Let it be this one.
As is probably obvious, I think there are good reasons to be excited about information technology’s capacity to improve governance. For the small stuff, it already has — I can get a new recycling bin delivered just by visiting a website!
But for the really big stuff — like keeping track of how we award multi-million dollar contracts, or how we regulate polluters, or how we track the political influence economy — there are significant barriers to applying automated analysis. A lack of usable corporate identifiers is among the biggest of these problems. Citizens are expected to have trackable identifiers for their cars, and firearms, and selves. I have to surrender my SSN to a private entity just to see if I’m allowed to buy a new phone!
But for companies, the situation is quite different. Corporate registrations are held by states, which face bad incentives around making this data usable both because inaccessible registration information makes the state more attractive to registrants, and because states collect substantial sums for use of the data. Meanwhile, the federal government is wrapped in a nightmare tangle of purpose-built ID fiefdoms and one overarching public-private partnership that supplies identifier data that’s both unacceptably restricted and of unacceptably poor quality.
Attempts to improve the situation will no doubt be met with much gnashing of virtual teeth: competitiveness! Privacy! Personhood! None of these hold much water, though: not only has relevant tax return information (which would go a long way toward solving this problem) been completely public at various times in our nation’s history, but other countries impose a disclosure burden adequate to resolve the problem while still enjoying vibrant, competitive economies.
Anyway, my colleagues make this case more rigorously and persuasively both at this microsite and in this blog post (about the overall problem) and this one (about the technical considerations behind our presentation). I hope you’ll give it all a look.
Microsite doesn’t seem to be working (ie I type in a company name and the dynamic graph doesn’t do anything, although Flash is wonky on this machine, if that’s what’s driving it), so I can talk about the merits of what you guys are suggesting. Although I can’t tell if you’re just publicizing the problem or proposing a solution. I’d be interested in hearing more of the latter.
And I assume you’re familiar with Alex (Yorkshire Ranter) Harrowell’s work on corporate influence networks in the UK?
Hmm. The site *should* be working. It’s not Flash — what browser are you using? Some company records won’t display big, messy graphs. You should try using the autocomplete feature (“general dynamics” should work well). And then just be patient.
This site is designed primarily to demonstrate the problem, not to propose a specific solution. There are various approaches to the issue that could be part of an eventual solution — Chris Taggart’s OpenCorporates, the Dodd-Frank LEI rulemaking, Beth Noveck’s ORGpedia initiative — but I think it’s too early to say which ones will be most successful.
Personally, I feel that the current regime of business intelligence firms trying to keep up with the private sector will never produce satisfactory results. One way or another we need to expand the reporting burden, and to do so with a strong emphasis on modern open data norms. I suspect that the role Sunlight will play is to inject those norms into the process as a leviathan like ISO establishes a new, government-sanctioned identifier regime.
I’m not familiar with Harrowell’s work specifically — our work has only recently begun to hook into the international sphere. Is there a particularly good place for me to start reading about it?
I’m running Moz 3.6.23 on my work computer (XP); not sure why it’s not working. Oddly not working in IE, either. There it gives me an error in the status bar saying “Line: 90 Char: 5 Error: ‘Processing’ is undefined” Hope that helps.
Here’s as good a spot as any to read about Harrowell’s stuff:
http://yorkshire-ranter.blogspot.com/2011/06/meet-project-lobster.html
Pretty much he’s scraping feeds to do some basic link analysis stuff. He used to do it with plane tail number registries to figure out Victor Bout’s airline empire; now he’s using FOIA-equivalent schedule disclosures to find out who’s meeting with whom in the UK Govt. Interesting hobby.