Office for Democratic Institutions and Human Rights COMPARATIVE STUDY LAWS AND REGULATIONS RESTRICTING THE PUBLICATION OF ELECTORAL OPINION POLLS February 2003

TABLE OF CONTENTS I. INTRODUCTION II. COUNTRY PROFILES ALBANIA AUSTRALIA BULGARIA CANADA CZECH REPUBLIC FRANCE INDIA ITALY MONTENEGRO PERU RUSSIA SINGAPORE SOUTH AFRICA SWEDEN UNITED KINGDOM UNITED STATES COUNCIL OF EUROPE ARTICLE 19 III. CONCLUSION

COMPARATIVE STUDY OF LAWS AND REGULATIONS RESTRICTING THE PUBLICATION OF ELECTORAL OPINION POLLS  1 January 2003 I. INTRODUCTION The right of every citizen to participate in government through free and fair elections is well settled under international human rights law. 2 While Article 21 of the Universal Declaration of Human Rights (UDHR) 3 is the source most often cited for this point, numerous international instruments recognize electoral rights. Article 21 states, in part: (1) Everyone has the right to take part in the government of his country, directly or through freely chosen representatives… (3) The will of the people shall be the basis of the authority of government; this will shall be expressed in periodic and genuine elections which shall be by universal and equal suffrage and shall be held by secret vote or by equivalent free voting procedures. The International Covenant on Civil and Political Rights (ICCPR), 4 a legally binding treaty ratified by 148 States, as well the three main regional human rights instruments - the African Charter on Human and Peoples' Rights, 5 the American Convention on Human Rights, 6 and the European Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR) 7 - all guarantee electoral rights. These same instruments also guarantee the right to freedom of expression and the right of the public to receive information and ideas, as well as the right of the mass media to impart information and ideas. Article 19 of the ICCPR, for example, guarantees freedom of expression in the following terms: 1. Everyone shall have the right to hold opinions without interference. 2. Everyone shall have the right to freedom of expression; this right shall include freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers, either orally, in writing or in print, in the form of art, or through any other media of his choice. Democracy depends on all contesting points of view being fairly and equitably communicated so that the people may make informed choices. As noted by the European Court of Human Rights in the case of Bowman v. United Kingdom, the right to free elections (Article 3 of the ECHR) and freedom of expression (Article 10 of the ECHR) operate to reinforce each other and "together form the bedrock of any democratic system." 8 Freedom of expression, the Court continued: is one of the "conditions" necessary to "ensure the free expression of the opinion of the people in the choice of the legislature. For this reason, it is particularly important in the period preceding an election that opinions and information of all kinds are permitted to circulate freely. Similar statements have been made by courts all over the world. Nevertheless, there are circumstances in which the two rights may come into conflict and international law recognises that certain restrictions on freedom of expression may be legitimate in order to ensure that the political debate prior to an election is not distorted. Most countries have imposed some restrictions of this sort. For example, controls on election spending are very common, and these obviously affect freedom of expression. One specific area in which restrictions on freedom of expression have been imposed to protect the integrity and fairness of the electoral process is in relation to the publication of pre-election opinion and exit polls. Pre-election polls are public surveys which assess the views of the electorate on various election-related matters while exit polls take place immediately after people have voted and assess the level of support for the various parties and candidates. Media coverage of such information can, at times, be controversial. This is particularly true of polls and projections commissioned or conducted by a source that is not impartial. Furthermore, polls may be subject to manipulation at many levels: in the choice of questions, the choice of sample, the time that the questions are asked, and so on. 9 It is often perceived, therefore, that polls and projections may have a distorting effect on the vote, rather than simply reflecting public sentiments. Consequently, many governments have adopted laws or regulations that restrict the reporting of electoral opinion polls by the media (discussed in greater detail in the next Part). Such laws and regulations constitute a de facto restriction on the freedom of expression as they interfere with both the principle of editorial independence and the public's right to receive information. It is well recognized under international law that any limitation placed on freedom of expression must remain within strictly defined parameters. The universally accepted standard for restrictions is set by Article 19(3) of the ICCPR, which states: The exercise of the [right to freedom of expression] carries with it special duties and responsibilities. It may therefore be subject to certain restrictions, but these shall only be such as are provided by Act and are necessary: (a) For respect of the rights or reputations of others; (b) For the protection of national security or of public order (ordre public), or of public health or morals. This article subjects any restriction on the right to freedom of expression to a strict three-part test, requiring that any restriction must a) be provided by Act; b) be for the purpose of safeguarding a legitimate public interest; and c) be necessary to secure this interest. 10 The third part of this test means that even measures which seek to protect a legitimate interest must meet the requisite standard established by the term "necessity". Although absolute necessity is not required, a "pressing social need" must be demonstrated, the restriction must be propor-tion-ate to the legitimate aim pursued and the reasons given to justify the restriction must be relevant and sufficient. 11 As has been noted: "[The adjective "necessary"] is not synonymous with "indispensable", neither has it the flexibility of such expressions as "admissible", "ordinary", "useful", "reasonable" or "desirable". [It] implies the existence of a "pressing social need". 12 In general, there are two ways in which any risk of distortion from pre-election polls can be minimised and both means have been employed in countries around the world. First, many countries prohibit the publication of such polls in the period immediately preceding the vote. A controversial issue here is the timeframe over which such a prohibition applies and whether this can be justified as a restriction on freedom of expression. In practice, the length of such prohibitions range from Singapore where polls are prohibited for the whole of the election period, to countries such as Australia, South Africa and the United States, which do not impose such restrictions. In some countries, such as Canada and France, constitutional decisions have helped to clarify the legitimacy of such prohibitions and, in particular, the timeframe over which a prohibition may be legitimate. In both of these cases, the prohibitions were shortened, respectively, to polling day and to 24 hours before the vote, following challenges to longer prohibitions based on the guarantee of freedom of expression. Second, some countries require the publication or broadcasting of polls, both pre-election and exit, to be accompanied by certain information, such as the source of the poll, the margin of error, the date on which the poll was conducted and so on. This can help serve as a sort of "health warning" about the validity of the poll and help avoid situations where electors place undue reliance on the poll.

II. COUNTRY PROFILES What follows is a brief overview of the relevant legislation in 16 countries, as well as standards set out by the Council of Europe and ARTICLE 19, regarding the publication of election survey results. Commentary have not been provided on the different approaches taken by different countries and while some governments have intentionally sought to make their laws consistent with human rights standards, others have not and some of the systems outlined below are clearly in breach of the right to freedom of expression. ALBANIA Article 130 of the Electoral Code of the Republic of Albania, 2000, which governs the election of National Assembly members and the conduct of local government referendums, prohibits the publication of opinion poll results during the last five days of the election campaign. Exit polls are not specifically addressed. Poll results published before the proscribed period commences must include the names of both the pollster and survey sponsor, the sample size, the margin of error, and the time period during which the poll was taken. The penalty for violating these provisions is a fine of 100,000 - 500,000 lek (715 EUR - 3,571 EUR) for members of the media, and between 1,000 - 2,500 lek (about 7 EUR - 18 EUR) for everyone else. 13 AUSTRALIA Australia has no legal restrictions on the publication of either pre-election opinion polls or - with the exception of the state of Victoria - exit polls. Reportedly, the Australian media rarely use exit polls due to negative past experiences with erroneous results and the increased speed at which official election results are delivered. 14 BULGARIA Bulgarian law prohibits the publication of new electoral survey results at any point during the last 14 days of the election campaign, and also on election day. The penalty for violation is a fine of 50 - 5,000 leva (approximately US$28 - $2,750). 15 CANADA The reporting of poll results during federal elections is regulated by the Canada Elections Act, 2000. The Act prohibits the transmission of new election survey results to the public on polling day, before the close of all the polling stations in the electoral district. Exit polls may not be shown until the close of polls and transmission of election results from a district in one time zone to those in another are not permitted before the close of polls in the district where the polls close later, to take account of the numerous time zones in Canada. 16 The Act also stipulates that the first person to transmit the results of an election survey to the public, or any person who transmits the results within 24 hours after they are first transmitted, must provide the following additional information: the name of the sponsor of the survey; the name of the person or organization that conducted the survey; the date on which or the period during which the survey was conducted; the population from which the sample of respondents was drawn; the number of people who were contacted to participate in the survey ; and if applicable, the margin of error in respect of the data obtained. 17 These rules are new, and result from a Supreme Court decision holding that a 72-hour ban on the publication of opinion survey results prior to elections violated freedom of expression as protected by the Canadian Charter of Rights and Freedoms, and could not be justified as necessary to protect the integrity of the electoral process. 18 The Court applied a necessity test very similar to the three-part test contained in the ICCPR. The Canadian government argued that the three-day ban was required to protect against the threat to freedom of choice by inaccurate polls and to protect some voters from being excessively influenced by polls. The Supreme Court stated that the government "cannot take the most uninformed and naïve voter as the standard by which constitutionality is assessed." Rather, the ban sends the message to voters that the media can be constrained by government not to publish factual information. The Supreme Court was of the view that the tangible harms to freedom of expression caused by the ban were not outweighed by the intangible benefits, and that less restrictive measures are available to protect the population from inaccurate polls, including requiring that the media publish information on the survey's methodology. Notably, the Supreme Court alluded to circumstances, for example, in the context of unfettered paid political advertising, in which the nature of the interests of the speakers could make the expression "inimical to the exercise of free and informed choice by others." However, the court concluded that no such systemic or structural dangers currently exist in Canada. This might, however, be the case, for example in a country where major media outlets were controlled by vested political interests. CZECH REPUBLIC The publication of opinion polls is prohibited in the Czech Republic for the entire week preceding the day of elections, up until the close of voting. Additionally, no exit polls may be undertaken on election day in the building in which a polling station has been located. 19 Presumably exit polls may be conducted elsewhere. FRANCE Under Article 11 of the Loi 77-808 du 19 Juillet 1977, the publication and broadcasting of opinion polls was banned for the seven days preceding each of the two rounds of voting in the country's national elections. Exit polls were banned until the close of voting. In 1997, however, several newspapers either published opinion poll results within the blackout period or indicated to readers where such results could be found on the Internet, in violation of the 1977 law. The newspapers were prosecuted by the Commission des Sondages - the regulatory body charged with overseeing the law - and the case went to the country's highest judicial court, the Cour de Cassation. In a landmark decision, the French court held that the 1977 law violated Article 10 of the European Convention on Human Rights, protecting freedom of expression, and specifically the electorate's right to receive and communicate information. Following the Court's decision the French Senate conducted its own study of the law and concluded that the week-long ban violated freedom of information (in addition to the media's right to expression) because it permitted the media to rely on poll results to inform their reporting, but to keep the basis of that reporting - the poll results - secret from the public. The Senate also concluded that modern communication technologies undermine the viability of media blackouts, as information may be published in other countries, accessible to audiences via satellite or the Internet. A new law, adopted in February 2002, replaces the week-long prohibition with a 24-hour publication ban. With the exception of Internet sites, no person may publish or otherwise transmit the results of any opinion poll - whenever carried out - on the day before the vote. When opinion poll results are published, the law imposes an obligation on the media to provide details of the poll's methodology (this is unchanged from the 1977 law), and exit polls remain prohibited. 20 INDIA In 1998, the Election Commission of India issued its Guidelines for the Publication and Dissemination of Results of Opinion Polls/Exit Polls, prohibiting the publication of opinion poll results beginning 48 hours before the start of voting, and continuing until the polls closed. Organisations or agencies that conducted and published the results of an opinion poll before the blackout were required by the Guidelines to indicate the survey's sample size, the geographic spread of the survey, the margin of error, details regarding the methodology, and information about the organization. During the 1999 election, numerous media institutions ignored the ban and were subsequently prosecuted by the Election Commission. After a number of lower court decisions, the Supreme Court of India ruled that the Commission Guidelines "exceeded the power of 'superintendence, direction and control' granted to it by Article 324 of the [Indian] Constitution." 21 The Court also questioned the practicality of such a ban, given the presence of international and online media in the country. The Commission Guidelines have since been withdrawn, and no new legislation has been enacted. Renewed calls for some form of ban on the publication of opinion and exit poll results were heard last February after numerous exit polls were wrong, prompting accusations of media bias. 22 ITALY Under Law No. 28/2000, a prohibition on the publication of electoral polls begins 15 days before election day and continues until the close of voting. Results that are published prior to the blackout period must be accompanied by information about the polling methodology, sample size, date, and response rate. The same data must also be submitted to the Department of Information and Press, which will post the information on a dedicated website. Although Italy's regime is one of the most restrictive in Europe, previous legislation imposed an even longer prohibition period. 23 MONTENEGRO The Law on Election of Municipal Councilmen and National Assembly Representatives, 1998, prohibits the publicly owned media from publicizing electoral survey results the entire week prior to election day. The publicly owned media are also prohibited from carrying exit polls. 24 PERU Article 191 of the Lei Organica de Elecciones, 1997, (Electoral Law) imposes a 15-day blackout period on the publication of electoral survey results prior to election day. The sanction for violation is a fine, the amount of which is determined by the National Legal Election Oversight Commission. RUSSIA Article 47 of the Law on Elections for the Russian President, 2002, prohibits the publication of any electoral survey results for five days prior to election day and on election day itself. The law further requires that when opinion polls are published, the following information must be provided: the name of the organization publishing the survey results, the name of the organization that conducted the survey, the name of the organization that commissioned the survey, the time frame in which the survey was conducted, the sample size, the geographic region in which the survey was conducted, the exact wording of the question, the method of collecting the information, and the margin of error. 25 SINGAPORE The Parliamentary Elections Act, 2001, severely restricts the publication of electoral opinion poll results and imposes an outright prohibition on the publication of exit polls. The blackout period for the publication of opinion poll results begins with the call for elections, at the very beginning of the election campaign, and ends with the close of all polling stations on polling day. Thus the publication of poll results are effectively prohibited for the duration of the entire election period. The sanction for violation of the provisions is a fine not exceeding $1,000 (approximately US$580) and/or imprisonment for a maximum of 12 months. 26 SOUTH AFRICA There is no prohibition on the publication of electoral survey results prior to an election. Exit polls, however, are banned by the 1998 Electoral Act, which states at section 109: "During the prescribed hours for an election, no person may print, publish or distribute the result of any exit poll taken in that election." 27 Compliance with the Electoral Act is monitored and enforced by the independent Electoral Commission, which has the power to bring proceedings for non-compliance before a specially-created Electoral Court. SWEDEN In Sweden, as with the other Scandinavian countries, there are no formal legal restrictions against the publication of electoral survey results during an election campaign. In practice, however, no media organisation publishes poll results later than a day before the election, and exit poll results are not published until all polling stations have closed. 28 UNITED KINGDOM There are currently no restrictions on the publication of pre-election surveys, although the publication of exit polls taken before voting closes is prohibited by the Representation of the People Act, 2000. The sanction for contravention of the exit poll publication ban is the same as for other summary conviction offences: a fine or imprisonment of no more than six months. The absence of legislative prohibition has been explained by the British media's commitment to self-regulation and impartiality. 29 The British Broadcasting Corporation (BBC), for example, has internal guidelines on reporting opinion polls that have reportedly been effective for a number of years, following a fiasco during the 1992 elections in which almost all the reported polls were proven wrong. 30 A sample of the BBC's self-imposed reporting guidelines include: not leading a program or bulletin with the results of a pre-election poll; not including the results of an election survey in a headline; not relying on the interpretation given to a poll's result by the publication or organization which commissioned it; always reporting the expected margin of error, and where the gap between the two leading contenders is within the combined margin of error, saying so; and always reporting the dates of the poll, and who commissioned and carried out the poll. 31 UNITED STATES In spite of controversial Presidential elections in 2000, which included sharp public and official criticism of the media for the widespread reporting of an inaccurate exit poll, there are currently no legal restrictions on either the publication of pre-election opinion polls or exit polls. It may be assumed that a ban of this nature would be unlikely to withstand a constitutional challenge in the United States. COUNCIL OF EUROPE In September 1999, the Committee of Ministers of the Council of Europe adopted Recommendation No. R (99) 15 On Measures Concerning Media Coverage of Election Campaigns. While stressing the importance and fundamental nature of the principle of the editorial independence of the media, the Ministers noted that, "particular attention should be paid to specific features of the coverage of election campaigns, such as the dissemination of opinion polls, paid political advertising, the right of reply, days of reflection and provision for pre-electoral time." The Ministers recommended that member State governments examine ways of ensuring respect for the principles of fairness, balance and impartiality in the coverage of election campaigns by the media. In that regard, the Recommendation makes the following statement regarding opinion polls: Regulatory or self-regulatory frameworks should ensure that the media, when disseminating results of opinion polls, provide the public with sufficient information to make a judgment on the value of the polls. Such information could, in particular: name the political party or other organisation or person which commissioned and paid for the poll; identify the organisation conducting the poll and the methodology employed; indicate the sample and margin of error of the poll; indicate the date and/or period when the poll was conducted. All other matters concerning the way in which the media present the results of opinion polls should be decided by the media themselves. Any restriction on member States forbidding the publication/broadcasting of opinion polls (on voting intentions) on voting day or a number of days before the election should comply with Article 10 of the European Convention on Human Rights [freedom of expression], as interpreted by the European Court of Human Rights. Similarly, in respect of exit polls, member States may consider prohibiting reporting by the media on results of such polls until all polling stations in the country have closed. ARTICLE 19 In 1994, ARTICLE 19 published Guidelines for Election Broadcasting in Transitional Democracies, an analysis of a wide range of issues relating to broadcasting in relation to elections, along with guidelines for broadcasters and decision-makers on these issues. Although these guidelines are directed specifically at broadcasters, they are of some relevance here. The relevant guideline on opinion polls states: Guideline 12: Opinion Polls and Election Projections Guidelines 12 If a broadcaster publishes the results of an opinion poll or election projection, it should strive to report the results fairly and, in particular, to publish all readily available information that would assist the listeners in understanding the poll's significance. Guideline 12.1 A broadcaster which publishes the results of an opinion poll should identify the organization that conducted the poll, the organization or party that commissioned and paid for the poll, the methodology employed, the sample size, the margin of error, and the fieldwork dates. In addition, the broadcaster should state that the poll reflects public opinion only at the time that the poll was taken.

III. CONCLUSION This survey of the regulatory approaches adopted by various democratic governments to the publication of pre-election opinion and exit polls reveals no set pattern from which general rules can be derived. Of the 16 jurisdictions surveyed, the following picture emerges: 6 countries - Australia, India, South Africa, Sweden, the United Kingdom and the United States - impose no restrictions; 2 countries - Canada and France - impose restrictions of 24 hours or less; 2 countries - Albania and Russia - impose restrictions of between 3 and 5 days; and 6 countries - Bulgaria, Czech Republic, Italy, Montenegro, Peru and Singapore - impose restrictions of 7 days or more. It is significant that from among the established democracies surveyed, only Italy imposes a ban of more than 24 hours and that there is a clear trend towards shorter bans. Courts in these countries have questioned the assumption implicit in bans that voters are uninformed and naïve, as well as the implications of bans of this nature - which prevent the media from disseminating true, factual material - for freedom of expression. These courts have also noted that in the modern world, where access to the Internet and satellite television is becoming evermore commonplace, bans of this sort may no longer be viable. Whether a country's law satisfies the three-part test for restrictions on freedom of expression does depend to some extent on the particular circumstances of that country. An important factor, in addition to the degree of Internet and satellite television access, is the degree of independence possessed by the national media and the willingness of these media to accept voluntary restrictions on reporting on polls. Where important parts of the national media, including the public media, are controlled by political figures, any risk of bias from opinions polls increases. Whether opinion polls actually make a difference to election results is highly debated. Most established democracies either avoid them altogether or have retained restrictions of very short duration, normally 24 hours or less. Where longer restrictions apply, it is up to the government to prove that they meet the three-part test for restrictions on freedom of expression. This involves, first, showing that the polls really do undermine the right of the electoral to freely chose their elected representatives. Second, a government wishing to uphold a ban must show that other, less intrusive measures, in particular requiring information about the poll, which can serve as a health warning, are ineffective in combating the distorting influence of polls. Bans of longer than 24 hours will rarely, outside of special circumstances, such as the first multi-party election in a country, be able to be justified.

1This report was prepared by Article 19, Global Campaign for Free Expression, London. 2Material for this paper was drawn in part from Guidelines for Election Broadcasting in Transitional Democracies, (London: Article 19, 1994). 3UN General Assembly Resolution 217A(III), adopted 10 December 1948. 4UN General Assembly Resolution 2200A(XXI) of 16 December 1966, in force 23 March 1976. 5Adopted 26 June 1981, in force 21 October 1986. 6Adopted 22 November 1969, O.A.S. Treaty Series No. 36, 1144 U.N.T.S. 123, in force 18 July 1978. 7Adopted 4 November 1950, in force 3 September 1953. 819 February 1998, Application No.24839/94. 9See www.aceproject.org for the results of an international study on electoral opinion polls and their coverage in the media. 10For an elaboration of this test see Goodwin v. United Kingdom, 27 March 1996, Application No. 17488/90, 22 EHRR 123 (European Court of Human Rights), paras. 28-37. 11Sunday Times v. United Kingdom, 26 April 1979, Application No. 6538/74, 2 EHRR 245 (European Court of Human Rights), para. 62. These standards have been reiterated in a large number of cases. 12Sunday Times v. the United Kingdom, 26 April 1979, Application No. 6538/74, para. 59 (European Court of Human Rights). See also the UN Human Rights Committee case, Mukong v. Cameroon, Communication No. 458/1991, 21 July 1994, 49 GAOR Supp. No. 40, UN Doc. A/49/40, para. 9.7. 13The full text of the legislation is available at: www2.essex.ac.uk/elect/legi_index.htm. 14T. Bale, "Restricting the Broadcast and Publication of Pre-Election and Exit Polls: Some Selected Examples", in Representation (2002), Vol. 39, No.1, p. 16. 15Article 60 of the Election of Members of the National Assembly, Municipal Councillors and Mayors Act, 1991, amended in 1995. Full text of the legislation is available at: www2.essex.ac.uk/elect/legi_index.htm. 16Sections 328 and 329 of the Canada Elections Act. 17Full text of the legislation is available at: laws.justice.gc.ca/en/E-2.01/13937.html#rid-14037. 18Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877 ("Thomson"). Available at: www.lexum.umontreal.ca/csc-scc/en/pub/1998/vol1/html/1998scr1_0877.html. 19Article 16 of the law on Elections to the Parliament of the Czech Republic, and on Amendments of Certain Other Acts, 1995. Full text of the legislation is available at: www2.essex.ac.uk/elect/legi_index.htm. 20For a discussion of the new law and its full text, in French, see: www.senat.fr/leg/pp100-057.html. 21Election Commission Guidelines, Note 14, pp. 18-19. 22The Tribune (India), 17 February 2002. Online: www.tribuneindia.com/2002/20020217/punjab1.htm. 23T. Bale, op. it, p. 19. 24Article 63. Full text of the legislation is available at: www2.essex.ac.uk/elect/legi_index.htm. 25This information is based on an English translation of a nearly final draft of the Russian law, provided to ARTICLE 19 by the OSCE. The 1999 Federal Law on the Election of Deputies of the State Duma of the Federal Assembly of the Russian Federation has been recently amended, but ARTICLE 19 was unable to locate an English translation. 26Full text of the legislation is available at: agcvldb4.agc.gov.sg/non_version/cgi-bin/cgi_retrieve.pl?&actno=Reved-218&date=latest&method=part. 27Full text of the South African legislation is available at: www.polity.org.za/html/govdocs/legislation/1998/act73.pdf. 28Note 14, p. 19. 29Ibid., p. 20. 30See: www.aceproject.org/main/english/me/mec03c.htm. 31The BBC Guidelines are available online at: www.bbc.co.uk/info/genelection/section4.shtml#polls.