Course:CPSC312-2016-Project2-RankedBallot
Ranked Ballot
Authors: Gabriel Henderson, Dominic Kuang
What is the problem?
We would like to see how the results of the 42nd Canadian general election might have differed under another voting mechanism.
What is the something extra?
Many critics of the First Past the Post (FPTP) system argue that it is unfair under the basis that a candidate can be elected without majority support. Indeed, it is theoretically possible for a party to attain complete rule (338/338 seats) with exactly 33.4% support in a three-party system under FPTP. Popular alternatives to the FPTP system are proportional representation (PR) and Preferential voting (PV), which uses ranked ballots.
The results under the PR system are immediate -- with the Liberal Party gaining 39.47% of the popular vote, they would form a minority government with 134 out of 388 seats. However, estimating the results of the election under PV is not entirely straightforward. We will need to make assumptions on the secondary, tertiary, etc. choices of voters given only their primary preference. In our simulation, we will need to assign an "ideological position" to each political party and make the assumption that voters will prefer parties that are as close to their chosen ideology as possible. For example, the second choice of a conservative voter will most likely be Liberal and not NDP, whereas the second choice of an NDP voter would most likely be Liberal, and not Conservative.
The ranked ballot system works as follows: any candidate with greater than 50% of the popular vote is elected; otherwise, the candidate with the least amount of votes is removed and all their votes are reallocated to the next most preferable choice. This process continues until a candidate receives a majority of the votes. This can have a massive effect on the results of the election, particularly in key "swing" ridings. As an example, consider the 2015 general election results for the riding Richmond-Centre:
| Candidate | Affiliation | Votes |
|---|---|---|
| Alice Wong ✅ | Conservative | 17,693 |
| Lawrence Woo | Liberal | 17,230 |
| Jack Trovato | New Democratic | 4,613 |
| Vincent Chiu | Green | 1,152 |
Under the PV electoral system, one possible outcome would be:
| Candidate | Affiliation | Votes |
|---|---|---|
| Alice Wong | Conservative | 17,693 |
| Lawrence Woo | Liberal | 17,230 |
| Jack Trovato | New Democratic | 4,613 |
| Vincent Chiu | Green | 1,152 |
| Candidate | Affiliation | Votes |
|---|---|---|
| Alice Wong | Conservative | 17,693 |
| Lawrence Woo | Liberal | 17,230 |
| Jack Trovato | New Democratic | 5,765 |
| Candidate | Affiliation | Votes |
|---|---|---|
| Lawrence Woo ✅ | Liberal | 22,995 |
| Alice Wong | Conservative | 17,693 |
With PV, the "vote splitting" phenomena is mostly eliminated and tactical voting is disincentivized. Thus, moving to a PV-based electoral system can dramatically affect the outcome of an election. We are curious to see how the results of the 2015 Canadian general election might have differed under a PV model. In particular, we want to see whether or not the Liberal Party would still attain majority rule.
We also wish to investigate what happens when citizens use a technique known as "Tactical Voting" in the actual FPTP system. For this, we will use the guidelines given at http://www.strategicvoting.ca/.
What did we learn from doing this?
We learned that functional programming is extremely suitable for this sort of task. We had no problems using the Haskell language to simulate the results of Canadian elections under different voting systems. The biggest obstacles we faced were parsing the data using the format given by Elections Canada and learning to use the Chart API to produce the pie charts. In case you are curious, our findings are below. Note that progressive tactical is simply a FPTP strategy and not an electoral system like the other three.
FPTP (Actual)
Liberal Majority
Ranked Ballot
Liberal Majority
Proportional Rep
Liberal Minority
Progressive Tactical
Liberal Majority
We found that proportional representation was the only electoral system that produced a Gallagher Index of under 5.0, as recommended by the parliamentary electoral reform committee.
News from December 1, 2016: