We've received a few calls about the House disbursements data we published, questioning the total amount spent by various offices. The House Committee on Administration provides two sets of numbers: summary numbers for each office's spending by quarter, including a grand total and a breakdown by category, and detailed files listing every disbursement reported for that quarter. There is a discrepancy between the totals reported in the summary section and the totals arrived at by adding all of the individual disbursements reported in the detail view of our downloadable data. Despite the discrepancy, a review of the data shows that both sets of numbers are accurate, as far as they go.
After combing through several dozen pages of PDF documents to get to the bottom of this, we noticed that there are hundreds of rows of expenditures reported each quarter that do not appear to be included in the aggregation of the various categories of expenditures such as "personnel compensation" or travel that are used to calculate the summary numbers. When we did some additional research and made several calls to the House offices to find out why the numbers were not always adding up we learned that some of the line items from the PDF files that cover calendar year 2010 are actually disbursements made in 2009, or, in some cases, even earlier. Members can report disbursements up to three years after they were made.
According to the spokesperson at the House Committee on Administration, those expenses which come in late should be calculated as part of the member representational allowance (MRA) for that specific year. To reflect that, we will change numbers for some of the lawmakers, for instance, Rep. Nancy Pelosi's, D-Calif., amounts will be $1,659,836 while Rep. Jim Costa's, D-Calif., amounts will be $1,565,495.
Thus, for agreggate numbers, users should consult the summary table, but be aware that those summary numbers may understate how much a member's office spent in a quarter or a year. Any number reported in the quarterly summaries may understate the actual spending because 2010 disbursements can be reported as late as 2013. To look at specific expenditures–travel, salaries for individual congressional staffers and so on, the detailed files should be used.
While we took into account anything that is paid out in any of the quarters of 2010 to be aggregated, to get a more conservative account of the spending coming from the 2010 MRA alone, line items from 2009 or 2008 reported in the 2010 files should be excluded.
In the end, what we have here are numbers which are as close to the real ones as possible given the several hurdles created by the House's rules on how this information is reported and the manner in which the data is presented in non-machine readable PDF files.