Preface
There exist a number of well-known and widely employed electoral systems. The most defining principle of plurality systems is individual competition between candidates. Such systems can be broken down into two different types: the relative- and absolute majority vote. The latter of these requires more than half of all votes cast in any given constituency in order for a mandate to be won. In proportional voting systems, citizens cast their ballots on party-lists instead, while mixed systems simultaneously encompass the aforementioned characteristics.[1]
Let us take a look at more specific categories now, as they are found in both theory and practice.
Relative majority possesses centuries-old roots in Britain, although naturally, as a modern voting system, it took shape only upon the establishment of civil society. It is a highly disproportional system – even more so before 1832 –, with the geographic size and population count of districts both presenting great fluctuation vis-à-vis each other. In addition, representatives were elected by open voting.
Absolute majority electoral systems were created during the second half of the nineteenth century to remedy the shortcomings of the relative majority vote. (Refer to the chapter on voting systems, including the issues of wasted vote and ‘Condorcet winners’; a method named after Nicolas de Condorcet, French mathematician who lived between 1743 and 1794.) According to Borda (1733-1799),[2] objectivity and justice would be best served if voters were allowed to rank all of a district’s candidates in their order of preference, with such an order determining the winner after the appropriate calculations. Given the sheer number of candidates, this method, itself never tested at parliamentary elections, might appear difficult to implement; however, it lends great importance to secondary votes, potentially bringing about a ‘just’ form of popular representation based on compromise. The alternative vote[3] particular to Australia since 1919 stands on a similar principal footing, also requiring the ranking of candidates.[4] If the primary preferences fail to yield an absolute winner, the secondary preferences on the ballots cast for the least popular candidate are added to the primary votes, repeated, if necessary, by tertiary and lesser preferences until such time that the first candidate passes the fifty percent mark.[5] (This method bears a certain resemblance to the single transferable vote, distinguished, however, by its lack of proportionality, since alternative voting only awards a single mandate per district.)
It must be highlighted that in theory, alternative voting is more just than absolute majority methods, because the latter only take into account secondary preferences on the two (or more) candidates already having passed into the second round via primary preference votes. By contrast, alternative voting allows for the weighing of all preferences, including tertiary or even more distant expressions of support. (Ad absurdum, in an absolute-simple majority system, similar to the one previously employed in Hungary, a less rejected candidate eventually claiming victory through secondary and tertiary votes would not even have made it to the second round with just fifteen percent of primary votes!)
What this would mean is that a reserved and less maverick candidate hated by fewer voters than the frontrunners would not even get to the second round in an absolute majority system, even though such a candidate, given all the secondary and tertiary votes received, would be the overall most popular choice.[6] This ties in well with the argument that a government is not only the government of the voters actively electing it, but of all citizens of the country. Therefore, other than its popularity, its unpopularity is also a relevant factor.
Types of absolute majority voting
The aforementioned absolute majority voting systems can be placed into two categories. In one of these, the first and the second place candidate from the first round are entitled to run in the second round, requiring fifty percent of votes plus one to gain a mandate. This method is known as the absolute-absolute majority system, restricted, for the most part, to direct presidential elections. Under the rules of absolute-simple majority, acquiring a simple majority in the second round is sufficient, naturally allowing more than two candidates to run, although deals struck by the political parties involved often cause such additions to drop out, reducing the actual number of candidates to just the two most popular. (The Hungarian voting system previously in effect followed this method.[7] The procedure particular to France is placed under scrutiny in the final chapter dealing with the French system of government and electoral system.)
Experience dictates that due to voter abstention, candidates having attained at least forty percent of the vote during the first round will tend to retain their advantage in the second, while those in the lead with thirty percent or less stand on a more shaky ground: the chief factor in whether or not they will also emerge victorious in the second round is secondary preferences.[8]
The nuances of the relative majority system have already been discussed, the implementation of which would force existing political parties to join forces and run joint candidates together, thereby making factional alliances more transparent and evident to the voter. This factor, given the system’s single-round, party-list nature, would be more equitable to citizens, granting them a clear picture of just who else they are about to support alongside their own favourite parties. Because of this, an argument could be made for the implementation of relative majority single-member districts in Hungary.[9]
Current legislation also allows for the running of joint candidates.[10]
On a related note, the second round also allows voters to pass their judgements on any coalitions that might have been formed, taking into account party declarations and communications. This is their only way of doing so, given the indirect nature of the prime minister’s selection – with the exception of the Israeli system.
Block voting
As evident as it may sound, it is nevertheless important to point out that one district equals one mandate (therefore, the district magnitude is one). On the other hand, this is by no means a necessity in plurality voting; the classical English system typically allocated two mandates per district, allowing citizens to vote for two candidates.[11] Even as far as the mid-twentieth century, certain countries (such as the UK, India, Canada and the USA) featured a few (!) multiple mandate constituencies, while the rest of the country could only send one representative to parliament; or congress. This solution – even though certain sources disagree – is a form of plurality voting, rather than constituting a proportional model.
The same can be said of the singe-list system, formerly known as the small-list system, which is currently employed in Hungary during municipal elections, and which also featured in the draft of a number of French electoral legislation proposals. Also called block voting, the essence of this model lies in the relative majoritarian distribution of mandates. Each citizen has as many votes as there are mandates to acquire in any given district. It was first employed in the Australian Senate from 1901 to 1948.
The cumulative single-list system differs from the single-list system in that it allows voters to cumulate all their votes for a single candidate. (The cumulative party-list system, detailed later, can itself be differentiated from its single variant by the fact that its distribution of mandates is based upon the ratio of votes cast on the various party lists. Therefore, cumulation only enters the fray in determining the order of candidates on any concrete list. It is considered a semi-proportional model.)[12] It remains a system predominantly featured in the Anglosphere: its first occurrance was in the Illinois House of Representatives in the United States between 1870 and 1980. From the late nineteenth century, both the US and the UK have been using this method in various school board and corporate elections. In the latter case, individual US states (currently seven) would sometimes enforce it as mandatory.[13]
Preferential systems
According to the time-honoured view, in a proportional/party-list system, citizens are only entitled to vote on party-created lists, with no ability to influence who their vote personally benefits. Is this truly the case, even today?
Proportional electoral systems first ‘came into fashion’ in the early twentieth century: 1906 in Finland (then part of the Russian Empire), 1909 in Sweden, 1917 in the Netherlands, 1919 in Switzerland, Italy and Norway alike. In Denmark, Germany and Austria, 1920 marked the year of conversion from a majority model to a list-based, proportional one. Practically in the cradle, these methods began to branch out and develop more nuanced alternatives, such as the various preferential systems, their true advent made manifest in the second half of the twentieth century.
The question of vote structuring determines how many votes a single citizen is allowed to cast, and how. There exist single-vote and multiple-vote systems. The latter can be broken down further into preferential models, models that enable the cumulation of votes and models that allow for vote splitting. In the first case, voters are entitled to make modifications on the party-list, swaying them towards their preferred candidate. In the second case, it is possible to reinforce the position of one candidate by allocating them extra votes, further increasing their chance to gain a mandate. Lastly, the third model opens up the aforementioned possibilities in the case of lists as well.
In terms of the right of voters to bring about such modifications, the various preferential systems paint a colourful picture worldwide. Only one preferred candidate can be designated in Austria, while the same number is four in Italy. It is possible to split votes in Switzerland and Luxemburg, giving each voter as many votes as there are mandates to gain in a district, which can be cast for different parties. By now, in most Western European countries, allowing citizens to influence party list compositions has become common practice. In Belgium, the influence of preferential votes determines almost fifty percent of the order of candidates on party lists. The German system remains almost the only one disallowing this practice, its single districts purporting to substitute for the lack of choice between individual persons.
Since the fall of Communism, a number of Central-Eastern European countries also implemented ways to create more flexible party lists. In Poland, citizens must select a name from a regional party list, an act with which they also cast their vote for that candidate’s party itself. Similarly, the Slovene model divides electoral districts into single-member constituencies, in which single candidates are also the nominees of their respective party lists. Again, voting for a candidate is also voting for a party list.[14] Mandate distribution is based on the proportion of ballots cast for each party list, giving successfully acquired mandates to candidates having won the greatest number of votes on their respective lists. Czechia and Slovakia also allow their voters to influence the prearranged order of party lists. Each citizen may designate four candidates they have especial preference for. Such preferential votes will result in the acquisition of a mandate in the event that at least ten percent of voters in the district cast such a vote, and the candidate in question managed to seise at least ten percent of all the preferential votes his party received. Austria operates similarly, except it requires an amount of preferential votes equal to at least half of what is necessary for a single district mandate, or one-sixth their total number cast in their party’s favour.
Finland’s ‘flexible list’ enables party list rankings to be molded entirely by voter will; on each ballot, the candidates of parties are merely displayed in alphabetical order.
(European practice indicates that party list rankings provided by the parties themselves will seldom be altered by preferential votes. Still, it is a matter of principle to allow a modicum of choice between the individuals running on the same platform, preventing the ‘secondary recruitment’ within the parties to be the sole decider of who gets into parliament. In our view, such methods unite the best of both worlds from single-district and party-list voting systems, aiming for an equitable solution in mandate distribution, acceptable both in terms of directness and proportionality.)[15]
Premium list systems
In Hungary, Act XXII of 1947 (along with creating discriminative rules for nominations and disqualifications) introduced the premium system. In it, attaining 60% of the vote automatically awarded 80% of the seats in parliament, while gaining 75% would land that faction all (!) mandates. The quotient was raised to 14,000 and national lists to 60 candidates. Another early example for majority bonuses is found in Fascist Italy from 1924 on, retained but moderated in its effects by the 1946 republic.
Premium systems similar to the above are still in use today. For instance, the electoral system of Paraguay allocates two-thirds of all mandates by default to the party standing victorious in the elections, distributing the rest proportionally. In South Korea, 244 representatives are directly elected, 38 by the winners themselves and 37 by the other parties. The winner in Malta is granted an additional four premium mandates. (It is a popular misconception in Hungary that under the rules of the current system, the party carrying an election is given the mandates of its rivals that failed to meet the minimum electoral treshold.
Limited voting (LV) can be considered a unique,[16] semi-proportional voting system. Initially a British invention to be used in a number of individual constituencies between 1867 and 1885, it later appeared in the Kingdom of Italy, US-occupied Japan and post-Franco, democratic Spain. Similarly to block voting systems, this method essentially enables participants to select multiple single candidates. However, citizens have fewer votes than there are mandates distributed in any single district. (For example, there used to be two single constituency votes allocated to Japanese citizens, with four to ten attainable mandates.) Why would this method be more proportionate than block voting, itself widely classified as a majority system? The answer is made apparent through a simple deduction. In the latter system, whenever a party or a faction enjoys overwhelming popularity in any given district, all its supporters tend to throw their whole weight (votes) behind this group. This is made especially likely in today’s party-centred political atmosphere. In such a case, a candidate with 25% popularity will always defeat another with 20%. By contrast, the LV system distributes more mandates than the amount up for grabs by the most popular candidates, thereby allowing some form of representation for their opposition as well. (Under a two-vote model, this tends to lead to the first and second place belonging to Party A, with the third and fourth going to Party B. In such systems, parties usually run as many candidates as there are votes, otherwise, said candidates would also compete against each other.) The SNTV system, also formerly employed in Japan, is considered a subtype of LV, giving each voter one vote in a district with multiple mandates to win. However, it is necessary not to mesh them together, seeing as multiple votes enable the option of vote splitting, thereby constituting a factor of primary importance when classifying voting systems. In this sense, one- and two-vote systems can be sharply differentiated. Therefore, there is a far greater distance between one and two than between two and three. (Japan has since converted to a mixed voting system.)
SNTV carries a large risk factor for parties, especially major ones, necessitating in-depth surveys to determine their electoral support. If they run fewer candidates than it would be justified by their popularity, then a great many votes cast for them will end up wasted, not contributing towards the acquisition of mandates. If, on the other hand, they stand too many candidates, they run the risk of having the vote splintered among them, decreasing each other’s chances and potentially allowing outsiders to have the last laugh and snatch mandates, leading to results worse than the party’s electoral support would suggest.[17]
The vote-transfer system
The virtues of proportionality and choice between individuals are organically conjoined in the vote-transfer system, originally conceived by British mathematician and schoolmaster Thomas Wright Hill in 1819 and later expanded upon by political scientist Thomas Hare in 1857.[18] It saw its first use during the 1856 Danish Rigsdag elections, adopted for the second chamber within a year. The vote-transfer system is currently employed in Ireland and Malta[19]. In multiple mandate electoral districts, citizens are expected to vote for and rank as many candidates as there are potential mandates in the district.
- After the ballots have been cast, the Droop quota[20] serving as the basis for mandate distribution is calculated. It is arrived at by dividing all valid votes cast in the constituency by the total number of allocated mandates plus one, and then adding one again to the quotient thus gained. Any candidate whose number of primary votes reach the Droop formula is elected.
- If all mandates cannot be distributed using the above method (which, given trends of vote division, is rather likely), a new calculation is necessary. The votes of candidates already elected are divided again, based on the secondary preferences indicated on each ballot, but the votes of secondary candidates are only increased by a single rate. This rate is:
The number of votes of an elected candidate – Droop quota
The number of votes of an elected candidate
Whoever manages to reach the quota by adding this number to the number of his votes acquires a mandate.
- If all seats allocated to this district are still not filled, the votes of the candidate receiving the least amount of primary votes are annulled, distributing the candidate’s ballots based on the secondary preference written on each, between the candidates still lacking a mandate. Those reaching the Droop quota are elected.
- If even after this step, there remain unclaimed mandates, the ballots of those who gained a seat according to points 2 and 3 are redistributed along secondary and tertiary preferences following the above method. Again, candidates arriving at the Droop quota gain a mandate.
- In case the mandates are still not fully distributed, the entire process continues as long as necessary.
In order to guarantee proportionality, multiple layers of voter preference must be taken into account. In spite of it being a single-winner vote, party candidates – in fact, candidates in general – tend to inform the electorate about whose backing they would find desirable after their own. (It is expedient for parties to weigh their own popularity and run only an optimal number of candidates in order not to affect each other’s chances negatively.) In Ireland, for example, the district magnitude is 3.7, therefore, their lower-preferences candidates matter less. Placing less burden on voters, lower magnitudes are certainly more advisable under the aegis of such systems.
Bibliography
Dezső Márta: Képviselet és választás a parlamenti jogban, Közgazdasági és Jogi Könyvkiadó, Budapest, 1998.
Fábián György – Kovács László Imre: Voksok és mandátumok, Villányi úti könyvek, Budapest, 1998.
Humphreys, John H.: Proportional Representation, A Study in Methods of Election, Methuen & Co. Ltd., London, 1911.
Körösényi András: A magyar politikai rendszer, Osiris, Budapest 1998.
Lijphart, Arend: Választási rendszerek: típusok, formák, fejlődésvonalak, in Fábián György – Kovács László Imre (szerk): Választási rendszerek, Osiris-Láthatatlan Kollégium, Budapest, 1997.
References
- Dezső Márta: Képviselet és választás a parlamenti jogban, Közgazdasági és Jogi Könyvkiadó, Budapest, 1998, 35. (Hungarian) ↑
- Dezső: Képviselet és választás, 41. ↑
- Also known, particularly in the United States, as instant-runoff voting, developed by US architect William Robert Ware around 1870. ↑
- Interestingly, the matter of prioritisation never truly emerged in party-list voting systems. The author perceives its significance in two ways. First, the secondary preference of voters could come into play in case a party would prove unable to pass the parliamentary threshhold by primary votes alone. Second, if, by one of the existing vote distribution methods (the highest average method in particular), two parties would enter a tie, preference could be given to the party with more secondary votes. ↑
- In case no one managed to acquire the absolute majority of primary preferences, the process continues with the candidate possessing the least amount of ballots dropping out, their votes collected into bundles, this time with the secondary preferences in mind, adding these bundles to the ballots of remaining candidates. In the event that an absolute majority winner is still not found, the next-to-last candidate’s vote-bundles are similarly distributed among remaining candidates, and so on and so forth until somebody reaches the absolute majority threshhold. ↑
- (A similar mathemathical dilemma could arise the following way: while even in the second round, the absolute majority method only gives relevance to secondary preferences, it is entirely possible that a candidate would become a representative under the alternative voting system only through tertiary or lower votes – in other words, purely out of a lack of rejection by their opponents.) ↑
- Compared to the previous Hungarian system, the current French model presents a peculiar case. In single-member constituencies, if the first round yields no absolute winner, the second round is played out between the winner and the runner-up of the first round, as well as anyone else with at least 12.5% of votes. (This model could be deemed more just than the Hungarian system, because a third candidate — and any additional ones — only „interrupt” the race if they possess substantial support and backing from the population; instead of merely gaining a bargaining position against the ’big shots’. By principle, it is an absolute-absolute majority system, only becoming absolute-simple in exceptional cases.) ↑
- Körösényi András: A magyar politikai rendszer, Osiris, Budapest 1998, 153–154. (Hungarian) ↑
- Naturally, the fact that the single-round relative majority system produces disproportionate results remains an argument against it. Another of its drawbacks is that even though an election carries its consequences over the next four years, its outcome is determined in but a single day; comparatively a very short time period. Two rounds mean at least two days, thereby decreasing the „random” factor so heavily present in single-round voting. For example, weather may discourage certain voter demographics from casting their ballots and a single piece of sensationalist news may captivate and manipulate voters for just enough time. (Fair weather often incentivises farmers to work their land, families to go hiking, etc.) ↑
- Unlike in the case of joint- and affiliated lists, the current Hungarian electoral statute gives no allusion as to where fractional votes cast on joint candidates are allocated. ↑
- In 1967, Mauritius converted from a relative majority system with forty single mandates to a relative majority system with twenty triple mandates. Later, they also implemented ten balancing mandates. The Seychelles employed seven double mandate- and one single mandate districts. ↑
- Fábián György – Kovács László Imre: Voksok és mandátumok, Villányi úti könyvek, Budapest, 1998, 43. (Hungarian) ↑
- See: ’FairVote – State Regulations on Cumulative Voting for Corporate Boards’. https://archive.fairvote.org/?page=758 ↑
- In our view, given that it disallows selection between different candidates of any given party, this model isn’t a preferential-proportional system. Instead, it shows more similarity with the kind of personalised PR systems currently in use in Germany. ↑
- Further reading on the various methods of mandate distribution: Arend Lijphart: Választási rendszerek: típusok, formák, fejlődésvonalak, in Fábián György – Kovács László Imre (szerk): Választási rendszerek, Osiris-Láthatatlan Kollégium, Budapest, 1997, 69–71. (Hungarian) ↑
- Lijphart: Választási rendszerek, 85. ↑
- Fábián-Kovács: Voksok és mandátumok, 44. ↑
- John H. Humphreys: Proportional Representation, A Study in Methods of Election, Methuen & Co. Ltd., London, 1911. ↑
- Ever since the 1974 Maltese constitutional revision. ↑
- This is sometimes confused with the Hagenbach-Bischoff quota. The difference lies, first and foremost, in that the latter method works only with whole numbers gained through the quotient, instead of rounding up and adding one in case of fractions, as seen in Ireland. (Naturally, when working with a multitude of votes, the above amounts only to a very minor difference between the two quotas.) In addition, the Hagenbach-Bischoff formula is known not only as a quota system, but also as a mandate distribution method employed, among others, in Hungary. It breaks the distribution of mandates down to two phases; in the second one, unallocated mandates are meted out to the parties using the highest averages method. (In Hungary, a similar technique – the D’Hondt method – is employed as the second stage of a two-level system, rather than merely the distributory phase of a single-level formula. Therefore, even here, we can observe nuances of difference.) On the differentiation of specific terminology, see: Fábián-Kovács: Voksok és mandátumok, 37-38. ↑