I recently made a mistake that turned into an object lesson on the limits of technology but also – and more importantly – on the limits of government openness. Earlier this year, in trying to figure out how many House staffers had gone on to become lobbyists over a two-year period, I naively relied on the ability of computers to match names. We have some sophisticated matching software, but it is not perfect. As a result, my recent post, “Almost 400 House staffers registered to lobby in last two years” improperly identified a number of staffers as lobbyists because they had the same name or almost the same as a registered lobbyist when, in fact, these were two different people.
The good folks at Legistorm noticed this for us, and so we decided to do an internal Sunlight audit, and our capable intern Breanna Edwards, under the supervision of our reporting team, provided it, using the Center for Responsive Politics’ Revolving Door website, LinkedIn, and official lobbying reports.
Of the 377 House staffers we originally identified as having gone on to lobby, Breanna was able to be certain or almost certain that 219 (58%) were indeed matches. She also identified 21 (6%) as almost certainly false positives (that is, the name match was correct, but it was clearly a case of two people with the same name). The remaining 137 matches our computers identified (36%) could not be confirmed definitively either way, at least just based on those three sources.
Breanna explains the challenges:
We decided to go about this project in a simple and straightforward way, using three tools: the Center for Responsive Politics’ “Revolving Door”, LinkedIn and lobbying reports. I put all of the names through these three tools in almost all instances to be thoroughly sure of my results, though sometimes two out of the three tools, or at the very least one of the tools did not necessarily yield any results.
Center for Responsive Politics’ Revolving Door is a neat tool that records which federal employees have become lobbyists, consultants and strategists, usually providing information on where they worked, the positions they have held for their employers and how long they were there. It’s a really easy tool to use. One simply has to enter in the subject’s name into the given field and see if there is a record. If they did have a detailed record, that made my job easier. However, this database isn’t completely foolproof. Sometimes the information Sunlight had did not match what CRP had, sometimes they only had parts of the information I was given, sometimes they didn’t have any information on the subject at all. If the Center for Responsive Politics’ information didn’t exactly match what I had, or if there was something in the information that I was unclear about, my next step was to double-check LinkedIn.
The great thing about LinkedIn is that people self-advertise a lot there (of course, that is how it was meant to be used) and so there was little doubt that the individuals I was looking for would definitely put most, if not all of their employment history on there. The trouble of course, is actually finding them and being sure that it was exactly the same person. I found some individuals with ease and, as expected, they gave the information I needed which was a green light. Among other individuals there were multiple people to choose from, all with some part of the employment history I was looking for, but none with the exact sequence I was looking for. I either took that to mean that we were looking at completely different people, or, if it was so blurred I could not tell, marked it as uncertain. In other cases, I just simply could not find the individual. That was when I turned to the lobbying reports.
Though they sometimes gave some great results, the lobbying reports were the absolute last resort, Only people who make a certain salary are even listed on lobbying reports, which severely limits the number of individuals who even turned up in the search. Another issue: Even when these individuals were listed, a good number of them either did not know how to fill out the form correctly or chose not to fill it out correctly. So while I did manage to procure a few confirmations using this method, the reports really did not help that much.
The main problem was names. The funny thing about names is that there are only so many of them. There are few unique names, as names are not identification numbers. The fact that many individuals with the same or similar names seem to have worked in the same or extremely similar offices did not help our effort. It was nearly impossible to distinguish these people. Other times they simply did not appear, but of course, just because they were not in any databases, didn’t mean that the information we had was false. All these factors made us unable to confirm some of our findings.
The takeaway lesson ought to be clear: it’s much more difficult than it should be to match congressional staff names with lobbyist names. To do this kind of research, we are at the mercy of what individuals choose to report, especially when they have relatively common names.
This makes it very difficult to have a good understanding of how the revolving door works. There are good reasons the public might want to know which lobbyists have insider connections, and who they are using those connections to help. There are also good reasons the public might want to know which offices have former staff working for various outside groups that are trying to influence legislation.
A simple solution Here is a simple proposal to improve transparency and disclosure around the revolving door:
1. Lobbyist registration forms list the names of lobbyists working on a particular issue. In order to be able to determine whether they have spun through the revolving door, each lobbyist should be identified by a publicly accessible unique identification number.
2. Upon registering, each lobbyist should also be required to list every position that he or she has held as a federal, state, or city employee. Currently, registrants are required to list all “covered” legislative and executive branch positions their lobbyists have held for the last 20 years only when the registrant begins representing a new client, but not on subsequent forms.
Doing this would create a simple, searchable, and centralized way to identify which lobbyists formerly served in government, and which didn’t. There is no reason why it should be as difficult as it currently is to verify whether individual public employees have gone on to become lobbyists.
Of course, there is much more to do on the issue of lobbying reform. We support the Lobbyist Disclosure Enhancement Act, introduced by Rep. Mike Quigley as well as the Real-Time Online Lobbying Disclosure Act. For a full list of our lobbying reform proposals, click here.