[Note: The format of the specimen ballot could not
be duplicated exactly in this scanned document]
Proposal For Electoral Reform
In recent years a number of provinces in Canada have undertaken
reviews of their electoral system. Elections in Canada have often
resulted with one party dominating the elected body with
considerably less than 50% of the vote while other parties with a
significant vote percentage can have few or no representatives. A
major stumbling block has been the reconciliation of the election
of area (riding) representatives on a plurality basis (largest vote
even if less than 50%) and the overall popular vote.
Canada is one of the few democracies which has not done anything
to improve this situation. Other countries have implemented a
variety of corrective techniques such as ranking of candidates; the
use of party lists; multiple voting rounds; and multi-member
ridings. These are all attempts to improve proportional
representation. A problem with any adjustment is that the closer
you get to true proportional representation the less the likelihood
of electing a majority government (where there are greater than two
parties).
This proposal is a simple method for improving this situation.
It incorporates the features of electoral systems which are
generally recognized as desirable but in addition incorporates a
number of unique improvements. The election of individual riding
representatives on a simple plurality vote is retained. Improved
proportionality of seats awarded is achieved by adding members on a
regional basis as defined in the rules below. The unique features
of this system are : the number of proportionally-based regional
members is not fixed but rather is self-adjusting based on the need
to improve representation as determined by the vote; independent
candidates have the same opportunities as party candidates to be
elected on a riding or regional basis; and voters always decide all
members elected in a single vote - not subject to party-lists or
complex multi-candidate voting.
RULES:
1. All candidates to the
elected body have two ways of being elected:
i)
by attaining the largest vote in the riding on a
plurality basis; or
ii)
by attaining a designated threshold-percentage of the
vote in a region.
2. A region is a
geographic-sociological grouping of ridings.
3. The threshold-percentage
vote designated to determine a region-based seat is generally
based on the true proportion of the vote that the seat
would represent in the region. This figure can be
adjusted by legislators to impact the number of region-based seats
awarded and the test for smaller party success.
4. Greater than one party-representative
by region-based seat can be provided by
designating multi-threshold levels. The choice of number
of levels and their threshold-percentage(s) will
determine how close the system is to true proportional
representation. The closer to true proportional
representation the greater the number of region-based seats and the
less the likelihood of majority government.
5. Large differences in
relative size of regions (# of ridings) can be compensated for
by varying either the number of threshold-levels by
region or by varying both the number of threshold-levels
and their corresponding percentage-vote by
region. (See elaboration of this procedure,
below)
6. All candidates have a
declared home-riding and their name and party affiliation appears
on the ballot in their riding.
Independent-candidate names appear not only in their home-riding
but on the ballot of all other ridings in their region (in these
non-home ridings they are identified as region-based
candidates). A riding victory takes precedence over a
region-based victory on the percentage-threshold basis
(where a candidate has attained both).
7. If a party has no
candidate in a particular riding, the party name still appears on
the ballot. As in the case of independents, the party in
this riding is identified as being a region-based
entry. For the party to appear on the
ballot, however, there must be at least one
party-candidate running in the region.
8. Where a party achieves a
successful threshold-percentage vote in the region the
Seat(s) awarded is(are) to the party-candidate(s) with
the greatest number of votes from
among ellow party-candidates in the region
who have not won a riding-based seat.
9. Award of seats on a
region-basis is compensatory. No seat is awarded if the party
has achieved the corresponding threshold-level
representation through regular riding victories.
10. A partial double-ballot is required for this
system. Since a party without a candidate would appear
on the ballot as well as independent-candidates from other ridings
in the region - it is possible that one of these might
attain the largest vote. To resolve this problem a
secondary part of the ballot allows voters who have
chosen a region-based candidate to place a provisional
vote for a home-riding candidate. If this situation occurs these
provisional votes would be added to the other
home-riding candidate votes to determine the riding victor.
This ballot obviates the need of a run-off election.
SUGGESTED BALLOT
Primary Ballot
CANDIDATE PARTY
1. Marlon
Brando Theta
2. Claudia
Schiffer Zeta
3. Paul
Newman Gamma
4. Robert
Redford Epsilon
5. Naomi
Campbell Chi
6. Pamela
Anderson n/a
7. M@L (member at
large) Omega
(Party has no riding candidate)
8. Sean Connery
M@L (Independent candidate
running in another riding of region)
9. Cameron Diaz
M@L (Independent
candidate running in another riding of region)
10. Clint Eastwood M@L (Independent
candidate running in another riding of region)
Secondary Ballot
If you have voted for a M@L candidate in the primary ballot,
you may make a provisional vote for a home-riding
candidate.
Provisionally vote for a home-riding candidate here:
1 2
3
4
5 6
The secondary ballot will only be tallied when a M@L
candidate has received the largest vote in the primary ballot.
ADVANTAGES OF THIS PROPOSAL:
1. The double knock-out provision gives
all candidates a second opportunity for success
(not just party-elites on a list). In closely fought
contests between two strong candidates -- both may be
successful -- one as the riding victor; the other as a regional
representative.
2. The award of extra seats in the
fashion proposed remedies serious deficiencies
in proportional representation as they occur.
It’s regional basis is important -- most, if not
all, jurisdictions are composed of distinctive regions
from a geographic-sociological perspective which can be
defined as groups of ridings. If regions are well designed this
system ensures these distinctive regional views are
always represented in the elected body.
3. Independent candidates have a
realistic chance of success on the
threshold-percentage basis. Such candidates have the
opportunity to campaign throughout their region
and develop a recognized presence with their electorate.
These voices free of party discipline are desirable. In
turn, this improved opportunity of success as an independent would
help to loosen-up party discipline.
4. Smaller parties have an enhanced
opportunity for success under this system. This
gives these small parties a chance to get started; to
learn the ropes; and demonstrate their worth to the
electorate -- if any.
5. The fact that the number of regional
based seats is not fixed allows the system to
reflect voters wishes in an effective manner. Regional
based seats are awarded only when there is a serious
deficit of party representation -- not just to fill an arbitrary
quota of seats even when a party may already be well
represented. In addition, because the number of
extra seats is not fixed, candidates are judged on an
independent test of voter support -- rather than a
confusing proportional test relative to other candidate
support.
6. Voters always determine the individuals elected
without being subjected to undemocratic party-lists of
people selected for special consideration.
7. The capacity of this system to provide riding
representatives as well as proportionally-based regional
representatives all with a single vote is important. Other systems
involve multi-candidate selection processes. Every vote cast beyond
the first is progressively less meaningful to the voter (but just
as important to the candidates). In addition, multi-candidate
systems often involve intra-party competition for the same
electorate -- this system does not.
8. Improved voter participation can be
expected as the voter has greater opportunity to
elect his choice of candidate or party through either a
riding plurality basis or on a regional percentage-vote
basis. The election is more meaningful to all voters.
9. Clarity and simplicity. Other
‘mixed proportional’ electoral systems
elect additional members to improve proportionality by
undertaking complex proportional vote,
calculations after all the votes are tallied - the
electorate and candidates are left hanging while this
‘mathematical lottery’ is completed.
Unfortunately, since elected members cannot be halved or quartered
you still get a result which is an approximate of proportionality.
This system in effect completes these calculations before the
election and sets clear and easily understood percentage targets
and number of possible additional members by region. The
electorate, parties, and candidates can readily judge their own
circumstances before the election, while votes are tallied, and
immediately at the completion of election returns.
10. Systems similar to this have been
criticized for creating two classes of representatives.
This criticism is inappropriate for this system. The
election of a regional based representative is realized
when in fact there is a deficit of representation of this regional
view -- the member represents exactly what he pr
she is -- the view of a given percentage of voters in
the region. The riding representative represents his or
her riding because he or she has attained the largest
vote in the riding. Both have equal voting rights in the
elected body and both have good reason to
be there.
ELABORATION OF RULE 5 : Adjustment For Regions Differing In
Size.
Awarding extra seats as described to regions where parties are
under-represented is proportionally correct when regions are all of
similar size (# of ridings). A problem arises when they are not --
as an additional seat would be proportionally more significant to a
smaller region than a larger one. In this proposal, legislators may
solve this problem in two possible ways -- by varying the number of
threshold-levels corresponding to the size of the region or by
varying both the number of threshold-levels and their designated
percentage-vote by region. These two options are discussed below
with the use of two examples of region design.
Example
1. Consider a jurisdiction with 100 ridings with 5
regions of 20 ridings each.
Legislators may choose to add 4 threshold levels to improve
proportional representation. Since one additional seat in 20 is
about 5% proportionally of the region -- therefore 5% may be a
logical choice for the primary threshold level. If a party attains
5.0% or more of the regional vote a seat is awarded. The remaining
3 threshold levels may be designated: 12%, 18%, and 24% of the
regional vote. Parties would be awarded 2, 3, or 4 seats in the
region if they attained these vote-percentages and had not attained
these seats in normal riding-based success. The percentage test is
made a little higher since the party has already achieved a voice
in the region with the first seat and perfect proportionality is
not the goal. At 24% of the regional vote a party would be assured
about 16% of the seats.
When applied to actual electoral results indications are that
with this process the number of extra seats awarded in most cases
would be equal to the aggregate of the number of regional levels
designated ±. 50%. In this example there would be 20
± 10 additional seats or 17% ± 8%of the total
seats in the elected body. The number of additional seats is
entirely voter-driven and depends on many factors including the
number of credible smaller parties and the uniformity of larger
party support among regions.
Example
2. : Consider a jurisdiction of 100 ridings and 7 regions broken
down as follows:
#Ridings:
5
9
11 15
16
19 25
(Total: 100)
Legislators may choose to add 1 to 5 threshold-levels by region
as follows
#Levels:
1
2
2
3
3
4 5
The smallest region would get one threshold-level and the
largest would get five. In other words the smallest region could
attain one additional seat by party and the largest region five
possible additional seats by party. This is the best arrangement
that can be made to proportionally designate the five threshold
levels among these regions.
There are now two ways to assign the corresponding
percentage-vote for these threshold-levels
1. One way is to vary the percentage(s) by region to
account for the difference in proportion that an additional seat
would represent in the region. The resultant assigned percentages
for this example would be as follows:
#Levels Threshold-Percentages
1
17%
2
8% 15%
3
6% 11% 16%
4 5%
10% 14% 17%
5 4% 7% 11% 14% 17%
This arrangement is the best proportional fit of each additional
member in relation to the size of region. However, with this method
smaller regions require a higher percentage vote to attain
representation.
2. A second and perhaps superior way of dealing with unequal
size regions is to vary the number of threshold-levels but retain
the same percentage test for each level regardless of the size of
regions. Accordingly, the threshold percentages selected for these
five levels may be as follows:
6% 12% 18% 24% 30%.
The average size of the 7 regions is 14.3 ridings so one
additional seat represents 6.5% of the representation on average -
6% therefore may be a logical choice for the first threshold-level.
The remaining percentage levels are chosen for clarity and
simplicity even though there is a small decrease in proportional
representation as the level gets higher. Establishment of a
threshold-level beyond 30% is approaching redundancy as there is a
increasing likelihood that the additional seat would have been
attained anyway in regular riding success.
Smaller regions gain an advantage with this method as they
attain an additional seat with a lower number of votes than a
larger region. However this can be justified for two reasons - it
is fairer to have representation triggered by the same
percentage-vote regardless of the size of the region; and secondly,
this is a relatively small counter-balance to the
“tyranny of the majority” of larger regions
which are always going to have the greater number of seats in any
event.
Again with this second region design the aggregate of threshold
levels is 20 so the anticipated additional seats in the elected
body would be 20 ± 10.