Voting methods that take voters stability into account

Question

Consider the case of Brexit: a relatively brief moment of voting time that decided the state of UK in EU. One shot.

Is this a good idea? Let's consider these factors:

• Voter stability: were voters stable at the time when 51.9% of them voted to leave?

The vast majority of voting systems, almost all of them as deployed in democratic systems, do not consider voter stability.

But do we really need to take potentially big actions when the public is unstable?

In my view, moments of public opinion instability are moments when the public is simply unsure about a decision. In such cases governments must simply give their citizens more time to think and converge on a stable decision.

My questions are:

• Q1: what are the voting systems that consider the stability of the voters into account?
• Q2: are there any political systems that consider voter instability?
• Q3: what definition of stability is suitable?

Answer 1

The vast majority of voting systems, almost all of them as deployed in democratic systems, do not consider voter stability.

Actually, the vast majority do take voter instability into account. A little less than half the countries in the world (i.e. more than half of the democratic countries) use a Bicameral system where the two chambers of parliament are elected in a different way or at different times and both are needed to change laws. That setup is expressly designed to prevent major changes after a single election win by a party. Other systems have protections like a president's veto power or judicial reviews of laws.

A second way that voting systems account for voter (in)stability is by not having the option of a referendum at all, or a non-binding "advisory" one at best. Those result in some extreme political contortions as the government finds a way to work around the unwanted vote outcome or redo the vote, but very rarely result in real change.

A prime example is the attempt to introduce a European Constitution. It was rejected by referendum in France and the Netherlands and supposedly abandoned. In practice, most of the big changes proposed ended up in the Treaty of Lisbon word for word, which conveniently did not require a referendum in France or the Netherlands. That treaty did encounter another bump in the road, being rejected by the Irish in a referendum, but a repeat of the referendum was organised and then the 'Yes' vote won.

So, the answer to your Q1 and Q2 can be found in the structure of governments the world over. The people can vote, but they cannot force the elected officials to take any action or make any law. Officials or parties elected in a wave of emotion or discontent cannot make big changes right away, they need to last until they can get members elected to the other chamber of the parliament a few years later, by which time voters usually have come to their senses again.

For your Q3, voter stability is hard to define. I would argue it needs to include informed support for a platform, program or position rather than an expression of protest or discontent. I.e. a voter's preference is likely to be more stable when they say "That is where I want the country to go" as opposed to "I'm unhappy with where the country is going, it needs to change!".

By this (incomplete) definition, the vote in the UK would be considered extremely unstable, as there was no plan or platform available for "Leave" voters to support, only promises without any basis in fact. The "Remain" campaign did not do much better, mostly outlining why a "Leave" vote would be bad.

Answer 2

For Q2, I don't know of any political system that objectively considers the stability of their voters. They all seem to me to be one-shot votes.

For Q3, I suggest that the best definition for stability to be:

• The probability that the majority votes flip/change in the next x years. This would require time series analysis of the votes to predict the probability of having vote count curves fall/raise such that the majority vote changes. For matters of long-lasting effects (such as Brexit) x must be set to be large.

For Q1, I propose this:

1. Let voters vote for the course of a long period, such as 5 years.
2. Each voter is allowed to change their votes in the course of 5 years.
3. Government monitors the votes, and plots diagrams and curves showing how the public opinion is fluctuating over time (e.g. due to political campaigns, movies, etc which could possibly have a temporary impact).
4. Based on the fluctuations, the government can estimate whether the opinion of their citizens have converged.
   • If the opinion of their citizens is converged, government takes necessary actions as per the votes.
   • Else, the government simply grants their citizens more time to think for until they converge.

Consider the following hypothetical example where votes are counted over the course of (say) 5 years:

If we choose Brexit, one may ask this question:

• When the 51.9% majority voted "leave", were the voters converged? Or was this due to temporary fluctuations due to recent media reports?

I think the answer is unknown (at least not publicly). What's worrying is that, regardless of measuring the voter stability, a decision is concluded to represent the British people.

Side note: I am somehow surprised how critical votes/referendums (e.g. Brexit) are one-shot votes, with complete disregard of the stability/convergence of their voters. I think any critical voting situation with long-lasting effects must be given adequate time for until the voters converge.

Interestingly, it seems that there are time-lines for the Brexit votes as follows:

In my view, by eye-balling, the Brexit leave decision was clearly due to unstable votes.

Answer 3

Governments wishing to build in voter stability can do so by requiring a supermajority before a proposal can be passed.

A supermajority, or a qualified majority, is a requirement for a proposal to gain a specified level of support which is greater than the threshold of one half used for majority.

Related concepts regarding alternatives to the majority vote requirement include a majority of the entire membership and a majority of the fixed membership.

A supermajority can also be specified based on the entire membership or fixed membership rather than on those present and voting.

Parliamentary procedure requires that any action of a deliberative assembly that may alter the rights of a minority has a supermajority requirement, such as a two-thirds vote.

Changes to constitutions, especially those with entrenched clauses, commonly require supermajority support in a legislature.

The requirement to have a supermajority before major decisions are enacted can be required of the voters as a whole in a referendum, or of the MPs, senators or other representatives in a legislative body. The US Constitution requires supermajorities in both Houses of Congress both to propose and ratify an amendment. This was a deliberate decision of the framers in order to make the constitution stable.

The device of a supermajority works by making change impossible until there is a significant majority in its favour, a majority that is too big to be reversed by mere day-to-day events. That is its advantage but also its disadvantage. If, for instance, a nation is split politically into two groups where one is only slightly larger than the other for a long period, then requiring a supermajority could end up keeping the majority out of power for the foreseeable future, and fossilising the political system generally.

An argument related to the danger of political fossilisation is that in a fast-changing world some decisions simply have to be made, one way or the other. Any decision is better than indecision. So long as the ground rules of a voting method are clearly set out and agreed in advance by all sides then that deal should be adhered to, even if it is regretted later.