Voting Reform Summary

In Canada we elect representatives to each provincial legislature, as well as to the House of Commons federally, by first-past-the-post (FPTP) elections in each respective constituency.

The Problem

FPTP works perfectly well when there are only two candidates for a given position, but when there are more – and in our provincial and federal elections there usually are more – it tends to skewed, unpersuasive victories to the candidate merely having first-preference support of the largest minority, NOT a definitive majority win.

Such minority victories detract from the legitimacy and credibility of the decision.

They also diminish the winners’ sense of accountability beyond their own narrow voter bases, and feed into voter cynicism, disillusionment, and over-all disengagement.


There then ensues a hue and cry for voting reform – to replace FPTP with, among other things, Proportional Representation (PR).

Many people, by default it seems, see proportional representation (in some unspecified form) as the only way to address the FPTP problem. While it’s not a bad choice, necessarily, it’s also not the only, nor necessarily the best, practical and fair solution.

There are other alternatives as well: some places, Australia, for instance, use a preferential-ballot evaluated using an approach called the Alternative Vote (AV), also known as Instant Runoff Voting (IRV) (also known as Ranked Ballots).

IRV / AV / Ranked-Ballots is somewhat better than FPTP but nevertheless shares many of its worst flaws. (See Why not IRV?)

All is not lost, however, for there are still more ways of dealing with preferential ballots; much better ways, in my view, called Condorcet (“con-dor-say”) methods.

Condorcet Methods

Condorcet methods can be readily implemented with minor disruption, low cost, and major positive effect.

In (1) a single voting round, each voter casts (2) a single, simple, ballot, from which (3) a round-robin match-up of each candidate against each other candidate ensues — holisticallly considering all preferences from all ballots.

In the absence of a voter-preference cycle in the collective results, all Condorcet methods will determine a winner who most people would acknowledge as the legitimate, true, choice of the majority.

Condorcet methods are scrupulously unbiased, robust, and reliable.

Condorcet Completion

In those (arguably rare) cases where a preference cycle exists, so-called Condorcet “completion” methods break such cycles to achieve a linear ranking of the candidates.

Of these, after due consideration, I propose an over-all approach called Ranked Pairs (NOT the same as “ranked ballots”).


Ranked Pairs voting (1) is easy for voters to understand and to do and (2) can be implemented as a direct replacement for any FPTP or AV/IRV system to (3) dramatically improve democratic responsiveness. (See How it Works!).

In the end, the candidate who beats every other candidate, in one-on-one round-robin competitions, is the winner — and will be the candidate most-preferred by the majority.

Proportional Representation

But what about Proportional Representation (PR)?

The answer depends on the particular PR system proposed. The various PR systems are not all created equal, and, as always, the devil’s in the details. (See Proportional Representation)

Mixed-Member Proportional Representation

One particular method, called Mixed-Member Proportional Representation (MMPR) – in use in a number of places round the world – is of particular interest in this regard.

MMPR is an important bridge between single-member representation and fully party-list PR — bringing to bear the best of both.

Ranked-Pairs MMPR

As used currently, however, MMPR incorporates FPTP with all its fundamental problems to elect significant numbers of representatives — a huge flaw that a later top-up from a party-preference vote cannot correct. But, as for other FPTP situations, these elements can easily be replaced with Ranked-Pairs.

This melding delivers a powerful and robust PR implementation, a “best of breed,” perhaps, which I here call Ranked-Pairs MMPR.

Multiple Representation

Multiple-representation systems, such as the Single-Transferable-Vote (STV), and cases where the n-most popular candidates are elected by FPTP, since ranked-pairs determines a full ordering of all candidates, can also easily and simply be improved by replacing them with Ranked Pairs.


The goal here, then, is to demonstrate and clarify these features to promote the adoption of Ranked Pairs as a solution for Canadian elections, both federal and provincial, as well as encouraging similar electoral upgrades elsewhere.

This is also consistent with a subsequent, or even concurrent pursuit of a Ranked-Pairs MMPR system by those who hold PR as the ultimate objective.

It’s time these FPTP elections were history – Take Action, Now! Let’s get it done! Like! Share! Pass it on…

Next: The Problem at Hand

Share Button