Every now and then I find something online that is out of the ordinary and intriguing. Today, via a comment left on an article on The Independent’s website, I discovered a brand new voting system.
Has anyone else ever heard of Direct Party and Representative Voting (DPR)? I didn’t think so. Direct Party and Representative Voting is the invention of Stephen Johnson, put forward as an alternative to FPTP, AV and just about every other known voting system. It’s an interesting proposal, but I’m not certain how effectively it would work in practice.
DPR is a “weighted PR system”, according to Mr. Johnson. At the ballot box, voters would receive a ballot paper that would contain two sections. The first section would list the “qualifying parties” (more about qualifying parties below), and the second section the list of candidates running in that riding. Each candidate would be identified as either belonging to/associated with one of the parties listed in the party section, or as an Independent. Mr. Johnson uses the terminology of a party “adopting” the candidate, but I can’t find anything on his website that explains how parties adopt candidates. Voters would have one vote in each section. The two sections would be counted separately. The Section 2 count would determine the election of the MP, using the usual FPTP. The count of Section 1 wouldn’t have any impact at the local level, but would become important when aggregated with all the other party votes to arrive at a national total for each party. The point of this will be explained below.
As under FPTP, which party forms the government would depend on which party could command the confidence of the House. If no single party was able to command a majority, government would be formed by a coalition of parties, or by a party as a minority government.
As you can see, thus far at least, DPR works pretty much like FPTP – single member constituencies elected by FPTP. The differences become apparent when it comes to voting in the House, which is where the party votes comes into play.
Government bills would require a majority in the house to pass, as is currently the case. However, where DPR differs is that this majority would be determined by the votes won by each party, not the number of MPs. This is where Mr. Johnson argues that PR is achieved with DPR. The parliamentary voting strength of each party is scaled to reflect their total votes rather than their seats. This means that each MP is entrusted with a vote which has a fractional value, an equal share of the Party’s parliamentary votes. In other words, it would no longer be “one MP one vote”.
To use Mr. Johnson’s examples, if a party received 40% support in “Party” vote (section 1 of the ballot) but won 50% of the seats (due to the vagaries of FPTP), each of their MPs would have a vote value of 0.8%. MPs from a party which got 40% of the “Party” vote but only 30% of the seats would have a weighted vote of 1.333. Independents would have a vote value of one.
How would this work in practice? If we take the results of the 2008 general election here in Canada, the results were as follows:
|Party||% Popular Vote
|% of Seats Won||Weighted Vote|
* Since people weren’t asked to vote separately for a candidate and for a party, we’ll use each party’s actual share of the popular vote as a stand in for the “Party Vote” under DPR, even though it is clear that some people voted for a candidate from a party other than the one they’d prefer in an attempt at strategic voting.
What would this mean on an actual vote in the House of Commons? Let’s take the example of vote on Third Reading of the 2010 Budget Bill (C-9).
|Actual Vote||DPR Vote|
As you can see, in the DPR scenario, the Budget would have been defeated, meaning the Government would have lost the confidence of the House (the Budget vote being a confidence matter).
This weighted voting would apply only to government bills. Private members’ bills and other non-government measures would be voted on the traditional way – one MP one vote (or free votes, as Mr. Johnson calls them).
There are some problems with this system. The most obvious to me is that DPR does nothing to address the issue of parties which receive a significant number of votes, but don’t manage to elect any members thanks to FPTP voting at the constituency level. In Canada, the obvious example of that is the Green Party, which received almost 7% of the vote in 2008. From reading through Mr. Johnson’s site, I think he believes that because people will be able to separate their vote for MP and their vote for Party, people will be more likely to base their vote for MP on each candidate’s qualifications rather than their party affiliation:
With DPR Voting, the election of the Representative depends on the merit of the individual. That might be personal qualities, track record in public life locally or nationally, and perhaps charisma. Most small parties who win even a small share of the national vote have at least one outstanding individual who might well be elected on their personal qualities, public profile etc. With DPR voting the voter can both support his/her preferred party, and then freely choose the best person to represent the constituency.
Sadly, I think this is a bit naive. There are a lot of voters who are simply hardwired to vote for anything that wears their favourite party’s colours. For a significant percentage of voters, their vote for candidate and party will be identical. Their choice of candidate will be based solely on that candidate’s affiliation with the party they support. They won’t care if one or many of the other candidates are far more qualified, more charismatic, more outstanding, etc. They will vote party and party only, on both parts of the ballot.
Mr Johnson does put forward the idea of an “automatic representation threshold”:
If no candidate from a party was elected in the Representative ballot but the party won enough votes to exceed the chosen threshold percentage, the leader of the party would automatically be elected as an MP. In this way the party would be represented in the Parliament and the MP would be able to exercise the appropriately ‘heavy’ vote. Such an MP would have no constituency link, but this would be an exceptional circumstance.
This might solve the problems the Greens face here in Canada, but somehow the idea of one MP with no constituency link sitting in the House of Commons where all other MPs do represent a constituency strikes me as odd, and something to which the larger parties would probably object.
Mr. Johnson also claims that DPR would put an end to “wasted votes”, however, I don’t see that this would be the case. Yes, voters would have the second vote for party, but DPR would not address the issue of MPs being elected by a minority of voters (i.e. receiving less than 50% of votes cast in their riding). As I’ve previously written about, in both the UK and Canada, a majority of MPs win their seat with less than 50% of the vote, and many have much less than 50%. While the weighted vote would sort of balance out voting results on votes on Government bills, DPR would do nothing to address the fact that most MPs will have been rejected by a majority of voters in their riding. DPR might marginally alter the outcome of votes on government bills, but MPs do far more than simply vote on government bills. DPR does not affect any of these other activities. For example, committee chairmanships and membership are based on how many seats a party has in the House. DPR won’t do anything to address this issue: an underrepresented party will still be underrepresented on committees.
Another issue is the assumption that votes on private members’ bills are free votes. Here in Canada, the Conservative government has used Private Member’s bills as a means to push through party policies that it knows would be defeated if presented as a government bill (because the other parties traditionally whip the vote on government bills, but allow their members to vote as they wish on private members’ bills). A good example of this was Bill C-391, to repeal the long-gun registry, something the Conservative Party has wanted to do but that the other parties oppose. Bill C-391 was a private members’ bill introduced by a backbench Conservative MP and made it through second reading to committee stage. The vote on concurrence of the committee report was defeated, thus defeating the bill before it got to Third Reading. Two of the three opposition parties whipped their vote because they recognized that this was an attempt by the government to sneak in through the back door legislation it couldn’t pass as a government bill.
There is also the matter of tactical voting in the House. If we go back to the vote on the 2010 Budget mentioned above, Canadian readers may recall that a significant number (30) of Liberal MPs abstained from the vote. This was tactical. The party opposed the budget but didn’t want to defeat the government. Under minority government conditions, which is what Canada has known since 2004, such strategic voting is commonplace. It’s one thing to juggle the numbers to ensure that a protest vote is registered that won’t bring down the government when it’s one MP one vote; if parties have to start calculating vote outcomes based on weighted votes, I can only imagine the possibilities for miscues that might arise.
Of course, the flip side of this is that because the vote is weighted and the governing party is (probably) over-represented thanks to the vagaries of FPTP, that party, in theory at least, would have to work harder to accommodate the interests of other parties in order to ensure its survival, even when in a majority government situation, as I will demonstrate below.
I’ve used the results of the 2011 election, which resulted in a majority Conservative government and the vote on the 2012 Budget motion (Bill C-38) as examples of DPR under majority government conditions.
Election outcome and resulting weighted vote:
|Party||% Popular Vote
|% of Seats Won||Weighted Vote|
Vote on 2012 Budget Motion:
|Actual Vote||DPR Vote|
*There were two Independents, one who was elected as a Conservative, and who voted for the budget, the other was elected under the NDP banner and who voted against the budget. Since they weren’t elected as independents, I included them in the totals of their original parties.
As you can see, under a weighted vote, the budget would have been defeated, and the since that is a confidence vote, the government would have had to resign.
Mr Johnson posits that parties would have to qualify in order to be included on the ballot, but doesn’t elaborate:
conditions for Party qualification is not inherent to the system and would be a matter for separate debate. The process of qualification could be regional or national and might, for example, require a number of signatures to be obtained across a number of different constituencies.
I am not certain how parties are recognized in the UK, but since Canada’s Elections Act already has a well-established process for registering parties, this caveat probably wouldn’t have to apply in Canada. Mr. Johnson’s goal here is to limit the number of smaller parties participating in each election. I don’t think this is good thing since limiting peoples’ democratic choices is not something I am prepared to do. I would strongly resent it if someone tried to disenfranchise the party I prefer, therefore I would not want to dictate which parties other people are allowed to vote for.
Mr. Johnson’s proposal is certainly interesting, but obviously completely untried. I can’t see any government willingly adopting a completely unknown and untried voting system. It also does little to address some of the major failings of FPTP, despite Mr. Johnson’s claims to the contrary. Still, it’s always good to see people challenging the status quo with new and innovative ideas. Kudos to Mr. Johnson for the effort!
(Note: you can read far more about DPR on Mr. Johnson’s site (link at beginning of this post). He includes charts comparing DPR to other electoral systems, among other things. I’m certain he’d appreciate any feedback you might care to offer. He has also written a follow-up post for this blog addressing some of the issues raised in this post: Part 1, Part 2.)