Electoral Reform
(House of Representatives - Wolfendale System)
Overview
This objective is considered to be self-evident in the design of a democratic system of government:
That such a system is intended, to the best of its effect, to balance the intention of maximal representation with the practicality of communicating effectively with those who are represented, the method by which they are best able to exert their intent upon the system.
Consequently the challenge has always been one of balancing the capacity to make a system functional, and the desire to minimise the number of people who aren’t represented. This is key to maximising the inclusiveness of the political institution as representativeness suffers if the system is too complex to be usable and understandable by a broad cross-section of society, and functional usability suffers if the system is insufficiently representative.
Many attempts have been made to build systems which achieve this intent with varying success, and while some of these systems are indeed very good, or allow for significant degrees of representation, they each have specific deficiencies. It is not suggested that there would be no deficiencies in this new proposed system, but merely that this new system would have the least deficiencies so far.
It is important in politics, after all, not to let the ideal of the perfect get in the way of the good.
Comparative System Deficiencies
Various other systems have been designed which are considered “good”, and here their advantages and disdvantages are explored for context.
Proportional (Various)
Fundamentally Proportional representation is a nice idea. Everyone’s opinion gets heard, every party gets its place and the government always covers all views. However, in practice there are some significant deficiencies:
Reps are typically selected from lists created by each party, and this makes it impossible to ensure that particular regions are reliably represented by someone who shares their views. (The system lacks constituency.)
The above opens an easy avenue for corruption, as a party is necessarily responsible for the selection order of its own list. This leads to the possibility of wealthy people bribing their way higher onto the list to get into government, especially within big parties.
The tendency to select from those who are known, which is innately human nature, undermines the intention of a democracy to reach out and represent more people in more areas. This leads to the issue of an electoral ‘bubble’ where even though a party may be voted in from many corners of a nation, the views within that party may all be from as small an area as a single city. This centralisation of thought is counter to what makes a democracy representative.
This is a basic philosophy like first-past-the-post, and has as many issues.
Preferential (Aus)
A Preferential system is theoretically intended to address the deficiencies of the Proportional system while keeping its intent alive in the classical FPTP (first-past-the-post) electoral structure. You can vote for your favourite party but still make sure that the party you hate doesn’t get in just because you didn’t oppose them directly. Additionally, a party may look at the policies of those who preferenced them and adopt popular lessons from them. However, in practice there are still major deficiencies:
At the end of the process only one individual may represent a diverse set of people. Typically this means upwards of 40% will only have their opinions represented by proxy of the aligned party.
Even in victory, individuals who got their preferenced choice instead of their first will only be partially represented.
A de facto two-party system arises from the need to actually fill a seat. These two parties may in fact be coalitions of one major and a small number of other party members who are broadly aligned, but the outcome is the same. The minor members only maintain their position through regular alignment with their major partner, and a two-party result occurs in essence.
The parties that result are fundamentally at odds within themselves, as each must balance conflicting viewpoints to maintain their half of the system.
This is a more advanced philosophy, and the net impact of its deficiencies is lower than Proportional; however, rarely are more than one third of people properly represented.
Hare-Clark (Tas)
The Hare-Clark system attempts to address the inequities of the Preferential system, and it does so fairly effectively. It uses multi-member seats, allowing for proper electoral representation to theoretically reach around 70%, though in practice it falls a little short of this. It is fundamentally a good system. However, it is not without its deficiencies:
It is possible for one particular party to be screwed hard on very slim margins. While not likely, there’s no real balance other than statistical probability to prevent this.
Seated Proportional (NZ)
The Seated Proportional system is an alternative branch of development in contrast to Hare-Clark, attempting to address the tension between FPTP and Proportional. Seats are elected FPTP, and then Reps are added from party lists to fill up the % deficiencies in the overall representation. However, this too has its deficiencies:
The FPTP baseline encourages the situation where regions are represented by one specific party that always wins in those regions, leaving the opposition to be filled out from party lists. Consequently it suffers from both the deficiencies of FPTP and Proportional, but tends to keep them from overwhelming the process. The system as a whole holds up, though an individual party may suffer from the effects of those deficiencies.
This is a more robust system than Hare-Clark, but it has lower representativeness, and can definitely be improved.
Proposal
What follows below is a form of Mixed Multi-Member Seated Preferential-Proportional system which, for simplicity, I am simply calling The Wolfendale System.
Concept Outline
Stage 1: We begin with a system of multi-member electorates similar to Hare-Clark, except that they have 5 seats within them instead of 3. A Candidate must live in their electorate and we also modify the representative selection process in favour of communicability in the following ways:
Unlike in Hare-Clark, voters do not vote for individual candidates but instead for a party’s block of representatives. The order of these representatives must be public and announced at the declaration of the candidates for the ballot. (No reordering is permitted after declaration. If one candidate must drop out for any reason those below them move up and any replacement candidate must be added to the bottom of the list.)
This modification is performed for purposes of public clarity. In a bicameral system it may be acceptable for the house of review to have a counting method that is poorly understood by the public so long as general faith in it is maintained, but it is essential that the general population be clearly cognisant of the method by which the representative government is elected.
Pure Hare-Clark may be preferable at this level for systems where the number of total representatives and the number of potential representatives an individual may consider worthy are both within classical band society sizes (<30-50).
These are elected preferentially such that:
First all the first preferences for parties are counted and the base proportions of the primary vote recorded.
Any with over 20% primary vote are assigned 1 seat, and the remainder held aside.
If they have 40% they get 2 etc.
Using the primary vote only in the remainder, the party with the lowest proportion is eliminated from consideration and all votes reassigned to the next party that the voter has preferenced on those ballots.
The new vote totals are now assessed as a proportion of the original whole, and if a new party or the remainder of another now reaches 20% they are assigned a seat and any remainder held aside.
Repeat steps 3-4 until one of the following cases have been reached (in order of likelihood).
Case 1 - 4 out of 5 seats have been assigned:
The final seat must then select from the remainder and uneliminated parties. It will do so thusly:
If there is at least 1 uneliminated party that does not have a seat yet still in consideration, the elimination and preference allocation from the remaining parties beyond the first is performed.
The final unelected & uneliminated party shall be considered the ‘prospective’ party.
When all proportions left in consideration are either remainders or the prospective party, the results are assessed to see if the prospective would be successful on a FPTP basis. If so, the prospective party will be awarded the seat.
This is because it is considered more appropriate for a collection of aligned minority views to be marginally over-represented in a particular locale, than to be forced into representation by a more distant proxy.
This mechanism also acts as a counter to the degradation of representation and loss of public faith in the institution which occurs when established parties are made fundamentally more likely to take more seats than their proportional support. It does this by allowing for representation in government of those who are politically engaged but disenfranchised by the establishment and acts as a counter to political stressors which result in crisis.
If the prospective is not awarded the seat, the prospective votes then flow on to the remainders and the seat is awarded on the FPTP results of that final preference flow.
This ensures that minority protestations have the maximum chance to affect the outcome without compromising communicability of the system.
Case 2 - 3 parties have created a deadlock where the fourth seat cannot be assigned with 20%:
The largest 2 remainders will be assigned the final 2 seats.
Stage 2: Now that all seats have been assigned we proceed to the next stage of ensuring maximal representation with our assumed baseline of communication difficulty. It should be noted that this step may not be required in every election and, depending on the political culture, may in fact be required only rarely. In order to process this stage:
All parties must have a list of potential ‘proportion candidates’. This list may include individuals who are also being considered in the ballot for a particular electorate, but must include at least 1 non-balloted individual for every 3 who are included on an electoral ballot. The total number of listed individuals must also be at least equal to the number of electorates the party chose to run on the ballot for.
We must now look to the national scale, producing the base proportions of the primary vote for the entire nation and recording them.
We must also look at the resulting party makeup of the representative government and determine what proportion of the government they represent.
Since it is likely in any open democracy that a number of minor parties will have failed to achieve representation, those that have successfully achieved representation should have a slightly higher proportional representation in government than their proportion of the primary vote when taken as an average. The following steps are then taken:
Determine if any minor parties that failed to achieve representation possessed in excess of 4% of the primary vote total.
If yes, assign representatives to them from their list until they achieve a proportion within the government at least equal to their primary vote proportion.
Determine which, if any, parties are under-represented in the new proportion of government by at least 1 whole representative worth of the primary voting proportion.
If so assign them a representative from their list.
Repeat step 2 until no party is under-represented >1 whole representative.
Caveat: In small governments this may lead to a nearly endless loop. If every elected party each receives one rep as a result of this process before any party receives 2, negate step 2 entirely.
As a shortcut it may be possible to use the following rule of thumb during electoral coverage to predict the necessity of step 2 above:
Determine if the proportion of vote which went to minor parties that have still failed to achieve representation is in excess of 2% or 2 individuals worth of proportion in government (whichever is lower).
If not, take no further action. (note: this means 2% for governments with <100 reps, and 2 individuals for governments with >100 reps.)
Outcomes
The net effect is to ensure that there is no incentive for a party to adopt a “broad church” of competing views. It will always be more efficient to represent a particular viewpoint and agenda to the voters and then negotiate the compromises with the most favourable options in coalition government. The idea of a majority government essentially becomes a thing of the past, which is beneficial for peaceful governance as it ensures that no agenda is implemented without negotiating opposition unless the public has overwhelmingly aligned behind it. It also encourages faith in the institution of government, because it incentivises voters to seek a party that they actually support instead of choosing between a “lesser of two evils” when their personal views do not align with the division fostered in any actual or de facto two-party system.
Further the overall representativeness approaches 100% similar to standard proportional (subject to whatever cutoff limit is applied and how much support is afforded to truly fringe candidates). No representative system actually reaches it but it shares the top spot with Proportional in principle, and additionally the local representativeness increases from the 50-60% typical of a Preferential system to at least 87% while preventing the erosion of local representation possible in a straight Proportional system. These numbers assume preference flows as representative.
Direct representation (treating preference flows as not-representative) is correspondingly improved under comparable circumstances but the maximum lack of direct representation is an unbounded function for all systems that doesn’t bear modelling. (An outcome with 26 parties who all received just under 4% of the vote is simply not realistic and modelling beyond that is farcical.)
The maximum % of voters whose views may not be locally represented by preference flows in this system is limited to 14.9%, in contrast with 49.9% for seated Preferential.
Worked Examples
Detailed examples of how this could work are available at the links below (opens in a new tab):
Additional Content
Extra information is available in general and specific contexts.
An original google document version of this entire reform can also be found here: [Link]
(Some minor textual details may vary, the website published version should be considered authoritative in the case of any confusion.)