Isle of Man · Voting Patterns
Who votes like whom
Most members of the House of Keys sit as independents; only a few stand for a party. This matrix compares every member’s recorded votes against every other member’s in the 2021 to 2026 House, across 241 counted divisions, to see what informal voting blocs the record reveals. Hover or tap any square for the pair behind it, or pick a member below to pin their row and see who they vote closest to, and least like. How this is calculated.
Colour is relative to this matrix, not a fixed 0-100% scale — every cell’s exact percentage and shared-vote count is in the tooltip. Pale grey squares mark pairs with too few shared votes to compare confidently.
Voting blocs
Grouping members by how closely their whole pattern of agreement across the House tracks anyone else’s, rather than by raw agreement (which most pairs score fairly high on), turns up a consistent shape: one larger, more tightly aligned group and several smaller, looser ones. Read it as a description of how this House voted, not a declared allegiance.
Each bar is the average agreement within that group. The vertical marker on every bar is the whole-House average (62.3%). Only a bar clearing that line is more internally aligned than the House is on average; the rest are closer to typical background agreement than to a bloc of their own.
Group 1 · the largest
16 membersIts members agree with each other more than the House agrees on average: a group that votes together more tightly than the House as a whole.
Group 2
6 membersIts members agree with each other more than the House agrees on average: a group that votes together more tightly than the House as a whole.
Group 3
1 memberA single member whose voting record does not track any of the other groups closely.
Group 4
1 memberA single member whose voting record does not track any of the other groups closely.
This groups members by their recorded votes only, not by any position they’ve declared or any party they stood for. Names within each card are ordered by how distinctly they define that group, so one near the top of a card is closer to its centre than one near the bottom, whose placement is closer to a coin flip. How this is calculated.
How to read this
For each pair of members, we look at every counted division in which both are recorded as voting For or Against, and calculate the share of those shared votes where they voted the same way. A counted division is one where the names printed in Hansard or Votes and Proceedings match the tally the Speaker declared, and which was not later repeated. Divisions on House procedure are not screened out, because the record does not yet flag them.
The Keys prints only the names of those who voted, so a member missing from a division may have been absent, may have abstained, may have been in the chair or may have been out of the room, and the record does not say which. This page therefore compares members only on the divisions they both voted in. It says nothing about who turned up, and a missing name is never counted as agreement or disagreement. Where the record prints an abstention, it is left out of the comparison in the same way.
Members are arranged so that similar voting records sit near each other, making clusters easier to spot; the order itself has no meaning beyond that. A pale grey square means fewer than 15 shared votes were recorded for that pair, too few to read anything into. The Speaker appears only where the record attributes votes to them.
This measures voting alignment only. It does not test any member’s manifesto promises.