Wednesday, December 7, 2016

Why non-partisans don't like the Gallagher Index

The following is an edited version of a comment I added to a National Post story:  Liberals called it ‘incomprehensible,’ but professor flattered his formula was used in electoral reform debate


The Gallagher Index only measures the disproportionality of an electoral outcome based on presumed support for political parties. It assumes support for parties that doesn't always exist on a ballot that doesn't separate parties from candidates, it doesn't measure a voting system only estimates of the parliament potentially formed by a voting system, and it doesn't measure any other type of disproportionality.

Imagine you believed in secularism and someone came up with an index based on religious beliefs.  You would likely be quite offended by the index.

This is how non-partisans, people who don't vote along party lines, will feel about the Gallagher Index. It appears to be being wielded not as one tool among many in a toolbox, but as a sword to disenfranchise non-partisans. Being critical of the Gallagher Index isn't necessarily a criticism of the math, which isn't that complex.  In my case it is a criticism of holding up support for political parties as being something that should be optimized for to the exclusion of voters who vote for candidates despite, not because of, political affiliations.

In 2006, 2008 and 2011 (but not in 2015) I voted for David McGuinty in Ottawa South. While he was nominated by the Liberal party of Canada, I did not vote Liberal. I do not want my vote to be misinterpreted as support for the Liberal Party of Canada, or to go towards electing other Liberal nominated candidates.

The more I hear people talking about the Gallagher Index the less likely I'm going to be willing to vote for any party nominated candidate. I can't remember the last time there was an independent that ran in Ottawa South federally, so I guess that means I can't vote.



It is simply wrong to claim that anyone who votes for a party nominated candidate can be counted as support for that political party, but this is the nonsense activity which happens when party-PR activists measure Canada's parliament with the Gallagher Index.



I've spent more than 20 years as an electoral reform advocate who believes that plurality voting makes our parliament unrepresentative of the people that body is intended to represent.  Plurality distorts party politics, forcing the merger of dissimilar parties as happened with the Reform and Progressive Conservatives.  Left alone, a plurality system will eventually force us back into a 2-party system with only two names on the ballot.  The more diverse views that are shoved under ever larger big tents, the stronger party discipline becomes.  The harder it becomes for parliamentarians to represent constituents rather than obey party dictates, the less representative the parliament can be.

Opposition to plurality voting, not support for party proportionality, is my reason for wanting electoral modernization. While I believe that multi-member non-plurality based systems are better than single-member non-plurality systems, I believe this as I support proportionality to criteria other than support for political parties.

Of all the systems I've reviewed over the decades the only class I believe matches both those who are concerned with proportionality and those who want to fix the harm caused by plurality voting are systems based on STV: STV with fixed district magnitude, STV with mixed district magnitude, and ranked ballot Rural-Urban Proportional (ranked RUP) which is a mixed system with district magnitude of one (AKA: alternate vote in rural) and higher (STV multi-member in urban).

Think of STV as being a way to pick a team on the ballot similar to how sports fans pick a Fantasy Sports team: they pick the players they think are best regardless of what team they happen to be playing for at the moment.  STV is a proportional system, but allows both partisans and non-partisans to vote while party based proportional systems (based on party lists) benefits those who vote along party lines at the expense of those of us who do not.


Any other voting options beyond STV effectively disenfranchise some voters -- those who vote along party lines, or those who do not.  While parliaments formed by STV tend to have a higher Gallagher Index than party list systems, I do not believe this is a metric to be optimizing for any more than I believe a least squares index of religious beliefs is something to be optimized for.

We all need to remember: Under our Westminster parliamentary system Canadians elect parliamentarians, and parliamentarians form a government.  Parliamentarians can also change the government without calling an election.


Canadians are not directly electing the executive branch as they do in the USA and other countries, and it would take massive changes (including constitutional changes) to allow Canadians to directly electing a government.

Post a Comment