CHIEF JUSTICE ROBERTS delivered the opinion of the Court.
In order to jumpstart a fledgling campaign or finish strong in a tight race, candidates for federal office often loan money to their campaign committees. A provision of federal law regulates the repayment of such loans. Among other things, it bars campaigns from using more than $250,000 of funds raised after election day to repay a candidate's personal loans. This limit on the use of post-election funds increases the risk that candidate loans over $250,000 will not be repaid in full, inhibiting candidates from making such loans in the first place. The question is whether this restriction violates the First Amendment rights of candidates and their campaigns to engage in political speech.
Candidates for federal office may, consistent with federal law, use various sources to fund their campaigns. A candidate may spend an unlimited amount of his own money in support of his campaign. See Buckley v. Valeo, 424 U.S. 1, 52-54 (1976) (per curiam). His campaign — a legal entity distinct from the candidate himself — may borrow an unlimited amount from third-party lenders or from the candidate himself. See 11 CFR §110.10 (2017); 52 U. S. C. §30101(9)(A)(i); see also Buckley, 424 U. S., at 52-54. And campaigns may, of course, accept contributions directly from other organizations or from individuals, subject to monetary limitations. Individual contributions are capped at $2,900 for the primary and $2,900 for the general election. See §§30116(a), (c); 86 Fed. Reg. 7869 (2021). Campaigns may continue to receive contributions after election day, so long as those contributions go toward repaying campaign debts. See 11 CFR §110.1(b)(3)(i).
Section 304 of the Bipartisan Campaign Reform Act of 2002 (BCRA), 116 Stat. 98, 52 U. S. C. §30116(j), further restricts the use of post-election funds. Under that provision, a candidate who loans money to his campaign may not be repaid more than $250,000 of such loans from contributions made to the campaign after the date of the election. Ibid. To implement that limit, the Federal Election Commission (FEC) has promulgated regulations establishing three rules pertinent here: First, a campaign may repay up to $250,000 in candidate loans using contributions made "at any time before, on, or after the date of the election." 11 CFR §116.12(a). Second, to the extent the loans exceed $250,000, a campaign may use pre-election funds to repay the portion exceeding $250,000 only if the repayment occurs "within 20 days of the election." §116.11(c)(1). And third, if more than $250,000 remains unpaid when the 20-day post-election deadline expires, the campaign must treat the portion above $250,000 as a contribution to the campaign, precluding later repayment. §116.11(c)(2).
Appellee Ted Cruz represents Texas in the United States Senate. This case arises from his 2018 reelection campaign, which was, at the time, the most expensive Senate race in history. Before election day, Cruz loaned $260,000 to the other appellee here, Ted Cruz for Senate (Committee). At the end of election day, however, the Committee was in the red by approximately $340,000. App. 285. It eventually began repaying Cruz's loans, but by that time the 20-day post-election window for repaying amounts over $250,000 had closed. See 11 CFR §§116.11(c)(1), (2). The Committee accordingly repaid Cruz only $250,000, leaving $10,000 of his personal loans unpaid.
Cruz and the Committee filed this action in the United States District Court for the District of Columbia, alleging that Section 304 of BCRA violates the First Amendment. They also raised challenges to the FEC's implementing regulation, 11 CFR §116.11. A three-judge panel was convened to hear the case. See BCRA §403(a)(1), 116 Stat. 113; see also 28 U. S. C. §2284.
The three-judge District Court granted Cruz and his Committee summary judgment on their constitutional claim, holding that the loan-repayment limitation burdens political speech without sufficient justification. 542 F.Supp.3d 1 (2021). The District Court also ordered that appellees' challenges to the regulation, previously held in abeyance, be dismissed as moot. The Government appealed directly to this Court, as authorized by 28 U. S. C. §1253. We postponed consideration of our jurisdiction. 594 U. S. ___ (2021).
The Constitution limits federal courts to deciding "Cases" and "Controversies." Art. III, §2. Among other things, that limitation requires a plaintiff to have standing. The requisite elements of Article III standing are well established: A plaintiff must show (1) an injury in fact, (2) fairly traceable to the challenged conduct of the defendant, (3) that is likely to be redressed by the requested relief. Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-561 (1992).
As the Government recognizes, the Committee's present inability to repay the final $10,000 of Cruz's loans constitutes an injury in fact both to Cruz and to his Committee. See Reply Brief 8. Cruz, of course, suffers a $10,000 pocketbook harm. See Czyzewski v. Jevic Holding Corp., 580 U.S. 451, 464 (2017). And the bar on repayment injures the Committee by preventing it from discharging its obligation to repay its debt, which may inhibit that form of financing in the future. The Government maintains, however, that these injuries are not traceable to the threatened enforcement of Section 304, for two reasons: first, because the inability to repay Cruz's loans was "self-inflicted," and second, because it is the threatened enforcement of an agency regulation, not the statute itself, that causes the harm. We address each argument in turn.
First, the Government argues that appellees lack standing because their injuries were "self-inflicted." Brief for Appellant 20. Because appellees knowingly triggered the application of the loan-repayment limitation, the Government says, any resulting injury is in essence traceable to them, not the Government. The predicate for this argument is appellees' stipulation in the District Court that "the sole and exclusive motivation behind Senator Cruz's actions in making the 2018 loan[s] and the [C]ommittee's actions in waiting to repay them was to establish the factual basis for this challenge." App. 325. At bottom, the Government asks us to recognize an exception to traceability for injuries that a party purposely incurs.
We have never recognized a rule of this kind under Article III. To the contrary, we have made clear that an injury resulting from the application or threatened application of an unlawful enactment remains fairly traceable to such application, even if the injury could be described in some sense as willingly incurred. See Evers v. Dwyer, 358 U.S. 202, 204 (1958) (per curiam) (that the plaintiff subjected himself to discrimination "for the purpose of instituting th[e] litigation" did not defeat his standing); Havens Realty Corp. v. Coleman, 455 U.S. 363, 374 (1982) (a "tester" plaintiff posing as a renter for purposes of housing-discrimination litigation still suffered an injury under Article III).
The cases the Government cites do not alter our conclusion. In Clapper v. Amnesty Int'l USA, 568 U.S. 398 (2013), for example, the plaintiffs attempted to manufacture standing by voluntarily taking costly and burdensome measures that they said were necessary to protect the confidentiality of their communications in light of the Government surveillance policy they sought to challenge. Id., at 402. Their problem, however, was that they could not show that they had been or were likely to be subjected to that policy in any event. Id., at 416. Likewise, in Pennsylvania v. New Jersey, 426 U.S. 660 (1976) (per curiam), we held that the unilateral decisions by a group of States to reimburse their residents for taxes levied by other States was not a basis to attack the legality of those taxes. Nothing in the challenged taxes required the plaintiff States to offer reimbursements; accordingly, the financial injury those States suffered was due to their own independent response to taxes levied on others. Id., at 664. Here, by contrast, the appellees' injuries are directly inflicted by the FEC's threatened enforcement of the provisions they now challenge. That appellees chose to subject themselves to those provisions does not change the fact that they are subject to them, and will face genuine legal penalties if they do not comply. See 52 U. S. C. §30109(a)(5); 11 CFR §111.24.
One final point bears mentioning. The Government maintains that it should not be blamed for appellees' injuries because it provided the Committee with a legally available "alternative" that would have avoided any liability — repaying Cruz's loans in full with pre-election funds, within 20 days of the election. But even if such funds were available, the Government's argument largely misses the point. For standing purposes, we accept as valid the merits of appellees' legal claims, so we must assume that the loan-repayment limitation — including the 20-day rule — unconstitutionally burdens speech. See Warth v. Seldin, 422 U.S. 490, 500 (1975) ("standing in no way depends on the merits of the plaintiff's contention that particular conduct is illegal"). Demanding that the Committee comply with the Government's "alternative" would therefore require it to forgo the exercise of a First Amendment right we must assume it has — the right to repay its campaign debts in full, at any time. And this would require the Committee to subject itself to the very framework it says unconstitutionally burdens its speech. Such a principle finds no support in our standing jurisprudence. See, e.g., Susan B. Anthony List v. Driehaus, 573 U.S. 149, 158-159 (2014).
The Government next asserts that although appellees would have standing to challenge the FEC's implementing regulation, 11 CFR §116.11, they do not have standing to challenge Section 304 itself. As a reminder, Section 304 prohibits the use of post-election funds to repay a candidate's personal loans; it does not restrict the use of funds raised before the election. See 52 U. S. C. §30116(j). That restriction comes instead from Section 304's implementing regulation, 11 CFR §116.11. This regulation provides that neither pre-election nor post-election funds may be used to repay candidate loans above $250,000 outstanding 20 days after the election. §§116.11(c)(1)-(2). Such amounts must instead be treated as contributions to the campaign, barring their repayment.
Bearing that in mind, the Government contends that the record before the District Court reveals that the Committee used funds raised before the election to repay the first $250,000 of Cruz's loans. For support, it naturally points to appellees' stipulation that "none of the $250,000 of the loan that was repaid was from contributions raised after the election." App. 329. Thus, the Government says, the Committee has not yet reached the cap in Section 304 on the use of post-election funds, and can still repay the remaining balance without running afoul of that statutory restriction. It is instead the agency's regulation — with its 20-day limit — that prevents repayment of the final $10,000. This matters, the Government insists, because "[s]tanding is not dispensed in gross," and plaintiffs must establish standing separately for each claim that they press and each form of relief that they seek. Brief for Appellant 17 (quoting TransUnion LLC v. Ramirez, 594 U. S. ___, ___ (2021) (slip op., at 15)). A challenge to the regulation, the Government argues, is separate from a challenge to the statute that authorized it.
For their part, appellees insist that the record, properly interpreted, shows that the Committee used post-election funds to repay Cruz. During the period between election day and when the Committee repaid Cruz's loans, the Committee received more than $250,000 in "redesignated" contributions to Cruz's 2024 campaign. Those contributions came from individuals who donated to the 2018 election in amounts exceeding their base limit and who, subsequent to the election, redesignated the overlimit amount to the 2024 campaign. See 11 CFR §110.1(b)(5). Such funds, appellees say, qualify as "post-election contributions" for purposes of Section 304, and may have been used to repay the first $250,000 of Cruz's loans. See §116.12(a).
These arguments have an Alice in Wonderland air about them, with the Government arguing that appellees would not violate the statute by repaying Cruz, and the appellees arguing that they would. But this case has unfolded in an unusual way. After all, Cruz and the Committee likely would have had standing to bring a pre-enforcement challenge (as they do now) to Section 304 in a much easier manner — by simply alleging and credibly demonstrating that Cruz wished to loan his campaign an amount larger than $250,000, but would not do so only because the loan-repayment limitation made it unlikely that such amount would be repaid. See Susan B. Anthony List, 573 U. S., at 158-159. In addition, it ordinarily would not matter whether a plaintiff was challenging the statute's enforcement or instead the enforcement of a regulation and, in doing so, raising arguments about the validity of the statute that authorized the regulation. Cf. Collins v. Yellen, 594 U. S. ___, ___-___ (2021) (slip op., at 18-19). The parties here, however, assume that the distinction makes a difference because the subject-matter jurisdiction of the three-judge District Court is limited to actions challenging the enforcement of the statute. See BCRA §403(a) (authorizing a three-judge court to hear any "action . . . brought for declaratory or injunctive relief to challenge the constitutionality of any provision of this Act or any amendment made by this Act").
It seems to us that the Government is likely correct that appellees have not shown that they exhausted Section 304's cap on the use of post-election funds. The loan-repayment limitation applies to contributions "made" after the date of the election. 52 U. S. C. §30116(j). And a contribution is "considered to be made when the contributor relinquishes control" over it, which occurs when the contribution is "delivered" to the Committee or the candidate. 11 CFR §110.1(b)(6). The redesignated contributions on which appellees now rely, however, involve funds that were delivered to the Committee before the 2018 election. And those funds have remained under the Committee's control from that date, even if they were later redesignated to a different campaign.
But we need not go further down this rabbit hole. Even under the Government's account, appellees have standing to challenge the threatened enforcement of Section 304. The present inability of the Committee to repay and Cruz to recover the final $10,000 Cruz loaned his campaign is, even if brought about by the agency's threatened enforcement of its regulation, traceable to the operation of Section 304 itself. An agency, after all, "literally has no power to act" — including under its regulations — unless and until Congress authorizes it to do so by statute. Louisiana Pub. Serv. Comm'n v. FCC, 476 U.S. 355, 374 (1986); see also FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 161 (2000). An agency's regulation cannot "operate independently of " the statute that authorized it. California v. Texas, 593 U. S. ___, ___ (2021) (slip op., at 15). And here, the FEC's 20-day rule was expressly promulgated to implement Section 304. See 68 Fed. Reg. 3973 (2003). Indeed, the Government admitted at oral argument that it could find no other basis to authorize enforcement of this regulation, Tr. of Oral Arg. 5, and "concede[d]" that "the most likely result, if the statute were declared invalid, is that the regulation would cease to be on the books or would cease to be enforceable," ibid. Thus, if Section 304 is invalid and unenforceable — as Cruz and the Committee contend — the agency's 20-day rule is as well. And the remedy appellees sought in the District Court — an order enjoining the Government from taking any action to enforce the loan-repayment limitation, App. 27 — would redress appellees' harm by preventing enforcement of the agency's 20-day rule. See Lujan, 504 U. S., at 561.
Contrary to the Government's suggestion, the foregoing analysis does not call into question the principle that "a plaintiff injured by one law does not thereby acquire standing to challenge a different law." Brief for Appellant 17. It is true that a litigant cannot, "by virtue of his standing to challenge one government action, challenge other governmental actions that did not injure him." DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 353, n. 5 (2006). Here, however, appellees seek to challenge the one Government action that causes their harm: the FEC's threatened enforcement of the loan-repayment limitation, through its implementing regulation. In doing so, they may raise constitutional claims against Section 304, the statutory provision that, through the agency's regulation, is being enforced. Cf. Collins, 594 U. S., at ___-___ (slip op., at 18-19). Even on the Government's version of the facts, then, we are satisfied that appellees have standing to challenge the threatened enforcement of Section 304. And because they are challenging "the constitutionality of [a] provision of [BCRA]," §403(a), jurisdiction was proper in the three-judge District Court. We thus proceed to the merits.
The First Amendment "has its fullest and most urgent application precisely to the conduct of campaigns for political office." Monitor Patriot Co. v. Roy, 401 U.S. 265, 272 (1971). It safeguards the ability of a candidate to use personal funds to finance campaign speech, protecting his freedom "to speak without legislative limit on behalf of his own candidacy." Buckley, 424 U. S., at 54. This broad protection, we have explained, "reflects our profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open." Id., at 14 (internal quotation marks omitted).
The Government seems to agree with appellees that the loan-repayment limitation abridges First Amendment rights, at least to some extent, see Brief for Appellant 27-32, and we reach the same conclusion. This provision, by design and effect, burdens candidates who wish to make expenditures on behalf of their own candidacy through personal loans. See 52 U. S. C. §30101(9)(A)(i) (defining "expenditure" to include loans); see also Buckley, 424 U. S., at 52. By restricting the sources of funds that campaigns may use to repay candidate loans, Section 304 increases the risk that such loans will not be repaid. That in turn inhibits candidates from loaning money to their campaigns in the first place, burdening core speech.
The data bear out the deterrent effect of Section 304. After BCRA was passed, there appeared a "clear clustering of [candidate] loans right at the $250,000 threshold." A. Ovtchinnikov & P. Valta, Debt in Political Campaigns 26 (2020), Record 65-1 (Ovtchinnikov, Debt); see also Brief for United States Senator Roy Blunt et al. as Amici Curiae 6-7. There was no such clustering before the loan-repayment limitation went into effect. The Government's evidence in the District Court, moreover, reflects that the percentage of loans by Senate candidates for exactly $250,000 has increased tenfold since BCRA was passed. See App. 312-313. Section 304, then, has altered "the propensity of many politicians to make large loans." Ovtchinnikov, Debt 26; see also Brief for Protect the First Foundation as Amicus Curiae 10-11. In doing so, it has predictably restricted a candidate's speech on behalf of his own candidacy. See Buckley, 424 U. S., at 54.
Quite apart from this record evidence, the burden on First Amendment expression is "evident and inherent" in the choice that candidates and their campaigns must confront. Arizona Free Enterprise Club's Freedom Club PAC v. Bennett, 564 U.S. 721, 745 (2011); see also id., at 746 ("we do not need empirical evidence to determinate that the law at issue is burdensome"); Davis v. Federal Election Comm'n, 554 U.S. 724, 738-740 (2008) (requiring no empirical evidence of a burden). Although Section 304 "does not impose a cap on a candidate's expenditure of personal funds, it imposes an unprecedented penalty on any candidate who robustly exercises that First Amendment right." Id., at 738-739. That penalty, of course, is the significant risk that a candidate will not be repaid if he chooses to loan his campaign more than $250,000. And that risk in turn may deter some candidates from loaning money to their campaigns when they otherwise would, reducing the amount of political speech. This "drag" on a candidate's First Amendment right to use his own money to facilitate political speech is no less burdensome "simply because it attaches as a consequence of a statutorily imposed choice." Id., at 739.
The "drag," moreover, is no small matter. Debt is a ubiquitous tool for financing electoral campaigns. The raw dollar amount of loans made to campaigns in any one election cycle is in the nine figures, "significantly exceeding" the amount of independent expenditures. Ovtchinnikov, Debt 11. And personal loans from candidates themselves constitute the bulk of this financing. See Brief for Appellant 35 ("more than 90% of campaign debt consists of candidate loans"). In fact, candidates who self-fund usually do so using personal loans. See J. Steen, Self-Financed Candidates in Congressional Elections 21 (2006).
The ability to lend money to a campaign is especially important for new candidates and challengers. As a practical matter, personal loans will sometimes be the only way for an unknown challenger with limited connections to frontload campaign spending. See G. Jacobson, Money in Congressional Elections 97-101 (1980). And early spending — and thus early expression — is critical to a newcomer's success. See Steen, Self-Financed Candidates in Congressional Elections, at 35, 171. A large personal loan also may be a useful tool to signal that the political outsider is confident enough in his campaign to have skin in the game, attracting the attention of donors and voters alike. See R. Biersack, P. Herrnson, C. Wilcox, Seeds for Success: Early Money in Congressional Elections, 18 Leg. Studies Q. 535, 537 (1993); see also Brief for United States Senator Roy Blunt et al. as Amici Curiae 13. By inhibiting a candidate from using this critical source of campaign funding, however, Section 304 raises a barrier to entry — thus abridging political speech.
The dissent cannot and does not claim that Section 304 imposes no burden on candidate speech. See post, at 5 (opinion of KAGAN, J.) ("every contribution regulation has some kind of indirect effect on electoral speech"). The dissent instead dismisses that burden as minor and insignificant. Post, at 4-6. As just explained, the extent of the burden may vary depending on the circumstances of a particular candidate and particular election. But there is no doubt that the law does burden First Amendment electoral speech, and any such law must at least be justified by a permissible interest. See McCutcheon v. Federal Election Comm'n, 572 U.S. 185, 210 (2014) (plurality opinion) ("When the Government restricts speech, the Government bears the burden of proving the constitutionality of its actions.").
With those First Amendment costs in mind, we turn to whether the loan-repayment limitation is justified. The parties debate whether strict or "closely drawn" scrutiny should apply in answering that question. Buckley, 424 U. S., at 25. We need not resolve this dispute because, under either standard, the Government must prove at the outset that it is in fact pursuing a legitimate objective. See McCutcheon, 572 U. S., at 210. It has not done so here.
This Court has recognized only one permissible ground for restricting political speech: the prevention of "quid pro quo" corruption or its appearance. See id., at 207; see also Federal Election Comm'n v. National Conservative Political Action Comm., 470 U.S. 480, 497 (1985). We have consistently rejected attempts to restrict campaign speech based on other legislative aims. For example, we have denied attempts to reduce the amount of money in politics, see McCutcheon, 572 U. S., at 191, to level electoral opportunities by equalizing candidate resources, see Bennett, 564 U. S., at 749-750, and to limit the general influence a contributor may have over an elected official, see Citizens United v. Federal Election Comm'n, 558 U.S. 310, 359-360 (2010). However well intentioned such proposals may be, the First Amendment — as this Court has repeatedly emphasized — prohibits such attempts to tamper with the "right of citizens to choose who shall govern them." McCutcheon, 572 U. S., at 227; see also Davis, 554 U. S., at 742; Bennett, 564 U. S., at 750.
The Government argues that the contributions at issue raise a heightened risk of corruption because of the use to which they are put: repaying a candidate's personal loans. It also maintains that post-election contributions are particularly troubling because the contributor will know — not merely hope — that the recipient, having prevailed, will be in a position to do him some good.
We greet the assertion of an anticorruption interest here with a measure of skepticism, for the loan-repayment limitation is yet another in a long line of "prophylaxis-uponprophylaxis approach[es]" to regulating campaign finance. McCutcheon, 572 U. S., at 221 (quoting Federal Election Comm'n v. Wisconsin Right to Life, Inc., 551 U.S. 449, 479 (2007) (opinion of ROBERTS, C. J.)). Individual contributions to candidates for federal office, including those made after the candidate has won the election, are already regulated in order to prevent corruption or its appearance. Such contributions are capped at $2,900 per election, see 86 Fed. Reg. 7869, and nontrivial contributions must be publicly disclosed, see 52 U. S. C. §§30104(b)(3)(A), (c)(1). The dissent's dire predictions about the impact of today's decision elide the fact that the contributions at issue remain subject to these requirements. See post, at 3, 14-15. And the requirements are themselves prophylactic measures, given that "few if any contributions to candidates will involve quid pro quo arrangements." Citizens United, 558 U. S., at 357. Such a prophylaxis-upon-prophylaxis approach, we have explained, is a significant indicator that the regulation may not be necessary for the interest it seeks to protect. See McCutcheon, 572 U. S., at 221; see also Bennett, 564 U. S., at 752 ("In the face of [the State's] contribution limits [and] strict disclosure requirements . . . it is hard to imagine what marginal corruption deterrence could be generated by [an additional measure].").
There is no cause for a different conclusion here. Because the Government is defending a restriction on speech as necessary to prevent an anticipated harm, it must do more than "simply posit the existence of the disease sought to be cured." Colorado Republican Federal Campaign Comm. v. Federal Election Comm'n, 518 U.S. 604, 618 (1996). It must instead point to "record evidence or legislative findings" demonstrating the need to address a special problem. Ibid. We have "never accepted mere conjecture as adequate to carry a First Amendment burden." McCutcheon, 572 U. S., at 210 (quoting Nixon v. Shrink Missouri Government PAC, 528 U.S. 377, 392 (2000)).
Yet the Government is unable to identify a single case of quid pro quo corruption in this context — even though most States do not impose a limit on the use of post-election contributions to repay candidate loans. Cf. Brief for Campaign Legal Center et al. as Amici Curiae 17-18 (citing the 10 States that do impose such a prohibition). Our previous cases have found the absence of such evidence significant. See Citizens United, 558 U. S., at 357 (the Government did not claim that the political process was corrupted in the 26 States that allowed unrestricted independent expenditures by corporations); McCutcheon, 572 U. S., at 209, n. 7 (the Government presented no evidence of corruption in the 30 States that did not impose aggregate limits on individual contributions).
The Government instead puts forward a handful of media reports and anecdotes that it says illustrate the special risks associated with repaying candidate loans after an election. But as the District Court found, those reports "merely hypothesize that individuals who contribute after the election to help retire a candidate's debt might have greater influence with or access to the candidate." 542 F. Supp. 3d, at 15. That is not the type of quid pro quo corruption the Government may target consistent with the First Amendment. See McCutcheon, 572 U. S., at 207-208.
The dissent at points shrugs off this distinction, see post, at 2, 12, n. 3, 13, but our cases make clear that "the Government may not seek to limit the appearance of mere influence or access." McCutcheon, 572 U. S., at 208. As we have explained, influence and access "embody a central feature of democracy — that constituents support candidates who share their beliefs and interests, and candidates who are elected can be expected to be responsive to those concerns." Id., at 192.
To be sure, the "line between quid pro quo corruption and general influence may seem vague at times, but the distinction must be respected in order to safeguard basic First Amendment rights." Id., at 209. And in drawing that line, "the First Amendment requires us to err on the side of protecting political speech rather than suppressing it." Ibid. (quoting Wisconsin Right to Life, 551 U. S., at 457 (opinion of ROBERTS, C. J.)).
In the absence of direct evidence, the Government turns elsewhere. It contends that a scholarly article, a poll, and statements by Members of Congress show that these contributions carry a heightened risk of at least the appearance of corruption. Essentially all the Government's evidence, however, concerns the sort of "corruption," loosely conceived, that we have repeatedly explained is not legitimately regulated under the First Amendment.
The academic article — cited for various propositions by both sides — concludes that "indebted politicians" are "more likely to switch their votes" if they receive contributions from the banking or insurance industries. Ovtchinnikov, Debt 31. But the authors explicitly note that they cannot distinguish between voting pattern changes traceable to legitimate donor influence or access, and voting pattern changes as part of an illicit quid pro quo. See A. Ovtchinnikov & P. Valta, Self-Funding of Political Campaigns, Management Science, Articles in Advance 18 (April 7, 2022) (Ovtchinnikov, Self-Funding). As noted, our precedents demand adherence to that distinction. See, e.g., McCutcheon, 572 U. S., at 209. The authors also state that their analysis is merely a "first step" in understanding whether politicians' self-funding decisions impact voting behavior, because they cannot "pin down a causal link" yet. Ovtchinnikov, Self-Funding 21.
The online poll the Government asks us to consider similarly misses the mark. The poll, conducted at the Government's behest for this litigation, reports that most respondents thought it "very likely" or "likely" that a person who "donate[s] money to a candidate's campaign after the election expect[s] a political favor in return." App. 351-352. But it failed to ask whether those same respondents thought it likely that donors who contribute to a campaign before the election also are likely to expect political favors in return. Nor did the poll mention that the individual base limits still apply to such contributions. And it failed to define the term "political favor," leaving unclear the critical issue whether the respondents associated such contributions with the direct exchange of money for official acts, which Congress may regulate, or simply increased influence and access, which Congress may not.
Finally, the Government places great weight on statements made by certain Members of Congress during debates that preceded the enactment of BCRA. One Senator, for example, remarked that without the loan-repayment limitation, a winning candidate who loaned money to his campaign could "get it back from [his] constituents [at] fundraising events" where he could ask, "How would you like me to vote now that I am a Senator?" 147 Cong. Rec. S2462 (March 19, 2001) (remarks of Sen. Domenici). Another stated that candidates "have a constitutional right to try to buy the office, but they do not have a constitutional right to resell it." 147 Cong. Rec. S2541 (March 20, 2001) (remarks of Sen. Hutchison). Nothing these legislators said, however, constitutes actual evidence that the loan-repayment limitation was necessary to prevent quid pro quo corruption or its appearance. And a few stray floor statements are not the same as "legislative findings" that might suggest a special problem to be addressed. Colorado Republican Federal Campaign Comm., 518 U. S., at 618.
All the above is pretty meager, given that we are considering restrictions on "the most fundamental First Amendment activities" — the right of candidates for political office to make their case to the American people. Buckley, 434 U. S., at 14. In any event, the legislative record helps appellees just as much as the Government, given that some Senators evidently viewed the limit as designed to protect incumbents like themselves from wealthy challengers. See 147 Cong. Rec. S2465 (March 19, 2001) (remarks of Sen. Sessions) ("[Section 304] prohibits wealthy candidates, who incur personal loans in connection with their campaign that exceed $250,000, from repaying those loans from any contributions made to the candidate. . . . I am glad I didn't face a person who could write a check for $60 million, $10 million — or $5 million, for that matter. If so, I would like to be able to have a level playing field so I could stay in the ball game."); see also 147 Cong. Rec. S2541 (March 20, 2001) (remarks of Sen. Hutchison) ("Our purpose is to level the playing field.").
That the limit may have been designed to protect incumbents should come as no surprise. Section 304 was enacted as part of the "Millionaire's Amendment" to BCRA, designed to hobble wealthy candidates mounting self-financed campaigns. See Davis, 554 U. S., at 739. And it was debated together with another provision we have already held unconstitutional, in part because it pursued the same impermissible goal of "level[ing] electoral opportunities for candidates of different personal wealth." Id., at 741. The connection between these two provisions casts further doubt on the anticorruption interest the Government now asserts in this case.
Perhaps to make up for its evidentiary shortcomings, the Government falls back on what it calls a "common sense" analogy: Post-election contributions used to repay a candidate's loans are akin to a "gift" because they "add to the candidate's personal wealth" as opposed to the campaign's treasury. Brief for Appellant 33. The risk of corruption is thus greater, the Government argues, because the donor is lining the pockets of a legislator or legislator-elect.
The dissent at multiple points makes the same argument, contending that contributions that go toward repaying a candidate's loan "enrich the candidate personally," allowing him to "buy a car or make tuition payments or join a country club." Post, at 7, 14; see also post, at 2, 3, 8, 13. But this forgets that we are talking about repayment of a loan, not a gift. If the candidate did not have the money to buy a car before he made a loan to his campaign, repayment of the loan would not change that in any way.
On top of that, contributions that go toward retiring a candidate's debt could only arguably enrich the candidate if the candidate does not otherwise expect to be repaid. In other words, the Government's gift comparison is meaningful only if the baseline is that the campaign will default. The Government, however, provides no reason to believe that most or even many winning candidates — the only candidates with whom its anticorruption interest is concerned — expect not to be repaid by their campaigns. To the contrary, the Government has recognized throughout this litigation that winning candidates are commonly repaid in full. See App. 31-32 (citing the former FEC Commissioner's statement that "only winners have an easy time dealing with debt"); id., at 317 (same); see also Ovtchinnikov, Self-Funding 11 (concluding that, even with BCRA's limitations on loan repayment in place, two out of three winning campaigns were able to repay a candidate's loans in full). For such a candidate, then, post-election contributions bear little resemblance to a gift, because there is less of a chance that his campaign will default. Such contributions instead restore the candidate to the status quo ante, a position to which he legitimately expected to return. As for losing candidates, they are of course in no position to grant official favors, and the Government does not provide any anticorruption rationale to explain why post-election contributions to those candidates should be restricted. See Brief for Appellant 45-46.
The analogy also proves too much. By the Government's logic, post-election contributions to retire candidate loans are little different from gifts given directly to the candidate. But that logic is belied by how the Government treats the two categories of purported "gifts." On the one hand, federal law flatly prohibits candidates from using campaign contributions for personal purposes. See 52 U. S. C. §30114(b)(2). And it forbids Senators from accepting gifts worth $250 or more. See 2 U. S. C. §4725(a)(1). By contrast, the postulated "gift-by-loan-repayment" limits are simply the individual contribution limits, which are now more than ten times higher than the gift limit: $2,900 per election. And Section 304 allows over 86 such "gifts" before a campaign hits the Act's $250,000 cap. Either the Government is openly tolerating a significant number of "gifts" far more generous than what it would normally think fit to allow, or post-election contributions that go toward retiring campaign debt are in no real sense "gifts" to a candidate. We find the latter answer more persuasive.
As a final argument, the Government claims that if the matter is otherwise in doubt, we should defer to Congress's "legislative judgment" that Section 304 furthers an anticorruption goal. Brief for Appellant 39; see also post, at 8 (KAGAN, J., dissenting) (also arguing that we have no "reason to second-guess Congress's experience-based judgment"). Such deference, the Government contends, is grounded "in part on the understanding that Congress `is far better equipped than the judiciary to amass and evaluate the vast amounts of data bearing upon legislative questions.'" Brief for Appellant 40 (quoting Turner Broadcasting System, Inc. v. FCC, 520 U.S. 180, 195 (1997) (some internal quotation marks omitted)). But as explained, the evidence here is scant, and Congress's judgment is hardly based on "vast amounts of data." Id., at 195. Moreover, deference to Congress would be especially inappropriate where, as here, the legislative act may have been an effort to "insulate[ ] legislators from effective electoral challenge." Shrink Missouri Government PAC, 528 U. S., at 404 (BREYER, J., concurring); see also Randall v. Sorrell, 548 U.S. 230, 248-249 (2006) (plurality opinion).
In the end, it remains our role to decide whether a particular legislative choice is constitutional. See Sable Communications of Cal., Inc. v. FCC, 492 U.S. 115, 129 (1989); see also Randall, 548 U. S., at 248-249 (stressing need for "the exercise of independent judicial judgment" in case raising concern that "contribution limits that are too low [may] harm the electoral process by preventing challengers from mounting effective campaigns against incumbent officeholders"). And here the Government has not shown that Section 304 furthers a permissible anticorruption goal, rather than the impermissible objective of simply limiting the amount of money in politics.
* * *
For the reasons set forth, we conclude that Cruz and the Committee have standing to challenge the threatened enforcement of Section 304 of BCRA. We also conclude that this provision burdens core political speech without proper justification. The judgment of the District Court is affirmed.
It is so ordered.
JUSTICE KAGAN, with whom JUSTICE BREYER and JUSTICE SOTOMAYOR join, dissenting.
A candidate for public office extends a $500,000 loan to his campaign organization, hoping to recoup the amount from benefactors' post-election contributions. Once elected, he devotes himself assiduously to recovering the money; his personal bank account, after all, now has a gaping half-million-dollar hole. The politician solicits donations from wealthy individuals and corporate lobbyists, making clear that the money they give will go straight from the campaign to him, as repayment for his loan. He is deeply grateful to those who help, as they know he will be — more grateful than for ordinary campaign contributions (which do not increase his personal wealth). And as they paid him, so he will pay them. In the coming months and years, they receive government benefits — maybe favorable legislation, maybe prized appointments, maybe lucrative contracts. The politician is happy; the donors are happy. The only loser is the public. It inevitably suffers from government corruption.
The campaign finance measure at issue here has for two decades checked the crooked exchanges just described. The provision, Section 304 of the Bipartisan Campaign Reform Act of 2002, prohibited a candidate from using post-election donations to repay loans exceeding $250,000 that he made to his campaign. The theory of the legislation is easy to grasp. Political contributions that will line a candidate's own pockets, given after his election to office, pose a special danger of corruption. The candidate has a more-than-usual interest in obtaining the money (to replenish his personal finances), and is now in a position to give something in return. The donors well understand his situation, and are eager to take advantage of it. In short, everyone's incentives are stacked to enhance the risk of dirty dealing. At the very least — even if an illicit exchange does not occur — the public will predictably perceive corruption in post-election payments directly enriching an officeholder. Congress enacted Section 304 to protect against those harms.
In striking down the law today, the Court greenlights all the sordid bargains Congress thought right to stop. The theory of the decision (unlike of the statute) is hard to fathom. The majority says that Section 304 violates the candidate's First Amendment rights by interfering with his ability to "self-fund" his campaign. Ante, at 12. But the candidate can in fact self-fund all he likes. The law impedes only his ability to use other people's money to finance his campaign — much as standard (and permissible) contribution limits do. And even that third-party restriction is a modest one, applying only to post- (not pre-) election donations to repay sizable (not small) loans. So the majority overstates the First Amendment burdens Section 304 imposes. At the same time, the majority understates the anticorruption values Section 304 serves. In the majority's view, there is "scant" danger here of quid pro quo corruption; loan repayments produce only the "sort of `corruption'" in which contributors wield "greater influence" over candidates than they otherwise would. Ante, at 16-17, 21. Assume away all objections to that distinction, which even the majority concedes is "vague," ante, at 16; for better or worse, it underlies this Court's recent campaign finance decisions. Still, the conduct targeted by Section 304 threatens, if anything does, both corruption and the appearance of corruption of the quid pro quo kind. That is because the regulated transactions — as Members of Congress well knew from experience — personally enrich those already elected to office. In allowing those payments to go forward unrestrained, today's decision can only bring this country's political system into further disrepute.
In assessing a law's burden on speech, this Court's decisions all distinguish between restricting expenditures and restricting contributions. See, e.g., Buckley v. Valeo, 424 U.S. 1, 19-23 (1976) (per curiam). (The majority glosses over that core distinction, for reasons that will soon become clear.) According to settled precedent, expenditure restrictions — caps on a campaign's or candidate's electoral spending — impose the greatest burdens on expression. The First Amendment, as the majority notes, "has its fullest and most urgent application" when a "legislative limit" prevents a candidate from "us[ing] personal funds to finance campaign speech" — that is, speech "on behalf of his own candidacy." Ante, at 10 (internal quotation marks omitted). By contrast, laws focused on third-party contributions to a campaign (a category the majority mostly prefers to ignore) typically "entail only a marginal restriction" on First Amendment interests. Buckley, 424 U. S., at 20. Take, for example, a simple limit on the amount someone can donate to a campaign, like the federal $2,900 ceiling. That kind of restriction, we have reasoned, in no way interferes with the donor's "freedom to discuss candidates and issues" through independent spending. Id., at 21. And it has only an indirect effect on the campaign itself. To be sure, the cap makes raising money (for speech and other things) harder: It forces candidates "to raise funds from a greater number" of people and generally results in the campaign taking in less money than it otherwise would. Id., at 22. But the Court has viewed such limits as troublesome only if they are so low as to prevent candidates from raising "the resources necessary for effective advocacy." Randall v. Sorrell, 548 U.S. 230, 247 (2006) (plurality opinion) (quoting Buckley, 424 U. S., at 21). In the usual case, the incidental effect of a contribution restriction on a campaign's speech does not count as a significant First Amendment burden. See Randall, 548 U. S., at 246-247.
Under that precedent, Section 304 "entails only a marginal restriction" on speech, because it regulates contributions alone. Buckley, 424 U. S., at 20. The provision leaves a campaign free to spend any amount of money for speech. Likewise, it leaves the candidate himself — here, Senator Ted Cruz — free to do so. The candidate can (in the majority's words) "use personal funds to finance campaign speech" without limit; if he wishes, he can devote his whole fortune to "speech on behalf of his own candidacy." Ante, at 10-11. Section 304 restricts only the use of third-party contributions to support his efforts — which, as just shown, imposes a far more modest First Amendment burden. Recall how Section 304 works: It prevents post-election campaign contributions from going to repay large loans that the candidate has made to his campaign. So the provision limits — much as standard contribution caps do — only the candidate's ability to shift the costs of his electoral speech to others. Or said a bit differently, it addresses not a candidate's "self-fund[ing]," ante, at 12, but only his reliance on third-party financing.
And even that regulation of third-party contributions is a narrow one. Under Section 304, a campaign can always accept donations for small loans a candidate makes. And it can use pre-election donations to retire even his sizable loans. The statute just insists that donations for that purpose occur when speech is ongoing, and before everyone knows which candidate won (and so is in a position to return the favor by delivering government benefits). Consistent with our caselaw, that minor restriction on a candidate's use of other people's money does not severely burden his (or anyone else's) expression.
The majority's argument to the contrary focuses not on the restriction Section 304 actually imposes, but on the indirect effects the provision might have. The majority does not dispute that Section 304 places no limits on the amount a candidate can spend for expression. See ante, at 11. Nor does (or could) the majority even claim that the provision caps what a candidate can lend his campaign. Instead, the majority argues that the law "may deter" a candidate from making large loans because it curtails a potential source of repayment — i.e., post-election donations. Ante, at 12. In that way, the majority insists, the law — though concededly regulating only the use of contributions — functions to "restrict a candidate's speech." Ante, at 11; see ante, at 13.
But every contribution regulation has some kind of indirect effect on electoral speech, and we have still understood them to impose only minimal burdens. Consider again a standard contribution ceiling, like the federal $2,900 cap. That limit, as we have acknowledged, makes raising money harder. See Randall, 548 U. S., at 247; Buckley, 424 U. S., at 20-21. And so it predictably gives a campaign less money to spend. (In fact, a lot less: Just think of a world in which a candidate could raise an unlimited sum from every supporter.) With the contribution cap in effect, the campaign cannot pay for (nearly) as many advertisements, mailings, signs, and so forth. And likewise, to return to the fact pattern here, the campaign has less money available than it otherwise would to repay a candidate's (or any other) loans. By the majority's logic, that downstream effect would mean the contribution cap imposes a significant First Amendment burden. But as noted above, we have always held to the contrary, save for the rare case in which the limit is so low as to preclude effective advocacy. See supra, at 3-4. There is no reason to treat Section 304 differently. In fact, its restriction on post-election contributions for loan repayment probably has much smaller indirect effects on a campaign's or candidate's speech than the contribution ceilings this Court has approved. (Again, just think of all the multi-million-dollar donations those ceilings prevent.) So the majority's view cannot be right.
And more fundamentally, the majority fails to appreciate what Section 304 has an indirect effect on: lending, rather than spending, money. In the majority's view, those two activities count as one and the same. See ante, at 10-11. But they are not, in an obvious way. The expenditure of "personal funds" for speech, this Court has observed, "reduces the candidate's dependence" on donors — precisely because he is not trying to speak on their dime. Buckley, 424 U. S., at 53. The loan of personal funds has the opposite effect, as further shown in this opinion's next part. When a candidate lends substantial funds to his campaign, he wants (maybe desperately needs) them returned; he thus risks — indeed, invites — dependence on donors, who alone can make him financially whole. Section 304 responds to that difference in whether a candidate is speaking independently, or instead relying on others' largesse. The provision at most deters a single mechanism for financing electoral activities, because it carries a heightened threat of corruption.
Preventing quid pro quo corruption or its appearance is a compelling interest by any measure. See Federal Election Comm'n v. National Conservative Political Action Comm., 470 U.S. 480, 496-497 (1985). Quid pro quo corruption — which extends beyond criminal bribery to "less blatant and specific" arrangements — "subver[ts] the political process" and threatens "the integrity of our system of representative democracy." Nixon v. Shrink Missouri Government PAC, 528 U.S. 377, 388-389 (2000) (internal quotation marks omitted). And the appearance of that corruption (though scarcely mentioned in the majority opinion) is "[o]f almost equal concern." Id., at 388. Avoiding that appearance is "critical" if public "confidence in the system of representative Government is not to be eroded to a disastrous extent." Id., at 389.
Serious dangers of actual and apparent quid pro quo corruption attend the transactions Section 304 regulates — again, the use of post-election contributions to repay a candidate's personal loans. Consider a simple comparison. When a campaign uses a donation to fund routine electoral activities (including speech), the money marginally aids the candidate's electoral odds, but in no way adds to his personal wealth. By contrast, when a campaign uses a donation to repay the candidate's loan, every dollar given goes straight into the candidate's pocket. With each such contribution, his assets increase; he can now buy a car or make tuition payments or join a country club — all with his donors' dollars. So contributions going to loan repayment have exceptional value to the candidate — which his donors of course realize. And when the contributions occur after the election, their corrupting potential further increases. At that time, a campaign can use donations only to repay loans, of which some 97% come from candidates. See 11 CFR 110.1(b)(3)(i) (2017); A. Ovtchinnikov & P. Valta, Self-Funding of Political Campaigns, Management Science, Articles in Advance 5 (Apr. 7, 2022) (Ovtchinnikov, Self-Funding). So post-election donors can be confident their money will enrich a candidate personally. And those donors have of course learned which candidate won. When they give money to repay the victor's loan, they know — not merely hope — he will be in a position to perform official favors. The recipe for quid pro quo corruption is thus in place: a donation to enhance the candidate's own wealth (the quid), made when he has become able to use the power of public office to the donor's advantage (the quo). The heightened threat of corruption — and, even more, of its appearance — is self-evident (except, it seems, to observers allergic to all campaign finance regulation).
In addressing that special danger, Section 304 is anything but a "prophylaxis-upon-prophylaxis," as the majority labels it. Ante, at 14. The idea behind that fancy-sounding epithet is just that the statute is a needless precaution: The $2,900 contribution ceiling, the majority asserts, already provides generous protection against the corrupting potential of donations, so the loan-repayment provision is unnecessary. See ibid. But that claim ignores that Section 304 targets only a subset of contributions, which raise (as just described) unique corruption risks. When an added protection addresses an added danger, the existence of a basic protection (however ordinarily ample) fails to show the supplement's pointlessness. Regular seatbelts might suffice to protect drivers on the interstate, but special belts — and roll cages to boot — are essential measures on the racetrack. So too, a $2,900 cap might suffice to prevent corruption from normal campaign contributions — but not from post-election contributions to repay a candidate's loan, and thus to enrich him personally. When Congress, as here, responds to a heightened threat with a heightened safeguard, the majority has no call to "greet" it "with a measure of skepticism." Ibid.
Nor does the majority have reason to second-guess Congress's experience-based judgment about the specially corrupting effects of post-election donations to repay candidate loans. The majority's first attempt to counter that judgment is that "we are only talking about repayment of a loan": "If the candidate did not have the money to buy a car before he made a loan to his campaign, repayment of the loan would not change that in any way." Ante, at 19. But that altogether misses the point. However much money the candidate had before he makes a loan to his campaign, he has less after it: The amount of the loan is the size of the hole in his bank account. So whatever he could buy with, say, $250,000 — surely a car, but that's beside the point — he cannot buy any longer. Until, that is, donors pay him back. Then, the hole is filled, the bank account replenished, and the purchasing power restored. That is a significant financial gain to the officeholder, courtesy of donors. If they had not stepped up, the officeholder would have been $250,000 poorer.
The majority's second theory fares no better. Contributions to repay loans, the majority argues, do not really enrich an officeholder, because he has, from the beginning, "expect[ed] to be repaid." Ante, at 20. But the record provides no support for that self-assured statement. Contra the majority, the Government "has recognized throughout this litigation" not that winning candidates are usually repaid, but only that they are repaid more often than losing ones. Ibid.; see App. 31-32, 317.
The common sense of Section 304 — the obviousness of the theory behind it — lessens the need for the Government to identify past cases of quid pro quo corruption involving candidate loan repayments. As this Court has made clear, "[t]he quantum of empirical evidence needed" to sustain a campaign finance law "var[ies] up or down with the novelty and plausibility of the [law's] justification." McConnell v. Federal Election Comm'n, 540 U.S. 93, 144 (2003). There is nothing novel or implausible about Section 304's rationale — once again, that payments going to line an elected official's pockets pose an especial risk of corruption. It is in fact what everyone knows to be true — because everyone knows people (including politicians) will often do things for money. The majority suggests that we should discard our understanding of how the world works because the Government has not come forward with adjudicated instances of corruption in the loan-repayment context. See ante, at 15-16. But quid pro quo exchanges, in that and every other setting, are nigh-impossible to detect and prove. That is indeed why we have campaign finance laws like Section 304. They prohibit conduct posing a heightened risk of corruption, so that the Government does not have to ferret out illicit exchanges case by case by case. To strike down Section 304 because the Government has not proved to a certainty some number of loan-repayments-for-political-paybacks is to miss the provision's essential point.
In any event, the Government and its amici have marshalled significant evidence showing that the loan repayments Section 304 targets have exactly the dangers Congress thought. See Brief for Appellant 37-40; Brief for Campaign Legal Center et al. 27-29. Here is a sampling from the record, involving jurisdictions unprotected by either Section 304 or a state equivalent. In Ohio, various law firms donated almost $200,000 to help the newly elected attorney general recoup his personal loans. Those donors later received more than 200 state contracts worth nearly $10 million in legal fees. See L. Bischoff, Donations Helping DeWine Pay Down Campaign Loan, Springfield News-Sun, Feb. 2, 2012, p. A1. In Alaska, a lobbyist collected almost $100,000 for post-election repayment of the Governor's personal loans. A business in which he held an interest later received a $9 million state contract. See B. Curry, Alaska Gov. Sheffield's Impeachment Inquiry Has Overtones of Watergate Scandal, L. A. Times, July 19, 1985, p. 11. In Kentucky, two Governors loaned their campaigns millions of dollars, "only to be repaid after the election by contributors seeking no-bid contracts." J. Moore, Campaign Finance Reform in Kentucky: The Race for Governor, 85 Ky. L. J. 723, 746 (1997). The scandal those transactions created led to a new state campaign-finance law similar to Section 304. In upholding that statute, a court more cognizant than this one about how corruption works explained that "heavily indebted candidates" were "easy bedfellows for quid pro quo contributors." Wilkinson v. Jones, 876 F.Supp. 916, 930 (WD Ky. 1995). That is also true on the local level. In San Diego, to take just one instance, three city council members cast critical votes benefiting lobbyists who had raised funds to retire their campaign debts. See C. Gustafson, Lobbyists See Benefit From Three City Officials, San Diego Union-Tribune, June 13, 2009, p. A1.
An empirical study in the record confirms the dangers of corruption shown in those examples. The study first found, based on data preceding Section 304's enactment, that politicians carrying campaign debt were "significantly more likely" than their "debt-free counterparts" to "switch their votes" after receiving contributions from special interests. A. Ovtchinnikov & P. Valta, Debt in Political Campaigns (2020), in No. 1:19-cv-00908 (D DC, July 14, 2020), ECF Doc. 65-1, p. 31. In other words, officeholders did more in exchange for donations repaying their personal loans than for other donations. The analysis next looked at Section 304's effect. Here, the data showed that politicians with debt exceeding the law's $250,000 threshold became "significantly less responsive" to contributions than before: They began to "behave remarkably similar to their debt free counterparts." Id., at 28; see Ovtchinnikov, Self-Funding 3 (similarly stating that those politicians became more "independent of contributions from special interest[s]"). In other words, Section 304 did just what Congress thought it would. By preventing post-election contributions from personally enriching politicians, the provision diminished donor-responsive voting. The majority tries to undermine those findings by quoting the kind of careful caveats always accompanying good social science. See ante, at 17; Ovtchinnikov, Self-Funding 21 (noting that the study is a "first step in understanding" and that more work is needed to "fully pin down" all aspects of causation). But the authors are confident — and rightly so — in the findings just described: that Section 304 markedly decreased the frequency with which officeholders voted as donors would like. And although the authors could not responsibly claim that all the shifted votes they tallied were part of quid pro quo deals — they are, after all, professors, not the FBI — they deduce from the data that politicians carrying campaign debt were "less likely to [be] sell[ing] access" than to be "sell[ing] votes." Id., at 18.
Finally, the record evidence addresses the "almost equal[ly]" important matter of the appearance of corruption. Shrink Missouri, 528 U. S., at 390; see supra, at 6-7. A Government-commissioned survey of public opinion found that 81% of respondents believed it "very likely" or "likely" that a person who "donate[s] money to a candidate's campaign after the election expect[s] a political favor in return." App. 351-353. That bears repeating: 81% — an overwhelming perception across all demographic categories, as well as across all party affiliations and political ideologies. See ibid. As the court reviewing the Kentucky version of Section 304 explained: "[T]here is an impression" when a contribution repays a loan after an election that the contributor is simply "lining the candidate's pocket, as there is no ongoing campaign to which the contribution may be made." Wilkinson, 876 F. Supp., at 930; see supra, at 12. The majority flyspecks the polling questions: Why didn't the poll define "political favor"? Did the poll mention that the contributions had to comply with the $2,900 cap? And so forth. See ante, at 17-18. But really — is it likely that such tinkering would have made a real difference? The poll results were so lopsided because the post-election contributions Section 304 targets — ones adding to the candidate's personal wealth — have so conspicuous a potential to corrupt. The public knows that to be true. The public's representatives in Congress knew it to be true. Only this Court — somehow — does not.
* * *
"Democracy works only if the people have faith in those who govern." Shrink Missouri, 528 U. S., at 390 (internal quotation marks omitted). And the people cannot have faith in representatives who trade official acts for financial gain. Section 304 prevents that kind of corruption, at barely discernable cost to First Amendment freedoms. The provision limits one narrow use of third-party contributions to a campaign, thus "entail[ing] only a marginal restriction" on speech. Buckley, 424 U. S., at 20. And the provision targets a practice posing exceptional risks of quid pro quo deals. Repaying a candidate's loan after he has won election cannot serve the usual purposes of a contribution: The money comes too late to aid in any of his campaign activities. All the money does is enrich the candidate personally at a time when he can return the favor — by a vote, a contract, an appointment. It takes no political genius to see the heightened risk of corruption — the danger of "I'll make you richer and you'll make me richer" arrangements between donors and officeholders. Section 304 has guarded against that threat for two decades, but no longer. In discarding the statute, the Court fuels non-public-serving, self-interested governance. It injures the integrity, both actual and apparent, of the political process. I respectfully dissent.