Simulation Summary

Design name:Single Mbr

Single member plurality for Canada's elections between 1874-01-22 and 1878-09-17.

Election id:4
Election date:1878-09-17
Transformations:None

The identity transformation. Doesn't actually transform the design.

Ballot Generator:SMP Ballot Gen

Generate ballots for a Single Member Plurality election (e.g. First Past the Post). Each voter is allowed a single preference. The candidate with the most votes wins.

Vote Counting Algorithm:First-Past-The-PostTO DO

Riding Results

Canada-Wide Stats

Total voters: 543,950

PartyParty NameIdeology Code#1stPref1#Votes2#ExhVotes#Elected%1stPref%Vote%Elected%Diff
LibLiberalLib174,523174,52306132.1%32.1%31.3%-0.8%
ConConservativeCon143,165143,16508126.3%26.3%41.5%15.2%
UnknownUnknownOth111,743111,7430020.5%20.5%0.0%-20.5%
PCProgressive Conservative PartyCon88,40588,40504516.3%16.3%23.1%6.8%
IndIndependentOth14,78314,783042.7%2.7%2.1%-0.7%
IndLibIndependent LiberalLib7,0627,062011.3%1.3%0.5%-0.8%
IndConIndependent ConservativeCon3,4313,431020.6%0.6%1.0%0.4%
NationalistConNationalist ConservativeOth838838010.2%0.2%0.5%0.4%
Total543,950543,9500195100.0%100.0%100.0%

Disproportionality Indices3

Gallagher Index:18.74
Weighted Gallagher Index:419.55
Loosemore-Hanby Index:22.80
Lijphart Index:20.54

Legislative Power

FptpLPS

A voter's base legislative power score (LPS) is 1 if the candidate they voted for was electd MP and 0 if the candidate was not elected.

The base LPS is adjusted by V/v where V is total number of votes cast in their district and v is the number of votes the elected candidate received. For example, if there were 100,000 votes cast in the district and the winner received 50,000 then the LPS is doubled (100,000/50,000 = 2) for those who voted for the elected candidate. That extra legislative power comes at the expense of those who did not vote for the elected candidate and have 0 legislative power.

The base LPS is further adjusted according to the malapportionment or deviation in riding size. The malapportionment factor is W = A/s where A is the average size (number of voters) of all electoral districts in the province and s is the size (number of voters) in the voter's district. For example, if the average district has 100,000 voters but the voter's district has only 80,000 then the voter has 100,000/80,000 = 1.25 times the legislative power of someone in an average sized district.

The voter's Legislative Power Score is then V/v * W if they voted for the elected candidate and 0 if they did not.

Top 6 Preferences w/ declining wt

A voter's legislative power is derived from their top 6 preferences. The first preference has a weight of 1.0; each subsequent preference has a weight that is 1/j less where j is the minimum of 6 and the number of preferences on the ballot. Any preference that isn't elected receives a weight of 0.

The weight contributed by each candidate is adjusted by comparing the number of votes they received to the provincial average. If the average MP in the province received 50,000 votes but this candidate won with only 40,000 votes, then the people who voted for this MP have a greater than average legislative power (50,000/40,000 = 1.25).

Finally, all Legislative Power scores are normalized by province so the average is 1.

In an FPTP election there is only one preference and thus voters who voted for the elected MP have an LPS of 1 and 0 otherwise.

Riding Size Inequity

Some of the inequities in voting can be attributed to disparities in riding sizes. Voters in larger than average ridings will have less legislative power than voters in smaller than average ridings.

This utility function gives a weight of 1 to every voter. The resulting legislative power index varies only according to the number of voters in the riding relative to the average riding in the province.

Same party ranked equal

In a party leader driven culture, as Canada has, the local candidate often is less important than the party they are affiliated with. This utility function gives the same weight to next prefences provided they belong to the same party. If a party switch is involved the score decreases by a parameterized amount (0.5 in this simulation).

For example, suppose the voter votes for A, B, C, D, and E. Candidates A, B, C, and D all belong to the same party while E belongs to a different party. Assume that B, C, and E are elected.

Candidate A starts with a weight of 1 (but wasn't elected and so contributes nothing). B is from the same party as A and so has the same weight. B was elected and thus contributes 1.0. C was elected and is from the same party as B and thus also contributes 1.0. E involves a party switch and thus contributes 1-0.5=0.5 to the score.

Legislative Power Score Distribution

Legislative Power Score Distribution

Legislative Power Score Distribution

Legislative Power Score Distribution

Legislative Power Index

Gini coefficient = 57.48

Legislative Power Index

Gini coefficient = 57.48

Legislative Power Index

Gini coefficient = 22.50

Legislative Power Index

Gini coefficient = 57.48

Regional Stats

Maritimes

Total voters: 128,639

PartyParty NameIdeology Code#1stPref#Votes#ExhVotes#Elected%1stPref%Vote%Elected%Diff
LibLiberalLib46,49746,49701636.1%36.1%39.0%2.9%
PCProgressive Conservative PartyCon25,10225,10201119.5%19.5%26.8%7.3%
ConConservativeCon25,04725,04701219.5%19.5%29.3%9.8%
Total96,64696,64603975.1%75.1%95.1%

Disproportionality Indices

Gallagher Index:14.32
Loosemore-Hanby Index:19.99
Lijphart Index:15.42

Quebec

Total voters: 141,119

PartyParty NameIdeology Code#1stPref#Votes#ExhVotes#Elected%1stPref%Vote%Elected%Diff
ConConservativeCon48,55948,55903234.4%34.4%52.5%18.0%
LibLiberalLib29,88229,88201621.2%21.2%26.2%5.1%
PCProgressive Conservative PartyCon18,63818,63801013.2%13.2%16.4%3.2%
Total97,07997,07905868.8%68.8%95.1%

Disproportionality Indices

Gallagher Index:23.36
Loosemore-Hanby Index:27.40
Lijphart Index:27.00

Ontario

Total voters: 269,773

PartyParty NameIdeology Code#1stPref#Votes#ExhVotes#Elected%1stPref%Vote%Elected%Diff
LibLiberalLib97,27097,27002736.1%36.1%30.7%-5.4%
ConConservativeCon69,01369,01303725.6%25.6%42.0%16.5%
PCProgressive Conservative PartyCon43,50243,50202316.1%16.1%26.1%10.0%
Total209,785209,78508777.8%77.8%98.9%

Disproportionality Indices

Gallagher Index:19.83
Loosemore-Hanby Index:26.47
Lijphart Index:19.62

Prairies

Total voters: 1,101

PartyParty NameIdeology Code#1stPref#Votes#ExhVotes#Elected%1stPref%Vote%Elected%Diff
ConConservativeCon5465460049.6%49.6%0.0%-49.6%
Total5465460049.6%49.6%0.0%

Disproportionality Indices

Gallagher Index:49.59
Loosemore-Hanby Index:49.59
Lijphart Index:49.59

BC

Total voters: 3,318

PartyParty NameIdeology Code#1stPref#Votes#ExhVotes#Elected%1stPref%Vote%Elected%Diff
PCProgressive Conservative PartyCon1,1631,1630135.1%35.1%25.0%-10.1%
LibLiberalLib8748740226.3%26.3%50.0%23.7%
Total2,0372,0370361.4%61.4%75.0%

Disproportionality Indices

Gallagher Index:27.95
Loosemore-Hanby Index:36.97
Lijphart Index:26.91

Ballot Metrics

Number of candidate permutations:5408
Average number of voters per permutation:1,333.21
Number of ballots with x preferences:
PrefsBallots
1543,950
Number of exhausted ballots:0
  1. First preference votes are the votes actually recorded in the election. The sum of this columnis equal to the total number of voters.

  2. #Votes lists the number of votes attributed to a candidate after all transfers have taken place. Exhausted ballots are not included.

  3. Disproportionality indicies provide a measure of disproportionality of the electoral system. For more details, see this summary.

  4. One issue with the Gallagher Index is that one region of the country can cancel out another. For example, the index for prairie provinces is 14.55 due to the over-representation of the Conservatives. In other regions of the country the over-representation of the Liberals contributes to the index. When the country as a whole is considered, these moderate each other making the overall proportionality look better than it is. The weighted Gallagher Index addresses this by summing the (simple) Gallagher Index for each region, weighted by the number of voters in that region.

  5. Two voters have the same candidate permutation if they voted for the same candidates in the same preference order. The number of voters per permutation depends on the number of preferences allowed, the number of candidates running, and the minVotingBlock parameter. The last factor is an efficiency consideration for simulations with large multi-member ridings.