I just thought of a way to model multi-party systems and voter turnout at the same time.
What you can do is place every voter unit on a spectrum with regard to every matter you now call a group. For those definately opposed groups, you’d have a spectrum with one group on one end, and the other on the other (capitalist-socialist, wealthy-middle class-poor, or conservative-liberal for example). Other groups would have to have the group on one end, and “antis” on the other (such as religious-antireligious, environmentalist-antienvironmentalist, etc.) Every voter unit would be rated on these scales.
Most would be neutral on many issues, extreme on others. And every possible combination could be represented, assuming such persons exist in the country in question. (Yes, it is possible to be both capitalist and environmentalist, or pro-labor and capitalist, or any number of possible situations, and there are policies that would please such people. For example, the capitalist-environmentalists here in California think hemp legalization would work wonders both for the economty, the environment, and local farmers.)
Political parties would have two traits: an “ideal partisan” model, and an acceptable member radius. The “ideal partisan” model would be a set of “group” stats that typifies that particular party. For example, the “Ideal Democrat (US)” would be socialist, environmentalist, pro-labor, poor or rich (not middle-class), liberal, pro-commuter, and so on. The “Ideal Republican” would be conservative, religious, capitalist, patriot, middle-class or rich (not poor), pro-motorist, and so on. As both are what we call “big tent parties,” both would contain voters, and attract swing voters, who deviate substantially from the “ideal.”
The Libertarians would have the following trait. The “Ideal Libertarian” would be rabidly capitalist, liberal, middle-class (not rich or poor), and nonpatriot. Being much more “anti-heretic” it would have a very small acceptability radius, meaning most Libertarians would conform very closely with the “Ideal.” (I and others are trying to increase that radius–see reformthelp.org).
In countries with “winner take all” style elections, two-party systems would dominate. It would be preferable to model ideological shifts in political parties (as is occurring in the US today). Voters who don’t fit in either of the major parties would tend to stay at home unless the leader himself is particularly palatable to them. Voters at the point of overlap (swing voters) between the parties would vote purely on the state of the country; “party ideals” could be counted on to support a leader who doesn’t piss them off by resembling the other party too much.
Countries with proportional representation would probably have better coverage as far as voter-type and party matches go, I think. How is it in Britain? How does it compare to turnout in the US?
You could even model single-party rule (for nondemocratic countries). The leader would have the advantage of an enforced voter turnout that matches his party’s agenda. He would have the disadvantage or high violent crime rates (constant political unrest, proportional to the number of disenfranchised voters), high corruption (power-seekers who game the ideological test for political power), constant concern about assasinations and terrorism, and great international opposition.