As raised in the previous edition, the Australian Capital Territory used the D’Hondt Method prior to changing to Hare-Clarke in 1992. While D’Hondt seems unlikely to be implemented in Australia on a large scale any time soon, Australia tending more towards the Single Transferable Vote (STV) models, it is the most popular method when implementing a party-list selection system. It can however be implemented in many different ways, with some using electorates, others implementing it alongside an electorate vote (this method being Mixed-Member Proportional). The system itself is more prominent in Europe, however an almost identical system, the Jefferson method, did arise in the United States in the 1790’s.
The Jefferson Method, named for future president Thomas Jefferson, was introduced following Washington’s Presidential Veto on the Hamilton Method (Yes THAT Hamilton). This method was then approved to be used to determine how many in the US House of Representatives each State would receive. Ultimately this method hasn’t been maintained for apportioning seats in the US. Rather humorously the Hamilton method was eventually used after the US briefly used the Webster Method, but again ultimately all we supplanted by the Huntington-Hill method. The US has used Huntington-Hill since the 1920’s. As interesting as all the minor differences between these methods are, they are decidedly not the topic we are here to discuss.
We start with who D’Hondt was, Victor D’Hondt was a Belgian lawyer and mathematician, living in the mid to late 1800’s. He laid out the method for the purposes we see it applied now for party-list elections, unlike Jefferson which was envisioned only for the purposes of apportioning seats in the House of Representatives. Seeming the method has been adopted in some form in more than twenty countries, including his home country of Belgium, one would have to consider the method to have been quite successful and persuasive.
Now that we’ve talked about a similar method and the bloke who laid the method out, let’s talk about how D’Hondt allocates seats. D’Hondt is described as a “Highest averages” method of selection. D’Hondt takes the total number of votes for a party, the divides that by the number of seats already won +1. Therefore a party receiving 10,000 votes and having already won three seats will receive the quotient 2,500, which is then used as that parties vote share to determine who receives the next seat. This is seen in the equation Quotient=total votesCurrent seats won+1 . If there are five members in a district using D’Hondt then we can create a grid for each party and their quotients after each member. The five highest quotients will be awarded a seat. Of the multiple proportional systems available, it is one of the less proportional models, however any proportional system is better than a majoritarian system. This is due to its slight favouring of major parties; some modifications of the system do correct this favouring of major parties. However the slightly weighting to major parties may be why it is one of the more prevalent proportional systems, as most parties act to protect their own interests of getting elected to government.
There are different applications of D’Hondt, an example coming from one of this author’s favourite European countries, Estonia. Estonia applies D’Hondt through twelve multi-member electorates. Each electorate has a different number of members elected from it, based on population, and has no single member electorates meaning it is not mixed-member proportional. Estonia also institutes two kinds of voting quotas, or thresholds, individual candidates may be elected if their vote share meets a simple quota, but parties must reach at least 5% nationwide to receive a seat.
The UK uses D’Hondt for some of its legislatures, and previously elected members to the European Parliament via this method (Northern Ireland however uses STV). It’s unlikely this will continue, as the UK is expected to leave the EU post Brexit. However a case that is likely to continue is Scotland’s use of the system, in which it is implemented under a mixed member proportional system. The Scottish Parliament is elected with a mix of members elected under first past the post in single member electorates, and multi-membered regions in which D’Hondt is used. This tends to create a parliament with roughly proportional allocation of seats for parties to their vote share.
Northern Ireland takes a rather novel approach to D’Hondt. While their parliament is elected via STV, Ministers are appointed via D’Hondt. This is done by applying D’Hondt to the total number of members elected to each party, with the parties selecting which of their members takes the department they have won. They then of course have the First Minister and Deputy First Minister, each holding equal power despite the names, who come from the largest Unionist Party and the largest Republican Party respectively. This is all due to the power sharing arrangements within the Good Friday Agreement, which do results in a rather interesting case of a forced coalitions, rather than the more commonly seen voluntary.
D’Hondt, despite being the most popular form of party-list voting, isn’t without its problems, namely the favouring of major parties. Of course this all depends on what your preference is when it comes to the makeup of government, coalitions of multiple parties? Or those dominated by a few major parties, propped up by minor parties. I personally prefer the slightly more minor party favouring Sainte-Laguë method, but that’s a different method for a different time.
Words by Mike Anderson
This article first appeared in print volume 88 edition 7 SOFT