The cost of trade-off is phrased in terms of representation and accountability. Various scholars have various perspectives on that trade-off. You had Arend Lijphart before you at some point. In Professor Lijphart's view, there's no trade-off whatsoever if you put zero weight on accountability. Another view, by G. Bingham Powell, is that there tends to be a fairly tight trade-off between these two qualities. A third view, advanced by John Carey, is that the relationship between representation and accountability could potentially be subject to optimization. That is, there's some sort of sweet spot that would allow us to choose an electoral system that gives us maybe not as much representation as you could hope for, or as much accountability as you could hope for, but a good blend of both.
My sense is that the literature is relatively inconclusive. I think there are fairly good metrics for representation. I think André Blais talked to the committee about measures of ideological congruence between the population and the legislature under various electoral systems. I think the balance of evidence is that you tend to get a bit better congruence under PR, but not by much, though the variance on congruence is smaller under PR.
There's very little good work on the performance of different electoral systems in terms of accountability, because, as I was saying previously, the metric that tends to be used, cabinet stability, is not something we can say is...we can't create a cardinal metric that says there's an optimal point of cabinet stability. Zero is bad, and complete cabinet stability, where you had no turnover, would be bad, too. It's not clear where the optimal point is, so our measures for accountability tend not to be very sophisticated, whereas the measures that scholars use to measure representation tend to be better.
This debate about trade-off remains ongoing. My sense is that there's no real evidence one way or another to suggest that some electoral systems provide a sweet spot. I think you're facing really hard and fast choices in this respect. I've already stated at the outset that I think the first guiding principle you've been given, to find a system that better reflects the democratic will, is tantamount to hunting for a unicorn. There's a lot of good theoretical work to suggest that's the case.
I think you will notice, as members of Parliament, that if you scrupulously followed the majority will in your constituency issue by issue, taking one issue at a time, you would soon find yourself supported by a tiny minority in your constituency, because the intersection of all those majorities could end up to be a very small set of people. I think that's the fundamental nature of the problem.