3 ms·
I think there are a few fundamental misunderstandings here. Firstly, it's not 51% of the voting populace that flip-flop. It's actually more like 2-3% "swinging
by phs318u 5y ago
I think there are a few fundamental misunderstandings here.
Firstly, it's not 51% of the voting populace that flip-flop. It's actually more like 2-3% "swinging voters" on average - maybe 5% tops. The rest are rusted on to their "team" for decades. The implication is that those on the "losing" side, do indeed consent to be governed by the other team. What's changed in recent decades is that the delta between the two teams has grown - largely on non-economic grounds (imo) - to the extent that people cannot stomach the thought of the other team being in charge. This is exacerbated by the fact that in countries without compulsory voting, participation is far less than 100%. There is greater advantage to be had driving policies away from the largely "centrist" mass possible voters, towards those on the edges who are more motivated to get out there and vote. In countries with compulsory voting, moving radically away from the centre gets punished such that the parties likely to win government are more alike than different. This means that social and economic change is likely to take longer, and when big social shifts happen, they eventually spread to "both" teams (clearly with some temporal difference).