The Center for Public Integrity’s ongoing analysis of 527 committees provides comprehensive documentation of a little-known, but well-oiled backdoor to political financing. Here is how our researchers do it.
Vetting the committees
Beginning in August 2000, the Internal Revenue Service was required to obtain detailed registration and financial activity reports from groups structured under Section 527 of the tax code regulating not-for-profit political organizations.
Since then, more than 20,000 groups registered as 527 committees. The vast majority of those were not actually 527 committees because they were already filing with state or federal election authorities. Center researchers narrowed this field in search of money raised and spent outside of federal and state campaign finance regulation. By focusing on organizations possessing one or more of the following traits, the Center was able to identify 628 committees to include in its study of “true” 527s:
- The committee is not required to report financial activities to state or federal election authorities.
- The committee is tied to or formed by a federal lawmaker.
- The committee is active in many states and spends most of its money on election-related activities such as broadcast advertisements, mailings and political research.
Some of the committees included in this database have registered with state election authorities but were added either because they have connections to prominent state and federal politicians or because they spent money in states that require outside entities to file with elections authorities.
Obtaining the data
The IRS Web site houses images of paper and electronic filings submitted by the committees. In order to make this reported information accessible, the Center had to get the data into one format. Before the IRS required 527s to file electronically, paper filings were downloaded from the IRS Web site and then keypunched by an outside firm, Secure Paper Solutions of Fredericksburg, Va. Center researchers still make every effort to find and keypunch paper records that may be filed in error. Data from the electronic filings are downloaded and incorporated into the database by Center staff.
As of Oct. 14, 2005, the database contained more than 500,000 contribution and expenditure records reported in 5,000-plus filings submitted by 527 committees throughout the past five years.
The Center will continue to add records to the database as filings come in, so the number of records and filings, as well as results of database queries on our Web site, may contain more records than referenced in previous reports.
Organizing the records
Center researchers encountered the following quirks in the IRS disclosure information and tried to account for and correct such difficulties:
- The IRS assigns an Employer Identification Number, or EIN, to each committee, but in many instances assigned more than one EIN to a single committee. The Center has consolidated multiple EINs under their proper committee.
- The IRS disclosure system contains a number of duplicate and amended reports, so the Center checked for redundant filings and accounted for later amendments in its research. When two identical filings covering the same time period were found, the Center used the one most recently filed in its study. If, on rare occasions, the amendments contained only partial information, then both reports were included.
- Some reports included amounts that were not fully broken down, or itemized. Summary figures included in this study are calculated using the totals reported on the first page of the 527 activity reports, while calculations to determine major contributors and recipients used the itemized transactions. In some cases, paper reports did not include any summary numbers; the Center calculated those using the itemized records.
- In some filings, the summary numbers reported by committees did not match the aggregate totals of contributions or expenditures. In many cases, the reason for this was a mathematical error by the committee, and in those cases the numbers were corrected. In cases where the totals could not be reconciled, researches used the best number available.
Analyzing the information
Although most 527 committees are single entities, several organizations maintain multiple 527 accounts. In order to assess the full impact of one organization, the Center created an identification coding system that groups all relevant committees under the umbrella organization.
For example, the American Federation of State, County and Municipal Employees registered two accounts with the IRS: the AFSCME Special Account and the AFSCME-People Non-Federal Account. When relevant, some of the Center’s analysis features the two committees’ activity totaled up, while some features the two committees’ activity separately. This distinction is noted on the appropriate Web pages.
Although the IRS has no mechanism to track transfers of money between affiliated committees or from one 527 to another, the Center has attempted where possible to track the movement of money between affiliated 527s. Currently, the Center treats as “transfers” — rather than as contributions or expenditures — funds moved between Joint Victory Committee, America Coming Together, The Media Fund and Democratic Victory 2004 when determining aggregate totals. For example, when Joint Victory Committee sent money it raised to America Coming Together and The Media Fund, those transactions are not counted towards aggregate contribution totals.
Using another type of coding system, as of Oct. 14, 2005, the Center successfully coded 83 percent of the money itemized in filings to better identify donors or recipients. This coding is used to group all records relating to a specific donor or recipient (which can be an organization or a company or individual), regardless of how the name of each donor or recipient appears within each actual 527 report.
The Center created a fourth level of coding to place each 527 committee in larger context to get an overall snapshot of who might be benefiting from 527 activity. Each of the 628 committees was given a descriptive code noting what type of entity sponsors it. A committee category may have several committee types within it, in order to provide further detail about the type of committee. Below, we list the broad groupings of committee types, followed by the specific categories of committee:
- Democratic Party, candidates and supporters category includes these committee types:
- National Democratic Party
- General Democratic/Liberal
- Democratic leadership committee-federal
- State Democratic Party and affiliate organizations
- Democratic candidate or officeholder state committees
- Federal Democratic Recount committee
- State Democratic Recount committee
- Federal Democratic exploratory committee
- Democratic leadership committee — state and local
- Democratic redistricting committee
- Republican Party, candidates and supporters category includes these committee types:
- National Republican Party
- General Republican/Conservative
- Republican leadership committee &emdash; federal
- State Republican Party and affiliate organizations
- Republican candidate or officeholder state committees
- Federal Republican Recount committee
- State Republican Recount committee
- Federal Republican exploratory committee
- Republican leadership committee — state and local
- Republican redistricting committee
- Single Issue category includes these committee types:
- National Single Issue
- State Single Issue
- Minority/Ethnic Issues
- Abortion category includes these committee types:
- Abortion Rights
- Corporate/Business category includes this committee type:
- Business and trade associations
- Labor category includes this committee type:
- Labor unions
These categories include the same committee type: Trial Lawyers, Environmental, Other Party and Unknown/Unaffiliated.
Help support this work
Public Integrity doesn’t have paywalls and doesn’t accept advertising so that our investigative reporting can have the widest possible impact on addressing inequality in the U.S. Our work is possible thanks to support from people like you.