Methodology

WASHINGTON, June 5, 2006 — In a nine-month analysis of privately financed travel data based on sometimes incomplete congressional disclosure forms, the Center for Public Integrity and its partners took the following steps to ensure that the resulting data would be as true as possible to the forms filed by members of Congress and their staffers.

What we did:

• Drew up detailed templates for data entry teams at Secure Paper Solutions of Fredericksburg, Va.
• Scanned batches of trip forms by member (and, where possible, by year) into PDF files, which were transmitted electronically to servers at SPS.
• Once data entry was complete, downloaded data files from SPS and imported them into the Center's internal database server.
• Reviewed 30,000 pages of raw documents — including forms and attachments — and compared them with the data entered. Each apparent error or inconsistency was reviewed by a database editor.
• Analyzed traveler names to standardize them to 715 members and 5,945 staffers.
• Compared start dates and end dates of trips taken by people with the same or similar names to identify amendments and duplicate filings.
• Built a calendar of all trips taken by people with the same or similar names to identify travel that overlapped to identify amendments and duplicate filings.
• Analyzed start, end and signature dates and examined cases in which they were non-sequential or inconsistent, identifying additional amendments and data entry errors.
• Identified the sole sponsor reported on 23,380 forms out of 26,577 received.
• Attributed the trips to 3,208 organizations and identified an additional 265 forms listing no identifiable sponsor.
• Identified the remaining trips as having been co-sponsored by multiple organizations or sponsored by organizations that could not be identified as having financed multiple trips.
• Removed from analysis 82 forms reported to be sponsored by lawmakers' offices, Cabinet agencies or other arms of the federal government.
• Excluded from analysis all trips that did not begin between Jan.1, 2000, and June 30, 2005 (one trip which began within that period ended outside of it, on July 1, 2005).
• Adjusted totals to account for forms on which travelers incorrectly added subtotals in the space reserved for "Other" expenses. Leaving them in could have led to double counting.
• Set to zero all totals reported in foreign currency.
• Compared trips with identical costs to identify duplicates and amendments with name variations not previously identified.
• Consulted five years of congressional directories to resolve inconsistencies in the entry of staffers' names and identify whether chiefs of staff, administrative assistants and others whose signatures appeared approving forms were in fact supervised by the member under whose name the forms were filed.
• Compared signature data to the name of the member under which congressional officials filed the trip. This corrected 524 forms filed under the name of someone other than the member responsible for approving them.
• Checked incomplete dates and assigned them to a year. By default, they were assigned to the year in which they were filed (in the House) and then that date was adjusted based on the signature and date stamp dates on the forms themselves.
• Reviewed trips reported as having no cost to identify "advanced authorization forms" that reported approval of a trip rather than that a trip had been taken and paid for by an outside sponsor.
• Called congressional offices about hundreds of trips that appeared to have violated ethics rules. Changes were only made to these trips when specific data entry errors were identified in the process.
• Researched reported destinations to identify whether the destination was foreign or domestic.

What we did not do:

• Attempt to determine whether committee staff members' forms were signed bv the proper supervisor.
• Attempt to harmonize sponsorships (except in a few limited cases in which it was necessitated by additional reporting) when different staffers listed conflicting sponsors for what appeared to be the same trip.
• Attempt to determine the precise job titles and roles of travelers other than members.
• Attempt to modify the data based on interviews; the goal was to ensure that the data accurately reflected what was disclosed on the forms, not what legislators and staffers later said they had intended to disclose.
• Attempt to attribute sponsorship proportionally to the various sponsors of co-sponsored trips.
• Attempt to determine whether ethics rules required the filing of specific forms, such as those without substantial costs or with no travel cost listed.