Of course, exact results would depend greatly on what the population distribution looked like and whether Auckland went for single or multi-member districts. So I assigned my Public Choice class to weigh things up as an assignment: assuming that the objective is to maximize Maori influence on council decisions, are separate seats a good idea?
Best argument I've seen for separate seats:
Having the option to enroll on a city-level Maori roll would allow Maori in wards where Maori population would otherwise be too small to be considered to have their votes weighed separately; in that case, the median city councilor can be made more pro-Maori. So lots of wards shift from 5% Maori to zero percent as those folks flip to the Maori roll, while folks in wards that are above 10% don't. The median councilor could then well have a higher proportion of Maori in his district than would be the case otherwise. This requires then that folks only (or massively disproportionately) choose the Maori roll if they're in low-Maori wards: that way, one of those wards disappears and the median shifts to the next higher Maori proportion district rather than the other way. Of course, it's pretty unrealistic to expect that level of strategic voting.
Best arguments I've seen thus far against separate seats:
- Maori and non-Maori political preferences are not so polarized as to mean that Maori voters in districts would be written off by candidates seeking a majority
- Having multi-member districts could increase the chances of Maori being elected without having the need for separate seats
- Caveat here: some folks found the nice old Bernie Grofman paper showing that multimember districts in US local elections reduced rather than increased black electoral prospects. Indeed, "at large" elections were worst for black representation.
So this really requires some kind of STV implementation in a multimember district to work. The Royal Commission report seemed to view multimember districts as unambiguously good for minority representation; the US evidence suggests otherwise.
Best argument I've been able to come up with for separate seats (which none of the students gave, but one came close):
Maori wards would be designated based on electoral enrollment numbers; actual influence within any ward depends on voter turnout. And Maori turnout is much much lower than the population average: The number of votes cast in the Maori electorates in national elections is roughly 58% of the number cast in general electorates. Consequently, even if my earlier worries were well founded, there could still be strong argument for separate seats (given my initial "assuming" that the goal is to design a system that maximizes Maori influence) because each Maori voter who turns out is given weight proportionate to the number who have enrolled on the Maori roll: each vote effectively counts as 1.73 votes as compared to that voter's influence in a general electorate.
While lots of students found US evidence on the effect of majority-minority districts, perhaps because I pointed them to some of it, none of them found the rather nice paper by former Kiwi Jeff Karp that argues that combined seats seem to do best for minority interests. He there shows that while minority groups tend to be disengaged from the political process, this dissatisfaction is minimized in systems where parties are encouraged to compete for votes across diverse groups (like first past the post elections with single member districts, or the less proportionate forms of PR).
On balance, I don't think separate seats would do much to improve substantial representation of Maori interests in Auckland and could work to harm substantive representation in order to ensure some descriptive representation.
So which matters more: having a couple of councilors who look like you and care passionately about your interests, or having equilibrium policy be a bit closer to what you'd want? I tend to go for the latter, but the former is certainly more expressively appealing.