[chriswarbo]

> The point that new parties have not actually been elected is not a sign that democracy doesn't work, it is a sign that people in summary don't actually think any of them would be an improvement.

Whilst this is true, it also matters how the people's will is summarised, and (perhaps even more importantly) how that summing method affects the behaviour of the people.

The obvious example being that first-past-the-post systems have a tendency to get ~2 dominant parties (e.g. US and UK), whilst proportional systems tend to get more parties or coalitions (e.g. Germany and Scotland).

Hence, in a first-past-the-post system, the fact that new parties don't tend to get elected is a sign that people don't think those parties offer enough of an improvement to convince a majority of voters to sacrifice their influence on the race between the big 2. It's like a giant prisoner's dilemma, where it's possible for every individual to think that some new party is a big improvement, but for nobody to vote for that party as they don't trust others to cooperate.

Although there are second-order effects, e.g. it may be worth chasing a seat in parliament even if there's no hope of forming a government; or getting second-order effects from an issue/protest vote (e.g. voting UKIP to influence whichever majority party wins to focus on immigration)