The country – with some 8 million population – is divided into regions each with its government and severely under-occupied bureaucracy It has an air force with modern fighter planes.

It is financially convenient to the political class to maintain this incredible situation.

Votes are essentially purchased by offering ever more goodies to the lower income groups and the unemployed, paid for by a highly regressive income tax and other, that reach a maximum of 40% at a low level of income.

There is no escape other than FPTP.

2. “The voting system interacts with the other elements of the political system and with the underlying society in very complicated ways”.
I see more complex the relations of the political system with social structure (which includes economy).
Social structure–>Social problems–>Ideologies–>Political Parties–> Party Manifestos–>Elections / Electoral Systems–>Assembly / Government–>Law/public policies–>Desired Changes in social structures.
IMO the social structure is the key point of the whole cycle.

In any case all we are doing is electing local mps. Coalition makes it more likely that a voter has an mp whose party is governing, its more important that people are represented than some arbitrary thing we call a political party.

Consider the current situation, according to the BBC’s poll of polls at the present time, it could be that the vote is split: Con: 35%, Lab: 29%, LD: 26%, Others: 10%. With such a spread it would be unlikely for a single party to hold an outright majority of seats on the current system. The LDs may well have very few seats (much fewer than than their 26% share would suggest) but their large popular vote makes it seem extremely fair to me that they could form a government with Lab, together representing 55% of the voting public.

I think in many ways this is the most interesting election of my adult life.

Nonetheless, the graphs near the begining to illustrate how stark the distorting and unfair effects of having a FPTP system are. This was really the main point I was trying to highlight.

The robustness of abstract models with respect to modifications of their assumptions is an interesting theoretical question. It might also be helpful to construct realistic models of electoral systems. The failure to have, or at least to apply, realistic models seems to have caused certain difficulties in economics lately. It is far from clear to what extent the public policies of, say, Israel, are an artifact of its electoral system, rather than a product of its history, culture and location. More generally, perhaps the number of essentially distinct voting systems is very small under sufficiently realistic assumptions.

But this opens up a big can of worms, and in the end one would probably have to have a full multi-party, proportionate system, which I don’t want at all, because then half the time the election doesn’t decide anything (case in point: Iraq).

The Swedish system has already been mentioned, I would like to add that most European, working democracies have a similar system.
Germany’s system is a little more complicated than full multi-party, proportionate, but as a first order approximation that will do. Germany has a 5% barrier for the parliament (a party has to get at least 5% of all votes to get into the parliament at all).
The situation that a small party held two parties hostage was basically the case during the 1960ies and 1970ies, when there were two big parties (social democrats, SPD and conservatives, CDU) and one small (liberals, FDP). Actually the small party overturned the government once by changing sides (that’s how Helmut Kohl became chancellor).
But all in all the system worked quite well.

My point being: If this system works or does not work depends much more on how the people participating handle it, than game theoretic discussions.

Having said that, I’m not sure that this argument is entirely borne out by experience. For instance, in Israel, some small and very hard-line religious parties appear to have a lot of power.

Let’s assume in the UK you had Labour on 49%, Conservatives on 48% and BNP supposedly “holding the balance”. To suggest that the BNP holds the others to ransom means that Conservative and Labour would much rather work with the BNP than each other. I don’t believe this to be true.

I think what would happen is that you would get a minority Labour Government who would find a way to engage support from the only slightly weaker Conservatives.

This might result in a lot of compromises, which could well be a Good Thing. Some things wouldn’t change at all, some things wouldn’t happen. In order to put through legislation people would have to try to come to an agreement. This is also a Good Thing.

Together with Duverger´s law, Michels “iron law of oligarchy” is one of the most celebrated of classical sociology. One of the fourth steps in the Clegg liberal democrat manifesto is to try to change this (“A fair deal, by cleaning up politics”).

2. For those interested in the empirical side of elections and the “evolution of electoral systems across countries i forgot to cite another good source:

Elections and electoral systems by Dieter Nohlen.

Here he uses more case analysis than statistical analysis. There is an interesting discussion about the real difference within majority-proportional systems and how to evaluate them.

3. It seems that british system evolved to a pure relative majority uninominal system gradualy. Until the 30´s Liberals were the second political force and were supplanted by Labour due to some bad moves regarding the irish issue and probably also due to sociological changes and international environment (cold war). Then in the 50´s the system was closed (relative majority uninominal) and as a result it was impossible for liberals to govern anymore. Now there are new sociological changes (globalisation, green thinking) and i think the liberals has good chance to become the second force again. If the sociological change is robust, this will happen in one election or the other, independently of the electoral system. By reading the Liberal manifesto it is clear that they are trying to gain territory against the Labour while increasing abstention from conservatives.

4. IMO, the British system might be unfair but taking into account the multinational character of UK, it is the best possible. I would not change to a proportional one; i´m not sure about the alternative system.

So it seems the bottom line is that its not too clear what the impact is and other factors come into play.

I will try and check out some of those references.

For example, someone told me that they were uncertain because LibDem was no longer a safe protest vote – they might get elected.

Someone else quoted a survey which had apparently suggested that if people believed the LibDems could win, they’d get 49% of the vote (unchecked) – so are we in the realm of catastrophe theory? Does it take a “catastrophe” to change a government, or does the ideal voting system give a smooth transition (modelling votes/opinion over time as a dynamic system of some kind)?