This appeal arose in connection with demonstrations planned for the 2004 Democratic National Convention (the Convention). The appellant, the Bl(a)ck Tea Society, seeks review of an order denying its request to modify a designated demonstration zone (DZ) set aside by municipal officials. We summarily affirmed the order on July 26, 2004 (coincident with the start of the Convention). This opinion limns the basis for our ruling.
The facts surrounding this litigation are thoroughly canvassed in the district court’s comprehensive opinion, see Coalition to Protest the Democratic Nat’l Conv. v. City of Boston, 327 F.Supp.2d 61 (D.Mass.2004), and it would serve no useful purpose to rehearse them in exegetic detail. Suffice it to say that the Convention was held at the Fleet Center, in Boston, Massachusetts, on July 26-29, 2004. Security at national political conventions is always tight and that was especially so this year in light of heightened sensitivity to security concerns following the terrorist attacks of September 11, 2001.
Security precautions at the Convention operated on two different levels. The City established a highly secure hard zone in the area immediately surrounding the Fleet Center (a zone for which the United States Secret Service assumed principal responsibility) and a less secure soft zone extending several blocks south in the area commonly known as Bullfinch Triangle. Only candidates, delegates, staff, press, and other specially authorized classes of persons were permitted into the hard zone — and even they had to pass through magnetometers before entering. By contrast, pedestrian access to and through the soft zone was generally unrestricted (al*11though vehicles were not allowed to enter). This dual arrangement left little opportunity for groups wishing to demonstrate to do so within sight and sound of the delegates (especially since chartered buses, which loaded and unloaded within the hard zone, ferried the delegates to and from the Fleet Center).
In an effort to facilitate demonstrators’ access to the delegates, the City established the DZ on the edge of the hard zone and allowed demonstrations within it. The DZ itself was far from a perfect solution. It comprised a heavily secured space, approximately 90 feet by 300 feet, located for the most part underneath unused rail tracks. It was surrounded by two rows of jersey barriers topped with eight-foot chain-link fencing; the perimeter was further surrounded by a semitransparent liquid dispersion mesh fabric; and a widely-woven mesh fabric was hung above the DZ between the rail tracks and the fence. Finally, the City placed coiled razor wire along the edges of the rail tracks in the vicinity of the Fleet Center (including the area above the DZ) in order to inhibit access to the tracks. Although there were three routes of ingress and egress to and from the DZ, the aggregate effect of the security measures was to create an enclosed space that the appellant likens to a pen.
The appellant filed suit in the United States District Court for the District of Massachusetts on July 21, 2004, seeking, inter alia, a preliminary injunction requiring the City to modify the DZ in certain respects.1 The next day, the district judge personally inspected the DZ; held a hearing; entertained an ex parte proffer of evidence from federal authorities concerning security matters;2 and, ruling ore sponte, denied the requested injunction. On July 23, he filed the memorandum opinion previously cited. The Bl(a)ck Tea Society appealed. We have jurisdiction under 28 U.S.C. § 1292(a)(1).
The district court determines whether to issue a preliminary injunction by weighing four factors: “(1) the likelihood of success on the merits; (2) the potential for irreparable harm if the injunction is denied; (3) the balance of relevant impositions, i.e., the hardship to the nonmovant if enjoined as contrasted with the hardship to the movant if no injunction issues; and (4) the effect (if any) of the court’s ruling on the public interest.” Charlesbank Equity Fund II v. Blinds to Go, Inc., 370 F.3d 151, 162 (1st Cir.2004) (quoting Ross-Simons of Warwick, Inc. v. Baccarat, Inc., 102 F.3d 12, 15 (1st Cir.1996)). We review the district court’s grant or denial of a preliminary injunction for abuse of discretion. Id. at 158. “This is a deferential standard of review, and the deference that it entails is most appropriate with respect to issues of judgment and the balancing of conflicting factors.” Id. Within this sphere, the district court’s conclusions of law are reviewed de novo and its findings of fact are reviewed for clear error. New Comm Wireless Servs., Inc. v. SprintCom, Inc., 287 F.3d 1, 9 (1st Cir.2002).
Freedom of expression, especially expression of political views, ranks near the top of the hierarchy of constitutional *12rights. See Cohen v. California, 403 U.S. 15, 24, 91 S.Ct. 1780, 29 L.Ed.2d 284 (1971). That freedom “is designed and intended to remove governmental restraints from the arena of public discussion, putting the decision as to what views shall be voiced largely into the hands of each of us, in the hope that use of such freedom will ultimately produce a more capable citizenry and more perfect polity and in the belief that no other approach would comport with the premise of individual dignity and choice upon which our political system rests.” Id.
The right to freedom of expression is secured principally by the First Amendment. U.S. Const, amend. I. Despite the importance of that right, the prophylaxis of the First Amendment is not without limits. Reasonable restrictions as to the time, place, and manner of speech in public fora are permissible, provided that those restrictions “are justified without reference to the content of the regulated speech, ... are narrowly tailored to serve a significant governmental interest, and ... leave open ample alternative channels for communication of the information.” Ward v. Rock Against Racism, 491 U.S. 781, 791, 109 S.Ct. 2746, 105 L.Ed.2d 661 (1989) (quoting Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 293, 104 S.Ct. 3065, 82 L.Ed.2d 221 (1984)).
A regulation is narrowly tailored if “the means chosen are not substantially broader than necessary to achieve the government’s interest.” Id. at 800, 109 S.Ct. 2746. To satisfy this benchmark, a regulation need not be the least restrictive alternative available to the government. Id. at 798-99, 109 S.Ct. 2746. Put another way, the validity of time, place, or manner regulations is not subject to “ ‘a judge’s agreement with the responsible decisionmaker concerning the most appropriate method for promoting significant government interests’ or the degree to which those interests should be promoted.” Id. at 800, 109 S.Ct. 2746 (quoting United States v. Albertini, 472 U.S. 675, 689, 105 S.Ct. 2897, 86 L.Ed.2d 536 (1985)).
An inquiry into the validity of time-place-manner regulations generally commands what we have termed “intermediate scrutiny.” Nat’l Amusements, Inc. v. Town of Dedham, 43 F.3d 731, 736 (1st Cir.1995). In this case, however, the appellant endeavors to ratchet up this level of scrutiny by characterizing the security measures (particularly the total prohibition of demonstrations in the hard zone) as a prior restraint on speech. We reject this approach: here, the City has not sought to prevent speech, but, rather, to regulate the place and manner of its expression. The Supreme Court has explicitly rejected attempts to analyze security-based timeplaee-manner restrictions as prior restraints, see, e.g., Hill v. Colorado, 530 U.S. 703, 733-34, 120 S.Ct. 2480, 147 L.Ed.2d 597 (2000); Schenck v. Pro-Choice Network, 519 U.S. 357, 374 n. 6, 117 S.Ct. 855, 137 L.Ed.2d 1 (1997); Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753, 763 n. 2, 114 S.Ct. 2516, 129 L.Ed.2d 593 (1994), and those cases are controlling here. If content-neutral prohibitions on speech at certain places were deemed pri- or restraints, the intermediate standard of review prescribed in the time-place-manner jurisprudence would be eviscerated.
Having settled upon the appropriate level of scrutiny, we move the inquiry forward. The answers to some issues are indisputable: the challenged security precautions are plainly content-neutral and there can be no doubting the substantial government interest in the maintenance of security at political conventions. What remains, therefore, is whether the City’s array of security precautions were narrowly *13tailored and whether those precautions left ample alternative avenues of communication.
We deem it appropriate to start a narrow tailoring analysis by examining the speech-related burdens that the challenged regulation imposes. It cannot be gainsaid that the security measures attendant to the Convention dramatically limited the possibilities for communicative intercourse between the demonstrators and the delegates. The measures allowed no opportunity for physical interaction (such as the distribution of leaflets) and severely curtailed any chance for one-on-one conversation. Visual communication using signs or other media was not prevented but was hampered to some extent by the cramped space and the mesh screening. And while the direct limits on aural communication seem minor, even this form of interaction may have been less effective because of the restrictions on other modes of expression. In sum, the challenged regulation imposed a substantial burden on free expression.
We turn next to the City’s goal, mindful that the government’s judgment as to the best means for achieving its legitimate objectives deserves considerable respect. Rock Against Racism, 491 U.S. at 798-99, 109 S.Ct. 2746. Here, the City’s overall goal was to maintain security at the Convention. But security simpliciter is too broad a rubric to be useful in this analysis. Security is not a talisman that the government may invoke to justify any burden on speech (no matter how oppressive). Thus, the question of narrow tailoring must be decided against the backdrop of the harms that a particular set of security measures are designed to forfend.
The City claims that the risk of harm was substantial. It designed the elaborate security measures here at issue in light of recent past experience with large demonstrations, including those at the 2000 Democratic National Convention in Los Ange-les. The double ranks of fencing were meant to deter attempts to break through the fence; the liquid dispersal mesh was intended to protect the delegates from being sprayed with liquids; and the overhead netting was added to prevent demonstrators from hurling projectiles. Conduct of this type admittedly has occurred at a number of recent protests.
The appellant points out, correctly, that there is no evidence in the record that the City had information indicating that demonstrators intended to use such tactics at the Convention.3 Building on this foundation, the appellant maintains that the City may not implement security requirements that substantially burden speech on the basis of unrelated past experiences. It further argues that most protesters do not engage in such conduct, and that the First Amendment rights of the majority of protesters should not be curtailed because of the potential for unlawful actions by a rowdy minority. In the absence of event-specific threat evidence, the appellant says, the City should have been limited to arresting miscreants and punishing unlawful conduct after it occurred.
We do not believe a per se rule barring the government from using past experience to plan for future events is consistent with the approach adopted in the Court’s time-place-manner jurisprudence. See Hill, 530 U.S. at 728-29, 120 S.Ct. 2480 (relying in part on past experi*14ence to find time-place-manner restrictions narrowly tailored); Rock Against Racism, 491 U.S. at 796-97, 109 S.Ct. 2746 (upholding restrictions enacted on the basis of earlier experiences with noise pollution in Central Park). The question is not whether the government may make use of past experience — it most assuredly can — but the degree to which inferences drawn from past experience are plausible. While a government agency charged with public safety responsibilities ought not turn a blind eye to past experience, it likewise ought not impose harsh burdens on the basis of isolated past events. And in striking this balance, trial courts should remember that heavier burdens on speech must, in general, be justified by more cogent evidentiary predicates.
On this hastily assembled record, the quantum of “threat” evidence was sufficient to allow the trier to weigh it in the balance. For now, we do not purpose to determine how that factor should be weighed either in future cases or in connection with better-developed factual proffers. We instead assess only whether the district court’s balance of this and other factors was so unreasonable as to constitute an abuse of discretion. We conclude that it did not, and, thus, we uphold the district court’s determination that the security measures undertaken by the City, though extreme, were nonetheless narrowly tailored.
Having left intact the district court’s conclusions anent narrow tailoring, we briefly address the availability of alternative avenues for expression. The district court perspicaciously noted that many other opportunities for demonstrations existed in the vicinity of the Fleet Center and throughout Boston. The City allowed informal demonstrations within the soft zone without a permit so long as those demonstrations involved fewer than 20 people (and allowed up to 50 people with a permit). Several other public spaces throughout Boston remained available for demonstrations, subject to existing regulations and content-neutral permitting requirements.
The appellant’s chief rejoinder is that these alternatives were not sufficient because none of them were within sight and sound of the delegates assembled at the Fleet Center. We disagree with that premise: the DZ did provide an opportunity for expression within sight and sound of the delegates, albeit an imperfect one. There are, moreover, two other pertinent considerations. First, although the opportunity to interact directly with the body of delegates by, say, moving among them and distributing literature, would doubtless have facilitated the demonstrators’ ability to reach their intended audience, there is no constitutional requirement that demonstrators be granted that sort of particularized access. Second, we think that the appellant’s argument greatly underestimates the nature of modern communications. At a high-profile event, such as the Convention, messages expressed beyond the first-hand sight and sound of the delegates nonetheless have a propensity to reach the delegates through television, radio, the press, the internet, and other outlets. On this record, then, we cannot say that the district court erred in concluding that viable alternative means existed to enable protesters to communicate their messages to the delegates.
Let us be perfectly clear: this is a close and difficult case. The district court, however, dealt with matters at first hand and concluded that the appellant had not shown a likelihood of success on the merits. See Coalition to Protest, 327 F.Supp.2d at 69. On this record and at this preliminary litigation stage, we find that conclusion reasonable.
*15We have frequently said that likelihood of success is an essential prerequisite for the issuance of a preliminary injunction. See New Comm Wireless Servs., 287 F.3d at 9 (“The sine qua non of this four-part inquiry is likelihood of success on the merits: if the moving party cannot demonstrate that he is likely to succeed in his quest, the remaining factors become matters of idle curiosity.”); Weaver v. Henderson, 984 F.2d 11, 12 (1st Cir.1993) (similar). In the interests of completeness, however, we briefly mention the remaining three parts of the preliminary injunction calculus.
A burden on protected speech always causes some degree of irreparable harm. See Elrod v. Burns, 427 U.S. 347, 373-74, 96 S.Ct. 2673, 49 L.Ed.2d 547 (1976). Here, however, the safety, security, and logistical concerns voiced by the City were real, and the district court was correct in giving those concerns due consideration. Thus, the balance of harms is inconclusive in this case. Similarly, the public interest cuts both ways. On the one hand, freedom of expression, especially freedom of political expression, is vital to the health of our democracy. On the other hand, making public safety a reality and ensuring that important political events are able to proceed normally are also valuable. Moreover, a determination of the public interest necessarily encompasses the practical effects of granting or denying preliminary injunctive relief. Here, the district court was constrained by the physical limitations of the Fleet Center venue4 and by the timing of the suit (which, despite considerable advance notice of the planned security measures, was brought less than a week before the Convention was scheduled to open).
We mention these constraints because it is readily evident that they sharply limited the remedial options available to the district court. Considering these factors in the aggregate, we find the equities closely matched, and we come away sharing the district court’s concern with the impracticability of eleventh-hour injunctive relief. See Coalition to Protest, 327 F.Supp.2d at 77.
We need go no further. With the Convention looming and with few options at its disposal, we think the district court’s resolution of the preliminary injunction request was fully supportable. The court did yeoman’s work in holding a prompt hearing, mastering the complexity of the issues and the physical shortcomings of the site, deciding the motion in a timeous fashion, and writing a thoughtful rescript that explained its findings and its rationale. Dealing with the record as it stands, the temporal constraints under which the district court labored, and the deferential standard of review, we have no principled choice but to uphold the challenged order.5
Affirmed.
. The litigation below also challenged certain other security measures (e.g., restrictions on marching, etc.) that are not at issue in this appeal. By the same token, the original lead plaintiff, the Coalition to Protest the Democratic National Convention, is not a party here.
. The district court eventually determined that the information it had received ex parte was not necessary to its decision and, thus, did not consider it. See Coalition to Protest, 111 F.Supp.2d at 75.
. That statement is true, so long as one ignores the ex parte presentation by the federal government. See supra note 2. Like the district court, we are content, for purposes of this case, to leave that evidence to one side. The difficult issues raised by the use of such proffers will have to be considered in future cases.
. The district court supportably found that these physical constraints made it impossible to either relocate or substantially enlarge the DZ. Coalition to Protest, 327 F.Supp.2d at 68. The DZ could not be expanded eastward into the loading area for buses without jeopardizing the ability to move people and vehicles safely and efficiently. Id. at 74. Finally, there was no practical way either to remove the overhead rail tracks or to relocate the DZ from beneath the tracks. Id. at 76.
. We are not faced with the question of whether the same result would be supportable in less tumultuous times or on a better-developed record, cf. Cohen v. Brown Univ., 991 F.2d 888, 902 (1st Cir.1993) (explaining that "a district court's conclusions at the preliminary injunction stage are only attempts to predict probable outcomes”), and, therefore, we express no view on these matters.