpteen candidates in order of preference on a metre-long ballot paper. Ranking many candidates is onerous, if not embedded obfuscation: like tax systems which try to be perfectly fair and equitable, but end as minefields for the straightforward and goldmines for the astute.

So many voters have recourse to just accepting whichever ranking their preferred party has decided. The purpose is to create a two-party system, not (as in FuzzyD) to provide a filter for all well-considered viewpoints to be represented. It is imagined that the parties can do this. As may be observed, they cannot.

Here there is a similarity of outcome with the German system of party lists and a lead candidate, which was also recently proposed for some matters by the EU Parliament. In Germany you have an individual vote and, separately, a party vote. Here again the system is weighted in favour of parties and indeed, with the 5% threshold for representation, small parties are eliminated. The purpose is to generate stability. FuzzyD has entirely different mechanisms for ensuring reasonable stability. Stability eventually turns sclerotic.

In 2011 a simple version of AV was rejected in a UK-wide referendum. I believe this was foolish, since the very imperfect scheme was still far better than unmodified First-past-the-post. The vote was likely distorted by public disdain for a major advocate of AV and a vicious campaign by the Conservative Party.

The most farcical system is the second-round mechanism, as practised not only in presidential elections in France, where the highest ranking go through to a run-off election. Here it is easy for over half the electorate to be frustrated and for the winner to suffer from limited legitimacy. Second rounds enable sentiment to be tested in the first round. In the eyes of many who vote for him (or her?) the winner is merely the lesser of two evils. Hardly an endorsement of the democratic ideal.

In conclusion: Under FuzzyD you have one vote, which you award to a person whose judgement and character you have reason to trust. If not directly elected, they transfer the vote to a like-minded candidate whose judgement and character they trust in turn. Under AV you have multiple votes, which you may find as complicated as the subsequent counting algorithms.

©2019 Paul Charles Gregory