See this press release about this Campaign Finance Institute working paper that is sure to play a prominent role in the debate over regulating 527s and the possibility that clamping down on 527s will simply shift election-related activities to 501(c)s. From the release:

The study shows that the groups deploy three different types of organization — PACs, 527s, and 501(c) advocacy entities — in their efforts to influence federal elections and public policy. These cumulative, coordinated efforts increase the groups’ financial influence in elections, and that of their large donors, far more than is generally realized. The CFI analysis presents much new information about the major role played by 501(c)(4) social welfare, (c)(5) labor union and (c)(6) trade association organizations in elections, and the different ways in which they and related 527 organizations are used by Republican and Democratic-oriented groups. The report highlights major weaknesses in the IRS’s public disclosure system for 501c groups’ “political expenditures,” including inadequate guidance and monitoring of reports. This report raises fundamental questions about current campaign finance practices, and about the arguments used by both sides in the ongoing political controversy over regulation of 527s.
CFI analyzed a variety of data, including interviews, concerning the following interest groups: American Federation of State, County and Municipal Employees, Americans for Job Security, Chamber of Commerce of the U.S.A., Club for Growth, MoveOn.org, National Rifle Association, National Right to Life Committee, New Democrat Network, Planned Parenthood, Progress for America, Service Employees International Union, and Sierra Club. Among the findings:
* Eleven of the twelve interest groups had either PACs or 527s, but seven used both in federal elections, cumulating financial influence.
* Nine of the groups also mobilized their 501(c)(4) social welfare, (c)(5) labor union or (c)(6) business league “advocacy” organizations for elections; but only one group consistently reported its political expenditures to the IRS due to inadequate IRS guidance for and monitoring of reports.
* Republican-oriented groups mostly used 501(c)s instead of 527s; Democratic-oriented ones mainly used 501(c)s to supplement 527 activities.
* Groups generally have been able in their public communications to distinguish their “election communications” from their “issue advocacy” and “lobbying” ones
* Most groups could transfer their 527 activities to 501(c) advocacy organizations, if necessary, although at a potential cost since political activities conducted by a 501(c) organization are subject to taxation.
* The overwhelming majority of large individual donors to the groups’ 527s contributed large amounts of money directly to candidates, parties and PACs, also cumulating financial influence in elections.
This study suggests that reforms that do not comprehensively address the whole framework of disclosure and regulation for interest groups will inevitably have unintended consequences. Risks of 527 regulation producing a flow of funds to 501(c) advocacy organizations are assessed as high, especially if gaps in IRS reporting of political expenditures are not addressed. Specific recommendations are provided for closing major holes in the public disclosure system administered by the IRS, FEC and FCC.