[EM] Individual freedom and voting methods

[Juho Laatu]

On 11.7.2011, at 19.55, Michael Allan wrote:

> The two topics might be related.  If individual freedoms are as
> important as Russ says (and Kant), then would it make sense to
> evaluate our voting methods in terms of which affords the greatest
> freedom to the voter?  Or might the voter be systematically
> constrained by the voting method, and yet still show a deference to
> the freedom of others in his/her decisions?

Those methods that we call good usually do maximize voter freedom. For example Condorcet methods try to maximize the information that they collect from the voters and then make decisions that respect the voter opinions as well as possible. Ranked methods do limit the voters to give only rankings. But the reason behind not using e.g. ratings is that one fears that the methods can not collect reliable ratings, and elections would become strategic, and in the end they would collect less information from the voters and respect less their opinions.

Another example is list based proportional methods vs. STV like proportional methods. STV methods allow greater freedom in the sense that voters can give more information (rankings), even without any party borderline limitations. List based methods collect just bullet votes (to a candidate or to a party) and thereby less information, but those simple votes contain some quite concentrated information if we assume that a party structure is a natural and simple way for human beings to model the world and how it should be managed. Despite of the better freedom in STV, the simpler list based methods work quite well and are nice to use especially if there are many candidates and numerous seats per district. There is thus some sort of balance between freedom of expression and simple modelling (and voter understanding) here. This is an idealized and simplified view on these two categories, but maybe these methods may serve as a working example on how the balance between freedom and simple modelling might work.

I believe there are still many possible improvements to be made in both areas.

Improvements in how we model the society are endless (although often people seem to think that they or their party already possesses the best possible model of the society). That is however a slow process and it often must work its way forward through trials and errors. It seems that people / societies must first fail a couple of times before they learn not to make those mistakes again. It takes time before people can find simple enough proverbs or other rules or philosophies that become generally accepted and thereby generally stop people making stupid things again and again.

Improvements in election methods are not as numerous since this field of study is simpler (well, often mathematically complex, but still these are just mechanical problems). One can also give more freedom to the voters when they grow wiser. I mean that when modelling of the society takes steps forward, people will be more capable of making sensible decisions at more detailed levels based on that improved understanding.

Currently the freedom of the voters is typically limited so that democracies are not direct democracies but representative democracies where regular voters are not allowed to decide, but they must elect some wise men to decide on their behalf. It is possible that when the education level and overall understanding of the society improves, one could give the voters also more power to decide. But maybe not too much and not too often since in order to collect as much information as possible (and thereby supporting true freedom to actually decide on political matters) one should not disturb the (often lazy) voters so often and with so detailed questions that they will lose interest. Parties are also a good traditional tool here. The minimum requirement to the voters is to just decide which team they will support.

There are also many other possible approaches to giving voters more decision power as the methods evolve and as the understanding of the voters evolves. The already mentioned STV style methods are one approach to allowing voters to decide in more detail. Parties could also be divided in smaller opinion segments in elections in order to allow voters to influence also in secondary key questions, not only in the primary ones. New computer based technologies could allow many new more interactive ways to communicate and influence decision making. Information flow in both directions is an important part of democracy and we could probably do better than use yellow press as the primary tool of communication.

In summary, it is possible to allow even more freedom (and also gain true freedom to decide), but one should not forget that people need also constraints and agreements (both voluntary and mandated) in the society to make it work well (and thereby also optimize the level of freedom that we can offer to people). Good models are essential since they allow people to work better together with minimum friction and maximum freedoms (and better understanding on where to voluntarily not do anything stupid).

Juho