Norway's system regarding constituencies's seat numbers is a work of genius, really. If your goal is presenting a clear constituency link and somewhat overrepresenting remote rural locales (done in a lot of countries, remember, more usually using some random cutoff) without distorting party strengths
too badly, not much can be done to improve it.
Constituencies are allotted a number of seats based on their population and area (basically, a square kilometer is treated as a person). But the number of seats filled based on constituency results is one fewer than the total number of seats in the constituency; the remaining seats are distributed nationally as equalization seats (with a threshold), then reattributed to the parties' constituency lists based on priority. The last few individual MPs will be elected somewhat at random. As a result, thinly-populated areas are overrepresented without also overrepresenting their parties of choice.
Evidently, as in every system (that is not Hare; IIRC Norway uses Ste Lague) that distributes seats by constituencies rather than nationally, large parties are somewhat overrepresented; and there really is only one large party in Norway. Hoyre and Fremskritt are about large enough to not be underrepresented, that's it. And there aren't quite enough equalization mandates around to fix that; you could improve the system in that respect. But the effect is not huge; only very marginally larger than in Sweden. Incidentally, the more parties narrowly cross the threshold, the worse the effect is going to be as they'll all be depending on equalization seats.
But yeah, when elections are routinely close, every little bit of systemic distortion is a problem. And that does seem to be the case in Norway at the moment.
Now, back to topic. Fremskritt and the children.
