Electoral Systems, Institutional Framework & Governance

Electoral systems have long been considered to have specific effects on issues of governance, policy making and political stability. Different electoral systems have marked implications for governance in parliamentary systems. In particular, there is an inbuilt tension between electoral systems which maximize the potential for one-party government (e.g. plurality/majority systems) and those which make multiparty coalitions more likely (e.g. proportional systems). Both constellations have clear policy impacts: single-party government makes decisive policy making and clarity of responsibility much easier, while coalitions are more likely to produce more representative policies and more inclusive decision making. Similarly, major shifts in government policy are easier to achieve under single-party government, while coalitions are more likely to see issues discussed and debated before any changes are made.

Almost all countries which have a presidential or semi-presidential constitution elect the president directly. In addition, some republics which have parliamentary constitutions nevertheless elect their head of state directly. In presidential systems, the extent to which an elected president can claim a popular mandate and legitimacy depends significantly on the means by which he or she is elected. Presidents who have clear majority support are likely to have much greater legitimacy and be in a stronger position to push their own policy agenda than those elected on a small plurality of the vote. This has an important impact on relations between the president and the legislature. A president elected by a clear absolute majority of the population can command a great deal of legitimacy in any conflict with the legislature. By contrast, Salvador Allende’s election in Chile in 1970 on 36 per cent of the vote, and opposed by a right-wing Congress, helped create the conditions for the 1973 military coup.

The relationship between the legislature and the executive differs between parliamentary, semi-presidential, and presidential systems. In a presidential or semi-presidential system, the president’s position does not depend on maintaining the confidence of the legislature: such a president cannot be removed from office on purely policy grounds. However, experience in Latin America in particular indicates that a directly elected president without a substantial block of support in the legislature will find successful government difficult. In presidential and semi-presidential democracies, the electoral systems for the presidency and the legislature therefore need to be considered together, although the different roles of the president and the legislature bring different factors into play in making the two choices of system. The synchronization or otherwise of the presidential and legislative elections and the provisions which may encourage or discourage fragmentation of parties and the relationship between parties and elected members should be considered at the same time.

Sub-sections of this chapter:

• Electing a President
• Electing an Upper House
• Different Tiers of Governance
• Electing Supranational Bodies
• Electing Federal/State Assemblies and Autonomous Jurisdictions
• Electing Local Authorities
  
• Electoral Systems and Political Parties
• Direct Democracy Options
  

Electing a President

In principle, any of the single-member district systems can be used for the direct election of a president. When a president is to be elected as the executive head of state, there is often a strong normative and practical preference for systems which ensure a victory by an absolute majority. The majority of all countries that have direct presidential elections use a Two-Round system.

The separation of the two rounds leads to efforts by the leading candidates to attract second-round support and endorsement from those eliminated after the first round. Such agreements are sometimes driven primarily by the desire for victory. They are thus perhaps less likely to reflect compatibility of policies and programmes than are pre-poll preference-swapping agreements reached between candidates in preferential systems with a single polling day. In addition, presidential elections held under TRS increase the cost of elections and the resources needed to run them, and the drop-off in turnout between the first and second rounds of voting can often be severe and damaging. For this reason, other options such as the Alternative Vote and the Supplementary Vote are increasingly being examined.

Electing a President - First Past The Post

The most straightforward way of electing a president is to simply award the office to the candidate who wins a plurality of the votes, even if this is less than an absolute majority. This is the case for presidential elections in Bosnia and Herzegovina, Cameroon, the Comoros Islands, Equatorial Guinea, Guyana, Honduras, Iceland, Kiribati, South Korea, Malawi, Mexico, Palestine, Panama, Paraguay, the Philippines, Rwanda, Singapore, Taiwan, Tunisia, Venezuela, and Zambia. Clearly, such a system is simple, cheap, and efficient, but in a strongly competitive multi-candidate contest, it leaves open the possibility that the president will be elected with so few votes that he or she is not seen as the choice of a substantial majority of the electorate—and indeed may specifically be opposed by a substantial majority: the majority voted against him or her. Examples include Venezuela in 1993, when Rafael Caldera won the presidency with 30.5 per cent of the popular vote, and the May 1992 election in the Philippines, when Fidel Ramos was elected from a seven-candidate field with only 24 per cent of the popular vote. Taiwan experienced a major political shift in 2000 when the challenger Chen Shuibian won the presidency with just 39 per cent of the vote, less than 3 per cent ahead of the next candidate.

The United States is unique in conducting its national presidential election by FPTP at federal state level. The FPTP winner in each federal state gains all the votes of that state in an electoral college, with two exceptions, Maine and Nebraska, where the votes of the state are allocated two (corresponding to the state’s two Senate seats) to the FPTP winner state-wide, and one to the FPTP winner of each individual congressional district in the state. The Electoral College then elects the president by absolute majority. This can lead to a situation in which the winning candidate polls fewer votes than the runner-up—as in 2000 when the Republican candidate George W. Bush won despite polling some half a million fewer votes than the Democrat candidate, Al Gore.

Electing a President - Two-Round Systems

As in legislative elections, one way to avoid candidates being elected with only a small proportion of the popular vote is to hold a second ballot if no one candidate wins an absolute majority on the first round. This can either be between the top two candidates (majority run-off) or between more than two candidates (majority-plurality). France, most Latin American countries, all the five post-Soviet Central Asian republics, and many countries in francophone Africa use TRS to elect their presidents. Elsewhere in Africa, the system is used by Angola, Cape Verde, Gambia, Ghana, Guinea-Bissau, Kenya, Liberia, Mozambique, Namibia, Nigeria, São Tomé and Principe, the Seychelles, Sierra Leone, Sudan, Tanzania, Uganda, and Zimbabwe; in Europe, apart from France, it is used by Armenia, Azerbaijan, Austria, Belarus, Bulgaria, Croatia, Cyprus, Finland, Georgia, Lithuania, Macedonia, Poland, Portugal, Romania, Russia, Slovakia, Slovenia, and Ukraine; and it is found in Afghanistan, Haiti, Indonesia, Iran, Timor-Leste, and Yemen.

There are a number of adaptations to straight majority run-off and majority-plurality rules. In Costa Rica, a candidate can win on the first round with only 40 per cent of the vote; conversely, in Sierra Leone, a second round is only avoided if one candidate gets 55 per cent in the first. In Argentina, a successful candidate must poll 45 per cent, or 40 per cent plus a lead of more than 10 per cent over the second-placed candidate. A similar 40 per cent threshold with a 10 per cent margin exists in Ecuador.

A number of countries also have minimum turnout rates for their presidential elections, typically 50 per cent, as is the case in Russia and many of the former Soviet republics; this is an additional mechanism for ensuring the legitimacy of the result but has substantial cost and logistical implications if the minimum turnout is not met and the election has to be re-run.

Apart from those countries where parties could create winning pre-election alliances so that presidential candidates could be elected in the first round (e.g. in Brazil in 1994 and Chile in 1989 and 1994), the experience of TRS has appeared problematic in Latin America. For example, in the 1990 elections in Peru, Alberto Fujimori obtained 56 per cent of the votes in the second round, but his party won only 14 of 60 seats in the Senate and 33 seats of 180 in the Chamber of Deputies. In Brazil in 1989, Fernando Collor de Mello was elected in the second round with just under half of the votes, but his party won, in non-concurrent legislative elections, only three of the 75 Senate seats and only 40 of 503 seats in the Chamber of Deputies.

The problems of governance which have resulted demonstrate the importance of considering interlinked institutional provisions together. Although TRS produced presidents who had the second-round support of a majority of the electorate, it existed alongside systems for election to the legislature which did not guarantee those presidents significant legislative support.

While the successful candidates gathered the support of other parties between the first and second rounds, there was little to enable them to keep that support in place after the elections.

Electing a President - Preferential Voting

One way of getting around the disadvantages of TRS is to merge the first and second round into one election. There are several ways of doing this. AV is one obvious solution; it is used to elect the president of the Republic of Ireland. A lower-placed candidate who picks up many second-preference votes can overtake higher-placed candidates. The most recent example of a president winning through the transfer of preferences in this manner was the 1990 election of Mary Robinson to the Irish presidency.

A second possibility is the preferential system used for presidential elections in Sri Lanka and for London mayoral elections, known as the Supplementary Vote. Voters are asked to mark not only their first-choice candidate but also their second (and, in Sri Lanka, their third) choices. The way in which this is done differs: in Sri Lanka, voters are asked to place the numbers ‘1’, ‘2’ and ‘3’ next to the names of the candidates, in the same manner as under AV and STV. In London, no numbers are required; the ballot paper contains two columns, for a first-choice vote and a second-choice vote, respectively. Voters are asked to mark their first-choice and second-choice candidates accordingly. This means that voters do not have to write in any numbers themselves.

Counting is the same in both cases: if a candidate gains an absolute majority of first-preference votes, he or she is immediately declared elected. However, if no candidate gains an absolute majority, all candidates other than the top two are eliminated and their second- (or, in Sri Lanka, second- and third-) choice votes are passed on to one or the other of the two leading candidates, according to the preference ordering marked. Whoever achieves the highest number of votes at the end of this process is declared elected.

The disadvantages of the Supplementary Vote system include its additional complexity and the fact that voters are effectively required to guess who the top two candidates will be in order to make full use of their vote. Despite these differences, both AV and the Supplementary Vote have the same core aim: to make sure that whoever wins the election will have the support of an absolute majority of the electorate. The use of preference votes to express a second choice means that a second round of voting is not required, and this results in significant cost savings as well as benefits in administrative, logistics, and security terms.

Electing a President - Distribution Requirements

A few countries combine their presidential elections with a so-called distribution requirement, which requires candidates to gain a regional spread of votes, in addition to an absolute majority, before they can be declared duly elected. In Indonesia, which held its first direct presidential elections in 2004, a successful presidential and vice-presidential candidate team needed to gain an absolute majority of the national vote and at least 20 per cent of the vote in over half of all provinces to avoid a second round of voting. This requirement was inspired by Nigeria, another large and regionally diverse country, where presidential candidates need not only to win an absolute majority of the vote nationally but also to secure at least one-third of the vote in at least two-thirds of the country’s provinces.

Distribution requirements do have the benefit of encouraging presidential candidates to make appeals outside their own regional or ethnic base, and if appropriately applied can work very well. However, the specification of two requirements for victory always carries the possibility that no candidate will fulfil both. It is important that designers note this possibility and include provisions to resolve it, because a system which produces no winner and no method of finding a winner could create a vacuum of power fraught with the dangers of instability. The second round in Indonesia merely requires a simple majority for the winner to be declared elected, but Nigeria retains the distribution requirement for the second round too, which creates the possibility of a third round. If this were to take place in practice, it could have implications both for the length of the election period and for the financial and administrative resources required. Distribution requirements introduce strategic imperatives for candidates. In Kenya, to be elected president, a candidate has to receive a plurality overall and at least 25 per cent of the vote in at least five out of the eight provinces. Even so, throughout the 1990s, a divided opposition allowed Daniel Arap Moi to remain president with less than an absolute majority of the vote. Disribution requirements can also be built into nomination requirements.

Electing an Upper House

Not all legislatures consist only of one chamber; particularly in larger countries, many are bicameral. Around the world, about two-thirds of all countries have unicameral legislatures, while the remaining one-third have some kind of second chamber.

Most second chambers (often called upper houses or senates) exist for one or both of two reasons. The first is to provide a different type of representation or represent different interests, most often the regions or provinces of a country. The second is to act as a ‘house of review’, to provide a brake or delay against impetuous decisions in a lower chamber. The powers of upper houses are often less than those of lower chambers, especially when they are chambers of review.

The structures of these vary widely, but in general the most common use of second chambers is in federal systems to represent the constituent units of the federation. For example, the states in the USA, Brazil, and Australia, the Länder in Germany, and the provinces in South Africa are all separately represented in an upper house. Typically, this involves a weighting in favour of the smaller states or provinces, as there tends to be an assumption of equality of representation between them. In addition, many second chambers feature staggered elections: half the chamber is elected every three years in Australia and Japan; one-third of the chamber is elected every second year in the USA and India, and so on.

Some countries whose upper houses are ‘houses of review’ place special restrictions on them. In Thailand, for example, the Senate is now elected, but senators are prohibited from belonging to a political party or campaigning for election. A less common type of alternative representation is the deliberate use of the second chamber to represent particular ethnic, linguistic, religious, or cultural groups. A second chamber may also deliberately contain representatives of civil society. In Malawi, for instance, the constitution provides for 32 of the 80 senators to be chosen by elected senators from a list of candidates nominated by social ‘interest groups’. These groups are identified as women’s organizations, the disabled, health and education groups, the business and farming sectors, the trade unions, eminent members of society, and religious leaders. The much-maligned British House of Lords is occasionally defended on the grounds that it contains individuals with specific policy expertise who can check government legislation drawn up by generalist politicians. Similarly, second chambers in countries like Fiji and Botswana are used to represent traditional chiefs, although these are appointed in the first case and elected in the second.

Because of these variations, many second chambers are partly elected, indirectly elected or unelected. Of those that are elected, most jurisdictions have chosen to reflect the different roles of the two houses by using different electoral systems for the upper house and the lower house. In Australia, for example, the lower house is elected by a majoritarian system (AV) while the upper house, which represents the various states, is elected using a proportional system (STV). This has meant that minority interests which would normally not be able to win election to the lower house still have a chance of gaining election, in the context of state representation, in the upper house. In Indonesia, the lower house is elected by List PR, while the upper house uses SNTV to elect four representatives from each province. In Colombia, while both houses are elected by PR, the Senate is elected from one nationwide district, thus making it more likely that small parties and minority interests will be represented in that chamber.

Different Tiers of Governance

As noted above, the requirements for designing an electoral system vary depending on the type of body to be elected and its function and powers. When a body is designed to serve supranational, provincial, or local interests, the considerations involved in the choice of system are different from those involved when designing national legislative bodies.

Electing Supranational Bodies

Supranational bodies with significant decision-making power encompassing a number of countries, such as the European Parliament, are as yet a rarity but may become more commonplace with the globalization of politics and the aggregation of interests at a regional level. The EU has adopted, and now made effective, a requirement for all member states to adopt a proportional system for elections to the European Parliament: 23 member states use List PR, and two (the Republic of Ireland and Malta) use STV. Seats are allocated to member states not purely in proportion to population but by a tiered system which gives equal numbers of representatives to countries of approximately equal size but also over-represents smaller countries.

The designers of such systems give greater priority to choosing systems which produce regional and partisan balance than to localized geographical representation. The European Parliament has 732 MEPs (Members of the European Parliament) representing over 500 million people, which makes small district connections between voter and representative impossible.

Electing Federal/State Assemblies and Autonomous Jurisdictions

The legislatures of regions or states within a federation may use the same electoral system as the national legislature (symmetry), as happens in South Africa’s closed-list PR system, or they may use different systems (asymmetry), as in the UK, where the Scottish Parliament and the National Assembly for Wales are elected by MMP and the national legislature is elected by FPTP. The system for a state legislature may give primacy to the inclusion of minority groups within its borders or balance between urban and rural interests. The more autonomy a region has, the less the pressure for its electoral arrangements to mirror those of other states or provinces. The very fact of its being an autonomous jurisdiction implies that its attributes and needs are quite distinct from those of other areas.

Electing Local Authorities

Any of the electoral systems outlined under this topic area can be used at the local or municipal government level, but often there are a number of special considerations arising from the particular role of local government. In particular, because local government is more about the ‘nuts and bolts’ issues of everyday life, geographical representation is more often given primacy.

Single-member districts can be used to give every neighbourhood a say in local affairs, especially where political parties are weak or non-existent. Where these districts are small, they are usually highly homogeneous. This is sometimes seen to be a good thing, but if diversity within a local government district is required, the ‘spokes of a wheel’ principle of districting can be applied. Here, district boundaries are not circles drawn around identifiable neighbourhoods but are segments of a circle centring on the city centre and ending in the suburbs. This means that one district includes both the urban and the suburban voters, and makes for a mix of economic class and ethnicity.

In contrast, the municipalities in some countries which use PR systems for local government have one single-list PR district which can proportionally reflect all the different political opinions in the municipality. In order to achieve this, however, specific space may need to be made for representatives of local associations who are not driven by party-political ideology to nominate lists, and perhaps also for independents to be nominated as single-person lists.

It is also true that the choice of local electoral system may be made as part of a compromise involving the system for the national legislature. For example, in some newly democratizing countries such as Congo (Brazzaville) and Mali, tradition and the French influence have resulted in a Two-Round System for the national legislature, while a desire to be inclusive and more fully reflect regional and ethnic loyalties resulted in the choice of PR for municipal elections.

The debate between parliamentarism and presidentialism in national constitutions has a counterpart in discussion of the structure of local government. Directly elected governors and mayors who head executive authorities that are separate from the elected local legislative body are becoming more popular worldwide, at the expense of elected authorities with collective committee structures directly responsible for services. The range of systems available for electing governors and mayors is in principle the same as that for the direct election of presidents, and parallels may also be drawn when considering the issues surrounding the relationship between the electoral system and the legislative–executive relationship at local level.

Electoral Systems and Political Parties

Different kinds of electoral system are likely to encourage different kinds of party organization and party system. While it is important for party systems to be as representative as possible, most experts favour systems which encourage the development of parties based on broad political values and ideologies and specific policy programmes, rather than narrow ethnic, racial, or regional concerns. As well as reducing the threat of societal conflict, parties which are based on these broad ‘crosscutting cleavages’ are more likely to reflect national opinion than those that are based predominantly on sectarian or regional concerns.

Highly centralized political systems using closed-list PR are the most likely to encourage strong party organizations; conversely, decentralized, district-based systems like FPTP may have the opposite effect. But there are many other electoral variables that can be used to influence the development of party systems.

For example, new democracies like Russia and Indonesia have attempted to shape the development of their nascent party systems by providing institutional incentives for the formation of national rather than regional political parties.

Other countries such as Ecuador and Papua New Guinea have used party registration and funding requirements to achieve similar objectives. Access to public and/or private funding is a key issue that cuts across electoral system design, and is often the single biggest constraint on the emergence of viable new parties.

Just as electoral system choice will affect the way in which the political party system develops, the political party system in place affects electoral system choice. Existing parties are unlikely to support changes that are likely to seriously disadvantage them, or changes that open the possibility of new, rival parties gaining entry to the political party system, unless there is a strong political imperative. The range of options for electoral system change may thus be constrained in practice.

Different kinds of electoral system also result in different relationships between individual candidates and their supporters. In general, systems which make use of single-member electoral districts, such as most plurality/majority systems, are seen as encouraging individual candidates to see themselves as the delegates of particular geographical areas and beholden to the interests of their local electorate. By contrast, systems which use large multi-member districts, such as most PR systems, are more likely to deliver representatives whose primary loyalty lies with their party on national issues. Both approaches have their merits, which is one of the reasons for the rise in popularity of mixed systems that combine both local and national-level representatives.

The question of accountability is often raised in discussions of political parties and electoral systems, especially in relation to individual elected members. The relationships between electors, elected members, and political parties are affected not only by the electoral system but also by other provisions of the political legislative framework such as term limits, provisions regulating the relationship between parties and their members who are also elected representatives, or provisions barring elected members from changing parties without resigning from the legislature.

The freedom for voters to choose between candidates as opposed to parties is another aspect of accountability. Many countries in recent years have therefore introduced a greater element of candidate-centred voting into their electoral systems, for example, by introducing open lists in PR elections.

Direct Democracy Options

Direct democracy is the term used to describe particular forms of vote within any democratic system. The term direct democracy is commonly used to refer to three distinct types of vote or instruments:

• referendum, which consists of a vote of the electorate required by the legal framework or requested by the Executive or Legislative on an issue of public policy such as a constitutional amendment or a proposed law;
• citizens’ initiative, which allows a certain number of citizens to initiate a vote of the electorate on a proposal outlined by those citizens to, for example, amend the constitution, or adopt, repeal, or amend an existing law; and
• recall, which allows a specified number of citizens to demand a vote of the electorate on whether an elected holder of public office should be removed from that office before the end of his/her term of office.

 

The common characteristic of these instruments is that they all give citizens the right to be directly involved in the political decision making process. Direct democracy is therefore often seen as conflicting with representative democracy, in which voters elect representatives to make decisions on their behalf. Alternatively, direct democracy can be viewed as a sometimes necessary means for the citizens to correct or limit the excesses of elected representatives and leaders, or by the government as a means of ensuring itself of a mandate to undertake what might otherwise be perceived as unpopular measures.

Broad arguments are advanced in favour of and against direct democracy.

Proponents argue that direct democracy can help to reduce the "democratic deficit", in which voters are losing confidence and interest in traditional models of representative democracy. They claim that by putting power back in the hands of the people, people will take a greater interest and role in issues of governance, thereby increasing the legitimacy of the democratic systems.

In contrast, critics argue that it undermines and weakens representative government, and that placing power in the hands of the people can threaten the rights of minorities in society. It is also argued that many voters do not have sufficient understanding to make informed decisions about referendum issues, especially in the case of complex constitutional issues. Voter education and campaign information are therefore significant issues in relation to direct democracy.

The following links look at the three different instruments of direct democracy in more detail, drawing attention to how they are designed and administered, and outlining the advantages and disadvantages of each.

Referendum

"Referendum" is the term given to a direct vote of the electorate required by the legal framework or requested by the Executive or Legislative on an issue of public policy, in contrast with votes cast at elections, which are made in relation to parties or individual candidates and generally reflect voters' preferences over a range of different issues. Referendums may be held in relation to particular circumstances (e.g., to amend a country's constitution) or in relation to particular political issues (e.g., whether or not to join an international organisation) but are in general held in relation to issues of major political significance. The terms used to define referendums may differ in different countries.

Common types of Referendums

The following are the most common types of referendums held in countries across the world:

Mandatory or obligatory referendum

A mandatory or obligatory referendum is a vote of the electorate which is called automatically under circumstances defined in the constitution or in the legislation. The consequences of the vote are usually binding. Therefore, if a proposal passes, the government or appropriate authority is compelled to implement it. Mandatory referendums may be required in relation to pre-determined issues. Typically, these are issues of major national significance, for example, adoption of international treaties, transfer of authority to international bodies, and taxes and public expenditure commitments. In addition, in many countries, proposed amendments to the constitution must be affirmed by a referendum.

Alternatively, mandatory referendums may be required in pre-determined situations. One example is in Presidential systems, where in the case of disagreement between the President and the Legislature, a referendum may be required to resolve the dispute.

Optional or facultative referendum

The second category of referendum is the optional or facultative referendum. These are votes of the electorate which are called by a formal demand, which may emanate from the executive, from a number of members of the legislature, from a number of citizens or from some other defined agent. The consequences of the vote may or may not be binding. A government can decide to initiate a referendum on a major political issue. It might do so because public pressure for a referendum forces it to hold one, or it might choose to hold a referendum because it is divided on the issue at hand. Optional referendums initiated by the government have been held frequently in Europe on the issue of European Union integration (although in some cases, such referendums have been mandatory because they involve an amendment to a country's constitution). Although these referendums may not be legally binding, it may be politically difficult for a government to ignore the outcome.

A further type of optional referendum is the abrogative referendum. An abrogative referendum is a vote of the electorate which may decide to either retain or repeal a law or decree that has been agreed and promulgated by the legislature and already implemented. Usually, citizens force a vote by collecting a certain number of signatures in support of a vote (see citizens’ initiative).

Design and Political issues of Referendums

Location of the referendum in the institutional framework

There are therefore key design features that help to define different types of referendums. The first and perhaps most important is how the referendum fits within a country's overall legal system; are referendums mandatory under a country's constitution or other laws? Is a referendum required to be held, or has it been initiated by the government, legislature, or citizens?

Issues

A second key feature is that of the issues in relation to which referendums can be held. If the circumstances in which referendums can be held are clearly specified in a country's legal framework, the issues on which referendums can be held will be known. Typically, these will relate to issues of major constitutional or political significance, such as a country's constitution, sovereignty, or international relations. A large proportion of referendums held in Europe, for example, have been connected to the issue of European integration. However, where referendums are optional, the subject matter on which they are held may vary. Without restrictions, abrogative referendums could relate to any issue that is the subject of legislation.

Some countries that provide for holding optional/abrogative referendums therefore place restrictions on the issues that can be the subject of referendums. In Uruguay, for example, referendums cannot be held in relation to fiscal policy or laws relating to the executive, whilst in Colombia, political amnesty is barred from being the subject of a referendum (see Case Studies).

Combination of polls

When a referendum is held, it must be decided whether or not it is to be combined with another poll (e.g., an ordinary general election), or whether the referendum is to be held separately. It is sometimes argued that combining polls can increase the risk that voters will confuse separate issues (e.g., the performance of the incumbent government can be confused with the issue on which the referendum is being held). However, from an administrative point of view, it may be more cost effective to hold a referendum at the same time as an election.

The referendum question or ballot text

A second important issue relates to the wording of the referendum question or ballot text put on the ballot paper. Studies suggest that the wording of the question can have an important effect on the outcome of a referendum. Who determines the exact question that appears on the ballot is therefore significant. Is the government responsible for framing the question, even in cases when the government initiates the referendum and therefore has an interest in designing the question to increase the chances of achieving its own desired outcome? Does the Electoral Management Body have oversight of the question? Whoever designs the question, it is important that the question put to voters is clear and straightforward. Studies have generally shown that when voters are in doubt about a referendum issue, they tend to be conservative and maintain the status quo.

Campaign regulations

In relation to the referendum campaign, it is vital to make sure that the voters hear the case put forward on both sides of an issue. Hence, campaign regulations may be implemented to try to ensure that there is a level playing field between organisations campaigning for and against the referendum. These might include limits on campaign expenditure (although in some countries limits may be deemed unconstitutional) and/or controls on the acceptance of campaign contributions, and control on the access to the media. In Quebec, all interested organisations must group themselves into two umbrella groups, while elsewhere any number of organisations can campaign independently for or against an issue being referred to the voters.

The role of the government and provision of information The role of the government in a referendum campaign can also be important. Is the government allowed to campaign for the outcome it supports? Does it distribute its own promotional material or run government broadcasts outlining its views? Similarly, is there a neutral source of information, separate from the government and referendum campaigners, and/or are there requirements providing for the dissemination of non-partisan information about the issue to voters? Does the Electoral Management Body have a role in providing information to voters? If information is provided through a neutral channel, who can put information into this channel? Voters may be more inclined to trust information from sources other than campaigners, and a source of information that is perceived to be neutral will be important to many voters.

Turnout/majority requirements

Finally, a critical issue is when a proposal is judged to have passed. In some countries, a specified approval quorum is necessary for a proposal to pass. That is, a minimum number or percentage of the entire electorate should support the proposal. In others, a referendum vote is only binding if a specified turnout threshold is reached. Finally, some countries require a double majority to pass (e.g., when a referendum must achieve a majority of the overall total votes cast and a majority of votes in at least a specified proportion of defined lower level electoral areas). Clearly, the requirements for a successful referendum have an impact on the likelihood of whether a referendum passes or not.

Legal consequences

The outcome of a referendum may either be legally binding or not. It is legally binding if the government or appropriate authority is compelled to implement it; it is non-binding if it is only advisory upon the government or appropriate authority. The legal consequences of a vote have to be clearly specified either in the constitution or in the ad-hoc decisions that call a referendum. In this context, it is important to distinguish between the legal and political consequences. Even though a referendum is legally consultative, it may for political reasons be difficult for a government or authority to evade the outcome.

Advantages and disadvantages of the referendum instrument

Several arguments are advanced in support of and in opposition to referendums.

Supporters of the use of referendums argue that, in the context of increasing voter apathy and disenchantment with traditional forms of democracy, direct democracy can help to re-engage voters with politics and democracy. Another argument advanced in favour of referendums is that they can be used to resolve political problems, particularly for incumbent governments; where a governing party is divided over an issue, for example, holding a referendum can help reach a solution on the issue without splitting the party (one example of this is the 1975 UK referendum on whether the UK should remain in the EC, an issue over which the ruling Labour government was deeply divided). There is also an argument that governments need a specific popular mandate for any transcendental changes that were not part of the original platform on which they campaigned. This is particularly the case when an amendment to a constitution which itself was approved by referendum is under consideration.

There are also a number of arguments made against the use of referendums. One is that it weakens representative democracy by undermining the role and importance of elected representatives. Further to this point, referendums are sometimes seen as a means available to elected representatives to avoid having to take an unpopular position on a controversial issue. Another is that voters do not always have the capacity or information to make informed decisions about the issue at stake, and instead may make ill-informed decisions based on partial knowledge or on the basis of unrelated factors such as the state of the economy. This trend may be exacerbated in the case of referendums on complex issues such as constitutional change or international treaties, with which voters are likely to be unfamiliar.

Opponents of referendums also argue that, if the executive has the power to determine when referendums are held, they can be used as a political tool to suit the needs of the governing party rather than the interests of democracy. They also claim that, since in many countries turnout at referendums is lower than at national elections, the argument that referendums increase the legitimacy of political decisions does not stand up. However, experts in Switzerland (where a number of direct democracy votes take place each year) believe that, although turnout at referendums is around 45%, more than 45% of electors participate in direct democracy, since different voters participate in the different votes that interest them.

Citizens' Initiatives

The citizens’ initiative instrument allows a certain number of citizens to initiate a vote of the electorate on a proposal outlined by those citizens. The proposal may, for example, amend the constitution, repeal or amend an existing law, or adopt (or require the legislature to adopt) an entirely new law. The number of signatures required to place an initiative on the ballot varies, but is usually a proportion of the number of voters who voted at the most recent election, or a fixed number or proportion of registered voters. Depending on the design of the initiative process, if the ballot measure is passed by voters, it may become part of the state or country's law. The initiative process therefore provides citizens with an opportunity to directly frame the laws and/or constitution under which they live. The use of the initiative instrument varies substantially across different countries.

Different types of Citizens' initiative

The following are examples of common types of citizens’ initiative:

Constitutional

Constitutional initiatives can be used to propose amendments to a country or state’s constitution. The number of signatures required to place a constitutional measure on the ballot is usually higher than the number required to place other types of measure on the ballot.

Statutory

Using a statutory initiative, citizens can propose statutory measures to be placed on the ballot. A lower signature threshold is usually required for this type of initiative.

Direct

If an initiative is a direct initiative, then the measure that is circulated in petition or placed on the ballot automatically becomes law if it is approved by voters, without any involvement by the legislature.

Indirect

In contrast, indirect initiatives allow for the involvement of the legislature in framing the laws that arise from the initiative process. This involvement might take a number of different forms. In some countries or states, when a petition has met the required threshold, the legislature is able to formulate its own proposal to go on the ballot as well, to offer voters an alternative between the citizens’ initiative and the legislature's response. Alternatively, the legislature may have a role after a measure has passed, e.g. if a ballot measure is approved by voters, the legislature may have some scope to amend the measure or to draft the actual law mandated by the successful initiative.

Agenda Initiatives and Abrogative Referendums

There are two other types of direct democracy mechanisms which are commonly labelled as initiatives, because they are invoked when citizens collect enough signatures in support of a measure. These are the agenda initiative and the abrogative referendum. An agenda initiative is a “special” direct democracy instrument because it enables a number of citizens to submit a proposal which must be considered by the legislature but which is not put to a vote of the electorate. Details about the abrogative referendum can be found in the section on referendums

Design and Political Issues

A number of important issues arise in relation to the process of preparing, circulating, and approving a citizens’ initiative. The significance of each of these issues varies, however, according to the relevant constitutional and political context.

Initiating the process

In order to start an initiative, proponents of the measure are required to collect the signatures of a specified number of supporters of the initiative and file them with the relevant government office. Ballot language must generally be proposed by the circulators and approved by the electoral management body or by the highest legal office of the relevant jurisdiction – e.g., Attorney General.

Ballot title

Once a proposition is filed, the first key issue is the wording of the ballot title to be circulated in petition and which will appear on the ballot. Typically, the ballot title will be a short summary of the proposed measure. The proposed measure will normally have been drafted by the individual or group proposing it, or by a legal team engaged by them, whilst the ballot title will normally be drafted by the office responsible for overseeing the administration of the initiative process. Agreeing the ballot title can be a long process, since whilst administrators aim to ensure that the ballot title clearly and accurately reflects the measure being proposed, proponents seek to put forward a title that maximises the chances of it being passed. Well funded initiative campaigns may file various versions of the same proposition then conduct opinion polls to assess the relative popularity of each, in order to ensure the proposition that is taken forward stands the highest chance of success.

Signature collection and verification

Once an initiative has been filed and the title agreed, its proponents must begin the process of collecting the relevant number of signatures required to get the initiative on the ballot. Signatures and identifying information such as voter registration number and/or address are often gathered on official forms provided by the electoral management body which include the full text or a summary of the proposal.

The number of signatures required can be expressed as a percentage of the electorate or as a fixed number of signatures from registered voters; these thresholds can vary from about one or two percent of the electorate up to 10 or even 20 percent. In addition, some geographical distribution of signatures, in a number or percentage of voters from all or most of the relevant geographical subdivisions, may be required for qualified support. Signature verification is undertaken by the department responsible for administering direct democracy and will normally be undertaken on a random sampling basis. Opponents are sometimes also given the opportunity to inspect the petitions and to file objections to individual signatures or to the entire submission on the grounds that it does not fulfil the legal requirements. Usually, a proportion of the signatures collected will be found to be invalid, so proponents generally seek to collect a number of signatures in excess of the actual threshold. If and when the relevant number of signatures is collected, the initiative is included on the ballot (usually in the next election in the appropriate jurisdiction).

Campaigning and the provision of information

During the petition circulation stage and in particular once an initiative has qualified for the ballot, proponents and opponents of initiatives will run campaigns for and against the measure. As with a referendum campaign, in some places, campaigning is regulated; contributions or expenditure may be capped. In others, however, there may be difficulties in implementing campaign controls because of the constitution; in many US states, for example, courts have ruled that expenditure limits are unconstitutional, because campaign expenditure has been equated with freedom of speech, which cannot be restricted. This has become a contentious issue because of concern in the US over the role of money and the initiative industry (see below).

In addition to campaigns run by organisations campaigning for or against the initiative, the administration responsible for overseeing the initiative process will often publish an information pamphlet providing voters with information about the proposal. Typically, this might include a statement from the pro- and anti- campaigners, as well as a non-partisan analysis of the measure produced by the government. It might also include statements from other individuals and organisations which support or oppose the measure.

Number of initiatives on the ballot

Because there are not normally any limits on the number of initiatives that can be included on a ballot, it may be the case that there are multiple initiatives on any given ballot. The number of initiatives on a ballot normally depends simply on the number of initiatives that have qualified for the ballot in the relevant period before the election (some initiative proponents will deliberately time petition circulation in order to ensure that the initiative appears on the ballot for one particular election rather than another). In the US state of Oregon, there were 26 different initiatives on the ballot at the 2000 Presidential election.

The "initiative industry"

In a number of US states in particular, concern has been expressed that the use of citizens’ initiatives is increasingly professionalized, and that the "initiative industry," rather than citizens, determines which measures make it onto the ballot. Professional firms can assist with virtually all aspects of the initiative process, including: initial drafting of the initiative; opinion polling and focus group research; negotiation over the ballot title; securing endorsements to be used in campaigning; petition circulation; proposing counter-initiatives; and campaigning for the initiative once it has qualified. One area of concern to some observers has been the use of paid signature collectors to circulate the initiative in an effort to qualify the measure for the ballot. This is perhaps because the use of companies acting for profit to collect signatures arguably seems most at odds with the concept of citizen-backed initiatives which are promoted by ordinary citizens who believe in a measure. It is almost accepted in some countries that, without the assistance of professional signature collectors, it will be virtually impossible to get an initiative on the ballot, meaning that only well-financed campaigners are able to get initiatives to the ballot. Some US states have tried to legislate against this by banning the use of paid signature collectors, or requiring professional firms to pay collectors by the hour rather than per signature. This has been ruled unconstitutional in some states, however. There have also been controversies regarding whether paid signature collectors must be registered voters in the jurisdiction where the petitions are circulated.

Counter-initiatives

In some places where citizens’ initiatives are proposed frequently, one trend is for opponents to propose a counter-initiative as a means of opposing the original initiative measure. This tactic can be successful in that the existence of two related but opposing initiatives increases voter uncertainty and confusion about an issue, increasing the likelihood that voters will simply oppose both measures. Use of this tactic may be effective for opponents of specific initiatives, but it also increases the number of initiatives in circulation and potentially on the ballot.

Role of the legislature, government officials and courts The role of the legislature, government officials, and courts in relation to citizens’ initiatives tends to be strictly defined and rather limited. As outlined above, there is a role for the legislature in relation to indirect initiatives, where there may be an opportunity for it to amend initiative measures, or propose their own alternatives. However, in cases where the design of the initiative process does not provide for the involvement of the legislature, the only reviews that are permitted by the state in relation to initiative proposals is whether or not they comply with the administrative requirements imposed, i.e., are there enough proponents, have enough valid signatures been gathered? A further administrative task is to draft and agree the summary of the ballot title. In some places, where there are strict controls on the scope of initiatives, administrators might also be responsible for checking that the initiative proposal deals only with one issue or very closely related issues. In systems where initiatives need to be translated into more than one language, administrators might also be responsible for confirming whether the translation of the initiative proposal means the same in all the relevant languages. However, other than these legally defined administrative tasks, there is often no role for officials or the courts in reviewing the constitutionality of initiative proposals, that is, the quality of being in accordance with and not contradictory to the constitution of a country.

Advantages and disadvantages of the citizens’ initiative instrument

Advantages

It is argued that the simple existence of the initiative mechanism acts as a check on the activities of the legislature. This is because legislators are more likely to introduce certain reforms and measures if the initiative mechanism exists, because it is likely that if they do not, an initiative on the issue will be launched. One example is that US researchers have shown that US states that use the initiative process are more likely than those that do not, to have introduced governance reform policies (e.g., term limits, campaign finance controls). Another indication of this is the number of initiatives that are introduced but subsequently withdrawn in Switzerland, because the introduction of the initiative has in itself forced the legislature to address the issue. It is therefore claimed that the initiative process makes legislatures more responsive.

Disadvantages

One often cited disadvantage of citizens’ initiatives is that they result in badly drafted law, since (except in the case of indirect initiatives) the wording of the measure as initially proposed ends up as statute if the measure is passed. It is argued that the failure to use the expertise provided by government lawyers and officials who are familiar with the drafting process leads to laws that can be meaningless or ineffective, or have to be re-drafted, because the individuals or lawyers who draft the measures are not experienced in legislative drafting. Additionally, in some cases, statute created by the initiative process is found to be unconstitutional.

A further disadvantage is the sheer number and complexity of issues that voters are expected to vote on. It is argued that it is impossible for voters to make informed decisions when there are a substantial number of initiatives on the ballot. A related argument is that citizens cannot be expected to make decisions on complex issues that they, unlike elected representatives, do not have the time to learn about. A frequent criticism of citizens’ initiatives is that they are only really accessible to well-resourced organisations and interests, and that the process is therefore hijacked by special interest groups promoting their own interests. Empirical evidence about the success of well-financed interest groups is mixed: some research has suggested that money does not in fact have a major effect on the outcome of citizens’ initiatives, whilst other studies have highlighted the role of money, in particular in defeating initiatives. However, there can be no doubting the importance of professional firms in the initial stage of the initiative process, in particular in terms of signature collection.

A common criticism of citizens’ initiatives (and other forms of direct democracy) is that they enable the "tyranny of the majority." This term refers to the electoral power of majority groups in society being used to restrict or hinder the rights of unpopular minority groups. Critics argue that without the moderating influence of the legislature, legislation may be passed which actively targets the rights of groups within society which are unpopular. However, research undertaken in the US has produced different conclusions about whether initiatives do actually restrict the rights of minorities or not.

Recall

Recall is the name given to an instrument by which a specified number of citizens can demand a vote of the electorate on whether an elected holder of public office should be removed from that office before the end of his/her term of office. Combining elements of the initiative instrument and a normal candidate election, a recall initiative is launched when a recall motion is filed with the relevant administration. Proponents are then required to gather a specified number of signatures in support of the recall measure. Typically, the number of signatures required will be a proportion of the votes cast for the officer who is the subject of the recall at the last ordinary election to that office. If and when the recall petition acquires enough valid signatures, the issue is put to voters at a ballot to determine firstly, whether or not the officer in question should be recalled and secondly, who should replace the officer if the recall measure is successful.

The recall mechanism is the least common of the three direct democracy mechanisms. Although many US states include provisions for the recall in their constitutions, the mechanism is not used at national level in the US. Provision for the recall mechanism outside the US and at national level is rare, even in countries where direct democracy is widely used (e.g., Switzerland). Having said this, however, in 2004 Venezuela held the first ever presidential recall. There are sometimes limits on when a recall can proceed – e.g., not in the first or last year of a term – and on who is eligible to vote in the recall; in Colombia, only those individuals who voted in the original election are eligible to vote in any recall.

Design and Political issues of Recalls

Restrictions on which officers the mechanism applies to

Where a country or state's constitution provides for use of the recall mechanism, relevant legal provisions will need to specify which elected officials the mechanism can be applied to. Only in Venezuela does the recall mechanism apply to a country's elected head of state. However, in most US states, the recall mechanism can be used to recall all elected state officials, from local and county officials up to the office of Governor. Judges may also be the subject of recall campaigns. In some states, some on-elected officials such as administrative officers can also be recalled.

Restrictions on applying the recall mechanism

In some places that have adopted the recall mechanism, officers can only be recalled if it can be demonstrated that they have acted improperly (e.g., the US state of Minnesota, where the grounds for recall of an elected official are serious malfeasance or nonfeasance during the term of office, or Colombia, where an official must be shown not to have carried out their campaign promises). In others, there are no requirements for officials to have acted improperly or incompetently, and the simple fact of enough signatures being collected is enough to force a recall vote.

Number of signatures required to force a ballot

The number of signatures required in order to hold a recall ballot clearly has a significant effect on the likelihood of being able to do so; the fewer the number of signatures required, the more likely it is that a vote on whether an officer should be recalled will take place. At the 2003 California recall, recall proponents were required to gather signatures of 12% of the vote for Governor at the last election in a period of 160 days. Many other US states require 25% of voters to support a recall; California's threshold of 12% is the lowest in the United States.

Verification of signatures

As with the citizens’ initiative mechanism, the requirement to produce a specified number of signatures in order to hold a ballot creates the need to collect and verify signatures. Signatures are normally only valid if they are collected from individuals who are registered to vote in the jurisdiction of the officer who is the subject of the recall. Verification of signatures is undertaken by the office administering the recall initiative. The importance of the signature collection and the verification process is demonstrated by the high-profile Chavez recall initiative in Venezuela, where the National Electoral Council initially ruled that a large number of signatures collected by recall proponents were ineligible.

Combining the votes on the recall and a successor

One feature of the recall mechanism which varies in different places is whether, once a recall petition has collected enough valid signatures, the recall ballot is combined with the vote for a replacement officer if the recall is successful. In some places, the votes are combined, meaning that voters have to vote on two issues: firstly, whether or not the officer in question should be removed from office, and secondly, who should replace the officer if the recall is successful. In such cases, if the recall vote is defeated, the vote on a successor is irrelevant and is ignored. However, if the recall vote passes, the candidate who achieves the most support on the second vote is elected as a successor to the recalled officer. Alternatively, an initial ballot on whether or not to recall the officer is held, and only if the recall vote is passed is a second vote on a successor held.

There are arguments for and against each of these alternatives. One argument against combining the votes is that the combination might confuse voters about the process, and that it prevents voters from focusing solely on the recall issue. Arguably, when voters make a decision about whether to support the recall, they should be able to focus on the issue at hand and the performance of the incumbent, without the distraction of possible successors. On the other hand, combining the two votes leads to cost savings and increases administrative efficiency. A further issue is that combining the recall and successor vote could mean that the vote for (i.e. to retain) an incumbent who is successfully recalled is actually higher than the plurality vote in favour of the successor, which could give a legitimacy problem. This scenario might be avoided by using voting systems such as the Alternative Vote or Supplementary Vote. However, this would mean that a combined vote would become increasingly complicated for voters.

The recall in a party-based electoral system More generally, there is a question of compatibility of recall with the electoral system. If the electoral system is candidate-based, there is no problem. If, however, it is party based, should the voters have the right to recall a specific representative nominated by a party? Would there be a difference between representatives chosen by the voters under open list proportional representation? In general, in list systems, who fills a vacancy caused by a successful recall: the party or the voters?

Advantages and disadvantages of the recall instrument

Proponents of the recall mechanism argue that it acts as a discipline on elected officials, in that elected representatives will be less likely to make unpopular decisions if it may make them more likely to be the subject of a recall campaign. However, the same argument is also used against the recall: opponents argue that the recall mechanism completely undermines representative government by making elected officials afraid to make unpopular but necessary decisions.

A further argument in favour of direct democracy is that it provides voters with the continued opportunity to make a democratic decision about who governs them, since they do not have only one opportunity every three to five years to elect the people who will represent them, but retain a degree of control over the decision for the duration of the office.

However, it is claimed by opponents of the recall that the mechanism could be used irresponsibly, and that it could be used by political parties as a political weapon against rival incumbents. The claim that the recall mechanism was being used as a political tool was made by many Democrats against Republican activists in relation to the 2003 California Recall.