Cite as: 558 U. S. ____ (2010) 73 Opinion of STEVENS, J. ration would do well to cultivate stable, long-term rela­ tionships of dependency. So we do not have a solid theoretical basis for condemn­ ing §203 as a front for incumbent self-protection, and it seems equally if not more plausible that restrictions on corporate electioneering will be self-denying. Nor do we have a good empirical case for skepticism, as the Court’s failure to cite any empirical research attests. Nor does the legislative history give reason for concern. Congress devoted years of careful study to the issues underlying BCRA; “[f]ew legislative proposals in recent years have received as much sustained public commentary or news coverage”; “[p]olitical scientists and academic experts . . . with no self-interest in incumbent protectio[n] were cen­ tral figures in pressing the case for BCRA”; and the legis­ lation commanded bipartisan support from the outset. Pildes, The Supreme Court 2003 Term Foreword: The Constitutionalization of Democratic Politics, 118 Harv. L. Rev. 28, 137 (2004). Finally, it is important to remem­ ber just how incumbent-friendly congressional races were prior to BCRA’s passage. As the Solicitor General aptly remarked at the time, “the evidence supports overwhelm­ ingly that incumbents were able to get re-elected under the old system just fine.” Tr. of Oral Arg. in McConnell v. FEC, O. T. 2003, No. 02–1674, p. 61. “It would be hard to develop a scheme that could be better for incumbents.” Id., at 63. In this case, then, “there is no convincing evidence that th[e] important interests favoring expenditure limits are fronts for incumbency protection.” Randall, 548 U. S., at 279 (STEVENS, J., dissenting). “In the meantime, a legisla­ tive judgment that ‘enough is enough’ should command the greatest possible deference from judges interpreting a constitutional provision that, at best, has an indirect relationship to activity that affects the quantity . . . of repetitive speech in the marketplace of ideas.” Id., at 279–