Part 1 (of multi-blog research piece)
Canada once used a variety of election systems, not just the single-winner First Past The Post that now is the rule in federal and provincial elections across the country.
The obvious sign of this is the use of multiple-member districts in federal elections (prior to 1968) and the use of MM districts in provincial elections in every province at one time or another. In Quebec they were used prior to 1867; in every other province MM districts were used at some point between 1867 and 1996 - in some cases even before 1867 as well. Every province used multi-member districts at one time or another, to elect all or only some of their members.
The systems used included Block Voting, Single transferable voting, Limited voting and a system where each seat was filled through a separate contest. STV and Limited Voting almost always produced mixed multi-party representation, according representation to both the majority and at least the largest minority. Block voting and single-winner contests usually (but not always) resulted in one-party sweeps of the district's seats.
Now federal elections —and all provincial and territorial elections— use only single-member districts, a situation that came about through electoral reform.
The British North America Act (1867) stipulates that the County of Halifax would have two members. All other ridings would have just one.
The method of voting is not mentioned in the Act but presumably it was assumed that each voter would have as many votes as the number of members to be elected. Anyway, that is the way that voting was conducted in federal elections, until the last MM ridings were disbanded in 1968. (The first break with this rule anywhere in Canada was in 1886 in the provincial district of Toronto, with the use of Limited Voting.)
All of the federal ridings and most of the provincial MM districts used Block Voting, where often - but not always - one party took all the district seats. The last time Block Voting was used in Canada in a provincial election was in 1986 in BC, when 34 -- about half the province's MLAs -- were elected in two-member districts. One district did elect mixed representation in that election.
PEI used multi-member districts for about a decade after its last use in BC, but there each seat was filled through a separate first past the post election contest. (so multi-member districts but not multi-member contests).
The last time a non-plurality system (a PR system or a majoritarian system) was used at the provincial level was in 1955 in Alberta when PR-STV was used to elect Edmonton and Calgary MLAs in 6- and 7-seat districts, and Alternative Voting was used in single-member districts in the rest of the province.
Even where district results are dis-proportional the overall party composition of the legislature may not be too dis-proportional. The balancing of opposite one-party sweeps in various districts that happens sometimes under First Past The Post means that parties may be represented proportionally as per their vote tallies, but meanwhile such a result entails that the votes cast by the supporters of the lesser party in each different region are wasted and those voters are unrepresented. It can also mean that the cabinet cannot include members from whole sections of the province (or country if we talking about the federal government) simply becasue the government party does not have members in those regions.
In every year from 1792 (or earlier) to 1838 and 1841 to 1861, there were always at least some government representatives elected in MM districts serving in the colonial legislative assembly of Upper Canada or Lower Canada or the United provinces (Quebec and Ontario).
From 1867 to 1996, there were always at least some government representatives in one or more provinces elected in MM districts.
From 1867 to 1968, there were always MPs elected in MM districts.
================
The Use of Multi-Member Districts in federal elections
Eleven ridings elected multiple MPs at one time or another, between 1867 and 1968. These were Halifax (NS), Ottawa, West Toronto, Hamilton (ON), Cape Breton (NS), Pictou (NS), St. John City and County (NB), Victoria (BC), and three in PEI: King's County, Prince County, and Queen's County (later again an MMD under the name Queen's).
(More on this in Part 2)
========================
All provinces and territories used MMDs at one time
Each of the provinces and territories (except Nunavut) once used multiple-member districts.
They switched to electing all their members in single-member districts elected through First Past The Post in these years:
Quebec 1867
Ontario 1893
North-West Territories 1894 (see the 1891 NWT election)
Yukon 1903
Manitoba 1954
Alberta 1956
New Brunswick 1967
Saskatchewan 1967
Newfoundland and Labrador 1975
Nova Scotia 1978
BC 1990
PEI 1996
(PEI's elections were special cases as each district elected two members, and voters who owned property in the district voted for the Councilman while voters resident in the district joined with the property-owners to vote for the Assemblyman. Later the exact same voters were allowed to vote for each of the two members in a district but still each seat was filled in separate contest.)
The systems used included Block Voting, Single Transferable Voting (STV), Limited Voting and a system where each seat was filled through a separate contest, either through FPTP or IRV.
Now federal elections -- and all provincial and territorial elections -- use only single-member districts, a situation that came about through electoral reform.
=======================================
What is the purpose of multi-member districts?
Multi-member districts under STV or district-level party-list PR, and
pooling of overall votes as under MMP (as in New Zealand) or party-list PR,
or both district PR and overall PR, as under Denmark's MMP/district PR,
means that a large proportion of the voters have actual representation that they voted for.
Multi-member districts means:
less fragmenting of the electorate (because there are fewer districts) so less opportunity to gerrymander
and allows the
use of natural pre-existing boundaries for the districts
such as city corporate limits or counties or provincial borders,
with the varying sizes of such natural districts being compensated for by varying number of members.
As long as each district has multiple members and each voter cast just one vote, balanced mixed representation is almost always produced --and always produced if seat count is more than three, if we look at past use of MM districts in Canada
or if voters may cast multiple votes, one party may not take all the seats in the district if voters have liberty to lump their multiple votes on one candidate (but mixed representation is less guaranteed than if single voting is the rule.)
As well, party-proportionality of results is more guaranteed if voters can mark back-up preferences, as in ranked voting (in MM districts of course - meaning STV).
Increased proportionality of representation in each district means
less incentive to gerrymander.
If party A gets all its due seats in district North and also in district South, why bother to shift the border southward or northward?
And the same holds true for Party B.
PR as can happen in multi-member districts means fairness to all, small and large parties and of all stripes.
Each party will get what the voters decide is its due.
Very fair...
Here's an essay on the use of MM districts in Canada, 1867-1990s, with the use of MM districts in the U.S. and U.K. also being briefly discussed.
================================================
Canada’s Past Use of Multi-Member Districts
Tom Monto, Edmonton January 2022
Recently I came across the 1976 thesis, Challenges to the Voting System in Canada 1874-1974, by Charles Harry John Phillips.* It inspired me to conduct a self-led research project on the use of multi-member districts in Canada’s past.
Phillips noted that challenges to the federal FPTP voting were more numerous than might be thought and in passing he mentions an old debate on the use of multi-member districts instead of the single-member districts that now are the rule in provincial and federal elections in Canada.
Multi-member districts (MMDs) are electoral districts that send two or more members to a legislative chamber. Although not used for election of legislators today in the U.S. or Canada, they were more common in both U.S. and Canadian political history than may be thought.
===================================
*Charles Harry John Phillips,
Challenges to the Voting System in Canada 1874-1974
University of Western Ontario, Ph.D. thesis, 1976
(available on-line https://ir.lib.uwo.ca/cgi/viewcontent.cgi?article=1860&context=digitizedtheses)
======================================
Phillips noted that among the many challenges to the federal FPTP voting in Canada in the past, very often new voting schemes included plans for multi-member districts. Phillips cautioned the reader that his thesis did not address the single-seat districts versus multi-seat districts question in any depth. This hole I seek to fill, at least to some degree, below, after a short discussion of the forms of challenge that Richards did discuss and the use of MM districts in the U.S.
Phillips in his thesis says early reform concerned questions of
- secret versus open voting (secret voting was settled as the best way to go) (secret voting first pursued as early as 1831 (p. 7);
- long elections where the vote in a district could take days or where elections in districts were held separately each in turn (to allow time for remedial and corrupt practices to be brought to bear)
- compulsory versus optional voting (should voters be forced to vote?) (Canada decided to keep voting voluntary. (In Australia, voting is compulsory - with negligent voters receiving something like parking ticket in some cases.)
- the method of aggregating votes to ascertain the successful candidates (the only topic of discussion in the electoral reform debate today). (plurality, majoritarian or quota-based?) (p. 4).
The use of the contingent vote (also known as ranked or preferential voting), the Ware formula, Bucklin vote, STV and Alternative Voting (IRV) began to emerge in the late 19th Century. As well, Limited Voting and Cumulative Voting also were discussed and sometimes brought into use in late 1800s/early 1900s. (Challenges to the Voting System in Canada, p. 14-15)
Of the systems listed, only STV is proportional. It is not party-list PR because it is candidate-based. It is considered Proportional because it elects members in a district roughly in due proportion to party vote share, at least as long as perhaps 10 to 20 percent of district voters vote for that candidate.) (p. 20)
Limited Voting worked like STV in that it produced both majority and minority representation; Cumulative Voting also did so as well but only if minorty goup voted carefully. Neither use transferable votes so more votes are wasted under those systems than under STV.
The other systems listed such as IRV are majoritarian, ensuring that the elected member is elected with the support of a majority of voters, with no attempt at proportionality. Formulas for counting of the vote vary [in part based on whether you think first choice is held much more highly than back-up preferences or if you think they ought to be valued about the same].
Gove system
Another early alternative to FPTP, not mentioned in Phillips’ list, is the Gove system, now known as Indirect STV. Voters cast single votes in multi-member district. Voters do not mark their ballots with rankings, but votes are transferred, as needed, based on the candidate’s pre-set instructions. A handy system to produce many of the benefits of STV in districts where it is difficult to collect all the ballots in one central place to conduct STV transfers or where X voting is preferred over ranked voting.
(See https://www.cs.jhu.edu/~jason/papers/eisner.istv91.pdf (un-dated [1990?]) for discussion of its potential use in South Africa.)
(Wikipedia “Single Transferable Voting” has a paragraph on the Gove system.)
=========================
Phillips presents highlights of the 1910 Report of the Royal Commission Appointed to Enquire into Electoral Systems (U.K. government), which investigated the alternative systems mentioned. By 1910, Alternative Voting had been put into use in Queensland (1892) and Western Australia (1907) and in some U.S. cities. STV had been used in Tasmanian cities and state elections (1896/1907) and perhaps in city elections in Johannesburg, South Africa.
Use of those systems came slightly later to Canada. Municipal-level AV was first used in Canada in Lethbridge in 1913 and municipal STV in Calgary in 1917. STV at the provincial level was first used in Canada in Winnipeg in 1920.
Urban STV/rural IRV, where all members in Alberta were elected either in STV or IRV, was first used in Alberta in 1926.
Phillips says the 1926 Alberta election “was the first election in Canada where every member in the chamber was elected through non-plurality systems.”
But it should be noted that that does not mean that single-member FPTP was the only electoral system used before 1926.
Previously, there had even been occasions where no member of a specific provincial legislature was elected through single-member FPTP. (And in fact STV may actually be considered a plurality system - each popular candidate is elected (in some cases only as long as he/she is popular enough to have quota) and no less-popular candidate is ever elected in place of someone more popular.)
PEI elected all its members in multi-member districts for its entire history until 1996.
Every Nova Scotia and New Brunswick provincial election from entry into Canada to the 1970s was held with members being elected in MM districts as well as single-member districts, or only in MM districts. (First election held in Nova Scotia - in Halifax in 1758 - used a MM District.)
Every BC provincial election from 1871 to 1990 also used a combination of MM districts and single-member districts.
The district contests electing multiple members were pure plurality contests - there was no quota or majority required to be elected. They were conducted using Block Voting, where each voter cast multiple votes (up to as many as the number of empty seats). The leading party often made a clean sweep of the district’s seats without a majority of the votes being required. But they had not been single-member FPTP contests.
Philips wrote that it is recognized that in Canada the voting system functions in conjunction with territorial constituencies varying greatly in geographic area and population, frequently justified by reference to a flexibly defined "sense of community". (p. 5) [This flexibility is such that widely-varying-sized single-member districts (federal ridings, provincial constituencies, city wards in some cases) all claim credit for producing "local representation.”]
Some voting schemes proposed would have geographical districts replaced by groups based on functional or occupational composition (such as workers and farmers). [Phillips said he would discuss these only when the new scheme was pursued seriously.] (p. 5)
He presents much on the early work of Canadian MP Richard Cartwright and the farmer-oriented Patrons of Industry, which laid the base for the advances that electoral reformers experienced during the "Progressive" period" of the inter-war years.
Phillips says early PR was seen by some as a way to get main parties to veer to address farmer and worker concerns; others saw it as a means for farmers and workers themselves to get their own candidates elected ("third" parties; direct representation in the House of Commons).
==================
Multi-Member Districts
A footnote in Philips’ thesis (p. 5) led me to this thesis:
Ruth C. Silva, "Compared Values of Single- and Multi-Member Legislative District", Western Political Quarterly, 1964
Ruth C. Silva, in “Compared Values of the Single- and the Multi-member Legislative Districts,” noted that multi-seat districts were used in the U.S. more than is generally recognized and that the effect of the MM districts depended on which voting system was used. (p. 5)
Much of my discussion below is centred on discussion of the variety of voting systems used in the different Canadian provinces during their use of MM districts.
====================================
Canada's MMDs
MM districts were used in Canada more than is commonly known.
Multi-member federal riding(s) were used in every federal election prior to 1965, starting with the Confederation election of 1867.
The last provincial multi-member electoral district was used in 1996. Prior to that, every province had used multi-seat districts at one time or another.
In 1952 every province except Ontario and Quebec was using at least one multi-member district, with New Brunswick and PEI electing all their MLAs in multi-member districts at that time. PEI elected all its members in multi-member districts for its entire history until 1996.
Little known is the use of MM districts in the Prairie Provinces. But each of those three western provinces used one or more MM districts for decades.
BC also used a mixture of MM and single-member districts for its entire history prior to 1990. That province's use of Alternative Voting is well known while Alberta and Manitoba's use of PR-STV (in five districts) is often overlooked. And the same happens with MM districts. Many overlook that New Brunswick and PEI (and the other two Atlantic provinces to a lesser degree) have a long and deep history of using MM districts.
The use of MMs dates back to the pre-Confederation period, even to the time when British North America included present-day U.S. and MM districts were even a feature of elections in the Old Country in the past.
==============================
Old England
Multi-member districts were used at the start of parliamentary democracy in old England. In 1265, Simon de Montfort convened a Parliament that contained two knights from each shire (county), and two burgesses from each borough (city) — one of the earliest examples of representative parliaments in the world. Single-winner First Past The Post did not come into use until later.
In 1793, Charles Grey (Earl Grey, II), a 28-year-old MP, brought the attention of the House of Commons to a petition formulated by the society named the Friends of the People. The petition described the iniquitous districting and allocation of seats in the country and called for "parliamentary reform." The petition related that at that time, The House of Commons was to a large degree elected in two-seat districts and many of those had rather few voters, thus creating safe seats for the Tory party.
It said there were 70 MPs elected in 35 very small boroughs ("burgage tenements" where the landlord determined how tenants would vote), 88 MPs elected in 44 boroughs of less than 50 voters each, 36 elected in 18 boroughs with less than 100 voters each, 52 elected in 26 boroughs with less than 200 voters in each. These plus 45 members elected in Scottish boroughs and burghs with less than 250 voters in each (the number of seats in each is unknown) made up the majority of seats in the House of Commons. Thus perhaps as few as 20,000 voters determined the composition of the majority in the House of Commons. Britain at the time had more than 10 million people.
The History of Reform (1884) said that Proportional Representation (minority representation) was not considered a pressing matter in the early 1800s and gives the reason "our forefathers suffered too much and too long from the representation of nothing but a minority, to think much about that in their plans of reform." (p. 83)
The House of Commons voted down Earl Grey's motion in 1793 but years later he rose to be Prime Minister and then he and his Whig party pushed through the Great Reform Bill of 1832.
=============================
In 1831, just before the passage of the Reform Bill, So only 106 MPs were elected in single-member districts.
540 MPs were elected in two-member districts.
12 MPs were elected in four-member districts.
Britain's MPs were elected in these various types of constitituencies:
BC = Borough constituencies
CC = County constituencies
UC = University constituencies
(number shows district magnitude)
BC-1 BC-2 BC-4 CC-1 CC-2 CC-4 UC-1 UC-2
England 4 195 2 0 38 1 2 242 dist.
E's MPs 4 390 8 0 76 4 4 486 MPs
Wales 13 12 1 26 dist.
W's MPs 13 0 0 12 2 27 MPs
Scotland 15 30 45 dist.
S's MPs 15 0 0 30 45 MPs
Ireland 31 2 0 0 32 1 66 dist.
Irish MPs 31 4 0 0 64 1 100 MPs
Total districts 63 197 2 42 71 1 1 2 379 dist. tot
Total MPs 63 394 8 42 142 4 1 4 658 MPs tot
(The Monmouthshire districts (a 2-seat County constituency and a single-member Borough constituency) is included in Wales.)
==========================
By passage of the Reform Bill,
56 "rotten boroughs" (all but one having two seats) were disbanded altogether;
30 "rotten boroughs" lost their second member (going down to being just a single-member district);
one district (Weymouth and Weyton Regis) went down from four seats to two.
Altogether 143 English MPs lost their seats.
But the point of the reform was not to change the number of MPs in the House so the following additions were made to cover the reduction.
The Act created 130 new seats in England and Wales:
-- 26 English counties were divided into two divisions with each division being represented by two members.
-- 8 English counties and 3 Welsh counties each received an additional representative.
-- Yorkshire, which was represented by four MPs before the Act (including the City of York, a separate district within the county), was given two additional MPs and split into three districts -- East Riding, West Riding and North Riding -- so that each of its three ridings was represented by two MPs).
-- 22 large towns were given two MPs.
(This information and much more detail can be found in Alexander Paul's 1884 book The History of Reform - A Record of the Struggle for the Representation of the People in Parliament, available for viewing online on the Hathi Trust website.)
Thus at least 88 districts elected two members each, or more (three or four were used occasionally). Usually block voting was used to elect the members in these, but sometimes other voting systems were used, as described next.
In 1867 some switched from Block Voting to Limited Voting. Co-incidentally this change was done at the time of Canada's Confederation. The Limited Voting system "was applied in England to constituencies returning more than two members from 1867 to 1885…. Limited Voting was applied to 13 constituencies each returning three or four members and included Glasgow, Birmingham and the City of London." (Phillips, Challenges to the Voting System, 1867-1974 (p. 17)
In 1885 the "Third Reform Bill" passed. In it or the related Redistribution of Seats Act (1885), the number of multi-member districts was cut back decisively. After 1885 only 54 MPs were elected in multi-member districts.
Major redistribution under the Act as follows:
Parliamentary boroughs (later known as borough constituencies):
All these units with a population of 15,000 or less ceased to have separate representation and were merged into a wider division (constituency) of their county – namely 79 constituencies were disenfranchised.
Six other boroughs were also merged into the county divisions: four that included large extents of countryside (Aylesbury, Cricklade, East Retford, Shoreham) and two that had been disenfranchised for corruption (Macclesfield and Sandwich).[10]
Those with populations between 15,000 and 50,000 were to have their representation reduced from two MPs to one, namely 76 constituencies.
Two-seat districts
Those with populations of more than 50,000 (23 in all) continued to be a set of two-member constituencies
The City of London would have its representation reduced to two MPs and remain undivided. (The city was represented by four MPs until 1885, when this was cut to two, and in 1950 the constituency, still sized at only a square mile of downtown space with very few residents in 1950, was abolished.)
The Universities of Oxford, Cambridge, and Dublin would each return two MPs.
The number of seats in the Commons was increased from 652 to 670, inclusive of Ireland.
so 54 MPs out of 670 elected in MMDs meant that after 1885 only 8 percent of MPs were elected in MMDs (two-seat districts).
As block voting was used to fill the seats, there was little fairness in the district results despite the potential that MMDs, even two-seat districts, had for fairness if only single voting is used.
Although later the University seats were filled using STV.
STV was used to elect some members of the House of Commons in MM districts for many years. In 1918 STV was adopted for the university constituencies of Cambridge, Oxford, and others. These constituencies, which it seems had two seats each, used STV until their abolition in 1948 (or 1922 in the case of Dublin University).
========
"Multi-member constituencies existed in the Parliament of the United Kingdom and its predecessor bodies in the component parts of the United Kingdom from the earliest era of elected representation until they were abolished by the Representation of the People Act, 1948."
(From https://en.wikipedia.org/wiki/Multi-member_constituencies_in_the_Parliament_ of_the_United_Kingdom).
In 2016 the Boundaries Review Commission reviewed the old-time use of multi-member districts and posited that it ws useful way to proceed. (But the ridings stayed being single-member districts anyway.)
Its report said
"As an example, Britain’s counties and major cities have the merit of being both grounded in history whilst retaining distinct administrative functions and common cultures. People identify with the county or major city they inhabit. These are also large enough that it is unlikely that redrawing boundaries will become necessary in the future. As the population expands or contracts, the number of MPs can be adapted accordingly.
And multi-member constituencies could easily be adapted to proportional voting systems in the future if the population wanted it. For the very largest counties and cities a subdivision could be made, rather like the historic Ridings of Yorkshire [when Yorkshire was broken up into East Riding, West Riding and North Riding]
And large constituencies returning five to ten MPs would give smaller parties a chance to mitigate the “wasted votes” syndrome felt by UKIP, the Greens – even the Liberal Democrats...."
see also Electoral Systems and Electoral Reform in Historical Perspective by David Klemperer
https://consoc.org.uk/publications/electoral-systems-and-electoral-reform-in-historical-perspective/
=========================
Multi-member districts in the U.S.A.
Currently today, ten U.S. states have at least one legislative chamber that includes members elected in multi-member districts. Many of them elect using the ticket voting system so no proportionality is usually produced despite use of MMDs.
The last MMDs used to elect national-level legislators were disbanded in 1842.
see my blog
for more information
The majority of states use single-member districts at both the federal and state levels, but Arizona, New Jersey, South Dakota and Washington use MMDs to elect all state House members.
Six other states use MMD(s) to elect some of their state legislators.
Ten other states allow the use of MMDs by law even when not used.
Five states have no law prohibiting or permitting MMDs.
Of the 7,383 seats in the 50 state legislatures, about 14 percent (1015) are elected from districts with more than one member.
(a slightly differnt figure is provided in a different source - see my blog)
==================
Districts and the different types of voting
A particular form of MMD is at-large election where all the members of a government body are elected in one district. In a few cases, a jurisdiction is divided into multiple multi-member districts.
The alternative to MMDs are single-member districts where just one member is elected - through First Past The Post, or Alternative Voting (AKA Instant-runoff Voting). More on this later.
Unlike single-member districts, the voting systems used in MMDs were more varied. Most of these MMDs used Block Voting, where each voter could cast as many votes as there are seats to fill. There are two variations of Block Voting. Both were used in Canadian history.
There were many other choices available as well - and Canada’s MMDs are known to have used five different voting systems.
================
Distinguishing Characteristics of MMDs
There are several distinguishing characteristics of the possible forms of voting in a MMD:
(A piece written by the Vermont Legislative Research Service provided the basic information presented here.
https://www.uvm.edu/sites/default/files/Department-of-Political-Science/vlrs/PoliticsGovernment/MMD.pdf)
The number of votes that each voter can or must cast
1. Block voting (bloc voting): Voters receive as many votes as there are open seats, and can vote no more than once for a particular candidate. All votes must be used.
The type of Block voting where voters had to cast the same number of votes, no more and no less, was also not used in Canada, judging by the number of valid votes tallied where known.
Only known instances of this in Canadian history are early election(s) in colonial Nova Scotia.
Instead Block Voting in Canada allowed for voters not to cast all their available votes, as described next.
2. Block voting as used in Canada (formal name: Bloc with partial abstention (BPA)
Voters may cast as many votes as there are open seats. and can vote no more than once for a particular candidate. Voters can elect not to use all of their votes.
3. Limited Voting
Each voter can cast more than one vote but not as many as the number of open seats.
Used in Toronto when it was a MMD, from 1886 to 1894.
4. Each voter casts just one vote.
Almost invariably this is the case where just one member is being elected but it is also the case in multi-seat contests, where it is called Single Voting.
Also, casting one vote is the rule where the seats in a MM district are filled through separate contests, as explained below in #9.
Type of votes cast
Where the voter is casting single votes, preferential votes are an option -- see #6.
5. Each voter casts just one vote - and it is non-transferable.
Single Non-Transferable Voting (SNTV) uses X voting but as no one group can (under normal circumstances) take all the seats in the district, mixed representation results. Mixed representation is the hallmark of proportional representation. Not all mixed rep is proportional but the result is more balanced and therefore more fair than a one-party sweep of multiple seats.
SNTV has not been used in any election in Canada.
6. Single Transferable Voting (PR-STV) has all the benefits of SNTV. And further it reduces the number of wasted votes/ignored voters that plague FPTP and other forms of X voting. It has the drawback of ranked voting, which some consider to be irksomely-complicated for voters.
PR-STV was used in five different provincial MMDs on the Prairies in four decades in the 1900s.
===========================
Limit of one vote per candidate or no such limit or something in between
Another distinguishing feature of voting is whether votes can be "lumped" on to a candidate or have to be cast with a limit of one for any specific candidate.
Mostly in Canadian elections, voters cannot give more than one vote to a candidate.
But in these two systems that was possible.
Neither of the systems were used in provincial or federal elections in Canada.
7. Cumulative Voting: Voters cast multiple votes and may use their votes however they wish, such as casting all of them on a single candidate or casting one or more of them for different candidates. Never used in Canada federal or provincial elections at any time. Illinois used this system previous to 1982. This system is not used in U.S. state legislative elections today.
8. Cumulative voting where each voter can give no more than say three votes for a specific candidate. This ensures a minimum size of the minority that can be assured of electing one candidate.
Cumulative Voting was used to elect some city officials in Toronto starting in 1903. Here’s how it was intended to work in that instance.
In the Toronto city election of 1903, 12 members of the Board of Education were being elected. Each voter could cast up to 12 votes, but with no more than three given to any candidate. This meant that if one-quarter of the voters gave all their three votes to a candidate, he or she would be elected. If they did that with four candidates, the four would be elected - four out of 12. Any minority smaller than a quarter of the electorate would not be assured of electing even one candidate, due to the three-vote limit.
All members' terms end at the same time or are staggered
An election may either fill all the seats of the MMD district or only a portion of the seats in a MMD. These two options can be used with almost any of the voting variants used in MMDs listed above. A single-member district cannot have elections that are staggered.
In Staggered elections in MMDs -: Two or more legislators represent the same district with elections happening in different years due to staggered terms.
If only two members, staggered elections become single-seat contests, but in MMDs with more than two seats, the staggered elections may be multi-seat contests.
Most often, in Canadian federal and provincial elections, all the seats are filled at the same time, with concurrent terms. Although on some occasions, specific districts hold their portion of the provincial vote days or weeks after others, due to local organizational, weather or transportation issues.
In Canadian experience, staggered elections were common at the municipal level. The cities of Edmonton and Calgary used staggered elections prior to the 1960s.
9. “Posts elections” - where MMDs did not mean multi-seat contests
Simultaneous separate elections
Even where districts have more than one seat filled at the same time (either no staggered terms or the District Magnitude is high enough that staggered terms do not prevent two or more seats being filled at one time), still the seats may be filled through single-seat contests. This can be accomplished by what is termed MMDs with posts. Instead of running in a single pool of candidates, candidates are divided up, and each separate group runs for just one of the district's seats, as in a single-member district.
Each voter may cast multiple votes, casting one for a candidate running for each post.
This was how BC filled the seats in its MMDs in the 1952 and 1953 elections, elections when BC was using IRV, not Block Voting, and in Toronto districts in 1908 and 1911 and in Winnipeg districts, 1914 and 1915.
10. Party Block Voting (also known as Group Ticket Voting)
The election of multiple members of the electoral college in each U.S. state (who go on to elect the president) are examples of MM districts.
But the members are not elected through Block Voting or any other system listed above.
Unusually for a MM district, each voter has just one vote. While the set-up would naturally produce mixed representation through SNTV, that is not the case.
The party with the highest popular vote (whether majority or not) is given the right to allocate all the EC seats belonging to the state.
Under Block Voting, mixed representation is not usually the case but it does happen occasionally. Under the Party Block method, mixed representation can never happen.
There are exceptions to the use of the Party Block method. In Maine and Nebraska, some of the EC seats are filled through plurality winners in portions of the states, with two seats given to the party that has the most votes across the whole state.
Like with the states using Party bloc vote or any first past the post elections or Block voting contests, in Maine and Nebraska there is no dependable proportionality of representation. Minority representation is produced only if the minority happens to be the largest group in one of the specified districts of the state. But on the other hand, sometimes a district of the state or the winner of the state-wide contest may be just a minority, with the majority of the voters (spread over one or more parties) given no representation at all.
=====================
CANADA'S HISTORICAL USE OF MMDs
Canada in its past use of MM districts, did not ever allocate seats based on party popular vote. Instead, true to the British tradition, votes were always cast - and applied - on a candidate-by-candidate basis.
Voting in MMDs in Canadian federal and provincial elections took these forms:
Block Voting - where voters could cast as many votes as there were seats to fill and where there was no minimum of votes needed to be cast by the voter. This is and was common in municipal elections, and was used in provincial (or colonial) elections in provinces from coast to coast at one time or another.
Block Voting where voters had to cast their full number of votes or none at all. This is known to have been used in the 1758 election in Nova Scotia, and possibly in one or more subsequent elections in that province.
Limited Voting - where voter cast two votes to elect three MLAs (Toronto - 1886, 1890)
STV - major cities in Alberta and Winnipeg from 1920s to 1950s
The Posts system - BC used this system to fill the seats in its MMDs in the 1952 and 1953 elections. In these elections, BC was not using Block Voting. Instead all members, whether in single-member districts or MMDs, were elected through Alternative Voting.
The Posts method was also used in connection to a First Past the Post system in the four Toronto districts (each electing two members) in the 1908 and 1911 provincial elections, and the three Winnipeg districts in the 1914 and 1915 provincial elections (each electing two members).
====================================
Historical MM districts in the U.S.
In the U.S., Block Voting was definitely the rule for election of legislators in MMDs, while Party Block Voting was - and is - used for election of most of the members of the Electoral College. As well, Cumulative Voting was used to elect legislators to some degree, especially in Illinois.
===========================================
MM District Votes in Pennsylvania and New York
Two clear examples of the use of MM districts in the U.S. are seen in the history of the states of Pennsylvania and New York.
Pennsylvania had multi-seat districts as well as single-member districts - 113 SMDs, 31 2-seat districts, five 3-seat districts and four 4-seat districts. This was before the Act of 1953 converted them into single-member districts.
1929 Philadelphia had 41 seats - 15 of these were in SMDs, 26 were in MMDs.
1953 Philadelphia had 39 seats - 24 of these were in SMDs, 15 were in MMDs.
Silva states "Pennsylvania’s Republican-controlled legislature appears to have favored the existence of more multi-member districts in Philadelphia when the Republican party was able to carry most or all of these districts. But, since the rise of Democratic supremacy in Philadelphia elections, the creation of more single-member districts allows the Republican party to salvage at least a few seats in some of the city’s carefully drawn districts." (so the way the districts are drawn - whether as SMDs or MMDs - was due to political considerations, not necessarily due to a question of fairness.)
In New York State, 1206 Assembly seats were filled during the state's first 76 years of existence as a state. 1086 seats (nine-tenths of them) were elected in multi-seat districts.
The situation was much the same for election of NY Senators.
From 1777 to 1846, all of NY's senatorial seats were filled in multi-seat districts.
Looking more generally, single-member districts (SMDs) became more common in the U.S. through 1800s.
By 1912, the Senates in 30 states were chosen wholly through SMDs, but a majority of lower-house members (House of Representatives) were still elected in multi-member districts.
In 1955, still 45 percent of the state lower-house members were elected in multi-member districts (excluding Alaska and Hawaii).
1960 three states elected all lower-house members through MM districts; 35 of the 50 states elected some of the lower-house members through MM districts and some through SMD.
Most MMDs in use had two seats but some had more than that - some had as much as five or more seats.
In the 1950s, Pennsylvania had 113 single-member, 31 2-seat, 5 3-seat and 4 4-seat districts.
1821 to 1846, New York was divided into 8 districts, each of which elected four senators. (p. 508)
Small single-member districts (SMDs) meant members were susceptible to local interests and were free-er from party discipline. (p. 506-507)
Large MMDs meant stronger party discipline, and that pushed members to concern themselves with wider more-general issues
For this reason, a move was made in NY to convert the 32 SMDs to 8 4-seat districts.
In the 1800s, it was said because SMDs are smaller, under SMDs not only were there fewer competent men to choose from but there was also greater opportunity for less able men to be elected. (p. 508)
================
CANADA
Canada’s electoral history is similar to that of the U.S. although it is shorter by about a hundred years - at least its post-Confederation history is. Multi-member districts were used in some place in Canada from Confederation (1867) until 1996, and were used even before 1867 as well.
In Canada, no multi-member districts are currently used in provincial or federal elections. But in the past, MM districts were used at both those levels - and in every province at one time or other.
The only province that appears not to have ever used a MM district since 1867 is Quebec.
In most of these cases, Block Voting was used, where each voter could cast as many votes as the number of seats to fill. But STV-PR was used in some cases; Limited Voting even made an appearance; and quite unusually, First Past The Post or Alternative Voting (another single-winner system) was used in MM districts through the device of holding separate contests for each seat in the district (the Posts system, #9 in the list above).
Block Voting in MM districts
The combination of multi-member districts, multiple voting and a plurality voting system often produces disproportional results. A party with the most votes in a multi-member district, even with only a third of the votes, can win all the seats.
But the use of multi-member seats is one way of responding to the growth of population in an electoral district without having to redraw electoral boundaries—just add one or more seats to those districts that show a large increase in the number of voters.
(This information was expressed in "Electoral Experimentation in BC and Canada," paper for the BC Citizens' Assembly, Weekend 2, Session 3 (available online)).
(Not mentioned is the fact that even when a district is given multiple members as an easy way to "re-district" seats, a voter in a multi-member district does not have to be given the ability to cast as many votes as the number of seats. With each voter casting only one vote in a multi-member district, the Single Non-Transferable Voting system (#5 in the list above) is created. SNTV can easily produce mixed roughly-proportional representation in the district.
=================
MMDs opened the door to mixed representation
Under Limited Voting and even under Block Voting, mixed representation resulted on occasion.
Sometimes it is due to none of the parties running candidates for all the seats. This may be caused by accident or dis-organization - the party just did not get the candidates lined up.
Sometimes it is due to one party running just one candidate and taking a seat, leaving the rest of the seats for other party or parties. It could be due to mis-understanding, not understanding that under Block Voting there is no vote splitting of a party vote (unless a party runs more candidates than the number of seats or votes that each voter can cast). Or a short party slate could be due to a working arrangement between parties. For example, where each voter has two votes. Farm-Labour co-operation indicates that each of those parties should run one candidate.
Sometimes mixed representation was due to two parties being fairly equal in popularity, at least as measured by votes received by the foremost candidate of each. In some of these cases, the candidate that was second-most in overall popularity belongs to a different party than the most popular one, so even in a two-member district two parties have one seat each. Where there are more seats than two, this can happen more often. But even in a five-seat district, one party sometimes takes all the seats under Block Voting. The difference in vote tallies between individual candidates of the leading party was not such as to allow a candidate of another party to get a seat.
Under Block Voting, voters can judge individual candidates differently (not just vote a party slate) with some individual candidates receiving more votes than other individual candidates of the same party, with voters plumping (voting less than their full complement of votes) or actually splitting their votes among two or more parties.
One reason for this splitting can be due to voter prizing local representation (representation of their particular corner of the district) over party allegiance.
The opportunity to judge individual candidates and to cross party lines with votes arises from the wide selection of parties and candidates offered to voters in MM districts. This wide selection is present in any MM district, to one degree or another, whether or not a fair voting system is used.
Just having multiple seats ensures that voters have a wide selection of candidates, even if the result in the end is a one-party sweep.
=========================
Proportional representation (STV)
Mixed representation is more common where STV is used than under Block Voting. In fact it was always the result in Canadian STV elections where there were more than three seats.
But the application of PR in Canada has been limited. We have never had a majority of the seats filled through PR in any province. When Alberta and Manitoba used PR-STV in their major cities, those cities had only a small fraction of the population of the respective province and only a small fraction of the province's seats. (However Ireland and Malta was using STV to elect all tis national members by the early 1920s.)
At the end of the use of PR (in the 1950s), the cities were under-represented (having fewer MLAs than their population warranted), but even if they had been represented at equal rate as rural areas, they would have elected only a fraction of the MLAs in the province.
At the most, 14 seats (out of 57) were filled through PR in any provincial election in Canadian history. But where used, STV proved itself well able to produce balanced mixed representation in each district where more than three seats were filled. Aside from electing many government party members, STV districts also elected far more than its portion of opposition members.
The most seats in a MM district using STV in Canadian history was 10. Winnipeg elected 10 MLAs through PR-STV in every election held from 1920 to 1945, and this representation was extremely fine-grained - sometimes having candidates of six different parties elected to represent the city's voters. (And it seems that Winnipeg's ten-member district was the first example of such a high District Magnitude (number of members per district). No district elected so many members in a single contest until New South Wales began to elect 21 members in single contests, in 1991.)
In Alberta, Edmonton and Calgary at the most elected seven and six MLAs respectively, and there too as many as three or four parties elected at least one MLA in a city's election contest.
General round-up of info on MMDS in Canada
in Canada there is no constitutional requirement of one or two seats districts in provincial elections nor in federal elections
federal ridings have used as many as two seats several times in our past.
Winnipeg used 10 seat district in prov elections
Edmonton 7
Calgary 6
Toronto three
BC used multi-seat districts of as much as 6-seat district
Vancouver City 1916-1933 elected six members
in most cases unfortunately block voting was used and one party usually took clean sweep.
But there was no constitutional requirement that Block voting be used and each of those four provinces listed -- MB, AB, ON, BC -- used other systems in MMDs sometime in their history.
as well other provinces (NS, NB, NL, SK) used districts of two seats for many years.
(NS in the first settler election in Canada (1758) elected 16 members in block voting.)
PEI too had two seat districts but there they went with FPTP in two separate contests in each district.
Quebec pre-Confederation used MMDs of as much as three members
example Montreal and QC elected three members in 1858.
I am not sure if article of provincehood restrains the province to single member district (maybe someone can investigate and see!) - I hope not.
City elections often are at-large (or MMDs) with as many as 12 members being elected at once, again usually with Block Voting although multi-winner STV was used in 20 cities in western Canada for a period of time
Two periods for Saskatoon!
Calgary used STV for one long period of 44 years, electing as many as 7 at one time, plus three other instances of electing two at a time.
so we are not restrained by an arbitrary one-seat max straitjacket.
and we can adopt MMDs of five or more seats (and/or top up conceivably) and fair voting if we choose.
=======================
End of Part 1
(Part 2 is a look at how MMDs were used in the Canadian provinces and territories, and in federal elections as well.)
______________________________
To read the next three parts of this multi-blog essay, here's the links:
===
also see
https://montopedia.wixsite.com/montopedia/post/discussion-of-voting-systems
=================================
++++++++++++++++++++++++++++
(housekeeping: url for this post
Comments