Petitioner, Social Weather Stations, Inc. (SWS), is a private non-stock, non-profit social research institution conducting surveys in various fields, including economics, politics, demography, and social development, and thereafter processing, analyzing, and publicly reporting the results thereof.  On the other hand, petitioner Kamahalan Publishing Corporation publishes the Manila Standard, a newspaper of general circulation, which features newsworthy items of information including election surveys.

Petitioners brought this action for prohibition to enjoin the Commission on Elections from enforcing §5.4 of R.A. No. 9006 (Fair Election Act), which provides:

Surveys affecting national candidates shall not be published fifteen (15) days before an election and surveys affecting local candidates shall not be published seven (7) days before an election.

The term "election surveys" is defined in §5.1 of the law as follows:

Election surveys refer to the measurement of opinions and perceptions of the voters as regards a candidate's popularity, qualifications, platforms or a matter of public discussion in relation to the election, including voters' preference for candidates or publicly discussed issues during the campaign period (hereafter referred to as "Survey").

To implement §5.4, Resolution 3636, §24(h), dated March 1, 2001, of the COMELEC enjoins ¾

Surveys affecting national candidates shall not be published fifteen (15) days before an election and surveys affecting local candidates shall not be published seven (7) days before an election.

Petitioner SWS states that it wishes to conduct an election survey throughout the period of the elections both at the national and local levels and release to the media the results of such survey as well as publish them directly. Petitioner Kamahalan Publishing Corporation, on the other hand, states that it intends to publish election survey results up to the last day of the elections on May 14, 2001.

Petitioners argue that the restriction on the publication of election survey results constitutes a prior restraint on the exercise of freedom of speech without any clear and present danger to justify such restraint.  They claim that SWS and other pollsters conducted and published the results of surveys prior to the 1992, 1995, and 1998 elections up to as close as two days before the election day without causing confusion among the voters and that there is neither empirical nor historical evidence to support the conclusion that there is an immediate and inevitable danger to the voting process posed by election surveys.  They point out that no similar restriction is imposed on politicians from explaining their opinion or on newspapers or broadcast media from writing and publishing articles concerning political issues up to the day of the election. Consequently, they contend that there is no reason for ordinary voters to be denied access to the results of election surveys which are relatively objective.

Respondent Commission on Elections justifies the restrictions in §5.4 of R.A. No. 9006 as necessary to prevent the manipulation and corruption of the electoral process by unscrupulous and erroneous surveys just before the election. It contends that (1) the prohibition on the publication of election survey results during the period proscribed by law bears a rational connection to the objective of the law, i.e., the prevention of the debasement of the electoral process resulting from  manipulated surveys, bandwagon effect, and absence of reply; (2) it is narrowly tailored to meet the "evils" sought to be prevented; and (3) the impairment of freedom of expression is minimal, the restriction being limited both in duration, i.e., the last 15 days before the national election and the last 7 days before a local election, and in scope as it does not prohibit election survey results but only require timeliness. Respondent claims that in National Press Club v. COMELEC,^[1] a total ban on political  advertisements, with candidates being merely allocated broadcast time during the so-called COMELEC space or COMELEC hour, was upheld by this Court.  In contrast, according to respondent, it states that the prohibition in §5.4 of R.A. No. 9006 is much more limited.

For reasons hereunder given, we hold that §5.4 of R.A. No. 9006 constitutes an unconstitutional abridgment of freedom of speech, expression, and the press.

To be sure, §5.4 lays a prior restraint on freedom of speech, expression, and the press by prohibiting the publication of election survey results affecting candidates within the prescribed periods of fifteen (15) days immediately preceding a national election and seven (7) days before a local election. Because of the preferred status of the constitutional rights of speech, expression, and the press, such a measure  is vitiated by a weighty presumption of invalidity.^[2] Indeed, "any system of prior restraints of expression comes to this Court bearing a heavy presumption against its constitutional validity. . . . The Government `thus carries a heavy burden of showing justification for the enforcement of such restraint.'"^[3] There is thus a reversal of the normal presumption of validity that inheres in every legislation.

Nor may it be argued that because of Art. IX-C, §4 of the Constitution, which gives the COMELEC supervisory power to regulate the enjoyment or utilization of franchise for the operation of media of communication, no presumption of invalidity attaches to a measure like §5.4.  For as we have pointed out in sustaining the ban on media political advertisements, the grant of power to the COMELEC under Art. IX-C, §4 is limited to ensuring "equal opportunity, time, space, and the right to reply" as well as uniform and reasonable rates of charges for the use of such media facilities for "public information campaigns and forums among candidates."^[4] This Court stated:

The technical effect of Article IX (C) (4) of the Constitution may be seen to be that no presumption of invalidity arises in respect of exercises of supervisory or regulatory authority on the part of the Comelec for the purpose of securing equal opportunity among candidates for political office, although such supervision or regulation may result in some limitation of the rights of free speech and free press.^[5]

MR. JUSTICE KAPUNAN dissents.  He rejects as inappropriate the test of clear and present danger for determining the validity of §5.4. Indeed, as has been pointed out in Osmeña v. COMELEC,^[6] this test was originally formulated for the criminal law and only later appropriated for free speech cases.  Hence, while it may be useful for determining the validity of laws dealing with inciting to sedition or incendiary speech, it may not be adequate for such regulations as the one in question.  For such a test is concerned with questions of the gravity and imminence of the danger as basis for curtailing free speech, which is not the case of §5.4 and similar regulations.

Instead, MR. JUSTICE KAPUNAN purports to engage in a form of balancing by "weighing and balancing the circumstances to determine whether public interest [in free, orderly, honest, peaceful and credible elections] is served by the regulation of the free enjoyment of the rights" (page 7). After canvassing the reasons for the prohibition, i.e., to prevent last-minute pressure on voters, the creation of bandwagon effect to favor candidates, misinformation, the "junking" of weak and "losing" candidates by their parties, and the form of election cheating called "dagdag-bawas" and invoking the State's power to supervise media of information during the election period (pages 11-16), the dissenting opinion simply concludes:

Viewed in the light of the legitimate and significant objectives of Section 5.4, it may be seen that its limiting impact on the rights of free speech and of the press is not unduly repressive or unreasonable. Indeed, it is a mere restriction, not an absolute prohibition, on the publication of election surveys. It is limited in duration; it applies only during the period when the voters are presumably contemplating whom they should elect and when they are most susceptible to such unwarranted persuasion. These surveys may be published thereafter. (Pages 17-18)

The dissent does not, however, show why, on balance, these considerations should outweigh the value of freedom of expression. Instead, reliance is placed on Art. IX-C, §4.  As already stated, the purpose of Art. IX-C, §4 is to "ensure equal opportunity, time, and space and the right of reply, including reasonable, equal rates therefor for public information campaigns and forums among candidates."  Hence the validity of the ban on media advertising. It is noteworthy that R.A. No. 9006, §14  has lifted the ban and now allows candidates to advertise their candidacies in print and broadcast media. Indeed, to sustain the ban on the publication of survey results would sanction the censorship of all speaking by candidates in an election on the ground that the usual bombasts and hyperbolic claims made during the campaigns can confuse voters and thus debase the electoral process.

In sum, the dissent has engaged only in a balancing at the margin. This form of ad hoc balancing predictably results in sustaining the challenged legislation and leaves freedom of speech, expression, and the press with little protection. For anyone who can bring a plausible justification forward can easily show a rational connection between the statute and a legitimate governmental purpose.  In contrast, the balancing of interest undertaken by then Justice Castro in Gonzales v. COMELEC,^[7] from which the dissent in this case takes its cue, was a strong one resulting in his conclusion that §50-B of R.A. No. 4880, which limited the period of election campaign and partisan political activity, was an unconstitutional abridgment of freedom of expression.

Nor can the ban on election surveys be justified on the ground that there are other countries ¾ 78, according to the Solicitor General, while the dissent cites 28 ¾ which similarly impose restrictions on the publication of election surveys.  At best this survey is inconclusive.  It is noteworthy that in the United States no restriction on the publication of election survey results exists.  It cannot be argued that this is because the United States is a mature democracy.  Neither are there laws imposing an embargo on survey results, even for a limited period, in other countries. As pointed out by petitioners, the United Kingdom, Austria, Belgium, Denmark, Estonia, Finland, Iceland, Ireland, Latvia, Malta, Macedonia, the Netherlands, Norway, Sweden, and Ukraine, some of which are no older nor more mature than the Philippines in political development, do not restrict the publication of election survey results.

What test should then be employed to determine the constitutional validity of §5.4? The United States Supreme Court, through Chief Justice Warren, held in United States v. O'Brien:

[A] government regulation is sufficiently justified ^[1] if it is within the constitutional power of the Government; ^[2] if it furthers an important or substantial governmental interest; ^[3] if the governmental interest is unrelated to the suppression of free expression; and ^[4] if the incidental restriction on alleged First Amendment freedoms [of speech, expression and press] is no greater than is essential to the furtherance of that interest.^[8]

This is so far the most influential test for distinguishing content-based from content-neutral regulations and is said to have "become canonical in the review of such laws."^[9] It is noteworthy that the O'Brien test has been applied by this Court in at least two cases.^[10]

Under this test, even if a law furthers an important or substantial governmental interest, it should be invalidated if such governmental interest is "not unrelated to the suppression of free expression." Moreover, even if the purpose is unrelated to the suppression of free speech, the law should nevertheless be invalidated if the restriction on freedom of expression is greater than is necessary to achieve the governmental purpose in question.

Our inquiry should accordingly focus on these two considerations as applied to §5.4.

First.  Sec. 5.4 fails to meet criterion ^[3] of the O'Brien test because the causal connection of expression to the asserted governmental interest makes such interest "not unrelated to the suppression of free expression." By prohibiting the publication of election survey results because of the possibility that such publication might undermine the integrity of the election, §5.4 actually suppresses a whole class of expression, while allowing the expression of opinion concerning the same subject matter by newspaper columnists, radio and TV commentators, armchair theorists, and other opinion makers.  In effect, §5.4 shows a bias for a particular subject matter, if not viewpoint, by preferring personal opinion to statistical results. The constitutional guarantee of freedom of expression means that "the government has no power to restrict expression because of its message, its ideas, its subject matter, or its content."^[11] The inhibition of speech should be upheld only if the expression falls within one of the few unprotected categories dealt with in Chaplinsky v. New Hampshire,^[12] thus:

There are certain well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional problem.  These include the lewd and obscene, the profane, the libelous, and the insulting or `fighting' words ¾ those which by their very utterance inflict injury or tend to incite an immediate breach of the peace. [S]uch utterances are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.

Nor is there justification for the prior restraint which §5.4 lays on protected speech.  In Near v. Minnesota,^[13] it was held:

[The] protection even as to previous restraint is not absolutely unlimited. But the limitation has been recognized only in exceptional cases. . . . No one would question but that a government might prevent actual obstruction to its recruiting service or the publication of the sailing dates of transports or the number and location of troops.  On similar grounds, the primary requirements of decency may be enforced against obscene publications. The security of the community life may be protected against incitements to acts of violence and the overthrow by force of orderly government . . . .

Thus, contrary to the claim of the Solicitor General, the prohibition imposed by §5.4 cannot be justified on the ground that it is only for a limited period and is only incidental. The prohibition may be for a limited time, but the curtailment of the right of expression is direct, absolute, and substantial.  It constitutes a total suppression of a category of speech and is not made less so because it is only for a period of fifteen (15) days immediately before a national election and seven (7) days  immediately before a local election.

This sufficiently distinguishes §5.4 from R.A. No. 6646, §11(b), which this Court found to be valid in National Press Club v. COMELEC^[14] and Osmeña v. COMELEC.^[15] For the ban imposed by R.A. No. 6646, §11(b) is not only authorized by a specific constitutional provision,^[16] but it also provided an alternative so that, as this Court pointed out in Osmeña, there was actually no ban but only a substitution of media advertisements by the COMELEC space and COMELEC hour.

Second.  Even if the governmental interest sought to be promoted is unrelated to the suppression of speech and the resulting restriction of free expression is only incidental, §5.4 nonetheless fails to meet criterion ^[4] of the O'Brien test, namely, that the restriction be not greater than is necessary to further the governmental interest.  As already stated, §5.4 aims at the prevention of last-minute pressure on voters, the creation of bandwagon effect, "junking" of weak or "losing" candidates, and resort to the form of election cheating called "dagdag-bawas."  Praiseworthy as these aims of the regulation might be, they cannot be attained at the sacrifice of the fundamental right of expression, when such aim can be more narrowly pursued by punishing unlawful acts, rather than speech because of apprehension that such speech creates the danger of such evils. Thus, under the Administrative Code of 1987,^[17] the COMELEC is given the power:

To stop any illegal activity, or confiscate, tear down, and stop any unlawful, libelous, misleading or false election propaganda, after due notice and hearing.

This is surely a less restrictive means than the prohibition contained in §5.4. Pursuant to this power of the COMELEC, it can confiscate bogus survey results calculated to mislead voters. Candidates can have their own surveys conducted. No right of reply can be invoked by others.  No principle of equality is involved. It is a free market to which each candidate brings his ideas.  As for the purpose of the law to prevent bandwagon effects, it is doubtful whether the Government can deal with this natural-enough tendency of some voters. Some voters want to be identified with the "winners." Some are susceptible to the herd mentality. Can these be legitimately prohibited by suppressing the publication of survey results which are a form of expression?  It has been held that "[mere] legislative preferences or beliefs respecting matters of public convenience may well support regulation directed at other personal activities, but be insufficient to justify such as diminishes the exercise of rights so vital to the maintenance of democratic institutions."^[18]

To summarize then, we hold that §5.4 is invalid because (1) it imposes a prior restraint on the freedom of expression, (2) it is a direct and total suppression of a category of expression even though such suppression is only for a limited period, and (3) the governmental interest sought to be promoted can be achieved by means other than the suppression of freedom of expression.

On the other hand, the COMELEC contends that under Art. IX-A, §7 of the Constitution, its decisions, orders, or resolutions may be reviewed by this Court only by certiorari.  The flaws in this argument is that it assumes that its Resolution 3636, dated March 1, 2001 is a "decision, order, or resolution" within the meaning of Art. IX-A, §7.  Indeed, counsel for COMELEC maintains that Resolution 3636 was "rendered" by the Commission.  However, the Resolution does not purport to adjudicate the right of any party.  It is not an exercise by the COMELEC of its adjudicatory power to settle the claims of parties.  To the contrary, Resolution 3636 clearly states that it is promulgated to implement the provisions of R.A. No. 9006.  Hence, there is no basis for the COMELEC's claim that this petition for prohibition is inappropriate.  Prohibition has been found appropriate for testing the constitutionality of various election laws, rules, and regulations.^[19]

WHEREFORE, the petition for prohibition is GRANTED and §5.4 of R.A. No. 9006 and §24(h) of COMELEC Resolution 3636, dated March 1, 2001, are declared unconstitutional.

SO ORDERED.

Davide, Jr., C.J., Vitug, and Gonzaga-Reyes, JJ., concur.
Bellosillo, Pardo,Ynares-Santiago, and Sandoval-Gutierrez, JJ., join the dissent of J. Kapunan.
Melo, Puno, and Panganiban, JJ., see concurring opinion.
Kapunan, J., see dissenting opinion.
Quisumbing, Buena, and De Leon, Jr., JJ., on leave.

---

^[1] 207 SCRA 1 (1992).

^[2] Ayer Productions Pty. Ltd. v. Capulong, 160 SCRA 861, 873 (1988).

^[3] New York Times v. United States, 403 U.S. 713,  714, 29 L.Ed. 2d 822, 824 (1971).

^[4] National Press Club v. COMELEC, 207 SCRA 1 (1992); Osmeña v. COMELEC, 288 SCRA 447 (1998).

^[5] National Press Club v. COMELEC, supra at 9.

^[6] 288 SCRA 447 (1998).

^[7] 27 SCRA 835, 888 (1969)  (Castro, J., concurring and dissenting).

^[8] 391 U.S. 367, 377,  20 L.Ed.2d 672, 680 (1968) (bracketed numbers added).

^[9] G. Gunther & K. Sullivan, Constitutional Law 1217 (13th ed. 1997).

^[10] Adiong v. COMELEC, 207 SCRA 712 (1992); Osmeña v. COMELEC, supra.

^[11] Police Dept. v. Moshley, 408 U.S. 92, 95, 33 L.Ed.2d 212, 216 (1972).

^[12] 315 U.S. 568, 571-572, 86 L.Ed. 1031, 1035 (1942). See John Hart Ely, Flag Desecration: A Case Study in the Roles of Categorization and Balancing in First Amendment Analysis, 88 Harv. L. Rev. 1482, 1497 (1975).

^[13] 283 U.S. 697, 715-16, 75 L.Ed. 1357, 1367  (1931); See also New York Times v. United States, 403 U.S. 7-13, 29 L.Ed.2d 822 (1971).

^[14] Supra.

^[15] Supra.

^[16] Art. IX-C, §4.

^[17] Bk. V, Tit. I, Subtit. C, Ch. 1, §3 (1) (emphasis added).

^[18] Schneider v. Irvington, 308 U.S. 147, 161, 84 L.Ed. 155 (1939).

^[19] See, e.g., Mutuc v. COMELEC, 36 SCRA 228 (1970); Gonzales v. COMELEC, 27 SCRA 835 (1969).

---

MELO, J.:

Petitioners Social Weather stations, Inc. and Kamahalan Publishing Corporation, publisher of Manila Standard, have brought this action to declare as unconstitutional Section 5.4 of Republic Act No. 9006. Petitioners claim that said provision, which prohibits the publication of surveys affecting national candidates fifteen days before an election, and surveys affecting local candidates seven days before an election, constitutes prior restraint on the exercise of the freedom of speech without any clear and present danger to justify such restraint.

Respondent Commission on Elections, on the other hand, justifies the restrictions on the ground that the same is necessary to prevent the manipulation and corruption of the electoral process by unscrupulous and erroneous surveys, it being claimed that the indiscriminate publication of surveys up to election day led to misinformation, junking of weak and losing candidates by parties, and the creation of a bandwagon effect in favor of certain candidates.

The majority opinion, written by Mr. Justice Mendoza concludes that the disputed provision constitutes an unconstitutional abridgment of the freedom of speech, expression and the press.

I have to agree.

Freedom of speech has been defined as the liberty to know, to utter, and to argue freely according to conscience, above all liberties.  It includes not only the right to express one's views, but also other cognate rights relevant to the free communication of ideas, including the right to be informed on matters of public concern. Indeed, the principle of free political discussion is one of the touchstones of democracy, it being a guarantee that the people will be kept informed at all times, thereby ensuring their intelligent discharge of the responsibilities of sovereignty.

However, despite the primacy of free expression in the hierarchy of fundamental civil liberties, the same is not absolute.  It can be validly regulated.  Regulation must, however, be reasonable.  It must be shown that the interest of the public, generally, as distinguished from that of a particular class, requires such regulation. Second, it must appear that the means used are reasonably necessary for the accomplishment of the purpose, and not unduly oppressive upon individuals.

The provision in dispute plainly constitute prior restraint on the freedom of expression.  As aptly stated by the noted constitutionalist Fr. Bernas, "any system of prior restraint of expression comes to this court bearing a heavy presumption against its constitutional validity, with the Government carrying a heavy burden of showing justification for the enforcement of such a restraint" (The Constitution of the Republic of the Philippines:  a Commentary, p. 142).

Any act that restrains speech is hobbled by the presumption of invalidity and should be greeted with furrowed brows.  It is the burden of the respondent... to overthrow this presumption.  If it fails to discharge this burden, its act of censorship will be struck down (Iglesia ni Kristo vs. CA, 259 SCRA 529 [1996].

Respondent COMELEC has fallen short of the required effort to overthrow this presumption, it having failed to show that the means used by Section 5.4 of Republic Act No. 9006 are reasonably necessary for the accomplishment of the purpose, and that the same are not unduly oppressive upon individuals.

It bears emphasizing that Section 5.4 limits itself to prohibiting the publication of surveys affecting national candidates fifteen days before an election, and surveys affecting local candidates seven days before an election.  It does not restrict reporting by tri-media of the merits or demerits of national and local candidates and their chances at the polls.  Neither does it prohibit commentaries by radio broadcasters and TV anchors, the expression of opinions by columnist and editors of newspapers.  In fact, the provision in dispute does not prohibit paid hacks from trumpeting the qualifications of their candidates.  In fine, while survey organizations who employ scientific methods and engage personnel trained in the statistical sciences to determine socio-political trends, are barred from publishing their results within the specified periods, any two-bit scribbler masquerading as a legitimate journalist can write about the purported strong showing of his candidate without any prohibition or restriction.  The means used to regulate free expression is thus, not reasonable necessary for the accomplishment of the purpose.  Worse, it is unduly oppressive upon survey organizations, which have been singled out for suppression, on the mere apprehension that their survey results will lead to misinformation, "junking," or contrived bandwagon effect.

Admittedly, not all organizations which generate surveys are legitimate.  Some publish surveys which are, at best, disingenuous.  Yet, the possibility of abuse does not authorize government to restrict the activities of survey organizations at the expense of the freedom of expression. The very foundation of democracy is, as stated in Abrams vs. U.S. (250 US 610), grounded on the belief

[T]hat the ultimate good desired is better reached by a free trade in ideas-that the best test of truth is the power of the thought to get itself accepted in the competition of the market; and that truth is the only ground upon which their wishes can be safely carried out.  That, at any rate, is the theory of our Constitution.  It is an experiment, as all life is an experiment.  Every year, if not every day, we have to wager our salvation upon some prophecy based upon imperfect knowledge.  While that experiment is part of our system I think that we should be eternally vigilant against attempts to check the expression of opinions that we loathe and believe to be fraught with death, unless they so imminently threaten immediate interference with the lawful and pressing purposes of the law that an immediate check is required to save the country.

To reiterate, the prohibition against surveys within the specified period is a prior and unreasonable restraint upon the freedom of expression which is not reasonable necessary to achieve the purpose of clean, honest, orderly and peaceful elections.

For the foregoing reasons, I vote to grant the petition for prohibition and to declare Section 5.4 of R.A. No. 9006 unconstitutional.

---

PUNO, J.:

Petitioners seek to declare as unconstitutional Section 5.4 of R.A. No. 9006, otherwise known as the "Fair Election Act," which states:

Sec. 5.4. Surveys affecting national candidates shall not be published fifteen (15) days before an election and surveys affecting local candidates shall not be published seven (7) days before an election."^[1]

"Surveys" refer to the "measurement of opinions and perceptions of the voters as regards a candidate's popularity, qualifications, platforms or a matter of public discussion in relation to the election, including voters' preference for candidates or publicly discussed issues during the campaign period xxx."^[2] Violation of the prohibition is punishable as an election offense under section 264 of B.P. 881, otherwise known as the Omnibus Election Code.^[3]

Petitioners assail the law as constitutionally infirmed on the ground that it is an abridgment of their freedom of speech and of the press.^[4] I concur with the majority opinion penned by Mr. Justice Mendoza which is protective of speech and file this separate opinion by way of supplement.

It is now deeply embedded in our jurisprudence that freedom of speech and of the press enjoys a preferred status in our hierarchy of rights.^[5] The rationale is that the preservation of other rights depend on how well we protect our freedom of speech and of the press.  In view of the preferred status of freedom of speech and of the press, several tests have been enunciated to protect it.  We have the dangerous tendency test which now commands little following.  We have the clear and present danger test, the most libertarian test, formulated by Justice Holmes in Schenk v. United States,^[6] viz: "The question in every case is whether the words used are used in such circumstances and are of such a nature as to create a clear and present danger that they will bring about the substantive evil that the State has a right to prevent." In Gonzales v. COMELEC,^[7] Mr. Chief Justice Fernando explained that "the term clear seems to point to a causal connection with the danger of the substantive evil arising from the utterance questioned.  Present refers to the time element.  It used to be identified with imminent and immediate danger.  The danger must not only be probable but very likely inevitable."  We have the various balancing tests typified by the Obrien test,^[8] to wit:

"[A] government regulation is sufficiently justified ^[1] if it is within the constitutional power of the Government; ^[2] if it furthers an important or substantial governmental interest; ^[3] if the governmental interest is unrelated to the suppression of free expression; and ^[4] if the incidental restriction on alleged First Amendment freedoms [of speech, expression and press] is no greater than is essential to the furtherance of that interest."

All of these tests have their own criticisms but I need not express any preference for any of these tests to resolve the case at bar, for regardless of the test used, the assailed provision is void on its face and patently unconstitutional.

The provision in question is unconstitutional because it constitutes a clear prior restraint on petitioners' freedom of speech and of the press.  I like to stress on the prohibition against prior restraint for two reasons: (1) a historical study of human rights will show that it is prior restraint that gave rise to freedom of speech and of the press; and (2) there is a growing tendency, as noted by legal observers, for governments to manipulate the free market of ideas in the guise of merely regulating the time, manner and place of exercising freedom of speech and of the press.  The tendency appears in various masks.  One of them is thru prior restraint or thru subsequent punishment of acts regulating the exercise of freedom of speech and of the press.

The invention of printing in the fifteenth century revolutionized the communication of ideas.  Soon it dawned on the temporal and spiritual authorities that printing should be controlled and thus prior restraint on freedom of speech and of the press was born.^[9] In 1501, Pope Alexander VI issued a Bull banning unlicensed printing.  In England, printing became a monopoly and was strictly dispensed and controlled by the Crown.  It was only in 1695 that the House of Commons declined to reenact its licensing statute. ^[10] In the 18^th century, however, the right of the press against prior licensing gained the important status of a natural right in England.^[11] In 1791, The First Amendment to the US Constitution, prohibiting the abridgement of freedom of speech and of the press, was ratified by the States.  Undoubtedly, the First Amendment is a bar against any prior restraint, especially the classic form of licensing by government authorities.  Thus, in the United States, the prohibition was elevated to a constitutional principle.  In 1931, in the leading case of Near v. Minnessota,^[12] the US Supreme Court, speaking thru Mr. Chief Justice Hughes expressly ruled that "xxx liberty of the press, historically considered and taken up by the Federal Constitution, has meant, principally although not exclusively, immunity from previous restraints or censorship." Our Constitutions of 1935, 1973, and 1987 guaranteed freedom of speech and of the press and undeniably, we adopted the US model and its rationale.  I therefore emphasize that prior restraints on freedom of speech and of the press should be given the strictest of scrutiny in light of their inherent and invasive impact.

In the case at bar, the law bans publication of surveys affecting national candidates 15 days before an election and surveys affecting local candidates 7 days before an election.  Violation of the ban carries a criminal sanction.  This is pure and simple prior restraint on the communication and free flow of ideas which should be made available to voters before they exercise their right of suffrage, the core of their political sovereignty.  Prior restraint can be justified only on the narrowest of ground like national security.  The prior restraint in the case at bar is not based on compelling reasons in the category of national security and hence is intolerable for government should not be encouraged to take any step to control the subject matter of speech, otherwise it will have the dangerous power to manipulate the form and shape of thoughts that will compete in the market of ideas.  In the free market of ideas, government is bound to follow the laissez faire policy to the maximum and not the paternalistic policy of government knows best.

The provision in question is also void for its overbreadth. The overbreadth doctrine prohibits government from achieving its purpose by "means that sweep unnecessarily broadly, reaching constitutionally protected as well as  unprotected activity."^[13] Stated otherwise, "the esence of overbreadth is that government has gone too far: its legitimate interest can be satisfied without reaching so broadly in to the area of protected freedom."^[14]

In the case at bar, the prohibited surveys are all inclusive. They include "measurement of opinions and perceptions of the voters as regards a candidate's popularity, qualification, platforms or a matter of public discussion in relation to the election, including voter's preference for candidates or publicly discussed issues during the campaign period xxx." Clearly, the provision bans not only popularity surveys which show the winning and losing candidates but all "measurement of opinions and perceptions of the voters as regards a candidate's xxx qualifications, platforms or a matter of public discussion in relation to the election xxx." The inhibitory effect of this ban on free speech and the free flow of information which voters need to guide their choice of candidates is too much.  Indeed, the ban does not distinguish between biased and unbiased surveys or between surveys conducted with scientific accuracy and surveys done sloppily or between surveys that help enlighten voters in exercising their right of suffrage and surveys that mislead.  Petitioner Social Weather Stations, Inc. is one of our more reliable private non-stock, non-profit social research institutions with a no-nonsense record.  Yet, the provision in question, because of its overbreadth, will bar petitioner from making its useful pre-election surveys.  Certainly, there are less drastic means which government can utilize to achieve its objective of protecting voters from false, misleading and unfair surveys.

I vote to grant the petition.

---

^[1] See also section 24 (H) of COMELEC Resolution No. 3636 implementing R.A. No. 9006.

^[2] Ibid., section 1 (9) (e).

^[3] See Section 13 of R.A. No. 9006.

^[4] See section 4 of Article III of the 1987 Constitution.

^[5] Philippine Blooming Mills Employees Organization v. Philippine Blooming Mills, 51 SCRA 189 (1973).

^[6] 249 US 47 (1919).

^[7] 27 SCRA 835 (1969).

^[8] See US v. Obrien, 391 US 367, 20 L. Ed. 672 (1968).

^[9] Press Control and Copyright in the 16^th and 17^th Centuries, 17 Yale L.J. 841 (1920).

^[10] Licensing Act of 1662, see Holdsworth, A History of English Law, 360-79 (2^nd Ed., 1937).

^[11] Emerson, The doctrine of Prior Restraint in Law and Contemporary Problems, vol. 20, pp. 651 (1955) citing Blackstone's Commentaries.

^[12] 283 US 697 (1931).

^[13] J. Norwak, R. Rotunda & J. Young, Handbook on Constitutional Law 868 (2^nd Ed., 1983).

^[14] Redish, The Warrent Court, the Burger Court and the First Amendment Overbreadth Doctrine, 78 Nw. U.L. Rev. 1035 (1983-4).

---

PANGANIBAN, J.:

I concur in the well-written ponencia of Mr. Justice Vicente V. Mendoza holding that Section 5.4^[1] of Republic Act (RA) No. 9006 is unconstitutional.  The provision is a patent infringement of the fundamental freedoms of expression and of the press.

In the recent case ABS-CBN Broadcasting Corporation v. Commission on Elections,^[2] the Court en banc junked Comelec Resolution No. 98-1419 dated April 21, 1998, which restrained the conduct of exit polls, a species of electoral surveys.  We held that "the holding of exit polls and the dissemination of their results through mass media constitute an essential part of the freedoms of speech and of the press."  They cannot be banned "totally in the guise of promoting clean, honest, orderly an credible elections.  Quite the contrary, exit polls -- properly conducted and publicized -- can be vital tools in eliminating the evils of election-fixing and fraud."  As mankind pushes the frontiers of science and technology in mass communications, so must the scope of free expression expand^[3] to cover the conduct and the publication of surveys.

In said case, we visited the long-standing fundamental principle underlying democracies that the freedom of expression is a preferred right, standing on a higher level than other substantive liberties.  Indeed, as this nation has recently witnessed once again, lessons of history, both political and leagal, illustrate that freedom of thought and speech is an indispensable condition of nearly every other form of freedom.^[4] Thus, our Constitution explicitly mandates that no law shall be passed abridging the freedoms of speech and of the press.^[5]

While the exercise of these basic rights could not be absolute - liberty is never absolute -- but may be subject to regulation by the state, any limitation should be justified by a clear and present danger of such substantive character that the state has a right to prevent.^[6] In other words, the evil sought to be avoided must be so substantive as to justify a clamp over one's mouth or a restraint of a writing instrument.^[7]

There is, however, no compelling or justifiable reason for the prohibition made by congress under the assailed law.  The Comelec also utterly fails to convince me that a substantive danger, which the state has a right to prevent, lies lurking and threatening to explode if ignited by the conduct and the dissemination of the prohibited surveys.

No lover of freedom, no guardian of the Constitution and no advocate of democracy can agree to this unreasonable restraint.

Indeed, Daily Herald Co. v. Munro^[8] held that the general interest of the state in insulating voters from outside influences is insufficient to justify speech regulation.

WHEREFORE, I vote to GRANT the Petition and to DECLARE SECTION 5.4 OF RA 9006 UNCONSTITUTIONAL.

---

^[1] "Surveys affecting national candidates shall not be published fifteen (15) days before an election and surveys affecting local candidates shall not be published seven (7) days before an election."

^[2] 323 SCRA 811, January 28, 2000.

^[3] Panganiban, Transparency, Unanimity & Diversity, 2000 ed., p. 376.

^[4] Supra, citing Salonga v. Cruz Pano, 134 SCRA 438, 458-459, February 18, 1985.  See also Gonzales v. Comelec, 27 SCRA 835, 849, 856-857, April 18, 1969; Philippine Blooming Mills Employees Organization v. Philippine Blooming Mills Co., Inc., 51 SCRA 191, June 5, 1973; National Press Club v. Comelec, 207 SCRA 1, 9, March 5, 1992; Blo Umpar Adiong v. Comelec, 207 SCRA 712, 715, March 31, 1992.

^[5] Sec. 4, Art. III, Constitution.

^[6] See Priomicias v. Fugoso, 80 Phil. 71 (1948); American Bible Society v. City of Manila; 101 Phil. 386 (1957); Iglesia ni Cristo v. MTRCB, 259 SCRA 529, July 26, 1996.

^[7] Adiong v. Comelec, supra.

^[8] 838 F 2d 380 (9^th Cir. 1988), cited in ABS-CBN v. Comelec, supra.

---

KAPUNAN, J.:

Two seemingly conflicting rights or interests, both integral to our democratic system, are involved in this case.

On the one hand are the freedoms of speech and of the press, which, as often stated, are accorded a preferred status in our constitutional hierarchy,^[1] essential as they are to preservation and vitality of our civil and political institutions.^[2] The primacy, the high estate of these freedoms is a fundamental postulate of our constitutional system.^[3]

On the other hand, the Constitution requires the State to "guarantee equal access to opportunities for public service,"^[4] and mandates Congress to "provide a system for securing the secrecy and sanctity of the ballot."^[5] The State's interest in holding "free, orderly, honest, peaceful and credible elections"^[6] cannot be denied.

At the heart of the controversy is Section 5.4 of Republic Act No. 9006,^[7] otherwise known as the "Fair Election Act,"^[8] which states that:

Surveys affecting national candidates shall not be published fifteen (15) days before an election and surveys affecting local candidates shall not be published seven (7) days before an election.

"Surveys," as used above, pertain to "election surveys," which in Section 5 thereof-

xxx refer to the measurement of opinions and perceptions of the voters as regards a candidate's popularity, qualifications, platforms or a matter of public discussion in relation to the election, including voters' preference for candidates or publicly discussed issues during the campaign period. xxx

The Fair Election Act was signed into law by the President on February 12, 2001.  Pursuant to its authority under Section 13 thereof, the Commission on Elections (COMELEC) on March 1, 2001 promulgated through Resolution No. 3636 the Implementing Rules and Regulations of the Fair Election Act.  Section 24 of the implementing rules is a verbatim reproduction of Section 5.4.

Petitioners contend that the subject provisions violate the freedoms of speech and of the press enshrined in Section 4, Article III of the Constitution thus:

No law shall be passed abridging the freedom of speech, of expression, or of the press xxx.

As publisher of a newspaper, Kamahalan maintains that its right to freedom of the press is unduly infringed by section 5.4.  Insofar as publication (of surveys) is a component of the freedom of speech, the freedom of SWS is also purportedly severely restricted.

Although among our most cherished rights, the freedoms of speech and of the press are not absolute or unlimited.  In certain instances, this Court has allowed the regulation of the exercise of these freedoms vis-a-vis election-related laws.  In Osmena vs. Commission on Elections^[9] and National Press Club vs. Commission on Elections,^[10] the law prohibiting newspapers, radio broadcasting and television station from selling or giving free of charge print space or air time for campaign or other political purposes was declared valid.  In Badoy vs. Commission of Elections,^[11] the prohibition on the publication of paid political advertisements outside the COMELEC space was likewise upheld.  In Gonzales vs. Commission on Elections,^[12] where the prohibition on the early nomination of candidates and the limitation on the period of election campaign or partisan political activity under Republic Act No. 4880 was assailed for being violative of the freedoms of speech, of the free press, of assembly and of association, the Court declared the law not unconstitutional.

Courts have employed certain tests to determine the validity of restrictions on the rights to free speech and free press.  The "dangerous tendency" rule provided that the State has the power to proscribe and punish speech which "creates a dangerous tendency which the State has a right to prevent."^[13] This formulation, however, had long been abandoned in the United States as well as in this jurisdiction.

The "clear and present danger" rule postulates that "the question in every case is whether the words are used in such circumstances and are of such nature as to create a clear an present danger that they will bring about the substantive evils that Congress has the right to prevent."^[14] This rule has been applied in our jurisdiction in a number of cases.^[15]

Nevertheless, Associate Justice Fred Ruiz Castro, Later Chief Justice, in his separate opinion in Gonzales vs. Commission on Elections,^[16] expressed the view that in determining the constitutionality of Republic Act No. 4880 assailed therein, another approach, the so-called "balancing-of-interests" test, was more appropriate.  He observed:

However useful the "clear and present danger" formulation was in the appraisal of a specific type of situation, there is fairly extensive recognition that it is not a rule of universal applicability and validity, not an automatic mechanism that relives a court of the need for careful scrutiny of the features of a given situation and evaluation of the competing interests involved.^[17]

Justice Castro cited American Communications Association v. Douds,^[18] where the "balancing-of-interests" test was supplied.  In said case, the United States Supreme Court stated that "in suggesting that the substantive evil must be serious and substantial, it was never the intention of [the U.S. Supreme Court] to lay down an absolutist test measured in terms of danger to the Nation."^[19] Chief Justice Vinzons, expounded:

When a particular conduct is regulated in the interest of public order, and the regulation results in an indirect, conditional, partial abridgment of speech, the duty of the courts is to determine which of the two conflicting interests demands the greater protection under the particular circumstances presented.  xxx In essence, the problem is one of weighing the probable effects of the statute upon  the free exercise of the right of speech and assembly against the congressional determination xxx We must, therefore, undertake the delicate and difficult task xxx to weigh the circumstances and to appraise the substantiality of the reasons advance in support of the regulation of the free enjoyment of the rights.^[20]

The test is further explained thus:

The theory of balance of interests represents a wholly pragmatic approach to the problem of First Amendment freedom, indeed, to the whole problem of constitutional interpretation. It rests on the theory that it is the Court's function in the case before it when it finds public interests served by legislation on the one hand, and First Amendment freedoms affected by it on the other, to balance the one against the other and to arrive at a judgment where the greater weight shall be place.  If on  balance it appears that public interest served by restrictive legislation is of such character that it outweighs the abridgment of freedom, then the court will find the legislation valid.  In short, the balance-of-interests theory rests on the basis that constitutional freedoms are not absolute, not even those stated in the First Amendment, and that they may be abridged to some extent to serve appropriate and important public interests.^[21]

In Zaldivar vs. Sandiganbayan,^[22] this Court reiterated that the clear-and-present danger test was not a cure-all to freedom of speech controversies:

The "clear and present danger doctrine," which test is invoked by respondent's counsel is not a magic incantation which dissolves all problems and dispenses with analysis and judgment in the testing of the legitimacy of claims to free speech, and which compels a court to exonerate a defendant the moment the doctrine is invoked, absent proof of impending apocalypse.  The "Clear and present danger" doctrine has been an accepted method for marking out the appropriate limits of freedom of speech and of assembly in certain contexts. It is not, however, the only test which has been recognized and applied by courts.^[23]

Zaldivar cited the case of Lagunzad vs. Soto Vda. De Gonzales,^[24] where the Court also referred to the shortcomings of the clear-and-present doctrine noted by Justice Castro in Gonzales.  Justice Melencio-Herrera further wrote:

xxx Another criterion for permissible limitation on freedom of speech and of the press, which includes such vehicles of the mass media as radio, television and the movies, is the "balancing-of-interests test."  The principle "requires a court to take conscious and detailed consideration of the interplay of interests observable in a given situation or type of situation."^[25] [Citations omitted.]

It is my considered opinion that given the apparent conflict between petitioners' rights of speech and press - rights enshrined in the Constitution, and the inherent power of Congress to legislate on matters public interest and welfare, and in pursuance of the constitutional policy of ensuring of "free, orderly, honest, peaceful and credible elections," it is ultimately this Court's function and duty to undertake the delicate and difficult task of weighing and balancing the circumstances to determine whether public interest is served by the regulation of the free enjoyment of the rights.

I believe that Congress did not exceed constitutional limitations in enacting Section 5.4.

Indisputably, the State has a legitimate interest in fostering an informed electorate.^[26] It has a compelling interest in protecting voters from confusion and undue influence^[27] and, generally, in preserving the integrity of its election process.^[28] In furtherance of these State interests, Congress is empowered to enact laws relative to the conduct of elections.  It may not only regulate the time, manner and place of the holding of the elections but may likewise regulate the election campaigns and other activities relative thereto.^[29]

In enacting the Fair Election Act, Congress declared that the State "shall, during the election period, supervise or regulate the enjoyment or utilization of all franchises or permits for the operation of media of communication or information to guarantee or ensure equal opportunity for public service, including access to media time and space, and the equitable right to reply for public information campaigns and fora among candidates and assure free, orderly, honest, peaceful and credible elections."^[30] Further, said law aims to "ensure that bona fide candidates for any public office shall be free from any form of harassment and discrimination."^[31]

Towards these ends, Section 5.4 was incorporated specifically to prevent the evils brought about by election surveys published immediately before an election.  The deliberations on the Senate Floor are revealing:

Adverting to the bill of Senator Tatad, Senator Defensor Santiago said that the country has no law regulating the conduct of surveys and the activities of survey stations and private groups particularly those relating to political opinions.  She said that some Western countries prohibit political opinion polls or surveys for certain periods before elections to avoid last minute pressure on voters as politicians and political parties often cause the conduct and dissemination of surveys to advance their political interests.

She informed the body that the Internet reported on a worldwide survey on the publication of poll results prior to elections where 30 of the 78 countries surveyed apply legal restrictions on the publication of public opinion survey results comprising an embargo prior to general elections, which ranges from 24 hours to six weeks before an election.  She said that among the reasons for the government restrictions cited by the Internet survey were protecting the integrity of the democratic process, the rights of privacy and national security considerations.

According to the report, she said, some countries in recent years have prolonged their embargo periods - for example, Italy, from seven to 28 days; Canada, from no ban to three days; and Chile, from one to seven days - while other countries have shortened or withdrawn their embargoes - for example, Croatia, from three days to 24 hours; Colombia, from 10 to seven days; and Argentina, from two weeks to no ban.

In this connection, Senator Defensor Santiago asked whether Senator Roco would consider an amendment providing for the criteria for the publication of opinion surveys as she expressed fear that an opinion survey firm might work diligently for some time in order to establish a reputation for credibility and then, at the ultimate hour, sell its services to the highest bidder.

Senator Roco recalled that earlier in the session, he had discussed with senator Defensor Santiago and some senators the idea of not just lifting the ban on election propaganda but also of giving fair protection to candidates, especially from the unfairness of reporting certain survey results during the campaign period.  He added that the committee had been studying the rule in the United States where poll surveyors mention who authorized and paid for the survey, and what method was used, and furnish raw data to anybody who feels aggrieved by the poll results.  He agreed that there must be a period when surveys should not be published because they influence elections through self-fulfilling predictions.

However, Senator Roco expressed concern that a full-blown debate on another issue might impede the approval of the bill, although he welcomed an amendment which would create a balance of fair reporting and fair opportunity for candidates.

Senator Defensor Santiago warned that the fate of the country's leadership should not be left in the hands of survey firms which are not accountable to the people and possess no amount of sovereign power.  Additionally, she expressed resentment that a public official like herself should be treated like a can of sardines because poll surveys have reduced political life to a mere matter of appearances.

Senator Roco commented that all professions which deal with communications are aware that the way a question is put can influence the answer; the more simplistic question can give rise to a host of interpretations.  On the other hand, he said, it is a matter of public interest if there is an attempt to measure validity or acceptability of issues; still, full disclosure and transparency should apply to poll surveyors and to all who try to promote and protect public welfare.^[32]

The original proposal was a 30-day restriction on the publication of surveys. Senator Flavier suggested the deletion of the restriction,^[33] while Senator Osmeña was amenable to a shorter period of 3 days.^[34] Senators Roco and Defensor-Santiago vigorously opposed the deletion. Senator Roco said that:

x x x the committee cannot accept the deletion of the prohibition as he observed that in the Philippines, the bandwagon effect is part of campaign planning.  He recalled that in 1969, the influence of propaganda was so evident: every single pole or space was plastered with "Marcos-Lopez" posters and for the duration of the one-year campaign period, the newspapers kept on repeating that Marcos-Lopez was unbeatable that after a while, the people believed it. He explained that it is the publication and not the conduct of surveys that would be prohibited in this Act. However, he pointed out that the surveys would be useful to senatorial candidates, especially those who wish to land in the top six slots, because their names would be repeatedly mentioned on TV so that the voters might be influenced to vote for them. He said that candidates particularly those who do not have access to TV and radio have no money to influence publications should be given equal break during the 30-day period. He appealed the Members to support the committee's position.^[35]

Senator Defensor-Santiago concurred with Senator Roco:

x x x She pointed out that at the start of the debate, the Body was of the consensus that the operating principles of the bill should be equality and impartiality. She opined that these principles would be violated if the Body would delete the prohibition. Moreover, she argued that a political neophyte who deserves exposure because of his honesty, competence and efficiency would probably not be in the winning circle until the crucial decisive few days before the election. She said that the publication of a survey at any point earlier than that would be detrimental to the candidate and to national interest. She expressed support for Senator Roco's appeal to maintain the present provision.  She said that the freedom of expression in a constitutional dimension was not relevant to the discussion because a candidate who can afford it can ask any agency to conduct a survey; however, out of compelling national interest in the Philippine culture context, the State prohibits the publication of surveys within a certain period so as to avoid manipulating the minds of the electorate and to preserve the principle of equality and impartiality.^[36]

Eventually, the position of Senators Roco and Defensor-Santiago prevailed^[37] although, after the Bicameral Conference, the original 30-day limitation was reduced to 15 days with respect to surveys affecting national candidates.

Evidently, Congress found that the publication of surveys within the prohibited period inordinately works against candidates who are shown to be "losing."  The assailed provision thus seeks to avert the "bandwagon effect" supposedly caused by the publication of election surveys. The bandwagon effect results when a voter opts for a candidate or candidates whom the surveys reveal as the leading contender or contenders, the voter believing, rightly or wrongly, that the candidate or candidates whom the voter actually prefers would lose anyway, as indicated in the surveys. The bandwagon effect produces more votes for the "winning" candidate ordained as such by the surveys and less votes for the "losing" candidate. Surveys add to the prospects of the "winner" and lessen that of the "loser," who is thereby deprived of an equal opportunity to get elected. Hence, the surveys take the form of a self-fulfilling prophecy.

Ideally, a citizen ought to vote for a candidate based on the latter's personal qualifications and platform for governance. This is the ideal that the law aims to achieve; surveys published during the prescribed period before the elections have been deemed by Congress to frustrate this objective.

The prospect of misinformation magnifies the dangers of the bandwagon effect. There is nothing to prevent unscrupulous interests from procuring the services of an enterprise masquerading as a "credible" research institution to conduct "surveys" with predetermined results, and cause their publication. Worse, there is nothing to prevent the simple publication of entirely false results. The evil of the bandwagon effect caused by election surveys, whether absolutely accurate or utterly untrue, is further enhanced by the pervasiveness of media. Advances in technology have widened the electorate's access to both information and, regrettably, to misinformation.

It may be argued that propaganda portraying a candidate as possessing certain virtues or espousing certain causes, regardless of the truth of these claims, also influence the voter in making his or her choice. The distinction lies in that a survey lulls the voter into thinking that the election is over but the counting, and that his vote for a losing candidate would not matter in the end. While election propaganda expressly urge the voter to choose a candidate because of his qualifications and causes, the surveys, clothed with the mantle of statistics and couched in esoteric terminology, implicitly urge the voter to choose a candidate because of his popularity. This persuasive effect is unique to surveys; it is a feature absent in election propaganda.

This congressional concern regarding the bandwagon effect is supported by a study cited by the Solicitor General:

It is noteworthy that it is easier to translate voting intentions into potential seats in a two-party system than in a multi-party arrangement. The accuracy of election polls is also determined by actual voter turnout; pre-election surveys can sometimes be out of date by the time they are reported. x x x. Last, polls can present an opportunity for deliberate misrepresentation or connivance by those who publish survey results; many examples of this practice by political parties have been cited. Advocacy groups seeking to influence the public agenda can also commission polls for public release and may draft questions to support their case or point of view. In short, public opinion surveys are blunt instruments of prediction and are susceptible to many forms of error.

Opponents of political polling point to notable failures like the predicted victories of Landon over Roosevelt in 1936, of Dewey over Truman in 1948, and of Wilson over Heath in Britain in 1970. Most pollsters considered the outcome of the 1980 presidential election in the United States too close to call, yet Ronald Reagan won by a landslide. The 1992 surprise victory of the Conservatives over Labour in Britain is another similar example. x x x

THE IMPACT OF POLLING ON THE ELECTORAL PROCESS

A. Direct Effects

Because polls are generally perceived to be accurate and scientific, the debate on polling centres largely whether it undermines the democratic process by influencing electoral behaviour and election results. Some political strategists and observers argue that the publication of polls gives an unfair advantage to parties or candidates whose fortunes are seen to be improving. The so-called "bandwagon" effect assumes that knowledge of a popular "tide" will likely change voting intentions in favour of the frontrunner, that many electors feel more comfortable supporting a popular choice or that people accept the perceived collective wisdom of others as being enough reason for supporting a candidate.

The bandwagon phenomenon, however, is dismissed by those who argue that voters do not pay much attention to poll results in the first place, that not everyone believes them, and that it is not important for everyone to be on the winning side. Furthermore, while some voters may want to be on the victorious side, at least a few will rally to support the expected loser out of sympathy - the so-called "underdog" effect - which would cancel out or annul any shifts in preference.

Although academics in the United States have long been divided over the impact of published polls on the outcome of elections, recent research supports the proposition that their publication can influence a close election, with the most impact occurring late in a campaign. Recent studies in Canada also support the notion that polls published during political campaigns can create the "politics of expectations," a situation that stimulates the bandwagon effect and promotes "strategic voting," in which voting is influenced by the chances of winning.  For example, citizens may cast ballots for their second-choice candidate who appears to have a better chance than the first choice of defeating a disliked candidate or party. Such behaviour is said to be increasing in Canada as close three-party races become more common. It is therefore argued that voters making such strategic choices have every right to expect that the results of opinion surveys are scientifically valid.^[38]

The same study also pointed out other "indirect effects" of surveys published during the election period, that it detracts from the "real" issues of the election and affects a candidate's momentum:

B. Indirect Effects

The indirect effects of polls during elections may be as important as their possible direct influence. Because of the multiplicity of published surveys and the attention they receive from the media, some charge that polls detract from discussion of the "real" issues. Indeed, many describe news coverage of Canadian elections as being analogous to that of a sporting event or "horse-race," with serious analysis of the issues or investigation into areas of voter concern being largely ignored. The media's emphasis on who is winning and who is losing (as well as on the campaign "style" of leaders and their parties) may also result in so-called "leader-fixation." As one scholar explains:

Polls conducted throughout the campaign . . . focus on leadership in an attempt to predict the outcome of the election and to explain it in terms of leader appeal. The polls are presented as measures to gauge how the leaders' campaigns are fairing. In this sense the media coverage misrepresents the political system, narrows the focus of public debate, and denigrates political leaders and institutions.^[39]

The deliberations during the Bicameral Conference also intimate another purpose in passing the challenged provisions, that is, to prevent the nefarious election scheme known as "dagdag-bawas." Dagdag-bawas, a phenomenon peculiar to Philippine elections, takes place when votes cast in favor of one candidate are deducted then credited to another. Senator Roco also observed that last-minute surveys generate "junking" of candidates at the tail end of the surveys by their very own party-mates or supporters.

CHAIRMAN ROCO.  I do not want to say it that way. I only said, that if you will target people to campaign against, you will target people who are outside 1 to 6 because it is a waste of time to try to drag No. 6 down to 13. Legitimate campaign.

Mapababa mo man ang No.1, umabot ng no. 6, he or she still occupies one post.  Hindi nawawala. Ang tatargetin mo, 9, 10, 11, parang junking doon sa sample ballots mawawala yong mga mabababa because yon ang puedeng mawala. Yong 1 to 6 or 1 to 8, ang hirap-hirap nang tanggalin.

So, in your sample ballots you don't care.  Sa sample ballots, kung sa surveys 10, 11, 12, eh, lalo kung 12, naku, candidate 'yon for disappearance.  Yon ang mga napapalitan ang mga favorite.  Hindi ba?  Sa Bicol tanggal yon.  Ang lalabas doon Bicolano lang.  Di ba? Kung mahina-hina ang No.12, tanggal na yon.  Mahina-hina ang No.11, tanggal na yon sa mga regions.  Every region has its own favorite.  Papasok na sa sample ballots.  Walang dayaan yon.  But you will not try to eliminate somebody who is impossible to eliminate.  What is your interest?  Loren is No.1, so, she becomes No.3.  Who cares! Maybe, she's a bit unhappy.  But she is still senator, you see!^[40]

Senator Legarda-Leviste also expressed that "it is the fear of some of the other senators that because they are perceived to be the last three or four slot occupants and they could be the target of a 'dagdag-bawas.'"^[41]

That the law, in Sections 5.2 and 5.3, prescribes certain requirements in the publication of surveys and allows the inspection thereof do not suffice to thwart the dangers sought to be avoided by Section 5.4. Election surveys are more in demand as the elections draw closer. The reason is obvious. The public rating of the candidates shifts from time to time over the months.  But a survey taken very close to the election might be taken as indicative of a firm and final tally of the results, giving more motive to fly-by-night pollsters or survey groups controlled by vested interests to manipulate the survey results.  It is conceded that Sections 5.2 and 5.3 affords interested parties an opportunity to examine and analyze the published surveys and to refute or confirm their accuracy.  However, these regulations lose their efficacy during the period contemplated by Section 5.4 because said interested parties would no longer have adequate time to test the veracity of said surveys, especially if they are published, say, a day before the elections.^[42]

Viewed in the light of the legitimate and significant objectives of Section 5.4, it may be seen that its limiting impact on the rights of free speech and of the press is not unduly repressive or unreasonable.  Indeed, it is a mere restriction, not an absolute prohibition, on the publication of election surveys.  It is limited in duration; it applies only during the period when the voters are presumably contemplating whom they should elect and when they are most susceptible to such unwarranted persuasion.  These surveys may be published thereafter.

Our electoral system and processes are not necessarily of the same level of political maturity that countries like the United States and other more developed countries have attained.  It is noteworthy that numerous other countries recognized the deleterious effects on the electoral process by the publication of surveys immediately before the elections.  Accordingly, they impose similar restrictions, although varying as to the periods:  Turkey and Luxembourg, 30 days; South Africa, 42 days; Italy, 28 days; Indonesia, 21 days; Peru, Venezuela and Uruguay, 15 days; Poland, 12 days; France, Hungary, Portugal, Switzerland, Chile, Columbia and Mexico, 7 days; Spain, 5 days; Russia, Australia and Bolivia, 2 days; Fiji, New Zealand, Armenia, Belarus, Bulgaria, Croatia, Khazakstan and Lithuania, 1 day.^[43]

The reasons advanced in support of Section 5.4, far from being matters of mere legislative preferences or beliefs regarding the evils sought to be remedied, sufficiently justify the restriction on such vital rights as the freedoms of speech and of the press.  It bears stressing that it is Congress, not this Court, which his primarily charged with the determination of the need for regulation of such activities. Thus, insofar as the need for regulation of the publication of election surveys within the periods laid down in Section 5.4 is concerned, this Court is in no position to substitute its judgment as to the necessity or desirability of the same for that of Congress.^[44]

IN VIEW OF THE FOREGOING, I vote to DISMISS the petition.

---

^[1] Blo Umpar Adiong vs. Commission on Elections, 207 SCRA 712 (1992); Mutuc vs. Commision on Elections, 36 SCRA 228 (1970).

^[2] Philippine Blooming Mills Employees Organization vs. Philippine Blooming Mills, 51 SCRA 189 (1973).

^[3] Gonzales vs. Commission of Elections, 27 SCRA 835 (1969).

^[4] Article II, Section 26.

^[5] Article V, Section 2.

^[6] Article IX-C, Section 4 reads:

Sec. 4.  The Commission may, during the election period, supervise or regulate the enjoyment or utilization of all franchises or permits for the operation of transportation and other public utilities, media of communication or information, all grants, special privileges, or concessions granted by the Government or any subdivision, agency, or instrumentality thereof, including any government-owned or controlled corporation or its subsidiary. Such supervision or regulation shall aim to ensure equal opportunity, time, and space, and the right to reply, including reasonable, equal rates therefor, for public information campaigns and forums among candidates in connection with the objective of holding free, orderly, honest, peaceful, and credible elections.

^[7] An Act to enhance the Holding of Free, Orderly, Honest, Peaceful and Credible Elections Through Fair Election Practices.

^[8] R.A. No. 9006, Section 1.

^[9] 288 SCRA 447 (1998).

^[10] 207 SCRA 1 (1992).

^[11] 35 SCRA 285 (1970).

^[12] Supra.

^[13] Gitlow v. New York, 268 U.S. 652, 69 L Ed 1138.

^[14] Schenck v. Untied states, 249 U.S. 47, 63 L Ed 470, 473-474.

^[15] ABS-CBN Broadcasting Corp. vs. Commission on Elections, 323 SCRA 811 (2000); Blo Umpar adiong vs. Commission on Elections, supra; Imbong vs. Ferrer, 35 SCRA 28 (1970).

^[16] Supra.

^[17] Id., at 898.

^[18] 339 U.S. 383, 94 L Ed 925.

^[19] Id., at 944.

^[20] Id., at 943.

^[21] KAUPER, CIVIL LIBERTIES AND THE CONSTITUTION, p. 113 cited in Separate Opinion, Castro, J., in Gonzales vs. Commission on Elections, supra.

^[22] 170 SCRA 1 (1989).

^[23] Id., at 8.

^[24] 92 SCRA 476 (1979).

^[25] Id., at 488.

^[26] Eu v. San Francisco Democratic Com., 489 US 214, 103 L Ed 2d 271, 109 S Ct. 1013.

^[27] Burson v. Freeman, 119 L Ed 2d 5.

^[28] Id.; ABS-CBN Broadcasting Corp vs. Commission on Elections, 323 SCRA 811 (2000).

^[29] Gonzales vs. Commission on Elections, 27 SCRA 835 (1969).

^[30] Republic Act No. 9006, Section 2.

^[31] Ibid.

^[32] Senate Journal, Session No. 13, August 21, 2000, pp. 189-190.

^[33] Senate Journal, Session No. 22, October 2, 2000, p. 266.

^[34] Ibid.

^[35] Id., at 267.

^[36] Id., at 267-268.

^[37] Id., at 268.

^[38] Comment of the Solicitor General, pp. 8-11, citing Public Polling in Canada by Claude Emery, at http://www.parl.gc.ca/information/library/PRBpubs.

^[39] Id.

^[40] Transcript of Committee Meetings, Bicameral Conference Committee on the Disagreeing Provisions of Senate Bill No. 1742 and House Bill No. 9000, November 23, 2000, p. 32.

^[41] Id., at 36.

^[42] Alvin Capino, in his column "counterpoint" (today, April 21, 2001) had this to say:

One more reason why survey results for senators should be taken with a grain of salt is the explanation of Felipe Miranda of Pulse Asia about the adjustment pollsters make because of the so-called command votes.

Command votes are the block votes of religious groups like Iglesia ni Cristo and El Shaddai. Members of these groups vote according to the instructions of their leaders.

According to those who attended a recent briefing of Miranda, the head of Pulse Asia places minor weight on the so-called command votes.  Pulse Asia places the command votes at a low 1.5 million votes.  The number, the say, would have no major impact on the election results.

The problem of pollsters is that members of the Iglesia ni Cristo with a voting strength of at least three million do not participate in surveys.  The fact that INC members are not covered by surveys could distort survey results.

A senatorial candidate, for example, who thinks that he safe in, say, his ranking of 8^th or 9^th might suddenly find himself outside the Magic 13 simply because the senatorial candidates below him were supported by the INC and he was not.

^[43] Senate Journal , Session No. 22, October 2, 2000, p. 267.

^[44] See American Communications Association vs. Douds, supra.