Christopher Robertson and three co-authors have posted this draft on SSRN (forthcoming, Journal of Legal Analysis).  Here is the abstract:

he Supreme Court says that campaign finance regulations are unconstitutional unless they target “quid pro quo” corruption or its appearance. The Court has used this doctrine to strike down many efforts at campaign finance reform. However, the court has merely speculated or reasoned in a conclusory way about when that criterion is satisfied. To operationalize and test the “appearances” standard, we fielded two empirical studies. First, in a highly realistic simulation, three grand juries deliberated on charges that ”independent” campaign spending in a Congressional race met the legal standard for bribery of the candidate. Second, 1276 nationally-representative online respondents considered whether to convict in such a scenario, with five variables manipulated randomly to enhance generalizability. In both studies, jurors found quid pro quo corruption for behaviors they believed to be common in contemporary politics. Because these tests use the procedural and substantive apparatus of Federal law to operationalize the quid pro quo corruption concept and draw from a diverse population of respondents, they are a stronger test of the “appearances” standard than mere opinion polling or judicial speculation. The data suggest that prior Supreme Court’s decisions were wrong, and that Congress and the states have greater authority to regulate campaign finance. This research also suggests that actual prosecutions under current bribery laws are surprisingly viable, but this risk is deeply problematic under the First Amendment, Due Process, and Separation of Powers doctrines. A regulatory system using safe harbors may be a solution.