Reform of the Electoral System in Canada and in Hungary – Towards a More Proportional Electoral System?
Reform of the Electoral System in Canada and in Hungary – Towards a More Proportional Electoral System?
Reform of the electoral system is and has been a timely topic in both Canada and Hungary. In Canada, the reform was part of Justin Trudeau’s election manifesto and, as his party won the majority of the seats in 2015, a reform process began. However, as we explain later, it was abandoned in February 2017. In Hungary, although the system was reformed in 2011, the electoral system is subject to heated debate for reasons discussed later in this paper. In autumn 2017, eight opposition political parties took part in a deliberation process aiming to propose a new electoral system.
To assess the chances, impact, and democratic consequences of a possible reform, many factors need to be taken into consideration. First, the current electoral and party systems need to be examined in order to gain a full understanding of the mechanisms and shortcomings of the current systems. Deficiencies emerged in both Canada and Hungary with respect to the distribution of seats compared to the percentage of votes, which, according to the reformist voices, require the reform of the current majoritarian systems towards more proportional ones.
Second, the rules regarding reform and the main institutional actors play an important role as well. These include who has the power to amend the system and whether the amendment of the constitution is needed. It also covers the reform procedure, and whether the reforming political powers (politicians, citizens’ initiatives, etc.) have referendum as a tool to force the change. As we show below, in the case of Canada it is not easy to assess whether any amendment of the constitution is needed, and what procedure should be applied, whereas in Hungary the constitutional background seems to be more clear-cut.
Third, the possible alternative systems need to be reviewed, in order to assess the ways in which reform may take place in future. As the paper intends to assess the institutional environment of a reform towards greater proportionality, in this part we discuss proportional representation (PR), single-transferable vote (STV) and mixed-member-proportionality (MMP) systems as alternatives.
Finally, we give a brief overview on the possible future of reforms; what chances reforms have, and what is the likelihood of a new and more proportional system to be adopted, with regard to the recent events in both countries.
II The Canadian and Hungarian Electoral and Party Systems
A party system is an enduring pattern of electoral competition between parties for public offices. A party system may be described from many aspects; one is the number of parties, or the effective number of parties in the system. With regard to the subject of this paper, party systems are of great importance as they ‘fill’ the electoral legislation; from the legal framework alone the outcome of elections cannot be predicted as the same electoral system may produce completely different results with different party systems. High disproportionality occurs usually when a multiparty system coincides with a majoritarian electoral system, as it is shown later in this paper. Hence, to understand the effective mechanism and outcomes of elections, and thus to decide whether reform is needed, two components, the electoral system (the legal framework) and the party system (the political reality), should be taken into consideration.
Canada is a federal state consisting of ten provinces and three territories. The electoral system of the country therefore may be scrutinised on two levels, at the provincial and on the national level. At the national level, Canada has a traditional ‘first-past-the-post’ (FPTP) electoral system, in which the country is divided into 338 constituencies (called ridings). In each riding, the candidate garnering the most votes becomes a member of the House of Commons. In the country with a population of nearly 37 million, voters have one categorical – i.e. not preferential – vote. As such, the share of seats in the House of Commons is the aggregation of the individual contests of the ridings – no compensational procedure exists. Although in many provinces there have been examples of reforming the electoral system at province level, at the federal level the electoral system – or at least its basic FPTP logic – has been unchanged, if not uncontested, since 1867.
The history of Canada’s party system can be divided into three main themes. The first, 1867–1921, can be labelled as a two party system as two main parties – the conservatives and the liberals – contested for the seats. The second, 1921–1993 is a ‘two and a half party system’, as, besides the two main parties, other smaller parties emerged. The third theme started with the elections in 1993, when the Conservatives went from the position of a majority party to a party with only three seats. Since then, the Canadian party system qualifies as a multi-party system. Since the latest elections, held in 2015, five parties have been represented in the House of Commons; the liberals, led by Justin Trudeau, won 184, the conservatives 99, the New Democrats 44, Bloc Québécois 10 and the greens one seat. The effective party number based on the results of the latest elections is 334, which shows that the country has a multi-party system.
Canada is an example of a so-called ‘non-Duvergian’ equilibrium. According to the famous political scientist Maurice Duverger, ‘the simple majority single ballot system favours the two party system’.
This means that – among other factors, as was admitted by Duverger himself – electoral systems shape party systems, and a system like that used in Canada favours a two party system. However, as we may see above, the Canadian example does not follow this rule, which means that not only the electoral system shapes the party system, but other factors as well. Thus, the coexistence of a multi-party system and simple-majority single ballot system produces highly disproportional results.
Hungary is a unitary state, where the electoral system is a mixed member proportionality (MMP) system, in which the Parliament, consisting of 199 members. is elected through an FPTP branch (106 seats) and a party list branch (93 seats). The voters therefore cast two ballots, one to the single constituency candidate, and one to the party list. Both votes are categorical; there is no possibility of preferential voting, or to determine directly the party list order. In the FPTP branch the candidate garnering most votes wins the seat; this branch is thus the same as the Canadian electoral system. In Hungary, however, there is a compensation mechanism: the votes cast for the candidates not winning the seat are transferred to the list branch. As a consequence of the reforms in 2010, the winner surplus vote has been introduced; the difference between the winner and the second candidate (minus one vote) qualifies as a surplus vote as well and these votes are therefore transferred to the party list. The votes cast to the party lists and the surplus votes then are divided by a proportional formula (d’Hondt) among those parties reaching the 5 percent threshold.
It is to be noted that the Hungarian electoral system has not been unchanged since the democratic transition of 1989–1990. After the winning coalition gained a majority of two thirds in 2010, the system was reformed. Although the new system described above preserved the MMP nature and the immanent logic of the previous one, significant changes were applied: the number of members of parliament was reduced from 386 to 199; in the FPTP branch the two-round vote was changed to a one-round vote, the winner surplus vote was introduced, and the proportion of FPTP branch was increased in relative terms so the weight of party list branch therefore declined.
Ever since the democratic transition, the Hungarian party system has been a multi-party system. As a consequence of the highly disproportionate electoral system from the end of the 1990s to 2010 two main political blocks emerged. The situation changed with the landslide victory of the right wing and the collapse of left wing, and, as a consequence, new parties have emerged. A party system with three main powers thus emerged: the right wing Fidesz-KDNP coalition, which formed the government, the far right Jobbik, and the remains of the left side, accompanied by new parties. The party system may be labelled a dominant party system. In the elections in 2014, eight political powers won seats in the Parliament: Fidesz-KDNP with 133 seats, MSZP (29), Jobbik (24), LMP (5), Együtt (3), DK 4, and PM (1). The effective party number is 3.2. As such, as in Canada, a highly disproportional system coexists with a multi-party system in Hungary.
III Necessity of Reforms
The reform of the Canadian electoral system has been demanded since 1921 and in Hungary reform was urged by political parties and also by scholars right from the transition in 1989-90. As mentioned above, the main reason in both cases is the discrepancy between the party system and the electoral system. The deficiency of the system may be approached from many aspects; just as there is no single and widely accepted function an electoral system has to fulfil, there is no widely and single accepted electoral system either. There are advantages and disadvantages to the proportional and to the disproportional systems as well. However, based on the Final Report and the Hungarian debates, we can identify two aspects; on the one hand disproportionality; the distribution of seats does not follow the proportion of votes [see points 1. and 2. below]. According to the critical opinions this results in a legitimation deficiency.
On the other hand, in these systems the political party forming the government often does not enjoy the support of the absolute majority, but only a relative majority of voters. Finally, as no preferential vote is allowed, this may lead to a situation in which the least preferred candidate may win [see C) below].
1 Proportional and Majoritarian Systems – Proportionality in General
Proportionality shows how accurately the system transforms voters’ preferences into seats, in other words, compared to their proportion of votes, the proportion of seats that are allocated to each party. If the two proportions are equal then the electoral system had a proportional result, i.e. the parties received seats according to their share of the vote.
Proportionality may be measured by different methods and formulas. It may be expressed with regard to one party at a certain election; the advantage ratio is the quotient of the percent of seats and percent of votes. For example, in 1993 the Conservatives in Canada garnered 18.84 percent of the total votes whereas they won 0.67 percent of the seats, thus the advantage ratio is 0.04. A more important quota is the Loosemore-Handby index (L-H index), which shows that, compared to a perfectly proportionate result, the percentage of the seats that are distributed to a different party. If A received 40 percent of votes and 60 percent of seats, B received 35 percent of votes and 30 percent of seats while C has 25 percent of votes and 10 percent of seats, then the L-H index is 20. In this paper we use the L-H index to express disproportionality of elections in Canada and in Hungary. We accept the classification according to which an election with a 0-5 L-H is proportionate, with 5-10 is moderately proportionate, and above 15 is disproportionate.
2 Disproportionality in Canada and in Hungary
FPTP systems without compensation usually give disproportional results. As we mentioned before, the effective mechanisms of electoral systems cannot be examined without the examination of the political environment, especially the party system, as these factors fill the rules of the system. As the Final Report reiterates, from 1867 until 1921, the Canadian FPTP system worked well, ‘there was a pretty good match between the distribution of seats in the House of Commons and the popular vote for political parties’. Since 1921, however, when the Progressive party emerged, with the multi-party system, the FPTP system produced disproportionality.
In the recent elections, held between 2004 and 2015 the system produced rather high L-H indexes (see table 1). According to the classification mentioned above, all these elections qualify as moderately disproportional.
The Hungarian MMP system produced disproportional results as well. It is to be noted that no conclusion may be drawn with regard to the proportionality just from the mixed nature of a system, as it depends greatly upon how effectively the PR branch can balance the FPTP branch. Both the German and Hungarian systems are MMP, for example, but while the former gives proportional results, as disproportionality is balanced at the level of the party list, it does not prevail in the Hungarian system. In the former Hungarian electoral system (from 1989 to 2010) the slightly disproportional territorial lists and the national list could not balance the single constituencies. Although the territorial lists disappeared in 2010, as a consequence of the winner surplus votes and the higher number of single constituencies, the new electoral system is more majoritarian than its predecessor. Table 2 contains the L-H index of elections from 1990: six elections (2010 and beforehand) were held with the former, and one election (2014) was held with the new system.
As we mentioned before, from 1998 the party system consisted of two blocks. It can be seen in the results of 2002 and 2006 that in this party system the electoral system produced more proportional results. With the return of the multi-party environment after 2010, however, it was disproportional again, and as proof of the more majoritarian inclination of the present system, in 2014 produced the higher L-H index in the history of democratic elections after the transition.
As we mentioned before, although we agree with Duverger that majoritarian electoral systems tend to form the party system that two block emerge, it is clear both from the Canadian and the Hungarian example, that other factors play important role in the determination of the party system. When these factors are stronger than the institutional factor, the party system becomes ‘incompatible’ with the electoral system, which means, that a multi-party environment develops in the framework of a majoritarian system, what leads to high disproportionality.
3 Absolute Majority Versus Relative Majority and the Question of Preference Ranking
Besides proportionality, another critical point is the question of the majority needed to obtain a majority of the seats. This question arises in two aspects. On the one hand, does the party or coalition obtaining a majority of the seats garner a majority of the votes as well? On the other hand, does the electoral system give the possibility of preferential voting, thus enabling the so-called ‘Condorcet winner’ to win the election?
As to the first point, it has been a serious complaint against the Canadian system since 1921 that the winner does not need to receive an absolute majority, but only a relative majority of the votes. This is the consequence of the co-occurrence of the multi-party system and FPTP system, where the candidate may win the seat with a relative majority. In Hungary the disproportionality of the system also enables a government to be formed without having an absolute majority of votes cast. We may state that, as far as the system is not proportional, there is always a chance that the coalition governs without the majority of votes.
The second point raises a more complex question. In categorical voting systems, voters cannot express their second, third, etc. preferences. With preferential voting, however, they can order their preferences, and it may lead to the avoidance of the so called ‘Condorcet paradox’, which can be illustrated with the following example. Suppose that there are three candidates, A, B, and C. Forty percent of the voters have the preferential ranking of A-B-C, which means that they prefer A to B and C, and B to C. 35 percent have a ranking of C-B-A, and 25 percent have a ranking of B-C-A (see Table 3)
|Proportion of voters||Candidate preferred in first place||Candidate preferred in second place||Candidate preferred in third place|
In a simple FPTP system without a preferential vote A wins, as she garners the most votes. But, as a matter of fact, A is the ‘Condorcet loser’, as she is defeated by both B and C if we consider how many voters preferred B or C to A, 60 percent (35 plus 25 as can be seen in the third and fourth rows) of the voters preferred B to A, while only 40 percent of the voters preferred A to B. It is the same with C. In this case, B is the Condorcet winner, as 65 percent of voters preferred her to C, while only 35 percent preferred C to B.
This example shows that, without preferential voting, it may occur that that candidate whose opponents were the least preferred wins the seat. This may cause disturbances in both the Canadian and Hungarian system. As to the latter, it is to be noted that before the reforms in 2010, with the two-round voting system in the single constituencies, voters could express their second preferences, and this could lessen the chance of the emergence of a situation like that described above. In the new system, however, with the one-round voting there is no chance that second preferences could prevail in this way.
IV The Constitutional and Institutional Background of Reforms
To estimate the likely success of a potential reform the constitutional background is needed to be taken into consideration, namely, whether an amendment of the constitution is needed and, if the answer is positive, what procedure is required. With respect to this we may distinguish between two types of constitutional restrictions regarding electoral reforms; formal ones, on the one hand, determine the procedure needed (majority, the co-decision of different bodies, time restrictions, etc.), and substantive ones, on the other hand, determine what standards need to be respected by the system for the reforms to be adopted (equal suffrage, guarantee of expression of will of voters, etc.). Moreover, other institutional factors may play an important role as well, for example whether reformers may initiate a referendum, or other factors that may make it easier or harder to accomplish the reform.
1 Canadian Framework
In Canada, the electoral system is not determined specifically by the Constitution Act of 1982 nor the Constitution Act of 1867. That means that the specific method of voting and the transformation of votes into seats are decided by the legislator when adopting the electoral system, in our case, most importantly the Canada Elections Act and proclamations issued under the Electoral Boundaries Act. Both constitutional acts, however, contain provisions that have to be respected if the reform is to be carried out without needing constitutional amendment. Constitution Act 1867 determines the number of House of Common seats allocated to each province and territory, the procedure and rules of this allocation, the ‘senatorial clause’ and proportional representation of the provinces (see below). Constitution Act 1982 contains provisions on the right to vote and on the amendment procedures.
Related to this, it is important to describe the rules of the constitutional amendment, which is regulated by part V of Constitution Act 1982. Under part V, four different procedures may be distinguished, first, the general ‘7/50’ procedure, in which the resolution of the House of Commons and Senate and at least two-thirds of the legislative assemblies of the provinces are needed; furthermore, these provinces need to represent at least fifty percent of the population. Proportional representation of the provinces in the House of Commons, for example, may be abolished by this procedure. Second, the unanimous consent of the provinces is needed on certain matters. Such a procedure would be required, for example, for the amendment of the so-called senatorial clause, under which each province may elect at least as many members to the House of Commons as it delegates to the Senate. Third, in matters affecting not all but some provinces, the consent of all provinces affected is needed. Fourth, unilateral amendment by provinces or the House of Commons is possible when purely provincial or federal interests are involved. In our case article 44 is paramount, under which the Parliament may unilaterally amend the constitution in relation to the executive government of Canada or the Senate and House of Commons. These formal procedures, with extra-textual amendment rules, make the Canadian Constitution one of the most difficult constitutions to amend in the world.
With regard to electoral reform towards proportionality, it is not clear whether amendment of the constitution is needed and, if so, under which procedure, the 7/50 or the unilateral procedure. Most experts heard by the Special Committee were of the opinion that if certain fundamental provisions are respected, the reform does not require the cooperation of the provinces, which means that either the unilateral amendment procedure is applicable, or no amendment is needed at all. This approach is strengthened by the case law on the relationship of the right to vote as declared in article 3 of the Charter and the FPTP system as well; in Figueroa v Canada the Supreme Court held that according to the Charter no specific electoral system is determined.
The answer became more ambiguous, however, with the decision of the Supreme Court in 2014 on the reform of the Senate (Reference re Senate Reform). In this case the Supreme Court had to answer more questions regarding the constitutional amendment consequences of the reform of the Senate, diverging from changing the term for Senators to abolishing the Senate. The Supreme Court stated that ‘the Constitution should not be viewed as a mere collection of discrete textual provisions. It has an architecture, a basic structure’. From this remark the conclusion may be drawn that changes in matters that are not regulated textually by the Constitution, but belong to its ‘basic architecture’ require the amendment of the Constitution. The Canadian Supreme Court therefore adopted the basic structure doctrine, most commonly associated with the Indian Supreme Court. This reduces the predictability of a Supreme Court decision on whether or not a certain matter requires constitutional amendment to a significant degree. There is a chance that, adopting this view, a proportional reform would affect the basic structure of the constitution, thus an amendment would be needed, even if the textual provisions mentioned above had not been affected.
Moreover, with regard to the amending procedures, the Court emphasised the exceptional and limited nature of unilateral amendment as laid down in article 44, and the general nature of the 7/50 procedure. What amendment procedure a reform introducing an STV, MMP or PR system would require is thus not certain. It makes the success of a possible reform at least questionable. The question is paramount, as the general 7/50 procedure is far less likely to be successful, as the cooperation of the provinces is needed, let alone the fact that some provinces introduced legislation under which the consent of the province may be given only after a referendum thereon in the given province.
Some scholars came to the conclusion that Reference re Senate Reform creates uncertainty and may imply that the general amending formula should be applied, as a reform moving from FPTP to some more proportional formula would alter Canadian politics to a great extent and hence would affect provincial interests. The Final Report noted, however, that ‘most experts suggested that the types of reforms contemplated by the Committee would not necessitate provincial support, provided certain requirements are met’. Emmett Macfarlane gives a convincing argument that if the proportional representation of the provinces is not affected then, as the House of Commons and the Senate are ‘functionally different’, ‘with the lower house emphasising representation of the national will by population and the upper house representation by state, province or region’, there is no reason for presuming that provincial interests would be affected to an extent that requires provincial consent.
The Canadian example shows that, in a federal state, the possible participation of the provinces in the reform of the national electoral system may significantly change the chances of success of any reforms. In a unitary state such as Hungary, this factor obviously cannot play a role in electoral reform processes.
Apart from constitutional amendment, the narratives in which the reforms take place are also worth mentioning. In his recent study, Michael Pal describes three main narratives of Canadian election law and its change, and argues that an ‘institutional’ approach may be held up, which puts emphasis on the partisan interests of those having the power to amend the system. He sees this narrative supported by the example of the failed Canadian reform in 2015. It is not the aim of this paper to identify the possible motivations of actors, but it is safe to say that, in the case of electoral system reform, partisan interests always play a prominent role. This may underline the functions of constitutional law in electoral system reforms, as there is a clear conflict of interests with regard to the fact that those who change the rules of the game are those who play according to the said rules.
2 The Hungarian Framework
In the sense of the applicable amendment and adoption procedure, the Hungarian legislation is far less ambiguous than the Canadian. The Fundamental Law, enacted in 2011, provides itself that the electoral system is laid down by a cardinal act (i.e. requiring the votes of at least two-thirds of the members of the Parliament who are present); furthermore, no referendum may be held on this matter and so the adoption of a different electoral system is the exclusive competence of the Parliament.
Within the framework of the Fundamental Law, the Parliament enjoys wide margin of appreciation, as it was upheld in numerous cases by the Constitutional Court. As the Hungarian Fundamental Law – similarly to the Canadian Constitution – does not specify the rules of the electoral system, the Parliament may reform the system to a great extent without the need to amend the Fundamental Law. However, it is to be noted that the amendment procedure of the Fundamental Law and a cardinal law differs only in the fact that for the former two-thirds of all members, while to the latter two-thirds of the members who are present is needed. With an organised and disciplined opposition, there is no effective difference between the two procedures with regard to the number of MPs needed. This means that the framework of the Fundamental Law may be changed with almost the same conditions as needed for electoral reform. This greatly reduces the effectivity of the constitutional framework.
Regarding the formal restrictions, besides the majority needed, it should be added that, in the case law of the Constitutional Court, the opinion appeared – however only in a dissenting opinion – that a time restriction could be derived from the Fundamental Law. This appears in the Venice Commission’s Code of Good Practice in Electoral Matters as well, which states that the fundamental elements of electoral law should not be open to amendment less than one year before an election. As the Fundamental Law does not contain this restriction textually, and neither was it applied by the Constitutional Court (although it had the opportunity to do so), the application of this restriction is far from obvious. We think, however, that it clearly follows from the standards of rule of law, and should be followed by both the legislator and the Court.
The substantive restrictions of a reform are the provisions on the right to vote and on the principles of elections. With regard to the electoral system in the narrow sense, principles of elections are of greater importance; under article 2 paragraph (1) of the Fundamental Law, elections shall be based on universal and equal suffrage and the ballot shall be direct and secret. To these four principles (universality, equal suffrage, direct ballot and secret ballot) a new one was added when the Fundamental Law was adopted in 2011, the principle of the guarantee of free expression of the will of the voters.
From these principles, equal suffrage and the guarantee of free expression of the will of voters could be important with regard to an electoral reform affecting the transformation of votes to seats. Under the case law of the Hungarian Constitutional Court, however, proportionality falls out of the scope of equal suffrage. The Court, in its previous case law, made a distinction between equal suffrage and proportionality as well – however not without the criticism of scholars – but in its decision on winner compensation it expressly held that ‘a majoritarian, proportionate, or a mixed system in itself does not raise the violation of equal suffrage. Consequently, a purely majoritarian system could be constitutional as well […].’ This means on the one hand that the current system, may it be as disproportionate as it is, qualifies as constitutional, and that the reforms would most probably not be affected by the constitutional background, even if an FPTP system were introduced.
Consequently, in Hungary, the amendment of the constitution is not needed for a proportional reform, and so the Parliament may adopt such a new system with the votes of two-thirds of the MPs present. The chances of reform, however, are lowered by the fact that no referendum may be held on the matter and therefore the opposition – which may have an absolute majority of the votes – cannot enforce the change.
V Effects of Reform
Apart from the possible constitutional textual changes following a potential reform, the change of the electoral system may trigger changes in the underlying democratic and constitutional mechanisms in a given state. Switching from an FPTP or majoritarian MMP system to a more proportional system brings about changes not only in the transformation of votes and electoral process, but in the structure of constitutional institutions as well. Choosing between a proportional or majoritarian system is basically choosing from two types of democracy described by Arend Lijphart as between the majoritarian (or Westminster) model and the consensus model.
This distinction may be summarised as follows: in the consensus model, a proportional electoral system prevails, multi-party coalitions emerge, the legislative and executive branch are not so tightly connected, the party system is a multi-party system and the concentration of power is lower, which means that there may be judicial reviews of legislation, federal-province division, etc. As opposed to this, in the Westminster model, a majoritarian electoral system prevails, the need to form a coalition is much lower, the party system often consists of two blocks, the legislative and the executive branches are more tightly connected and the concentration of power is higher.
Although it may seem that the electoral system is only one component among many, it is to be noted that these components are interrelated, and the electoral system determines other factors. As we mentioned, the electoral system affects the party system; furthermore, it affects the possibility of coalition governments, thus the connection between the legislative and executive branches. It follows that, with any reform proposal, we make a proposal for a model of democracy as well. The consensus model gives more legitimation and involves more agents in the decision-making process, while the Westminster model may provide stable governmental power and thus governmental crises may be avoided.
VI Possible Ways of Change – Alternative Systems
For both Canada and Hungary, different types of system are open as options. These options differ greatly both in complexity and in how widespread they are among democratic states. In this part we examine proportional representation (PR), single transferable vote (STV), and mixed member proportionality (MMP) systems. As part of the last of these, we examine how the FPTP branch could be amended with preferential voting.
1 PR Systems
In PR systems, the mandates are distributed in districts that are big enough to allow proportional representation. In the Netherlands and in Israel, for example, the whole country is an electoral district; the voters cast their ballot to lists, and then the seats are distributed using a proportional formula. The larger the district size (the so called district magnitude), the more proportionate the system. PR systems thus may provide near full proportionality.
The trade-off for proportionality is the greater distance between the elected and voters, in that it provides no local representation. In Canada, local representation was laid down as an important value of reform, and in Hungary it forms part of the electoral tradition as well, although not with the same force as in Canada. Moreover, in Canada the representation of provinces is laid down in the constitution. The amendment of the constitution would therefore be required if the whole country constituted a single district. Another trade-off is the stability of the government. The more proportional a system is, the more likely a coalitional government forms, instead of a single-party government. If social and political cleavages are deep enough to make coalitions impossible or fragile, a pure PR system may give less stability.
As a consequence of disappearing local representation, we may conclude that a purely PR system will not be and maybe should not be introduced in Canada or in Hungary.
In STV systems the given state is divided usually into districts with 3-8 mandates. The voters cast their ballot according to their preferential order, marking the candidate in the first place, second, and so on. The distribution of seats is rather difficult: first, a quota is determined and any candidate reaching this quota wins a mandate. The preferential votes for the winning mandates are distributed among the remaining candidates. If somebody reaches the quota then she wins the mandate or, if nobody reaches it, the candidate with the least votes is dropped out, and her preferential votes are distributed. This process runs until every mandate is distributed. The most notable example of an STV system is Ireland.
This system provides higher proportionality, while it does not abandon local representation, as the districts have 3-8 mandates. Furthermore, voters may express their preferential order, which reduces the necessity of strategic voting (i.e. abandoning the candidate preferred in the first place as the candidate does not have a chance to win). An apparent drawback is the complexity of the system; the distribution process is hard to understand. Another effect – may it be an advantage or disadvantage – is that candidates from the same parties run against each other; this, however, as political scientist Michael Gallagher said in his testimony before the Special Committee, does not really damage party cohesion.
As STV preserves local representation while giving more proportionality then, apart from its complexity, it is hard to raise any objections to this system with regard to Canada or Hungary. However, in case of Hungary the general objections related to instability caused by proportionality may be brought up.
In an MMP system a proportional branch balances the disproportional effects of the FPTP branch. As I mentioned before, there are many variants and they differ on the proportionality they provide. The advantage of these systems is that they provide both local – through the FPTP branch – and proportional representation. It is interesting, that in Canada one of the greatest objections to an MMP system was the fear that it creates two kinds of MPs; it creates a distinction between those elected in single districts and from the lists. According to the experiences with the Hungarian MMP system, however, this has not been found to be the case, as in the media and public discussion it is not widely known whether an MP was elected for a district or from a party list.
It is to be added that the FPTP branch may be amended in a way that enables preferential voting. In an alternative vote (AV) system, for example, the voters cast their ballot according to their preferential order. If a candidate garners an absolute majority according to the first preferences, she wins the seat. If not, the candidate with the least votes is eliminated, and her votes are distributed according to the second preferred candidate on her ballot papers. This method may not increase proportionality, but it may solve the Condorcet paradox described above.
VII The Future of Electoral Reforms in Canada and Hungary
Finally, we need to mention the possible future prospects of reforms in Canada and in Hungary. As we mentioned above, in Canada the electoral reform was a major campaign promise, but it was abandoned in early 2017. The abandonment draws attention to the paradox that only those in power can change the electoral system, but they are clearly benefiting from it, as they realise their power; once the Liberals won the elections, reform became more of a burden.  It is rare that the governmental political forces have an incentive to reduce the majoritarian nature of the current system. An example for the exception comes from France; where before the elections held in 1985 the socialists introduced a more proportionate system to mitigate their electoral defeat. This, however, is not an example for change motivated by commitment to electoral policy, but by sheer political interests.
In Hungary, as we mentioned above, the system has changed, but not in a proportionate direction. Although it is discussed in public, as the reform requires the majority of two-thirds, it is hardly conceivable that, prior to the elections held in 2018 or even in the term from 2018 to 2022, the system would be reformed. Maybe the most probable scenario for a reform would be a situation similar to that in France in 1985 described above; if a political party won the next election by two-thirds and at the end of the term the polls suggest a great defeat, a possible strategy for mitigating defeat could be to introduce a more proportional system.
We have seen that the coexistence of disproportional FPTP and MMP electoral systems and multi-party systems cause high disproportionality in both Canada and in Hungary, which may be the most important reason for reforming the systems. As we pointed out, the reform could provide a chance that preferential voting could be introduced, which gives an opportunity to have results displaying the actual preferences of voters.
In Canada, the constitutional background does not give a clear answer whether amendment of the constitution is needed, and if the answer is positive, then what procedure may apply. As we argued, the constitutional framework greatly affects the chances of successful reform, especially where constitutional amendment is as difficult as in Canada. The Hungarian constitutional background is more clear-cut; however, the required votes of two-thirds of MPs is rather hard to fulfil, especially if we take into consideration that any party or coalition with such a majority does not have the incentive to reform the system. In Hungary, the fact that no referendum may be held on the subject makes reform more difficult.
We listed the possible ways of changes, through PR, STV and MMP systems. PR systems may be ruled out as they do not promote a personal connection between voters and the elected, which is traditionally a feature of electoral systems in both countries. STV systems couple the advantages of voter-elected connection and proportionality; the disadvantage of these systems, however, is their complexity and the general objections against proportionality (i.e. less stable governments, especially in young democracies). MMP systems may be a solution in both countries, as they may be fully or partly proportional, and the coexistence of lists and single candidates ensures the voter-elected connection as well.
Finally, we pointed out the paradox of electoral reforms, namely that those elected usually do not have any incentive to reform the system towards more proportionality, as they themselves have been elected under the rules of it. As in both countries the system may be reformed by the national assembly (House of Commons in Canada, Parliament in Hungary), this questions whether in the near future any successful reform could take place.
* János Mécs is a doctoral student at Eötvös Loránd University (ELTE), Faculty of Law.
 As a consequence, a special parliamentary committee, the Special Committee on Electoral Reform (Special Committee) was formed. The Special Committee published its comprehensive final report (hereinafter: Final Report) in December, 2016. The Final Report is available at: <http://www.ourcommons.ca/Content/Committee/421/ERRE/Reports/RP8655791/errerp03/errerp03-e.pdf> accessed 25 September 2017.
 Reuters: Canada abandons electoral reform in reversal of Trudeau’s pledge – available at <https://www.reuters.com/article/us-canada-politics-reform/canada-abandons-electoral-reform-in-reversal-of-trudeau-pledge-idUSKBN15G5AD?il=0> accessed 25 September 2017.
 The Budapest Beacon: ‘Eight opposition parties sign joint declaration on election reform at Agóra’ <https://budapestbeacon.com/eight-opposition-parties-sign-joint-declaration-election-reform-agora/> accessed 25 September 2017.
 Pradeep Chhibber, Ken Kollman, The formation of National Party Systems (Princeton University Press 2011, Princeton) 4.
 In the formula of effective electoral party number, the parties are weighted with regard their share of votes, whereas effective number of parliamentary parties is weighted by the share of mandates. See Markku Laakso, Rein Taagepera, ‘Effective Number of Parties: A Measure with Application to West Europe’ (1979) 12 Comparative Political Studies 3–27. In this paper we use effective electoral party number. This gives a more accurate insight to the party system, as opposed to the mere number of parties. The effective party number is the inverse of the sum of the squared proportions of the vote share.
 The Hungarian electoral system, for example, with regard to the legislative background was mostly untouched from 1989 to 2010; however, it produced different outcomes with respect to proportionality.
 It is an accepted and evident concept in the Hungarian literature on electoral systems as well, that the examination of electoral system may not be conceived without the examination of the broader context. See, for example: Dezső Márta, ‘Választási rendszerek’ in Máthé Gábor (ed), Választójog (Változó Világ 50 – Press Publica 2004, Budapest) 58.
 The ten provinces are British Columbia, Alberta, Saskatchewan, Manitoba, Ontario, Quebec, Newfoundland and Labrador, Nova Scotia, New Brunswick and Prince Edward Island. The three territories are Yukon, Northwest Territories, and Nunavut.
 Nagy Levente, Arányossági számítások Kanada választási rendszerében (2015) 24 (1) Politikatudományi Szemle 145.
 Chhibber, Kollman (n 4) 2.
 The results (share of votes): Liberal Party – 39,5 percent; Conservative Party – 31,9 percent; New Democratic Party – 19,7 percent; Bloc Québécois – 4,7 percent; Green Party – 3,4 percent. See: <https://lop.parl.ca/ParlInfo/compilations/ElectionsAndRidings/ResultsParty.aspx> accessed 09 October 2017.
 Bernard Grofman, Shaun Bowler, André Blais, ‘Introduction’ in Bernard Grofman, Shaun Bowler, André Blais (eds), Duverger’s Law of Plurality Voting – The Logic of Party Competition in Canada, India, the United Kingdom and the United States (Springer 2009, New York) 2–3.
 Maurice Duverger, Political Parties, Their Organization and Activity in the Modern State (Wiley 1954, New York) 217.
 Duverger (n 13) 228.
 Duverger explains it by two mechanisms; the mechanical effect means that ‘small parties will tend to be squeezed out mechanically, simply because the operation of electoral system denies them seats’ whereas ‘psychological effects are centred in the reaction of instrumentally minded voters and elites to the expected working of the electoral system’. Grofman, Bowler, Blais (n 12) 2.
 It exceeds the boundaries of this paper to find a conclusive answer or to detail the relevant scientific findings further. For further information see: André Blais, Eugénie Dostie-Goulet, Marc André Bodet, ‘Voting strategically in Canada and Britain’ in Grofman, Bowler, Blais (n 12) 13–26; and Richard Johnston, Fred Cutler, ‘Canada: The Puzzle of Local Three-Party Competition’ in Grofman, Bowler, Blais (n 12) 83–96.
 Except for the voters without Hungarian residence; they may vote only for party lists.
 Thus, if the winner garners 10,000 votes and the second candidate 5000, then the second candidate has 5000 surplus votes, and the winner 4999, i.e. the votes ‘not needed to win’ – as the first candidate could have won with only 5001 votes. – For a critical review of this approach and winner surplus vote see: János Mécs, ‘A győzteskompenzáció alkotmányosságáról – a pozitív töredékszavazatok megítélése az egyenlő választójog tükrében’ in Kálmán Bence, Tar Adrienn (eds), Tudományos Diákköri Dolgozatok (ELTE Bibó István Szakkollégium 2015, Budapest) 45–84; Halmai Gábor, ‘Két (egyetlen) választás Magyarországon – Az AB a parlamenti és fővárosi önkormányzati választási rendszeréről’ (2014) 4 Fundamentum 83–88.
 For a more detailed description in English of the change of electoral laws in Hungary see: János Mécs, ‘Development of the electoral system in Hungary since 1991’ in Márta Dezső, Zoltán Pozsár-Szentmiklósy (eds), Development of electoral systems in Central and Eastern Europe since 1991 (Association of European Election Officials 2016, Budapest) 134–137.
 Horváth Attila, Soós Gábor, ‘Pártok és pártrendszer’ in Körösényi András (ed), A magyar politikai rendszer- negyedszázad után (Osiris – MTA Társadalomtudományi Kutatóközpont Politikatudományi Intézet 2015, Budapest) 273–274.
 The results (share of votes): Fidesz–KDNP – 44,87 percent; MSZP–EGYÜTT–DK–PM–MLP – 25,57 percent; Jobbik – 20,22 percent; LMP – 5,34 percent. See: <http://www.valasztas.hu//hu/ogyv2014/861/861_0_index.html> accessed 9 October 2017.
 Final Report (n 1) 23.
 For the arguments and speculations on the change in 2010 see (2010) 4 Fundamentum available at: <http://fundamentum.hu/sites/default/files/10-4-04.pdf> accessed 9 October 2017.
 Douglas W. Rae ‘A választási rendszer változó elemei’ in Fábián György (ed), Választási rendszerek (Osiris Kiadó 1997, Budapest) 55.
 We add the absolute difference between percentage of votes cast and mandates obtained in case of each party up, then divide the number by two.
 As A has 20 percent more seats, and B and C have fewer seats (by 5 and 15 percent respectively) compared what was due based on proportional distribution. The sum is 40, which should be divided by 2, because it is a zero sum game, thus what qualifies for A as overrepresentation is at the same time underrepresentation in the case of B and C, so by adding up the absolute differences we counted every difference twice.
 The Final Report, however, refers to the Gallagher index when referring to disproportionality. On the Gallagher index. see: Michael Gallagher, ‘Proportionality, Disproportionality, and Electoral Systems’ (1991) 10 (1) Electoral Studies 33–51.
 Fábián György, ‘Választási rendszerek az Európai Unió Tagállamaiban’ in Szoboszlai György (ed), A közjogi választások egyes elméleti és gyakorlati kérdései. Közigazgatási továbbképzési jegyzet (Magyar Közigazgatási Intézet 2004, Budapest) 59–96.
 Final Report (n 1) 22.
 Nagy (n 9) 150.
 Tóth Csaba, ‘A magyar választási rendszer működése’ in Garai Borbála, Takács Péter (eds), Multa, rogare, rogata tenere, retente docere – Tudományos Diákköri Dolgozatok 2001. (ELTE ÁJK 2002, Budapest) 127–145.
 Tóth Csaba, ‘Választási rendszer’ in Körösényi (n 20) 243.
 In 2014 the winner parties gained two-third-majority in the legislation with less than 50 percent of votes (45 percent).
 The following example was showed to the Special Committee by political scientist Eric Mascin. See Final Report (n 1) 62–63.
 The Final Report addresses – or at least mentions – this issue: See Final Report (n 1) 62–63.
 In our example, in the second round C or B could step back and thus by using voters’ second preferences it may be avoided that the least preferred candidate won.
 It is to be noted that under section 52 paragraph (2) of Constitution Act 1982, the Constitution of Canada includes the Constitution Ac, 1982, and acts and orders referred to in the schedules. The most important act this latter includes is Constitution Act 1867, which contains the basic structural and organisational provisions.
 Emmett Macfarlane also mentions Parliament of Canada Act, which contains provisions on the qualification and disqualification of members of the House of Commons. Emmett Macfarlane, ‘Constitutional Constraints on Electoral Reform in Canada: Why Parliament is Mostly Free to Implement a New Voting System’ (2016) 76 Supreme Court Law Review 402.
 Section 37, 51, 51A and 52 of Constitution Act 1867.
 Right to vote is stipulated by section 3 of the Canadian Charter of Rights (hereinafter: Charter), which is part of Constitution Act 1982.
 The literature often mentions five procedures, counting the unilateral provincial and House of Commons procedure as two. See, for example Richard Albert, ‘The Difficulty of Constitutional Amendment in Canada’ (2015) 53 (1) Alberta Law Review 86. In this paper we follow the Supreme Court of Canada, which, in its Reference re Senate Reform, distinguished between four procedures: ‘Part V reflects the political consensus that the provinces must have a say in constitutional changes that engage their interests. It contains four categories of amending procedures.’ Reference re Senate Reform,  ISCR 704, 2014 SSC 23.
 Section 38 of Constitution Act 1982. In the ‘7/50’ abbreviation seven stands for the number of provinces and 50 for the percentage of population needed.
 Section 42 paragraph (1) point a) of Constitution Act 1982. This provision refers to the proportionate representation of the provinces, which should not be confused with the proportionality of electoral system. The former refers to the fact, that each province should represent itself in proportion with its population, while the latter refers to the proportionate distribution of seats with regard to party preferences among voters.
 Section 41 paragraph (1) point b) of Constitution Act 1982. – Voters in Prince Edward Island are entitled to elect four members as a result of this provision, though the population of the Island would imply fewer members.
 Section 43 of Constitution Act 1982.
 Sections 44 and 45 of Constitution Act 1982.
 Albert (n 41) 112.
 Final Report (n 1) 11.
 Figueroa v Canada (Attorney General),  1 S.C.R. 912, 2003 SCC 37. para 37: ‘After all, the Charter is entirely neutral as to the type of electoral system in which the right to vote or to run for office is to be exercised. This suggests that the purpose of s. 3 is not to protect the values or objectives that might be embedded in our current electoral system, but, rather, to protect the right of each citizen to play a meaningful role in the electoral process, whatever that process might be.’
 Reference re Senate Reform,  1 SCR 704.
 Albert (n 41) 87.
 Emmett Macfarlane came to the conclusion that ‘It is clear from the Court’s articulation of the constitutional architecture concept that electoral reform could be regarded as a change of a constitutional nature.’ Macfarlane (n 38) 405.
 Richard Albert argues that ‘major constitutional amendment to the Constitution of Canada may today be so difficult as to render it impossible’. By major constitutional amendment he understands the 7/50 and the unanimity formula. Albert (n 41) 87–88.
 Richard Albert mentions Alberta as an example, and argues that this kind of provincial legislation is problematic, as presents extra-textual requirements to a possible amendment. Albert (n 41) 98–100.
 See, for example Michael Pal, ‘Constitutional Amendment After the Senate Reference and the Prospects of Electoral Reform’ (2016) 76 Supreme Court Law Review 377–398.
 Final Report (n 1) 11.
 Macfarlane (n 38) 408–409.
 Apart from the institutional approach, Michael Pal describes historical and theoretical narratives. First, the historical narrative emphasises the historical ‘development’ towards a more inclusive right to vote. Second, according to the theoretical approach, it adopts an egalitarian model, according to which the campaign finance regulation aims to level the playing field. – Michael Pal, ‘Three Narratives About Canadian Election Law’ (2017) 2 Election Law Journal 255–262.
 Ibid 260.
 Article 2 paragraph (1) and article 8 paragraph (3) point c) of Fundamental Law.
 See for example: Decision 3141/2014. (V. 9.) AB of the Constitutional Court of Hungary section ; and Decision 63/B/1995 AB of the Constitutional Court of Hungary, ABH 1996, 509, 513.
 Article S) paragraph (2) of the Fundamental Law.
 See Judge István Stumpf’s dissenting opinion to a decision, which examined the constitutionality of a change of the local government electoral system, which took place just months before the elections. From the fact that this opinion appeared in a dissenting opinion we may not draw the conclusion that the Court rejected this argument, as this was not on the merits of the case. See Decision 26/2014. (VIII. 23.) AB of the Constitutional Court of Hungary sections –.
 European Commission For Democracy Through Law (Venice Commission) ‘Code of Good Practice in Electoral Matters’ Point II. 2., available at: <http://www.venice.coe.int/webforms/documents/default.aspx?pdffile=CDL-AD(2002)023rev-e accessed> 9 October 2017.
 Article XXIII and article 2 paragraph (1) of Fundamental Law.
 Varsányi Benedek, ‘Egyenlő választójog – arányos választási rendszer’ (2006) 3 Fundamentum 107–114, and Mécs (n 19).
 Decision 3141/2014. (V. 9.) AB of the Constitutional Court of Hungary section .
 This, fortunately, has not been articulated by anyone as a desirable way of reform.
 Arend Lijphart, Patterns of democracy: Government forms and Performance in Thirty Six Countries (Yale University Press 1999, New Haven) 9–48.
 Ibid 3–4.
 In IDEA Handbook on Electoral System Design (hereinafter ‘IDEA Handbook’) PR and STV systems are categorised as proportional systems. Following this categorisation a possible reform towards a more proportional system could take the form of one of these systems. MMP systems are categorised as mixed systems, and in the case of Canada it would be a step forward in terms of proportionality as well. For the categorisation and description of these systems see: Institute for Democracy and Electoral Assistance ‘Electoral System Design: The New IDEA Handbook’ 27–129, available at: <https://www.idea.int/sites/default/files/publications/electoral-system-design-the-new-international-idea-handbook.pdf> accessed 9 October 2017.
 Ibid 57.
 For further information on district magnitude see: Rae (n 24) 49–51.
 I.e. producing L-H index less than 5 percent.
 Final Report (n 1) 42.
 Under section 41 point (b) of Constitution Act 1982, the resolution of each province is needed if the senatorial clause is affected, under which each province is entitled to ‘a number of members in the House of Commons not less than the number of Senators by which the province is entitled […].’ However, if the whole country were a single district then this article would certainly be affected, as no province would be entitled to a specific number of mandates in the House of Commons. For the same reason, section 41 point (a) would be triggered as well, which stipulates the proportionate representation of the provinces in the House of Commons. This section triggers the general (‘7/50’) amending formula.
 For further explanation of the STV system see: IDEA Handbook (n 71) 71–77.
 Final Report (n 1) 79.
 For the description of the process of the failure of reform in 2015–2017 see: Henry Milner, ‘Electoral System Reform, the Canadian Experience’ (2017) 16 (3) Election Law Journal 349–356.; see also: Pal (n 58).
 Alan Renwick, The politics of electoral reform (Cambridge University Press 2010, Cambridge) 89–111.