Sunday, August 19, 2007

Why I can't support the Citizens' Assembly on Electoral Reform report

This October 15th, Ontarians will be asked in a referendum whether they wish to adopt a "Mixed Member Proportional System (MMP) as a means of electing our provincial legislature and government. Put simply, if adopted Ontarians would be asked to cast two ballots in every subsequent provincial election, which are now held at fixed dates; one vote for their local member and a second one for their political party of choice ("list" candidates). The Ontario Legislature would be made up of 90 local members and 39 list members, whose purpose would be to ensure that a party's popular vote in the election actually reflects their proportion of seats in the provincial parliament. Therefore if a party garners 25% of the popular vote, but only elects 15 members, that party would be allotted 17 of the list members to ensure their entitled 32 seat (25% of 129) composition in the Legislature.

However, there are two important caveats to this proposed restructuring of our electoral system. The first is that a party must garner at least 3% of the popular vote province-wide in order to be eligible for any list seats. The second is that, even if a party's local member composition exceeds their seat entitlement based on percentage of popular vote, they do not lose any seats. For example, if a party earns 45% of the popular vote and wins 70 of the 90 local seats, their overall entitlement including list seats would be 58 out of the 129 in total, based on percentage of popular vote. This party would not have 12 seats taken away; they would keep their 70 seats even though their percentage of seats in the legislature exceeds their percentage of popular vote. These two limitations contribute to many's belief that the proposed changes to our electoral system would be ineffective and perhaps cause more confusion than they are worth.

At the bottom are election data from the last four Ontario elections, taken from the Citizen's Assembly report itself, and which are applied to the proposed MMP scenario. The first three columns are the political parties who won seats, the percentage of vote and the percentage of seats actually won, based on the first past the post system of electing one member per riding. The fourth column represents how the seats would have broken down in a legislature with 90 local seats, as proposed under MMP. The fifth column indicates the number of list seats that would be allocated under MMP, while the sixth indicates the total number of seats in the proposed MMP. The final column suggests the number of seats to which each party would be entitled if the above two restrictions were not in place.

The first thing one notes is that it would be virtually impossible for a party to obtain majority standing in the Ontario Legislature, wherein a party would require a majority of the popular vote province wide. Indeed, the report itself points out that the last time this occurred in Ontario was 1937. The report suggests a bias against majority governments in that they are undemocratic. Unless the actual allocation of seats matches the percentage of popular vote, we somehow have a breach of democratic principles.

However, the application of MMP principles to actual election data shows that there are flaws with the proposed model. In the first place, the elections of 1990 and 2003 are telling, in that both the NDP and Liberals, respectively, benefitted from a gross over-representation of seats based on popular vote, to a point where, under an MMP model, they won more seats than their percentage of popular vote entitled them to (the Conservatives came close in 1995). In the earlier election, Bob Rae's government would have won 51 local seats when, as a function of proportion, they only ought to have had 49. In the latter, Dalton McGuinty's Liberals won 63 with an entitlement of only 59. MMP, in spite of its seemingly dogmatic adherence to the principles of proportionality, can still result in parties "stealing" seats to which they are not proportionally entitled.

What is also telling is the fact that in each of the elections, some of the list seats would not even be allocated! This is a result of the restriction on list seats to those parties who gain more than 3% of the popular vote. The number of unallocated seats ranges from 2 in the last two elections, where independents and other parties garnered 4 and 2 percent of the popular vote, to as many as 6 in 1995, where they earned 6 percent of the vote. In some cases, the winning parties are skimming seats to the detriment of the smaller parties, while in others seats are just being wasted. Overrepresentation of seats certainly occurs with the status quo. However, why would we want to embark upon a course of potential complication and confusion if we are not going to eliminate the unfairness of the allocation of "list" seats? Ultimately, what is the point?

There are those for whom an election at least every five years just isn't democratic enough. Unless every government is a minority one, critics assume that there is something insidious at work. All parties must have their part in the formulation of public policy and if one party has a majority, that is just not possible. It is for these people that MMP is designed to assuage. However, the scenarios foreseen by the don't lend themselves to much of anything different from the status quo.

On page 11 of their report, the Citizens' Assembly asserts that "coalition majority governments" (ostensibly a mixture of two of the main parties, neither of whom has an outright majority on their own) are common under MMP and are both stable and effective. What they don't point out is that such coalitions are possible and common in the current system. Not only that, the ones we have seen recently at the federal level, where Harper's Conservatives have danced alternatively with both the Bloc and the N.D.P., may even be more to the liking of MMP supporters, as they frequently change within the lifetime of a parliament. This is the opposite of what MMP supporters hate most; an unchanging, unflinching monolith of a majority government that can't be moved for at least four years.

What is more strange, and potentially disturbing, is the report's claim that MMP "may actually enhance stability by discouraging abrupt shifts in policy that can occur in the current system when the government changes from one party to another...". Huh? I thought one of the benefits of democracy was that if the public grew tired of a party in power too long, that they could throw the bums out?! It isn't the place of an electoral system to dictate the pace of policy change, but the voters themselves. Many would argue that after eight years of Conservative rule under Harris and Eves (or even five years under Rae and the N.D.P.), incremental change wasn't the order of the day, but a wholesale change in policy direction. That's democracy.

In terms of list seats, those allocated solely on the basis of "topping up" parties who are underrepresented following the allocation of the local seats, there is a potential for problems. As it is proposed, each party would nominate their list candidates and rank them in order of preference from 1 to 39. If, for example, the Conservatives were entitled to 24 of the list seats (as in the 2003 election scenario), then the top 24 on their list would find their way onto the floor of Queen's Park. The Ontario Citizens' Assembly Report seems naively optimistic that the process of selecting and ranking of these 39 candidates for each party will not be problematic, suggesting on page 8 that "voters will be able to assess whether parties created their lists in a fair and transparent way". Thirty-nine politicians. For each party. Recent history suggests that parties have difficulty with transparency in nominating a single candidate in some ridings, if the recent spate of direct party appointments (ie. those without a formal nomination process) is any indication. The often unsavoury business of the party nomination process has the possibility of multiplying exponentially if the 39 list seats are included in a revised electoral format.

All MMP really does is make us feel better about the results being more statistically fair.....to the three established parties. It does precious little for smaller parties, nor will it have much of an effect on how parties cooperate over and above the status quo. If the intention, coupled with the imposition of fixed election dates, is to move us towards a more Americanized system of gridlock and stalemate (what the architects of MMP really mean by "stability"), then perhaps one would consider voting for a system of MMP. For my part, I will stick with the devil I know.

2003 results

% of vote

% of seats

applied to 90 seat legislature of "local" members

allotment of "list" seats

Total number of seats in MMP legislature

Entitlement of seats in MMP legislature (based on % of party vote)

PC

35

23

21

24

45

45

Liberal

46

70

63

0

63

59

NDP

15

7

6

13

19

19

other parties

4

0

0

0

0

5

TOTALS

100

100

90

37

127

128

1999 results

PC

45

57

51

7

58

58

Liberal

40

34

31

21

52

52

NDP

13

9

8

9

17

17

other parties

2

0

0

0

0

3

TOTALS

100

100

90

37

127

130

1995 results

PC

45

63

57

1

58

58

Liberal

31

23

21

19

40

40

NDP

21

13

12

15

27

27

other parties

3

0

0

0

0

4

TOTALS

100

99

90

35

125

129

1990 results

PC

24

15

14

17

31

31

Liberal

32

28

25

16

41

41

NDP

38

57

51

0

51

49

other parties

6

0

0

0

0

8

TOTALS

100

100

90

33

123

129

2 comments:

  1. From what I understand, the 3% threshold is to stop single-issue parties, like say a solely pro-life or solely pro-choice party around polarised times like the abortion debate in 1990.

    MMP was designed to prevent the Nazis from gaining power in Germany again, after WW2 in Reconstruction Germany.

    New Zealand has had a good experience with MMP. Their gov't can only govern with some other parties (because they don't have a real majority), so they have to keep promises. Another point I read in the Macleans article: http://www.macleans.ca/article.jsp?content=20070904_155329_1720

    Another interesting point for MMP and the 'list MPPs" I hadn't thought about, I read in a comment here:

    Steve Withers: Great article! Having lived in New Zealand for 11 years under a very similar MMP system to that recommended for Ontario, I can't recommend it highly enough. You get a vote that actually counts! That is just the beginning. In my riding of Otaki, we have at least 3 active MPs (1 local and 2 list) ,from 3 different parties, actively representing us. Darren Hughes of the Labour Party (local MP), Nathan Guy of the National Party (list MP) and Sue Kedgley of the Green Party (list MP). Nathan Guy has his riding offices not far from Darren's.

    So we are much better represented now than we were under the old system.


    I'd like at least to get MMP in Ontario. Then hopefully we can talk about getting rid of first-past-the-post for federal elections and institute either STV or MMP. So to look beyond just Ontario, basically; all the other provinces have not been able to pass electoral reform (mostly due to the 60% supermajority gov'ts have set rather than the normal 50%), so if we can do it, it could have great ramifications in the rest of the country.

    Some other good articles:
    http://www.thestar.com/OntarioElection/article/253963
    http://www.thestar.com/OntarioElection/article/257555

    In reference to 'stealing' seats, my argument to that would be that since FPTP stops counting votes at a certain point, MPP gives us all more rep.

    Thanks for this post btw, its so nice to see this stuff discussed somewhere.

    ReplyDelete
  2. Basically, your objections boil down to why MMPR *might* be flawed, as opposed to all that we know about how how FPTP *is* seriously flawed. The fact that as much as 10% of the electorate (well above the 3% threshold) currently do vote for a fourth party which *never* gets a single seat under our current system is an affront to any notion of democratic representation. This travesty needs to be fixed, for the precisely the reasons (voter choice beyond just two or three parties) that you raise.
    The fear-mongering about "gridlock" has no basis in reality, see the New Zealand and German cases (among many other).

    ReplyDelete