The arithmetic of a fake majority

I once tried to explain to my grandfather, who has spent his career inside a political party, why I couldn’t answer a question that ought to be simple: which party do I belong to? He heard it as indecision. I meant it as a diagnosis.

Picture a parliament split so that the right holds a bare majority of seats. Every member on the left agrees, unanimously, on some question in front of them. On the right, the same question splits the caucus — badly, down to 55/45. Under party discipline, the leadership calls it, the caucus votes as a bloc, and the 55% that barely agreed with itself outvotes the 100% that did. The chamber then reports a clean, decisive result. Nothing about that result was clean. A minority opinion, dressed up as unity, beat a supermajority opinion, dressed down as division.

A minority opinion, dressed up as unity, beat a supermajority opinion, dressed down as division.

This isn’t a hypothetical embarrassment restricted to one country. It’s what happens whenever you compress many individual, non-identical judgments into a single yes/no signal before counting them — gerrymandered maps, bloc votes, a party choosing not to show up so the abstention itself functions as a vote. Each is a place where the compression step, not the electorate, decides the outcome.

My grandfather’s defense of the whip was that without it, 349 independent legislators voting their own conscience would be chaos — effectively a new national referendum on every bill. I don’t think that’s the failure mode he thinks it is. Voting your conscience against your party isn’t a breakdown of democracy; structurally, it’s closer to the opposite — it’s the one moment where the represented and the representative are allowed to actually match. What discipline buys you is coordination. What it costs you is information: the chamber becomes very good at acting decisively and very bad at telling you what it actually thinks.

None of this argues for pure, unmediated majoritarianism either — the second you erect any barrier to raw majority rule (a court, a constitution, a second chamber), you’ve created a new question: who controls the people who control the barrier? Every fix for one distortion is a new place to hide another one. So this isn’t a case for abolishing parties, or barriers, or compromise. It’s a narrower claim: whatever structure you build, ask whether it’s reporting what people believe or manufacturing what looks like consensus. Most of the ugliness in politics isn’t malice. It’s arithmetic, laundered through a chamber that was never designed to show its work.