The Process of Change

The process through which an electoral system is designed or altered has a great effect on the type of the system which results, its appropriateness for the political situation, and the degree of legitimacy and popular support it will ultimately enjoy.

Electoral systems are very rarely designed on a blank slate where no precedents exist. Even the design efforts in Afghanistan and Iraq have historical multiparty competitive precedents to draw on (albeit distant in time and casting little light on what may work in the future). Some key questions of electoral system design are:

• Who designs? That is, who puts the idea of electoral system change onto the political agenda, and who has the responsibility for drawing up a proposed new or amended system and through what type of process?
• What are the mechanisms built into the political and legal framework for reform and amendment?
• What process of discussion and dialogue is necessary to ensure that a proposed new or amended system is accepted as legitimate? Once change has been decided upon, how is it implemented?

Sub-sections of this chapter:

• Who Designs?
• What are the Mechanisms for Reform and Amendment?
  
• Advice on Debate and Dialogue
• Advice on Implementation
• Assessing the Impact of Change
• Trends in Electoral System Reform
  

Who designs?

There are several ways by which electoral systems come into being.

• First, they can be inherited without significant alteration from colonial or occupying administrations (Malawi, Mali, the Solomon Islands and Palau being examples).
• Second, they can result from peace process negotiations between communal groups seeking to bring an end to division or war (e.g. Lesotho, South Africa and Lebanon). In these circumstances the electoral system choice may not be open to full public scrutiny or debate.
• Third, the system may be effectively imposed by the groups responsible for post conflict political reconstruction (e.g. the Coalition authorities in Iraq and the appointed Transitional National Council in Afghanistan).
• Fourth, elements of a previous authoritarian regime may have a strong role in designing a new electoral system during the period when they are being divested of power (as in Chile).
• Fifth, an expert commission may be set up to investigate the electoral system alone (as in the UK or Mauritius) or as part of the broader constitutional context (as in Fiji). This may lead to recommendations being put to a national referendum (as was the case in New Zealand) or to a legislative vote on the commission’s recommendations (as in Fiji).
• Sixth, citizens may be involved more widely in the design process by the establishment of a non-expert citizens’ assembly on the electoral system. This was the approach adopted by the Canadian province of British Columbia; it led to a recommendation for a change from FPTP to STV that would be put to a province-wide referendum for decision.

What are the Mechanisms for Reform and Amendment?

While electoral systems are an extremely important institution affecting the way in which a country’s system of government works, traditionally they have not been formally specified in constitutions, the highest source of law. In recent years, however, this has started to change.

Today, a number of countries have ‘embedded’ details about the electoral system in their constitution or in a separate schedule to the constitution. The significance of this for electoral reformers is that constitutionally entrenched laws are usually much harder to change than ordinary laws, usually requiring a special majority in the legislature, a national referendum abuses or some other confirmatory mechanism, which shields such systems from easy alteration.

For example, the South African constitution states that the electoral system for the National Assembly elections shall result ‘in general in proportionality’ and so reform options are limited to PR-type systems unless a constitutional amendment is made.

However, the details of the electoral system are still more often to be found in regular law and thus can be changed by a simple majority in the legislature. This may have the advantage of making the system more responsive to changes in public opinion and political needs, but it also contains the danger of majorities in a legislature unilaterally altering systems to give themselves political advantage.

The opportunities for reform rely on both the legal mechanisms for change and the political context within which calls for change are made. Not all movements for electoral system change are successful. Almost all recent examples of major change have occurred in one of two sets of circumstances.

The first is in the course of a transition to democracy or shortly afterwards, when the whole political framework is ‘up for grabs’.

The second is when there is a crisis of governance in an established democracy. Two examples are the perceived illegitimacy of two successive majority governments elected with fewer votes than their major opponents in New Zealand, and the perception that high levels of corruption in Italy and Japan were endemic to the political system rather than the results of the actions of particular individuals.

Even when there is huge popular distrust and dissatisfaction with the political system, change still needs to be agreed by the current holders of power. Political elites are only likely to act if they can see benefit to themselves from change or if they are frightened of the electoral consequences to themselves of failing to change. Even when convinced, they will, unsurprisingly and almost inevitably, seek to choose a system that maximizes the benefit to themselves. If they are unsure how this can be achieved or if different interests seek different solutions, negotiated compromises may be likely—perhaps involving mixed systems.

However, agreements and changes may not turn out to have the effects intended by their proponents or may produce other, unintended effects. In Mexico, reforms in 1994 designed by the governing party to make concessions to the opposition led to the most disproportional result in recent years. The cases of South Africa and Chile illustrate the fact that political realities and the desire of ruling parties to maintain their power and influence can be just as much a block on electoral system reform as legal hurdles. In South Africa there have been widespread calls for an element of local accountability to be built in to the closed-list PR system of large electoral districts under which elected representatives are perceived as detached from their electors. These were reinforced by the majority findings of a presidential commission which reported in January 2003, but the government shied away from changes that would reduce its control over candidate selection and caucus voting behaviour, and declined to entertain reform. In Chile, General Pinochet’s legacy was to rig the electoral system to advantage his allies. More than a decade after his removal from power, that system remains effectively unchanged.

In New Zealand, the use of referendums during the process of change resulted initially from a political move—an attempt by the leader of one major party to wrong-foot the other major party during a general election campaign. In the first referendum, the electorate was asked whether it wanted change at all and to indicate its preferred new system from four options. In the second, the chosen new system was pitted against the retention of the previous system. As a result, the new multi-member proportional system was adopted with a clear expression of public legitimacy.

Electoral systems will inevitably need to adapt over time if they are to respond adequately to new political, demographic and legislative trends and needs. However, once a system is in place, those who have benefited from it are likely to resist change. Without a transition or a major political crisis as catalyst, it appears that change at the margins may well be more likely than fundamental reform. In post-conflict transitions, this creates a tension between the practical constraints that may affect the implementation of elections driven for example by the political imperatives of a peace agreement, and the desirability of getting the system right at the beginning. To try to engineer improvements within existing systems, reformers may consider changing district magnitude, threshold levels or quota formulae. Many significant reforms proposed in the past few years have involved adding a List PR element on to an existing FPTP system to create a mixed, more proportional system (e.g. the changes enacted in Lesotho and Thailand).

Advice on Debate and Dialogue

It is the task of reformers not only to understand the legal form of the technical arguments for and the implications of potential change but also to understand and be able to explain the political arguments and the implications for the wider political framework of the country. Significant voices in civil society, academia, and the media may contribute to developing a public perception that change is necessary. But a sufficient number of those in power will need to be convinced of the benefits, including the benefits to them.

Even with the current increased interest in electoral systems, the number of people, both in elite circles and in society generally, who understand the likely impact of changes may be very limited. This is further complicated by the fact that the operation of electoral systems in practice may be heavily dependent on apparently minor points of detail. Reformers may need not only to fully work through and explain the legal detail that would be necessary to implement change, but also to make technical projections and simulations (often using data from previous elections) to show, for example, the shape and implications of proposals on electoral districts or the potential impact on the representation of political parties. Technical simulations can also be used to ensure that all contingencies are covered and to evaluate apparently unlikely outcomes: it is better to answer questions while change is being promoted than in the middle of a crisis later!

Of course, not all effects of a new electoral system will be known in advance. In fact, designing an electoral system when it is unclear what the results of an election held under it will be in terms of actual vote and seat distribution may even be helpful when it comes to arriving at a system that will be seen to be the most fair to all actors involved.

As much as possible of the technical details and mechanical effects should however be known to all and understood in advance of a change. Voter involvement programmes, for example, inviting members of the public to participate in mock elections under a potential new system, may attract media attention and increase familiarity with proposals for change. They may also help to identify the problems—for example, voter difficulty with ballot papers—which a new system may generate.

Advice on Implementation

Voters, election administrators, politicians and commentators all tend to be comfortable with what is familiar. Years of use may have smoothed the rough edges of established systems. A new system can thus be a leap into the unknown, and problems in implementation can arise from its unfamiliarity. This cannot be avoided completely, and the planners of change cannot sit back when legislative changes are in place. A process of change is complete only with intensive voter education programmes to explain to all participants how the new system works and with the design and agreement of user-friendly implementing regulations.

The most effective voter education—and election administrator education—takes time and its therefore important to start early. Unfortunately, in reality time is often in short supply to an electoral management body (EMB) organizing an election under a new system, but it need not to be if good planning is applied. All good negotiators use time pressure before a final agreement is reached, and this is particularly true when the new system is the product of hard negotiation between political actors. An effective EMB will nonetheless prepare as much as possible as early as possible.

Assessing the Impact of Change

Having discussed the process of change in some depth, a word of caution is needed. Because electoral systems have psychological as well as mechanical effects, the long-term effect of changes may take some time to work through. Parties, candidates and voters may take two or three elections to fully observe and respond to the effects and incentives of particular changes. The tendency towards mixed systems may accentuate this, as the overall effect on candidates and voters of mixed incentives may be less clear.

Judgement may be necessary as to whether problems in a new or amended electoral system are merely transitional or whether they show that the system is fundamentally flawed and requires urgent amendment or replacement. In the aftermath of George Speight’s 2000 coup, such a debate is currently taking place in Fiji: will the Alternative Vote settle down so that parties and voters respond to incentives for inter-ethnic moderation, or does the course of events since its adoption in 1997 indicate that it is fundamentally unsound in the Fijian context?

Trends in Electoral System Reform

The Italian referendum in 1993, leading to a change to a Mixed Member Proportional System for the elections the following year, marked the beginning of a series of significant changes in electoral systems all over the world. In the vast majority of the cases, changes have been made on the margins, with a new seat allocation formula, a new number of electoral districts, or an extra few appointed members in the legislature; but many countries have also gone through reform processes that have altered their electoral system completely.

As Table 1 shows, the trend is rather clear. Most countries that have changed electoral systems have done so in the direction of more proportionality, either by adding a PR element to a plurality system (making it a Parallel or MMP system) or by completely replacing their old system with List PR. The most common switch has been from a plurality/majority system to a mixed system, and there is not one example of a change in the opposite direction. The new plurality/majority systems all come from within the same family except for the case of Madagascar, which moved from a List PR system, not to a pure plurality/majority system, but to a hybrid where the FPTP share is larger than the List PR share.