Unlike the plurality system, the proportional list system provides a wealth of information at constituency level only, which results in the breaking of the link between the local and national levels and, therefore, in a weak influence of the government - opposition dichtomy on the local situation.
As far as the practical consequences, the plurality system presents us with two tendencies within a bipolar situation at the level of single-member constituencies leading to quasi-bipartisan situations at the upper level (generally speaking, the national level). The regular nature of the behavioral patterns of plurality systems gives way, in proportional list systems, to a considerable type of situations that make possible all forms of multiparty systems.
If we take our theoretical model consisting of a multiparty parliamentary system with six parties we conclude that both electoral systems evolve in a different way. On the one hand, the plurality system leads to a quasi-bipartisan system because of the close link (dependency) between the situation at the constituency level and that of government - opposition in Parliament. On the other hand, the proportional list system allows for all types of multiparty systems because of the variable quantity of information coming from both levels (constituency, Parliament). The original six party system can develop in three different ways: toward a six-plus party system, toward a bipartisan system or remain stable in its initial six party form.
ote: even if the parliamentary strength of parties is the same, their nature will be different according to the electoral system used since their electoral life expectancy and development are based on totally different information.
Introduction
The influence an electoral system has on the political life of a nation has given rise to debate since the middle half of the last century. There are two schools of thought with opposing views since the end of the century. The first one believes that electoral systems only act mathematically in the transfer of votes into seats and on the basis of proportionality only; for the second, electoral systems are a determining factor in the way the votes are directed towards the different political parties. The latter school of thought has a difficult task at hand because of the large number of factors that appear to intervene before a voter chooses his party or candidate.
Our approach is theoretical and builds upon patterns of behavior. The electoral process is first considered in itself, in an abstract manner and is then brought into the political contest of a legislative election. Ideology, programs and personality of candidates will not be considered in this paper. Neither will the different groups in which can be classified the electorate: religious, ethnical or organized labor.
The voter's behavior will be considered in the following way:
° voter's evaluation of the candidates in different types of contests,
° evaluation of the governmental or opposition tendencies divided into a certain number of parties.
The act of voting is considered as belonging to political action in its universal acceptance: that of retaining power and governing.
The research goes from the most simple case to the more complex: an election with no incumbent considered by itself to an election in its national political setting, constituency and national political environment. Parties asre then introduced. Candidates are party members in or out of Parliament andt he national and local influences are considered.
Many electoral systems have been invented and experimented throughout the eighteenth to twenty first century. Two of these have been chosen as the most representative: the plurality system and the list proportional system. The single-member relative majority systems have been used for long time periods by many Anglo -American countries and diverse types of list proportional systems have been largely used in many European countries since 1918.
Comparing two electoral systems based on very different concepts and techniques is not an easy task. All has been done to choose examples that fit into both electoral systems: the number of candidates, of parties.
The basic examples are moderate multiparty systems: six parties divided in support or opposition to government. The aim of this paper is to see the evolution of these six parties, for example: will there be a multiplying effect? Will there be stability? Will there be a reduction to five; four, three, or two parties?
- Plurality system (first-past-the-post).
- Isolated election without incumbent.
Let us consider the plurality election of the chairman of the board of a newly created association (non political). Les us now admit that the members (voters) of this association do not know the candidates.
Nothing at this point favors anyone of the candidates. Do the candidates have equivalent chances? In the case of five candidates there is a five-pole straight contest for only one seat.
- Isolated election with incumbent.
Let us now take that same association a year later at election time. The chairman, who has just finished his term of office, is one of the canidates. The contest is not centered anymore on the designation of anyone of the candidates to the chairman's seat but takes into account the incumbent chairman himself. The election is no more the designation of anyone of the candidates but, in the first place, a vote for or against the incumbent.
Being the incumbent is not always a privilege. The incumbent may have been a chairman but his term of office may not have been a success. This may have given rise to a negative appreciation of a part of the members but, whatever happened, the incumbent is never just another candidate.
The election is no more a five-pole contest but a dichotomy: vote for or against he incumbent. The election is now a two pole non symmetrical contest.
- Plurality election becoming the constituency of a parliamentary election.
The election we have considered has now become a single-member constituency in a parliamentary election. The chairman is now the incumbent member of Parliament.
In order to clarify our analysis, we will divide the national political life in two levels: the first level (central level) corresponds to the Parliament and the government. The local level is made up of the constituencies from which are elected the members of Parliament.
At the central level we have: the Parliament consisting of several hundred members who, when considered as individual members, do not have to be classified in a special way. The government depends or usually depends (in parliamentary systems) on the support of an absolute majority of Parliament.
The local level consists of national constituencies where there are two-pole symmetrical contests.
Let us now consider the central level. When parties are not taken into account, nothing in Parliament (considered by itself) enables the classification of the members:Things are different when Parliament is inserted into the political arena: support or opposition to government action, the legislative process. Support or opposition to government proposals can only be considered within the framework of decision-making and has to take place on the basis of a majority "for or against". We have thus an active dichotomy: for the proposal, against the proposal plus the bstentions.
Constitutions give different meanings to abstention in Parliament. It can either be counted or not in the case of a governmental confidence vote. When a government does not enjoy majority support in Parliament in order to obtain enough votes for passing legislation it seeks parties that do not have members in the government. For major legislation a vote against the government can be perceived as a vote of non confidence which will cause the government to fall. On minor issues it will be of no importance. In this paper we will then assume that abstention is similar to a vote against the government.
The two political levels are:
Central level (Parliament, vote on proposals), two pole non-symmetrical contest.
Local level(election of the members of Parliament multiplied by the number of constituencies): a two-pole non symmetrial contest.
The incumbent can support the government or the opposition in the following manner: larger number of incumbents giving government support rather than the opossition.
Comparing the two levels gives the following results: two-pole different types of contest.
- At the central level the contest is in Parliament between two forces composed similarly of a certain number of members of Parliament (MPs). Both poles of this contest are made up of multiple units although in unequal numbers (members supporting the government > members of the opposition).
- At the local level -(constituencies) the two-pole contests are non symmetrical between a one-unit pole and a multiple unit pole (the incumbent versus other candidates).
The two levels differ in a numerical fashion:
- First level: one Parliament, one government, one legislative process at a given time.
- Local level: multiple constituencies (several hundred).
° Difference in duration at the two levels:
- In Parliament there is a differing degree of political action continuity.an absolute continuity of government action coupled with long periods of parliamentry action.
- In constituencies the election occupies only a brief period of time and each election is separated by long intervals (years). The only continuity is in the local work of the elected member of Parliament.
From these comparisons we gather that the constituency level offers weaknesses in comparison to what occurs at the parliamentary level where we find greater continuity of unicity in the different branches haviing to do with elections or their direct consequences (Parliament and government).
Let us now consider, in the general context, the nature of the non-symmetrical dichotomy at constituency level. Of the two competing sides, the weakest is that of the candidates contesting the incumbent; this for two reasons: firstly, their number has no limits. The more numerous the candidates the less chance of them nearing the number of votes the incumbent will obtain. Secondly, the candidate contesting an incumbent can be a first time candidate whereas the incumbent has worked in the constituency during the whole of the ending legislature.
The incumbent can represent government support or its opposition. It then appears that the candidates on the same side as the incumbent lose their electoral credibility and that the essential contest lies between the incumbent and the candidates on the opposite side of the dichotomy.
The first case (gov support) puts the incumbent in a favorable position as he cumulates the advantage of being alone against the opposition (as far as electoral credibility is concerned). He also carries with him part of the legitimate political power that goes with government and takes advantage of belonging to a parliamentary majority.
The second case (gov opposition) reduces the advantage of the incumbent as he only belongs to the smallest and less organized part of Parliament; the opposition does not, in itslef, give political prestige.
One the one hand, we now have two non-symmetrical contests but with differing degrees of symetry:-
- High degree ... incumbent supporting government.
- Low degree ... incumbent in opposition
On the other hand, we now have only two out of the six candidates who have some degree of electoral credibility, and that, in all cases.
- List proportional system.
- Isolated election with no incumbent.
We will now procede in a similar way as we did previously for plurality voting. The basic list proportional system differs from plurality by its multi-member constituencies.
Let us take the election of some board in a newly created firm in which the members (voters) either know the candidates or the organizations presenting the lists. Let us admit that five seats on the board are to be filled and that four lists have five candidates each.
The seats are distributed to the candidates according to their order on the lists so that the candidates do not have equal chances of being elected. The voter chooses one of the lists and the electoral system and the technique it implied does the rest. This means that lists generally consist of three types of candidates: privileged candidates, on the balance candidates and candidates having no hope of being elected.
Two types of situations can be envisaged:
- Lists that can pretend getting one or all candidates elected. This can be found in two ways: by presenting lists with less candidates than are to be elected in the constituency or by obtaining almost all the votes in the constituency (which is unlikely).
The usual situation found is lists with the same number of candidates as there are members to be elected and a certain equilibrium of votes among lists. Out of five lists, one has the assurance of being elected, another list might be elected, two have little chance of being elected and one, no chance.
The voter will vote for a list but, at this stage, has no reason to favor one list more than another.
- Election (isolated) with incumbents.
Let us now consider the same election but with incumbents, for example: five incumbents and five seats to be filled. Four lists are competing and three of those lists have incumbents.
The dichotomy incumbent - candidates of the other tendency (that we found in majority electoral systems) cannot be present here because of the large number of incumbents. The example, we are working with three lists with incumbents and one without but many other possibilities could have been considered and it is for this very reason that no consistent type of contest can be envisaged at constituency level. A dichotomy could exist in the case where only two lists have incumbents but this is not the general rule.
A semi-dichotomy appears in the following example: two lists with an equal number of incumbents and two other lists with no incumbents. This two-list type of contest will probably occur in small constituencies (few members to elect) is just one of the multiple number of possibilities.
The basic electoral unit, considered by itself, has the following characteristics: the voter perceives the electoral contest differently according to the number of candidates to be elected and the way the incumbents are distributed among the different lists. One has to keep in mind the probable re-election of the incumbent heading their list.
- Basic list proportional election seen as a constituency of a parliamentary election.
We now follow the same procedure as in the plurality (majority) type of election by bringing the isolated election into a national election where it will be one of the constituencies.
We still have a dichotomy at the parliamentary level but there is no guarantee of finding one at the constituency level. The presence of a dichotomy depends on variable circumstances which are not dealt with in this paper.
Let us now consider what effect the dichotomy at the parliamentary level can have on the dichotomy of cases which will occur at theconstituency level. We will envisage the effect on the lists themselves, then on the incumbents.
The effect on the lists
Without taking into account the incumbents, the lists can be divided in support for government and in opposition to government.
We find then a dichotomy but whereas that obtained in the case of plurality (majority voting) lead to simplifying the perception of the electoral contest, in this case we only find reproduced the two-pole situation of the central level (Parliament)without a reducing (or simplifying) factor for the following reasons:
° the local level reproduces the positions of the central level dividing the lists into governmental support or opposition;
° the voter does not find this choice made easier as each of the two sides can be made up of a certain number of lists.
The effect on incumbents.
The closest we can get to obtain siilar conditions to those found in the case of plurality (majority) is that of a constituency of one or two lists with incumbents and one or more lists without any. Let us suppose that of the two lists with incumbents one supports government and the other is in opposition (the most probable case). The local (constituency) dichotomy of two lists with incumbents is of the same type of the central level and reinforces the electoral credibility (their political significance) of the two lists with incumbents. This is, of course, detrimental to the list without incumbents.
The reproduction at constituency level of the government-opposition dichotomy found in Parliament is possible while impossible in the case of plurality (majority) elections. This is a consequence of multi-member constituencies were incumbents of the two tendencies (government support - opposition) are one against he other in the same electoral unit. This is, of course, impossible in one-member constituencies of plurality systems.
Other probable cases are more complex:
- Incumbents in three of the four lists.
We will imagine four alternative cases of incumbents favoring or in opposition to government. For example: Two lists favor the government and two opposeit.
The presence of two lists of any of the two political tendencies with incumbents can be an advantage considered as a cumulative result (per tendency). Nevertheless, it must be borne in mind that nothing will designate the list to the voter's intention (except for the larger number of incumbents on one of the lists) and the voter's choice is in no way simplified.
We have, until now, considered the voter choosing one of the lists of his constituency. Let us now take the case of the voter taking into account the position of the incumbent in relation to the government or the opposition.
The political significance of an incumbent at pthe level of arliament in his support or opposition to government:The presence of more than one list with incumbents of both political tendencies complicates the voter's choice, who first has to choose between the government or the opposition and then choose one of the lists of the tendency he has first chosen.
The voter does not obtain sufficient information at the central level in order to make the final choice but has to consider the local lists themselves that can be examined from an angle other than their national political attitude.
- Incumbents on all of the lists.
The case previously considered had only two lists with incumbents (one list supporting government, the other in opposition) which gave those lists an advantage over those with no incumbents. In the present case, the voter's choice becomes more difficult as he has to choose between two or more lists of the same political tendency. The advantage of the lists with incumbents will decrease as their number increases.
A large number of incumbents spread over a large number of lists will tend to multiply the contests thus moving away from the two-pole contest at the central level. The election at constituency level will movce away from the national political contest which is: government support or°opposition to government.
- Introducing the Political Parties.
Until now, we have dealt with candidates, incumbents, members of Parliament, governments, opposition and lists. The lists,which are a must in list proportional systems, introduce political parties. The notion of government - opposition must be translated into parties in order to become a practical reality. A government can be but is not always composed of only one party, it can be a coalition government. A government generally has the support of an absolute majority of Parliament. The opposition is rarely composed of one party.
- Plurality (majority) election: introducing the parties.
Let us suppose that government suport and the opposition are both divided into three parties each. As long as we have considered the abstract notion of government - opposition, candidates or lists, there was no reason to introduce a scale for different elements of our model except for the numerical dominance of the parliamentary support of government over the opposition. The introduction of parties brings with it the need to determine their relative value or strength. It is probable that the six parties in Parliament will have different numbers of members of Parliament. It is possible that two may have the same numbers, three is less probable. The total of the three governmental parties is larger than the total of the opposition.
We thus have two groups of three parties each made up of an unequal number of members (although there remains the possibility of two parties being equal in numbers to members of different groups).
Let us admit that of those six parties each has a candidate in every constituency. We will suppose that one of those parties will present an incumbent in every constituency at a general election. All members of Parliament do not need to represent themselves but to simplify we will suppose that the members of our six parties are all incumbents at the following election.
So at in Parliament we will find three-party government support and three party opposition. In the constituency, there will be a sample of six constituencies, each constituency having an incumbent of a different party.
Before introducing political parties, the only differentiation of candidates was by separating the incumbents from other candidates and the political attitude of the candidates who either support the government or oppose it. The fact that the MPs in unequal members per party brings with it a possibility of differentiating the non incumbents. This happens to be the only way the voter, in our abstract political approach, can decide which candidate to choose against an incumbent.
We shall now reproduce at constituency level the indications of relative size that we have already seen in Parliament: At parliamentary level there are three parties that support the government against three others that are against the government. At the constituency level there is the example of six constituencies with incumbents from the six different parties where the level of support is not equal in each constituency. (The incumbents can be considered as being more electorally credible than the other candidates of the same tendency be they government or opposition.)
The link between the parliamentary power of the party is clearly perceived at constituency level where the party is represented by only one candidate be the incumbent or not. In other words, the one-member constituency gives a clear image of what happens at the parliamentary level that is political power (government) and main opposition.
As far as the number of candidates is concerned, we still have non symmetrical contests in favor of the incumbents . Nevertheless, now we have a way of giving an advantage to one of the other candidates in comparison to the others of the same tendency. It now appears that the electoral advantage of the tendency not having an incumbent is in favor of the candidate belonging to the most important party of this tendency (in most cases at Parliament level).
When we analyze the positions of the candidates we find that the incumbents and other candidates of one of the parties that suppports the government in Parliament are more numerous than the candidates and incumbents of the other parties that support the government. Considered from the perspective of the pro-government electorate, this party then appears to have more credibility every time the incumbent is from one of the opposition parties t. The pro-government party is then one of the parties in four out of six of the dichotomies. From a national point of view this means that his parliamentary pro government parrty is one of the two main contestants in over half of the constituencies.
A similar argument can be advanced as regards the opposition party. The oppostion party will have incumbents in a number of constituencies that is not necessarily smaller than those wih pro-government part incumbents. It is as if one of the main pro-government parties is not always bigger (in terms of parliamentary representation) than one of the opposition parties, exept in the case where the government corresponds to only one party with an absolute majority in Parliament. We can now say that one of the two main parties (pro-government and opposition) are present as the main contestants in all six constituencies and that they are one against the other in two of six.
We have not, as yet, taken account of the time factor. The parliamentary and constituency levels were considered at a given moment. Let us consider the effect of a succession of elections on the evolution of party strength in Parliament. One the one hand, every seat lost by one of the pro governmentincumbents will go to one the opposition incumbents while every seat lost by one of the incumbents will go to a pro government acandidate. On the other hand, each seat lost by a pro government incumbent will go to an opposition incumbent and each seat lost by an opposition party will go to a pro government candidate. So that all changes (whatever changes occur) will move towards a two-party system.
We also have to take into account that seats will change hands fairly frequently because of the advantage in seats gained by the winner in votes found in majority systems (% seats > % votes). It then appears that the two-party system originating from the constituency will be reproduced at in parliamentary. The dichotomy government - opposition will then tend to become a one-party government, that one party having an absolute majority of seats in Parliament and an opposition party having most of the remaing seats.
Regional conditions can, at times, partially dim national issues and enable a purely regional party to intervene in the dichotomy of a number of constituencies. This transfer of electoral credibility can occur at the point where we found (before introducing political parties) a weakness in the dichotomy (the case of the incumbent supporting government action). This type of incumbent has a definite advantage over the other candidates for two reasons:
° being the incumbent implies presence at parliamentary and constituency lfor a number of years;
° government support brings with it the the legitimacy of governmental power.
This can have a psychological effect on the voter wishing to vote against the government. The estimated strength of the opposition partes is then somewhat blurred by the apparent power of the incumbent and local actors can then appear electorally credible. It is at this point that regional parties can, in certain circumstances, gain seats.
The switching of a national party for a regional party one could be brought about this way: The pro government incumbent lets us suppose that the opposition incumbent is the opposition candidate belonging to the most important party in Parliament while the one of the incumbents belongs to the party well know locally. At this point there is opposition between parties L and N at local level, the first being more powerful at national level (Parliament), the other at the local level.
There is a turning point that depends on the relative importance of national or regional topics in the electoral platform. It is at this critical point where the regional importance of a government party representative overrides the national and parliamentary mportance of an opposition party representative.
- List proportional systems: introducing political parties.
Let us consider what effect the different size of the parties in Parliament has on the voter's choice in list proportional systems. We have already considred government support and opposition as consisting of lists. These lists can now be seen as party lists; this simplifies our work in comparison to majority systems using single-member constituencies.
The relative size of the parties in Parliament intevened (in the case of majority elections) in the non symmetrical dichotomy contest at constituency level (for example: an incumbent against a number of candidates with no differentiating factors in the constituencies). The size of the parties in parliamentary level brought useful information to the constituencies (local level). Insufficient information at local level was made up by that of the parliamentary level. In proportional list systems, the information at local level is of a diffeent nature and much more abundant.
Let us now introduce the relative size of the parties (in the constituency and in Parliament) in our previous example (main electoral contest between the two lists with incumbents and four lists of marginal electoral credibility). We have six parties in Parliament. The are three pro govvernement parties and three oppostion parties. As preiously, let us admit that the three pro government partis oppose each other in the constituencies and oppose the opposition parties. The respective position of the parties is sufficiently clear atconstituency level to push into the background their relative representation at the Parliament level. As we consider the conditions in which the voter has to make his choice, this is a good thing as the information given in the constituencies is only partly similar to that given at the parliamentary level. The A party is the most important party giving government support at both levels whereas on the side of the opposition, the one party which is number one at local level is only number two at the parliamentary level. Had there been insufficient information at the local level, this difference of relative size between the two levels would have been aconfusing factor in the voter's choice.
Let us now consider another possibility where, at local level, a less clear picture of relative party size will push the voter to look for further information at the parliamentary level An example of the constituency with six parties each presenting a list. each having a different umber of incumbents. What will a voter do if he wants to support the government? Which of the three lists will have his vote? He probably will be more inclined to vote for one of the lists that have incumbents. The list that has three incumbents has an advantage over list that has only two. Unfortunately, this information found at the constituency level is not in conformity with that at the parliamentary level. The list that has more incombents at the constituency level and less Parliament members than the other party, which has more in Parliament than at consituency lrbel. Many examples of non conformity between the two levels can be imagined and the fact that list proportional representation has a number of incumbents in direct competition at constituency level implies that the information which a voter must have at local level in order to decide whether to vote for or against the government can or cannot be in conformity with party representation Parliament.
It now appears that in the case of list proportional systems, the introduction of the relative size of parties in Parliament and in the constituencies does not simplify the voter's choice. A possible contradiction inth e information coming from the parliamentary and constituency levels will tend to isolate political life at the local or constituency level from that of the central or parliamentary level.
A generalization of political influences throughout the two levels which was possible for a majority, one-memberconstituency system, is not possible for a proportional list system. List proportional systems give examples of information coming from the two levels that are in complete contradiction. From a voter's point of view, information from the first level is of no practical use, complicates his choice or merely confirms the information he has from the local level. This implies that a certain number of constituencies can lead an electoral life in complete isolation to what happens at the Parliament level.
Concluding remarks
Single-member constituency plurality system: The information available to a citizen wanting to choosebetween the different parties of one of the two tendencies: government opposition is given in a complementary manner at at Parliament and at constituency levels..
List proportional systems: The information available at the local level (constituency) is sufficient for the voter to choose his party with little intervention from the first level (national, Parliament and government - opposition).
Practical implications
One-member constituency majority systems lead to bipolar contests (with two variables) at local level which lead to a quasi two-party system at national level (parliamentary representation). Plurality systems lead to stable all-round electoral attitudes which give rise to similar predictable electoral representation.
The large number of possible contests in the multimember constituencies of list proportional representation does not lead to a uniform predictable pattern of representation. It must be underlined that no prediction of electoral representation can be made at this precise level of inquiry as far as list proportional systems are concerned.
Evolution of the original six-party model. As far as our six-party model is concerned, its evolutionary assumptions are:
° plurality systems: they lead to quasi two-party systems (interdependence of the two levels, constituency and Parliament) for electoral choice information.
° list proportional systems: they can give rise to to all types of multiparty systems as there are variable amounts of information. Some information comes from the parliamentary level but it is overshadowed by the information available at constituency leve. This can leave in the background the essential government-opposition perspective. The six party model can then, in the long run, change into a two-party model, a six plus party model or stay in its original form.
'Written in collaboration with Maria Rodriguez-McKey)