assume the US voting populace was two sets of clones, 50% Mitch McConnell Republicans and 50% Nancy Pelosi Democrats - but the distribution always varied within

assume the US voting populace was two sets of clones, 50% Mitch McConnell Republicans and 50% Nancy Pelosi Democrats - but the distribution always varied within a standard deviation (eg year to year or district to district or state to state). Maybe there's even some generational biases, where one set of clones is more common for a decade. I think you'd find that pretty often, more than half the country disapproved of congress, even though, in the long run, each clone is equally likely as a voter regardless of geography, economic interest, etc. How could you design a voting system that's responsive to these variations? This challenge gets more complex the more political alignments you account for. Few people in an opinion poll of Congress are assessing their love of bicameral legislature. More commonly they're stating that they wish they could just have their way, regardless of opposing voices.