Hungarian Electoral System: A Study in Complicated Simplicity
Term paper for course 020008-A Mathematics and Democracy
14 April 2003
Juraj Draxler
Matúš Harvan
Table of contents:
Introduction 3
Hungarian electoral system 5
Why this system? 8
Influence on party politics, inter-party cooperation, and government formation 11
Conclusion 16
Bibliography 17
Appendix 1 19
Appendix 2 - Kenneth Benoit’s Model of Electoral Change 21
Appendix 3 26
Appendix 4 30
Introduction
“[Electoral systems] are strange devices – simultaneously cameras and projectors. They register images which they have partly created themselves.”
Maurice Duverger[1]
Why study an election system of a small, post-communist country? For one, the Hungarian system is often described as the most complicated in the world. If nothing else, then this trait makes it a fair game for any student of electoral practices. With its three tiers for the allocation of seats in the legislative body, the use of different formal thresholds for different voting rounds and with different rounding systems among the tiers, the system is enticing to look at purely by the virtue of its mechanics.
Secondly, the system is of the mixed type, combining elements of majority and proportional seat allocation. Mixed electoral systems used to be an oddity, but they started to proliferate in the 1990s.[2] Hungary is therefore interesting as a case in point regarding this sudden fashion for mixed-member systems, and can provide some practical example for the discussion of advantages and disadvantages of this system type.
Thirdly, one can look at the specific circumstances in which an electoral system is adopted. For decades, ever since Maurice Duverger came up with what later became known as “Duverger’s Law” postulating effect of electoral laws on party systems, political science seems to have been preoccupied with consequences of election systems at the expense of the study of the reasons that lead to the adoption of a particular system.[3] This balance has somewhat tilted with the adoption of democracy in Eastern Europe, which brought the question of electoral system adoption and change more strongly to the fore.
But of course the impact of an electoral system on the politics of a country is a desirable object of study - not only in terms of the Duverger effect on the number of decisive parties in the legislative body but also in terms of how it shapes the election contest, facilitates formation of party alliances and impacts the character of the political groups - simply, in all its intra- and interparty effects.
This paper will, then, attempt to describe the Hungarian electoral law as both an interesting invention and as a mechanism that powerfully shapes the fates of its creators. In the first section we will introduce the system and provide some analysis of its properties. Next, we will look at it in the light of the theories of electoral systems change, and finally we will proceed, in the third section, to a brief analysis of its effects on electoral contest and on the nature of the party system. The focus is therefore firmly on the seat allocation mechanism and its effects on the body politic.
The thrust of the paper is practical: to show that the Hungarian model, lying somewhere in the middle between the majoritarian and proportional poles, is an interesting object of comparison with systems that lie to either side of this divide. Empirical data shows that Eastern European countries tended to reduce the extremities in their systems in 1990s, often altering majority voting to mixed-member, and reducing proportionality of PR systems by changing the rounding up formulas.[4] By occupying the middle ground between these two tendencies, the Hungarian system can help us assess the limits of this process of convergence and show how desirable or undesirable it could be.
Hungarian electoral system
In terms of the importance of the seat allocation mechanism, one must bear in mind that Hungary has a one-chamber legislative body, and the power of this body is not powerfully checked by the president, who is elected by parliament and has a largely ceremonial function. This electoral law was adopted in 1989 following roundtable negotiations between the communist government and opposition groups and has remained essentially the same ever since (except for the change in the election hurdle for the PR component from 4 to 5 percent in 1994 – see below). The law is not part of the constitution, but it still requires a two-thirds majority in the parliament in order to be amended.[5]
The seat allocation mechanism consists of three components that partly influence each other. 176 seats are elected in single-member districts. The candidate who wins the most votes, but at least 50 percent plus one, is elected. If no one passes the formal threshold, the top three candidates and any candidate with more than 15 percent of the vote in the first round proceed to a second round, where the candidate with most votes wins. A second round is also held if less than a half of eligible voters cast their votes in the first round. In both cases where a second round is held, at least 25 percent of eligible voters must take part. Candidates can withdraw from the race between the two rounds, which increases the chances of electoral alliances, as we will see later.
The second component is the open party list vote, for the purposes of which Hungary is divided into 20 electoral districts. The number of seats in each district varies from 4 to 28, with a median of 6. A maximum of 152 candidates can be elected in this round, and there is a threshold of 5 percent in this component.[6] Seats are distributed according to a special largest-remainder formula. This formula excludes from the allocation of remainder seats those parties whose remainder votes are less than two-thirds of the quota calculated in that district. The undistributed seats are added to the mandates that are to be distributed in the next round. This third round thus allocates the minimum of 58 national list seats. Surplus votes – those that were cast for losing candidates in the SMD round and the remainder votes from the party list round – are used. D’Hondt (which is known to favor larger parties) is used. As the seats unallocated in the party list round are added to the 58 minimum number of mandates, the number of seats distributed in this round varies (in some elections it rose to as much as 90 seats). The total number of seats in the parliament stays at 386. The idea behind the third round is to compensate for votes “wasted” in the first two. The complete version of the electoral law is presented in Appendix 4.
The system, as we see, provides linkage between the different components.[7] This linkage is semi-compensatory. It does balance out the majoritarian outcome of the first round, but only partially. That is why many authors talk about “manufactured majorities” in regards to the Hungarian system.
As we will see later, this complicated system arose as a certain compromise in political negotiations in 1989. The influence of other election systems is therefore only indirect. The most striking resemblance is to the German system, which also has a single-member and a party list tier (but which is categorized as a PR system, since the linkage between the single-member and the party tier is designed to be compensatory towards achieving PR). Also, voting in the single-member tier has hurdles that are similar to the French double-ballot system.
The disproportionality of the system was of course clearly demonstrated in the first elections, when the Hungarian Democratic Forum, the election winner, received 24.7 percent of the list vote but 42.5 percent of parliamentary seats. The effect was very clear in every election since (see table in Appendix 1). In three out these four elections the two largest parties received a bigger share in seats than was their percentage of the list vote, while others less (in one election, 1994, only one party, the socialists, benefited from this magnifying effect).
Especially in the first three elections, when the number of subjects independently winning seats in the parliament was rather large (in 2002 the number fell dramatically to 3), we observe that the absolute percentage gain by the largest party or parties is conversely reflected by a magnified absolute percentage loss of the smallest part. In other words, as column K in the table in Appendix 1 shows, while the system behaves fairly proportionally in the middle, the absolute value of percentage loss of the smallest party tends to get rather large (one outlier being the MDF in 1998). We chose to call this effect ‘polar magnifying’.
This is a purely empirical observation. It depends on certain conditions to hold, such as a certain level of correlation of the share of both tiers where voters cast their votes. Nevertheless, we think the effect is worth noting, as it can, in the least, serve as a rule of a thumb for parties to assess their post-election strength.
To sum up, the seat distribution of the Hungarian electoral system has the following format:
Why this system?
The question why Hungary has its electoral system breaks down to three more specific inquiries: why did Hungary adopt a mixed-member system rather than the more standard proportional representation or majority system, why did it come up with its specific, highly idiosyncratic form, and, lastly, why is it keeping the system?
This cannot be truly answered without a look at historical facts on the one hand, and a theory of electoral change on the other. Formal descriptions of the workings of electoral systems do not explain why the system is adopted, for two reasons. The models do not tell us which choice is better, as they are based on different assumptions (governability and proportionality in representation). Moreover, theories that treat electoral models as an exogenous variable do not explain the interests and motivations of real-life actors that choose to adopt the models.
One theory that seeks to explain the existence of a certain electoral system, is the theory of “founding elections”, which “freeze” the rules and players for the subsequent democratic cycle. This theory holds well for Western Europe (by which are meant those European states that were non-communist until 1989), where the electoral systems tended to change very little after World War II, with the exceptions of France and Greece.[8] Some scholars apply this model to Eastern Europe, with the addendum that the system for the first election was usually designed as majoritarian or mixed if the communist party had a strong say in drafting the rules.[9] The logic behind this is that communist parties, while being rather unpopular at the period, could most easily field recognizable personalities in the district contests. In the non-personalized PR elections, the risk was that a large majority of people would be willing to vote for any party rather than the communist. This was certainly indicated by surveys in 1989 Hungary,[10] secretly conducted for the Hungarian Socialist Workers Party (HSWP).[11] Some party officials were well aware of the capabilities of electoral systems, as evidenced by a report addressed to a closed session of the HSWP Politburo by the Party Secretary for internal security György Fejti:
[A]n exclusively electoral list system is not at all favorable for the [HSWP]… Constituencies, undoubtedly, bring about a competition of individuals, personalities. And a competition of local programs. So, the focus of the debate is not really the general program of the party, but events on the local level. This is why I believe that this is the advantageous method for the [HSWP]…[T]he historical parties who had run in the elections in 1945 and 1947, and are remembered from then, will be in favor of lists, the exclusively [nationwide] electoral lists… I firmly propose to use the German model where half the seats are decided on by constituencies and the other half through county lists.[12]
The HSWP could afford to take a rather active stance in the negotiations with the opposition as it had not compromised itself in the recent past by rigid Stalinist tendencies – the communist government took a lenient stance towards formation of political opposition and economically Hungary was one of the most liberal countries in the socialist camp. Some of the reformist members of HSWP enjoyed high popularity with the public.[13]
It can be said that the party underestimated the unpopularity of the communist label, and at the end returned only one SMD candidate in 1990. Nevertheless, four years later its fortunes reversed dramatically (see previous section) and the party was swept to power with a clear majority in the parliament (54.1 percent of seats), based on a minority of votes (32.99 in the list section and 32.19 in the SMD section).[14]
While it is very well to look at “founding elections” for an intuitive foray into the issue of electoral system change, what this approach leaves wanting is a set of analytical tools for predicting more generally the conditions for an adoption of an electoral system, and, moreover, for a potential change of the system.
Kenneth Benoit sums up the theories of electoral change as comprised of three categories.[15] Let us look at these before briefly before assessing the Hungarian experience against them.
The first category he labels as Self-Interest Derived Preference explanations. These include policy-seeking preference theories, according to which the electoral system choice arises from a calculation aiming at maximizing the chances that a certain policy be adopted. This presupposes maximization of parliamentary seats and of coalition potential. The second set of theories in this category is called office-seeking. The office-seeking models postulate that key actors aim at maximizing their share in parliament regardless of the compatibility of goals with other parties. The last sub-category deals with personal-gain theories. Here, a party does not necessarily want to maximize a share in the legislative body. Benoit uses as an example the 1989 roundtable negotiations in Poland, in which the ruling Polish United Workers’ Party gave up on the issue of the free election of the senate in exchange for securing the tenure of presidency by General Jaruzelski.