concurring in the judgment in part and dissenting in part.
Although I agree that aspects of the honoraria ban run afoul of the First Amendment, I write separately for two reasons. First, I wish to emphasize my understanding of how our precedents, beginning with Pickering v. Board of Ed. of Township High School Dist. 205, Will Cty., 391 U. S. 563 (1968), and culminating in its most recent application, Waters v. Churchill, 511 U. S. 661 (1994), direct the Court’s conclusion. Second, I write to express my disagreement with the Court’s remedy, which in my view paints with too broad a brush.
I
The time-tested Pickering balance, most recently applied in Waters, provides the governing framework for analysis of all manner of restrictions on speech by the government as employer. Under Pickering, the Court must balance “the interests of the [employee], as a citizen, in commenting upon matters of public concern and the interest of the [government], as an employer, in promoting the efficiency of the public services it performs through its employees.” 391 U. S., at 568. In contrast to some of our prior decisions, this case presents no threshold question whether the speech is of public, or merely private, concern. Respondents challenge the ban as it applies to off-hour speech bearing no nexus to Government employment — speech that by definition does not relate to “internal office affairs” or the employee’s status as an employee. Cf. Connick v. Myers, 461 U. S. 138, 149 (1983).
*481In setting out the employees’ interests in this case, the Court draws a meaningful distinction between the ex ante prohibition of certain kinds of speech and the ex post punishment of discrete, unforeseeable disturbances. See ante, at 466-468. There is some force to the Court’s observation, because ex ante rules, in contrast to ex post punishments, carry risks of overinclusiveness and underinclusiveness. Nevertheless, reliance on the ex ante!ex post distinction is not a substitute for the case-by-case application of Pickering. There are many circumstances in which the Government as employer is likely to prefer the codification of its policies as workplace rules (which, incidentally, provide notice to employees) to the ad hoc, on-the-job reactions that have been standard fare in many of our employment cases. In most such circumstances, the Government will be acting well within its bounds. I see little constitutional difference, for example, between a rule prohibiting employees from being “‘rude to customers,’” see Waters, supra, at 673, and the upbraiding or sanctioning of an employee post hoc for isolated acts of impudence. To draw the line based on a distinction between ex ante rules and ex post punishments, in my view, overgeneralizes and threatens undue interference with “the government’s mission as employer,” 511 U. S., at 674.
Given the breadth and intrusiveness of the honoraria ban in this case, however, I agree with the Court that significant weight must be placed on the employees’ side of the scale in the Pickering balance. We recognized in Simon & Schuster, Inc. v. Members of N. Y. State Crime Victims Bd., 502 U. S. 105, 115 (1991), that the imposition of financial burdens may have a direct effect on incentives to speak. See also Minneapolis Star & Tribune Co. v. Minnesota Comm’r of Revenue, 460 U. S. 575, 585 (1983) (observing that the threat of burdensome taxes “can operate as effectively as a censor to check critical comment”). Although the honoraria ban certainly does not curtail all of the non-work-related speech *482of the nearly two million members of respondent class, it doubtless inhibits some speech on matters of substantial public interest. In my view, the impact of the honoraria ban upon this class of employees’ interests in speaking out as citizens, rather than as employees, cannot be gainsaid.
The Government advances two categories of interests in support of the honoraria ban. First, the Government submits its interests in promoting the efficiency of public service and in avoiding the appearance of impropriety created by abuse of the practice of receiving honoraria. We have credited these objectives as both salutary and significant on several occasions. See, e. g., Federal Election Comm’n v. National Right to Work Comm., 459 U. S. 197, 210 (1982); First Nat. Bank of Boston v. Bellotti, 435 U. S. 765, 788, n. 26 (1978); Buckley v. Valeo, 424 U. S. 1, 26-29 (1976). Although they lend support to the Government’s efforts to put a stop to honoraria paid for work-related speech, these interests have less force in justifying a ban that prohibits honoraria paid for speech on matters wholly unrelated to the workplace. Perhaps recognizing this, the Government maintains that it has an additional interest in resisting evasion of its rule and sparing administrative resources. According to the Government, apparently innocuous payments may be made for illicit purposes, and the difficulty inherent in distinguishing the innocuous from the illicit mandates a broad prophylactic ban.
Balancing is difficult to undertake unless one side of the scale is relatively insubstantial. The Government argues that the Court should defer broadly to its determination that the benefits of the ban outweigh its costs. The Government relies on Waters, in which a plurality of the Court observed that “we have consistently given greater deference to government predictions of harm used to justify restriction of employee speech than to predictions of harm used to justify restrictions on the speech of the public at large.” 511 U. S., at 673. But this principle has its limits, as the Waters plu*483rality went on to recognize. As the magnitude of intrusion on employees’ interests rises, so does the Government’s burden of justification. Cf. Connick, 461 U. S., at 150 (“[T]he State’s burden in justifying a particular discharge varies depending upon the nature of the employee’s expression”); id., at 151-152 (finding “a wide degree of deference” appropriate where employee’s speech touched only peripherally on matters of public concern). Thus, in Waters, the plurality noted that “[i]n many such situations the government may have to make a substantial showing that the speech is, in fact, likely to be disruptive before it may be punished.” 511 U. S., at 674. This case presents one such situation.
The Court makes a persuasive case that the Government has made no such showing and that the Government’s asserted interests are insufficiently weighty to justify a ban that curbs so much employee speech. See ante, at 470-477. In promulgating the ban, Congress relied on the reports of two blue-ribbon panels that suggested the prudence of a wide-ranging prohibition. Neither report noted any problems, anecdotal or otherwise, stemming from the receipt of honoraria by rank-and-file Executive Branch employees. Neither report, therefore, tends to substantiate the Government’s administrative efficiency argument, which presumes that abuses may be so widespread as to justify a prophylactic rule. Congress assuredly was inspired by a worthy interest, but it made no effort to establish a connection between its interest and the large-scale inhibition of non-work-related speech by the respondent class. Our cases do not support the notion that the bare assertion of a laudable purpose justifies wide-ranging intrusions on First Amendment liberties. In Civil Service Common v. Letter Carriers, 413 U. S. 548 (1973), perhaps the closest analogue to this case, we upheld provisions of the Hatch Act, 5 U. S. C. § 7324(a)(2), against a First Amendment challenge only after canvassing nearly a century of concrete experience with the evils of the political spoils system. Cf. FCC v. League of Women Voters of Cal., *484468 U. S. 364, 401, n. 27 (1984) (noting that the Hatch Act “evolved over a century of governmental experience with less restrictive alternatives that proved to be inadequate to maintain the effective operation of government”).
I also agree with the Court that loopholes in the current ethical regime tend to cast doubt upon the gravity of the problem of honoraria abuse, or at least doubt upon the weight of the problem as perceived by Congress and the Office of Government Ethics (OGE). Cf. City of Ladue v. Gilleo, 512 U. S. 43, 52-53 (1994). Thus, in a provision that detracts seriously from the Government’s administrative convenience rationale, the statute permits the author of a series of three speeches or publications (but not of a single speech or publication) to receive an honorarium if the series has no nexus to Government employment. See 5 U. S. C. App. § 505(3) (1988 ed., Supp. V). Under OGE regulations, employees may receive honoraria for poems, but not for speeches on poetry. See 5 CFR § 2636.203(d) (1992). Employees may be compensated for writing chapters in books, but not for writing the same piece if published as an article. Ibid. Congress is not required to address every aspect of a problem whenever it decides to act. But the patchwork nature of this regime might reasonably lead one to question the strength of the Government’s asserted interests in a broad, prophylactic ban.
The Government, when it acts as employer, possesses substantial leeway; in appropriate circumstances, it may restrain speech that the Constitution would otherwise protect. The Government’s prerogatives in this area stem from its public-serving mission as employer. As the plurality observed last year in Waters, “[w]hen someone who is paid a salary so that she will contribute to an agency’s effective operation begins to do or say things that detract from the agency’s effective operation, the government employer must have some power to restrain her.” 511 U. S., at 675. In this case, however, the Government has exceeded the limits *485of its latitude. The bare assertion of interest in a wide-ranging prophylactic ban here, without any showing that Congress considered empirical or anecdotal data pertaining to abuses by lower echelon Executive Branch employees, cannot suffice to outweigh the substantial burden on the 1.7 million affected employees. I agree with the Court that § 501 is unconstitutional to the extent that it bars this class of employees from receiving honoraria for expressive activities that bear no nexus to Government employment.
II
The class before us is defined by pay scale, not by its members’ propensity to write articles without nexus to Government employment. As to any member of the class, the hono-raria ban may have unconstitutional applications. But the ban may be susceptible of constitutional application to every member of the class, as well. We do not decide the question — a far harder case for respondents, in my view— whether it is constitutional to apply the honoraria ban to speech by this class that bears a relationship to Government employment. I believe that the Court overlooks this nuance when it enjoins all enforcement of § 501 against the class, any one of whose members may, in the future, receive honoraria for work-related activities. I would give respondents relief tailored to what they request: invalidation of the statute insofar as it applies to honoraria they receive for speech without nexus to Government employment. See Brief for Respondents 45-46 (noting that respondents’ central aim “could also be achieved by a remedy similar to the one urged by the government — by holding the ban invalid as applied to respondents’ writing and speaking activities [that] have no nexus to their federal employment”).
I agree with the Court’s assertion that the remedy in this case is properly limited to the parties before us. We have long characterized overbreadth analysis as “strong medicine,” to be “employed by the Court sparingly and only as a *486last resort.” Broadrick v. Oklahoma, 413 U. S. 601, 613 (1973). Accordingly, we have observed that “[i]t is not the usual judicial practice, . . . nor do we consider it generally desirable, to proceed to an overbreadth issue unnecessarily.” Board of Trustees of State Univ. of N. Y. v. Fox, 492 U. S. 469, 484-485 (1989); see also New York v. Ferber, 458 U. S. 747, 768 (1982) (footnotes omitted) (“By focusing on the factual situation before us, and similar cases necessary for development of a constitutional rule, we face ‘flesh-and-blood’ legal problems with data ‘relevant and adequate to an informed judgment’”). The class-before us, though defined broadly, consists of people who stand to benefit generally from the invalidation of this statute as applied. Because, as respondents freely admit, their central objective may be achieved by an as-applied challenge, see Brief for Respondents 45-46, I agree that the Court has little warrant to venture into the more difficult and uncertain overbreadth terrain.
Like the dissent, however, I believe that the Court imposes an unduly broad remedy when it enjoins enforcement of the entire provision as to respondent class. There is a commonsense appeal to the Government’s argument that, having deemed a particular application of a statute unconstitutional, a court should not then throw up its hands and despair of delineating the area of unconstitutionality. See Brief for United States 38; Reply Brief for United States 15. On its face, the statute contains an exception for a “series” of speeches, appearances, and articles unless “the subject matter is directly related to the individual’s official duties or the payment is made because of the individual’s status with the Government.” 5 U. S. C. App. §505(3) (1988 ed., Supp. V). I see no reason why the nexus principle underlying this general provision cannot serve as the appropriate remedial line.
The Court assumes that it would venture into judicial legislation were it to invalidate the provision as it applies to *487no-nexus speech. But it is equally, if not more, inconsistent with congressional intent to strike a greater portion of the statute than is necessary to remedy the problem at hand. Although our jurisprudence in this area is hardly a model of clarity, this Court has on several occasions declared a statute invalid as to a particular application without striking the entire provision that appears to encompass it. In United States v. Grace, 461 U. S. 171 (1983), for example, the Court addressed the constitutionality of a federal statute making it unlawful to “parade, stand, or move in processions or assemblages in the Supreme Court Building or grounds,” or “to display therein any flag, banner, or device designed or adapted to bring into public notice any party, organization, or movement.” 40 U. S. C. §13k. Finding the statute’s extension to public fora inadequately justified, the Court deemed the provision unconstitutional as applied to the sidewalks surrounding the Supreme Court. Grace, supra, at 183. In Tennessee v. Garner, 471 U. S. 1 (1985), the Court invalidated a state provision permitting police officers to use “ ‘all the necessary means to effect the arrest’ ” of a fleeing or forcibly resisting defendant only insofar as it authorized the use of deadly force against an unarmed, nondangerous suspect. Id., at 4, 22. In Brockett v. Spokane Arcades, Inc., 472 U. S. 491 (1985), the Court invalidated a state obscenity statute “only insofar as the word ‘lust’ is taken to include normal interest in sex.” Id., at 504-505.
In Brockett, the Court declared: “[W]here the parties challenging the statute are those who desire to engage in protected speech that the overbroad statute purports to punish, or who seek to publish both protected and unprotected mate-ria^,] . . . [t]he statute may forthwith be declared invalid to the extent that it reaches too far, but otherwise left intact.” Id., at 504. Of course, we also noted that “[pjartial invalidation would be improper if it were contrary to legislative intent in the sense that the legislature had passed an inseverable Act or would not have passed it had it known the *488challenged provision was invalid.” Id., at 506. In Brockett itself, the state statute contained a severability clause announcing that the remainder of the Act would continue in effect should “any provision of this act or its application to any person or circumstance” be held invalid. Id., at 506, n. 14. Because the opinions in Grace and Garner made no mention of the existence of statutory severability clauses, these cases can perhaps best be explained as having involved implied severability. After delineating the range of the statute’s impermissible applications, the Court implicitly concluded that the legislatures at issue — Congress in Grace and the Tennessee Legislature in Garner — would have preferred that the remainder of the statutes continue intact. These cases are entirely consistent with our severability precedents, in which we have held that “Congress’ silence is just that — silence—and does not raise a presumption against severability.” Alaska Airlines, Inc. v. Brock, 480 U. S. 678, 686 (1987).
As the Court of Appeals noted below, “[s]ection 501(b) does not contain a severability clause, and the legislative history yields no direct evidence of intent concerning severability.” 990 F. 2d 1271, 1278 (CADC 1993). Under Alaska Airlines, this fact is by no means dispositive; the operative question instead is whether Congress would have promulgated § 501(b) had it known that it could not lawfully proscribe hon-oraria of employees below GS-16 for activities unrelated to Government employment. See 480 U. S., at 684; Stern, Separability and Separability Clauses in the Supreme Court, 51 Harv. L. Rev. 76, 82-83 (1937) (in dealing with the question of severable applications, the Court asks “whether the legislative body would intend the law to be given effect to whatever extent was constitutionally possible”). I think this question can be answered in the affirmative here. In severing this particular application, we leave the provision intact as to high-level Executive Branch employees and to the Legislative and Judicial Branches. Common sense suggests, and *489legislative history confirms, that these Government employees, not the parties to this case, were the statute’s principal targets. See, e. g., 990 F. 2d, at 1278 (describing floor debates at which speakers continually referred to abuses by “Members of Congress” as the impetus for reform). As for employees below GS-16, it seems to me clear that Congress would have barred tax examiners from receiving honoraria for lectures on tax policy even if it could not bar the same examiners from receiving honoraria for articles on low-cholesterol cooking. The balance of the provision serves a laudable purpose and is capable of functioning independently once the improper application is excised. Cf. Alaska Airlines, supra, at 684.
In sum, I agree with the Court that §501 is unconstitutional insofar as it bars the respondent class of Executive Branch employees from receiving honoraria for non-work-related speeches, appearances, and articles. In contrast to the Court, I would hold § 501 invalid only to that extent.
Chief Justice Rehnquist, with whom Justice Scalia and Justice Thomas join,dissenting.
I believe that the Court’s opinion is seriously flawed in two respects. First, its application of the First Amendment understates the weight that should be accorded to the governmental justifications for the honoraria ban and overstates the amount of speech that actually will be deterred. Second, its discussion of the impact of the statute that it strikes down is carefully limited to only a handful of the most appealing individual situations, but when it deals with the remedy it suddenly shifts gears and strikes down the statute as applied to the entire class of Executive Branch employees below grade GS-16. I therefore dissent.
I
In 1991, in the aftermath of recommendations by two distinguished commissions, Congress adopted its present ban *490on the receipt of honoraria. Congress defined an “honorarium” as
“a payment of money or any thing of value for an appearance, speech or article (including a series of appearances, speeches, or articles if the subject matter is directly related to the individual’s official duties or the payment is made because of the individual’s status with the Government) by a Member, officer or employee, excluding, any actual and necessary travel expenses incurred by such individual (and one relative) to the extent that such expenses are paid or reimbursed by any other person, and the amount otherwise determined shall be reduced by the amount of any such expenses to the extent that such expenses are not paid or reimbursed.” 5 U. S. C. App. §505(3) (1988 ed., Supp. V).
The ban neither prohibits anyone from speaking or writing, nor does it penalize anyone who speaks or writes; the only stricture effected by the statute is a denial of compensation.
In Simon & Schuster, Inc. v. Members of N. Y. State Crime Victims Bd., 502 U. S. 105 (1991), we evaluated the constitutionality of New York’s “Son of Sam” law, which regulated an accused or convicted criminal’s receipt of income generated by works that described his crime. Id., at 108. We concluded that the law implicated First Amendment concerns because it “impose[d] a financial disincentive only on speech of a particular content.” Id., at 116. Because the “Son of Sam” law was content based, we required the State to demonstrate that the regulation was necessary to serve a compelling state interest and was narrowly drawn to achieve that end. Id., at 118. We determined that the State had failed to meet its burden because the statute was overbroad. Id., at 123.
Unlike the law at issue in Simon & Schuster, the honoraria ban is neither content nor viewpoint based. Ante, at 468; cf. Ward v. Rock Against Racism, 491 U. S. 781, 791 (1989); *491Renton v. Playtime Theatres, Inc., 475 U. S. 41, 47-48 (1986). As a result, the ban does not raise the specter of Government control over the marketplace of ideas. Cf. Simon & Schuster, supra, at 116. To the extent that the honoraria ban implicates First Amendment concerns, the proper standard of review is found in our cases dealing with the Government’s ability to regulate the First Amendment activities of its employees.
A public employee does not relinquish First Amendment rights to comment on matters of public interest by virtue of government employment. See Pickering v. Board of Ed. of Township High School Dist. 205, Will Cty., 391 U. S. 563, 568 (1968); Connick v. Myers, 461 U. S. 138, 140 (1983). We have emphasized, however, that “the State’s interests as an employer in regulating the speech of its employees ‘differ significantly from those it possesses in connection with regulation of the speech of the citizenry in general.’” Ibid. (quoting Pickering, supra, at 568). The proper resolution of these competing interests requires “ ‘a balance between the interests of the [employee], as a citizen, in commenting upon matters of public concern and the interest of the State, as an employer, in promoting the efficiency of the public services it performs through its employees.’ ” 461 U. S., at 140 (quoting Pickering, supra, at 568). Just last Term, a plurality of the Court explained:
“The key to First Amendment analysis of government employment decisions, then, is this: The government’s interest in achieving its goals as effectively and efficiently as possible is elevated from a relatively subordinate interest when it acts as sovereign to a significant one when it acts as employer. The government cannot restrict the speech of the public at large just in the name of efficiency. But where the government is employing someone for the very purpose of effectively achieving its goals, such restrictions may well be appropriate.” Waters v. Churchill, 511 U. S. 661, 675 (1994).
*492In conducting this balance, we consistently have given substantial weight to government employers’ reasonable predictions of disruption, even when the speech involved was on a matter of public concern. Id., at 673-674 (plurality opinion). As we noted in Connick, “ ‘the Government, as an employer, must have wide discretion and control over the management of its personnel and internal affairs.’” 461 U. S., at 151 (quoting Arnett v. Kennedy, 416 U. S. 134, 168 (1974) (Powell, J., concurring in part and concurring in result)).
These principles are reflected in our cases involving governmental restrictions on employees’ rights to engage in partisan political activity.1 In Public Workers v. Mitchell, 330 U. S. 75 (1947), we examined § 9(a) of the Hatch Act, which prohibited officers and employees in the Executive Branch of the Federal Government, with exceptions, from taking “ ‘any active part in political management or in political campaigns.’” Id., at 78. We analyzed §9(a)’s strictures as applied to the partisan political activities of an industrial employee at the United States Mint, and concluded that “[f]or regulation of employees it is not necessary that the act regulated be anything more than an act reasonably deemed by Congress to interfere with the efficiency of the public service.” Id., at 101. Despite the fact that §9(a) barred three-million public employees from taking “effective part in campaigns that may bring about changes in their lives, their fortunes, and their happiness,” id., at 107 (Black, J., dissenting), we held that if in Congress’ judgment “efficiency may be best obtained by prohibiting active participation by classi-*493fled employees in politics as party officers or workers,” there was no constitutional objection, id., at 99.
More than 25 years later, we again addressed the constitutionality of §9(a) of the Hatch Act. In Civil Service Comm’n v. Letter Carriers, 413 U. S. 548 (1973), we “unhesitatingly reaffirmed] the Mitchell holding,” id., at 556, because “neither the First Amendment nor any other provision of the Constitution invalidated] a law barring this kind of partisan political conduct by federal employees,” ibid. We applied the balancing approach set forth in Pickering to the Hatch Act’s sweeping limitation on partisan political activity, and determined that the balance struck by Congress was “sustainable by the obviously important interests sought to be served by the limitations on partisan political activities now contained in the Hatch Act.” 413 U. S., at 564. We concluded that “[p]erhaps Congress at some time w[ould] come to a different view of the realities of political life and Government service,” but we were in no position to dispute Congress’ current view of the matter. Id., at 567.
Although protection of employees from pressure to perform political chores certainly was a concern of the Hatch Act, see ante, at 471, it was by no means the only, or even the most important, concern.2 See Letter Carriers, supra, at 566. Rather, the Court recognized that a major thesis of the Hatch Act was that
“to serve this great end of Government — the impartial execution of the laws — it is essential that federal employees ... not take formal positions in political parties, not undertake to play substantial roles in partisan political campaigns, and not run for office on partisan political tickets. Forbidding activities like these will reduce the *494hazards to fair and effective government.” 413 U. S., at 565.
The Court emphasized that “it is not only important that the Government and its employees in fact avoid practicing political justice, but it is also critical that they appear to the public to be avoiding it, if confidence in the system of representative Government is not to be eroded to a disastrous extent.” Ibid. Thus, the Hatch Act served as a safeguard to both the actual and perceived impartiality and effectiveness of the Federal Government. See Mitchell, supra, at 95-96; Letter Carriers, supra, at 564-567.
Applying these standards to the honoraria ban, I cannot say that the balance that Congress has struck between its interests and the interests of its employees to receive compensation for their First Amendment expression is unreasonable. Cf. Letter Carriers, supra, at 564; Pickering, 391 U. S., at 568.
The Court largely ignores the Government’s foremost interest — prevention of impropriety and the appearance of impropriety — by focusing solely on the burdens of the statute as applied to several carefully selected Executive Branch employees whose situations present the application of the statute where the Government’s interests are at their lowest ebb: a mail handler employed by the Postal Service who lectured on the Quaker religion; an aerospace engineer who lectured on black history; a microbiologist who reviewed dance performances; and a tax examiner who wrote articles about the environment. Ante, at 461-462. Undoubtedly these are members of the class, but they by no means represent the breadth of the class which includes all “‘employee[s]’ . . . below grade GS-16, who—but for 5 U. S. C. app. 501(b)— would receive ‘honoraria’, as defined in 5 U. S. C. app. 505(3).” App. 124-125. Nothing in the class certification limits the receipt of honoraria to the activities engaged in by the several employees discussed by the Court. See, e. g., *495ante, at 463, n. 6. This artificially narrow prism of class members, however, is the focus of the Court’s entire First Amendment discussion.
The class definition speaks of anyone who would receive an honorarium but for the statute. App. 124-125. An unknown number of these individuals would receive honoraria where there is a nexus between their speech and their Government employment. There is little doubt that Congress reasonably could conclude that its interests in preventing impropriety and the appearance of impropriety in the federal work force outweigh the employees’ interests in receiving compensation for expression that has a nexus to their Government employment. Cf. Federal Election Comm’n v. National Right to Work Comm., 459 U. S. 197, 210 (1982) (“The governmental interest in preventing both actual corruption and the appearance of corruption of elected representatives has long been recognized”).
The Court relies on cases involving restrictions on the speech of private actors to argue that the Government is required to produce “evidence of misconduct related to hono-raria in the vast rank and file of federal employees below grade GS-16.” Ante, at 472; ante, at 475-476, and n. 21.3 The Court recognizes, however, that we “ ‘have consistently given greater deference to government predictions of harm used to justify restriction of employee speech than to predictions of harm used to justify restrictions on the speech of the public at large.’” Ante, at 475, n. 21 (quoting Waters, 511 U. S., at 673 (plurality opinion)).
*496Prior to enactment of the current honoraria ban, Congress was informed by two distinguished commissions that its previous limitations on honoraria were inadequate. The 1989 Quadrennial Commission recommended that “Congress enact legislation abolishing the practice of accepting honoraria in all three branches.” Fairness for Our Public Servants: Report of the 1989 Commission on Executive, Legislative and Judicial Salaries vi (Dec. 1988). To Serve With Honor: Report of the President’s Commission on Federal Ethics Law Reform (Mar. 1989) (hereinafter Wilkey Commission) echoed many of the Quadrennial Commission’s concerns:
“We recognize that speeches by federal officials can help inform the public or particular groups and may encourage interchange between the public and private sectors. Nevertheless, we can see no justification for perpetuating the current system of honoraria. Hono-raria paid to officials can be a camouflage for efforts by individuals or entities to gain the officials’ favor. The companies that pay honoraria and related travel expenses frequently deem these payments to be normal business expenses and likely believe that these payments enhance their access to public officials who receive them....
“Although we are aware of no special problems associated with receipt of honoraria within the judiciary, the Commission — in the interest of alleviating abuses in the legislative branch and in applying equitable limitations across the government — joins the Quadrennial Commission in recommending the enactment of legislation to ban the receipt of honoraria by all officials and employees in all three branches of government.” Id., at 35-36 (emphasis added).
The Wilkey Commission “recognize[d] that banning hono-raria would have a substantial financial cost to many officials,” id., at 36, but determined that “the current ailment is *497a serious one and that this medicine is no more bitter than is needed to cure the patient,” ibid. The Wilkey Commission also was aware that its recommendations covered not only high-level federal employees,4 but it “regarded] the current state of affairs as to honoraria in particular as unacceptable in the extreme, and believefd] that [the Government could not] wait until an unspecified date in the future to end this harmful practice.” Id., at 38.5
The Court concedes that in light of the abuses of honoraria by its Members, Congress could reasonably assume that “payments of honoraria to judges or high-ranking officials in the Executive Branch might generate a similar appearance of improper influence,” ante, at 473, but it concludes that Congress could not extend this presumption to federal em*498ployees below grade GS-16. The theory underlying the Court’s distinction — that federal employees below grade GS-16 have negligible power to confer favors on those who might pay to hear them speak or to read their articles — is seriously flawed. Tax examiners, bank examiners, enforcement officials, or any number of federal employees have substantial power to confer favors even though their compensation level is below grade GS-16.
Furthermore, we rejected the same distinction in Public Workers v. Mitchell:
“There is a suggestion that administrative workers may be barred, constitutionally, from political management and political campaigns while the industrial workers may not be barred, constitutionally, without an act ‘narrowly and selectively drawn to define and punish the specific conduct.’ . . . Congress has determined that the presence of government employees, whether industrial or administrative, in the ranks of political party workers is bad. Whatever differences there may be between administrative employees of the government and industrial workers in its employ are differences in detail so far as the constitutional power under review is concerned. Whether there are such differences and what weight to attach to them, are all matters of detail for Congress.” 330 U. S., at 102.
Congress was not obliged to draw an infinitely filigreed statute to deal with every subtle distinction between various groups of employees. See Letter Carriers, 413 U. S., at 556; Mitchell, supra, at 99.
The Court dismisses the Hatch Act experience as irrelevant, because it aimed to protect employees’ rights, notably their right to free expression, rather than to restrict those rights. Ante, at 471. This is, indeed, a strange characterization of § 9(a) of the Hatch Act. It prohibited officers *499and employees in the Executive Branch of the Federal Government from taking “ ‘any active part in political management or in political campaigns.’” Mitchell, supra, at 78. The penalty for violation was dismissal from office. 380 U. S., at 79. Since the right to participate in a political campaign is surely secured in the abstract by the First Amendment, see, e. g., Buckley v. Valeo, 424 U. S. 1, 15 (1976) (per curiam), it can hardly be said that the Act protected the rights of workers who wished to engage in partisan political activity. One of the purposes of the Act was assuredly to free employees who did not wish to become engaged in politics from requests by their superiors to contribute money or time, but to the extent the Act protected these employees it undoubtedly limited the First Amendment rights of those who did wish to take an active part in politics.
The Government’s related concern regarding the difficulties that would attach in administering a case-by-case analysis of the propriety of particular honoraria also supports the honoraria ban’s validity. As we emphasized in Waters, “[t]he government’s interest in achieving its goals as effectively and efficiently as possible is elevated from a relatively subordinate interest when it acts as sovereign to a significant one when it acts as employer.” 511 U. S., at 675. Congress reasonably determined that the prior ethics regime, which required these case-by-case determinations, was inadequate. See App. 257 (“[T]he current state of affairs as to honoraria [is] unacceptable in the extreme”). As a subsequent 1992 GAO Report confirmed, individual ethics officers and various agencies gave differing interpretations to the nexus requirement, and some “approved activities that were questionable as to the appropriateness of accepting compensation.” GAO Report, at 9.
The Court observes that because a nexus limitation is retained for a series of speeches, it cannot be that difficult to enforce. Ante, at 474. The exception that the honoraria ban *500makes for a “series of appearances, speeches, or articles,” far from undermining the statute’s basic purpose, demonstrates that Congress was sensitive to the need for inhibiting as little speech consistent with its responsibility of ensuring that its employees perform their duties impartially and that there is no appearance of impropriety. Reply Brief for United States 12-13. One is far less likely to undertake a “series” of speeches or articles without being paid than he is to make a single speech or write a single article without being paid. Congress reasonably could have concluded that the number of cases where an employee wished to deliver a “series” of speeches would be much smaller than the number of requests to give individual speeches or write individual articles.
Unlike our prototypical application of Pickering which normally involves a response to the content of employee speech, the honoraria ban prohibits no speech and is unrelated to the message or the viewpoint expressed by the Government employee.6 Cf. Waters, supra, at 664-666 (plurality opinion) (analyzing termination of an employee based upon statements critical of the employer); Rankin v. McPherson, 483 U. S. 378, 381-382 (1987) (analyzing termination of an employee based upon a comment about an attempted assassination of President Reagan); Pickering, 391 U. S., at 564 (analyzing termination of an employee based upon a letter critical of the school board). Furthermore, the honoraria ban exempts from its prohibition travel and other expenses re*501lated to employee speech. See 5 U. S. C. App. § 505(3) (1988 ed., Supp. V); 5 CFR §2636.203 (1994). Because there is only a limited burden on respondents’ First Amendment rights, Congress reasonably could have determined that its paramount interests in preventing impropriety and the appearance of impropriety in its work force justified the hono-raria ban. See Civil Service Comm’n v. Letter Carriers, supra; Public Workers v. Mitchell, 330 U. S. 75 (1947).
There is a special irony to the Court’s decision. In order to combat corruption and to regain the public’s trust, the Court essentially requires Congress to resurrect a bureaucracy that it previously felt compelled to replace and to equip it with resources sufficient to conduct case-by-case determinations regarding the actual and apparent propriety of hono-raria by all Executive Branch employees below grade GS-16. I believe that a proper application of the Pickering test to this content-neutral restriction on the receipt of compensation compels the conclusion that the honoraria ban is consistent with the First Amendment. See Civil Service Comm’n v. Letter Carriers, 413 U. S. 548 (1973).
II
One would expect, at the conclusion of its discussion in Parts I-IV, for the Court to hold the statute inapplicable on First Amendment grounds to persons such as the postal worker who lectures on the Quaker religion, and others of similar ilk. But the Court in Part V, in what may fairly be described as an O. Henry ending, holds the statute inapplicable to the entire class before the Court: all Executive Branch employees below grade GS-16 who would receive honoraria but for the statute. Under the Court’s “as applied” remedy, § 501(b) would not apply regardless of whether there was a nexus between the compensation and the individual’s employment. Even if I agreed that application of the honoraria ban to expressive activity unrelated to an employee’s Gov*502ernment employment violated the First Amendment, I could not agree with the Court’s remedy.7
In United States v. Grace, 461 U. S. 171 (1983), we analyzed the constitutionality of 40 U. S. C. § 13k (1982 ed.), as applied to the public sidewalks surrounding the Supreme Court. Section 13k prohibited, “among other things, the ‘display [of] any flag, banner, or device designed or adapted to bring into public notice any party, organization, or movement’ in the United States Supreme Court building and on its grounds.” 461 U. S., at 172-173 (quoting § 13k). We concluded that there was insufficient justification for § 13k’s prohibition against carrying signs, banners, or devices on the public sidewalks surrounding the building. Id., at 183. As a remedy, we held that § 13k was “unconstitutional as applied to those sidewalks.” Ibid.; see also Edenfield v. Fane, 507 U. S. 761, 763 (1993) (striking down a ban on solicitation by certified public accountants as applied to the “business context”).
Although the Court limits its analysis to only those applications of the honoraria ban where there is no nexus between the honoraria and Government employment, the Court prohibits application of the honoraria ban to all Executive Branch employees below grade GS-16 even where there is a nexus between the honoraria and the employees’ Government employment. Ante, at 479-480.8 Even respondents acknowledge that the central aim of their litigation could “be achieved by a remedy similar to the one urged by the government — by holding the ban invalid as applied to respond*503ents’ writing and speaking activities, which have no nexus to their federal employment.” Brief for Respondents 45-46.
Consistent with our approach in Grace, supra, if I were to conclude that § 501(b) violated the First Amendment, I would affirm the Court of Appeals only insofar as its judgment affirmed the injunction against the enforcement of § 501(b) as applied to Executive Branch employees below grade GS-16 who seek honoraria that are unrelated to their Government employment.
In Ex parte Curtis, 106 U. S. 371 (1882), we upheld a statute that prohibited certain Government employees from giving or receiving money for political purposes to or from other Government employees. Ibid. The evident purpose of the statute was to “promote efficiency and integrity in the discharge of official duties, and to maintain proper discipline in the public service.” Id., at 373. “The decisive principle was the power of Congress, within reasonable limits, to regulate, so far as it might deem necessary, the political conduct of its employees.” Public Workers v. Mitchell, 330 U. S. 75, 96 (1947) (analyzing Ex parte Curtis, supra).
Prior to the Hatch Act, Congress had prohibited Civil Service employees from “ ‘us[ing] [their] official authority or influence to coerce the political action of any person or body.’ ” Mitchell, supra, at 79-80, n. 4 (quoting Civil Service Act, ch. 27, §2, 22 Stat. 404).
Ironically, the Court engages in unsupported factfinding to justify its conclusion. Thus, its First Amendment analysis is replete with observations such as “[w]ith few exceptions, the content of respondents’ messages has nothing to do with their jobs and does not even arguably have any adverse impact on the efficiency of the offices in which they work,” ante, at 465; and “[bjecause the vast majority of the speech at issue in this case does not involve the subject matter of Government employment and takes place outside the workplace, the Government is unable to justify § 501(b) on the grounds of immediate workplace disruption.” Ante, at 470.
“The Commission also considered whether it was appropriate to impose a flat ban on outside earned income by all federal employees, or in the alternative, by the highest paid federal employees. In view of the diverse circumstances of federal employees, we felt that an across-the-board ban on outside earned income was unnecessary and too harsh.” Wilkey Commission 38.
The Court discusses a report of the General Accounting Office (GAO) to refute the argument that there is some evidence of misconduct related to honoraria in the rank and file of federal employees below grade GS-16. Ante, at 472, n. 18 (citing General Accounting Office, Report to the Chairman, Subcommittee on Federal Services, Post Office and Civil Service of the Senate Committee on Governmental Affairs, Employee Conduct Standards: Some Outside Activities Present Conflict-of-interest Issues (Feb. 1992) (hereinafter GAO Report)). The GAO audited 11 agencies’ controls over outside activities by employees. Id., at 2. The GAO Report reflects that under the prior regime, some agencies exhibited “overly permissive approval policies,” ibid., and “five agencies approved some outside activities, such as speaking and consulting, that appeared to violate the standard of conduct prohibiting the use of public office for private gain,” ibid. Nine of the eleven agencies reviewed had “approved outside activities in situations that involved potential violations of standard-of-conduct regulations or conflict-of-interest statutes.” Id., at 58. Nevertheless, the Court maintains that there is no evidence of even the appearance of impropriety by employees below grade GS-16. Cf. ante, at 472, n. 18.
The Court’s fanciful example of an employer terminating an employee because of the disruptive effect of the employee’s expression even where the employer agrees with the expression, ante, at 467, n. 11, does not detract from the fact that viewpoint and content neutrality are important factors in evaluating the reasonableness of the public employer’s action. See, e. g., Civil Service Comm’n v. Letter Carriers, 413 U. S. 548, 564 (1973) (“The restrictions so far imposed on federal employees are not aimed at particular parties, groups, or points of view.... Nor do they seek to control political opinions or beliefs, or to interfere with or influence anyone’s vote at the polls”).
Lost in the shuffle of the Court’s remedy is Peter Crane, a GS-16 lawyer from the Nuclear Regulatory Commission, and a respondent before the Court. Ante, at 461. Although the rationale behind the Court’s holding does not necessarily apply to Crane, see ante, at 478, n. 23, the Court’s holding apparently does.
Because the Court has rewritten the honoraria ban so that it no longer applies to Executive Branch employees below grade GS-16, I certainly could not condemn the Court for its refusal to rewrite the statute. Cf. ante, at 479, n. 26. I simply challenge the Court’s failure to tailor its remedy to match its selective analysis.