Sunlight Foundation Reporting Group

Stupak 11 post: What we got wrong

We set out to look at how leadership--of both parties--persuades rank and file members to vote their way. In the 111th Congress, we've seen enormous discipline on both sides of the aisle on a series of high profile votes. Our hunch is that the leadership of both parties has something to do with that, and understanding what levers they have--whether it's funding earmarks, supporting their campaigns with money, appearing at fundraisers or through other means that we still can't track with the current state of congressional disclosure--is something we want to follow in the coming months.

We began this exploration by looking at earmark requests made by key figures in the health care vote. We did a much less-noticed item about Rep. Joseph Cao, R-La., the lone Republican to vote for the original House health care bill and who is defying the House Republican's one-year moratorium on making earmark requests. We'll also be looking at how earmark requests of incumbents with serious challengers compare to members in safe districts. And there's long been anecdotal evidence that appropriations are one of the means used to gain votes for legislation--for example, the various pot sweeteners included in a 2007 Iraq war funding bill that would have set a hard time line for withdrawal. We'd like to see if, with the current state of disclosure, we can link those pot sweeteners to particular members.

But we clearly got the tone of the piece wrong, which is why we have revised the story to better reflect our intentions. Here's what we got wrong in our post tracking earmark requests made by the Stupak 11:

1) We should have been clearer in the headline and at the top of the story that all members, and not just  Rep. Bart Stupak and his ten colleagues, were due to release their requests the day after the health care vote. We implied that the requests and the vote were directly connected; we were wrong to--the requests are just one part of a process we are trying to follow. We should have stuck to the question of how the Appropriations Committee and leadership would handle those requests rather than try to answer it with the information we had at the time. And we should have been clearer that we're looking at how leadership gets rank and file lawmakers to toe the line, with the Stupak 11 as just a potential case study for that phenomenon.

2) The spreadsheet that accompanied the story--one we put together to accompany the piece--had four major errors: Earmarks for Rep. Charlie Wilson for 2010 and 2011 were increased by a factor of ten, as were FY2011 earmarks for Rep. Steve Dreihaus. So too were FY2011 earmarks fro Rep. Kathy Dahlkemper, which led to our erroneously listing her as one of five members who increased the dollar amount of earmarks requested from 2010 to 2011.

We apologize for these erroneous numbers. We went through this process using calculators, excel spreadsheets and Google spreadsheets, double and triple checked numbers, but did not catch the inaccuracies that occurred when we plugged our numbers into a table for publication. We should have been more rigorous with our fact checking process given the challenges of dealing with this difficult data. In any case, we issued corrections to the spreadsheet and story here and here.

3) When you get anything wrong, one of the first things you do is go back and make sure that you haven't made any other mistakes. We have revised the numbers in the original story to accurately reflect the amounts of earmarks members have requested. In some cases, it has been impossible to go back and reproduce all the numbers we originally published on Friday. We are not certain, but it appears some published lists of earmarks might have changed since we did the original story. One example: We used the following formula to import Rep. Paul Kanjorski's earmark requests into a Google spreadsheet:

=ImportHtml("http://kanjorski.house.gov/index.php?option=com_content&task=view&id=1720&Itemid=129","table",4)

When we checked the number on Tuesday, March 22, we got a total of 67124509, or $67 million, which appeared in the spreadsheet we published with the story. Today, using the exact same Google formula, the number comes in at 122220794 or $122 million. Kanjorski isn't alone--whether we tally earmarks by hand reading them to one another using a calculator, or go to great lengths to change the text into machine readable data, we get different results for some members. We've emailed Kanjorski's office for clarification but have not heard back. Similarly, we originally calculated Rep. Marcy Kaptur's total earmark requests for FY2011 at $290 million; when we recalculated the number on Wednesday, we found about $36 million more. Again, we're awaiting a response from Kaptur's office.