98 McCONNELL v. FEDERAL ELECTION COMM'N Syllabus but also through state committees, which function as an alternate avenue for precisely the same corrupting forces. Indeed, the evidence shows that both candidates and parties already ask donors who have reached their direct contribution limit to donate to state committees. Congress' reasonable prediction, based on the history of campaign finance regulation, was that donors would react to § 323(a) by directing soft-money contributions to state committees for the purpose of influencing federal candidates and elections, and that federal candidates would be just as indebted to these contributors as they had been to those who had formerly contributed to the national parties. Preventing corrupting activity from shifting wholesale to state committees and thereby eviscerating FECA clearly qualifies as an important governmental interest. Pp. 164-166. (3) Plaintiffs argue unpersuasively that, even if §323(b) serves a legitimate interest, its restrictions are so unjustifiably burdensome and overbroad that they cannot be considered "closely drawn" to match the Government's objectives. P. 166. (i) Section 323(b) is not substantially overbroad. Although §323(b) captures some activities that affect state campaigns for nonfederal offices, these are the same activities that were covered by the Federal Election Commission's (FEC) pre-BCRA allocation rules, and so had to be funded in part by hard money because they affected both federal and state elections. As a practical matter, BCRA merely codifies the FEC's allocation regime principles while justifiably adjusting the applicable formulas in order to restore the efficacy of FECA's longstanding restriction on contributions to state and local committees for the purpose of influencing federal elections. By limiting its reach to "Federal election activities," §323(b) is narrowly focused on regulating contributions that directly benefit federal candidates and thus pose the greatest risk of corruption or its appearance. The first two categories of "Federal election activity"-voter registration efforts and voter identification, GOTV, and generic campaign activities conducted in connection with a federal election--clearly capture activities that confer a substantial benefit on federal candidates by getting like-minded voters to the polls. If a voter registration drive does not specifically mention a federal candidate, state committees can take advantage of the Levin Amendment's higher contribution limits and relaxed source restrictions. Moreover, because the record demonstrates abundantly that the third category of "Federal election activity," "public communication[s]" that promote or attack a federal candidate, directly affects the election in which that candidate is participating, application of § 323(b)'s contribution caps to such communications is closely drawn to the anticorruption interest it is intended to address. Finally, Congress' interest in pre-