Open Data Creates Accountability


A series of recent blog posts raised questions on the value of open data and transparency.

While thoughtful skepticism is constructive, there appears to be some significant confusion about the meaning of “open data,” and about transparency and accountability. When activist developers like Aaron Swartz are concluding that “the case for opening up data to hold government accountable simply isn’t there,” or former government leaders like Beth Noveck are suggesting that there are “serious doubts” whether “open data” make government “more transparent or accountable,” then it’s time to engage.

We should clarify something straight away — this term “open data.” Open data wasn’t invented in 2009; open data isn’t born in a data portal. Construed most broadly, open data is people knowing things with technology. This information can be tabular, or not, structured, or not (though our preferences are clear.)

When people ask whether open data can create government accountability, they’re essentially asking whether it’s helpful to know things about the government, and, strangely, coming up with uncertain answers.

These answers are flawed, in part, because “open data” is being narrowly conceived of as the thing that fuels data contests and populates data portals, that is, the thing that sprang into vogue as Obama came into power.

While Sunlight has been deeply involved in the last 3 years of “open data,” we’re also deeply grounded in the last 50. Every bit of open data we have now to be mashed up, evangelized, or opened exists, in part, through the accountability laws and norms that decades of work have created, about where citizens stand before their governments, and vice versa.

If our first question is “does knowledge of government create accountability,” then the answer is clearly, definitively yes. Knowledge of the government creates accountability. As surely as ignorance and secrecy empower manipulation and abuse, information and knowledge empower self-determination.

This is baked into Sunlight’s mission — the idea that understanding the government changes how it works. The Brandeis quote that is the source of Sunlight’s name encapsulates that idea, and our work is intended to embody it.

To suggest that open data can’t create accountability is to ignore the open data that helps create the accountability we already enjoy, and work to strengthen.

Let’s start with the Freedom of Information Act. It’s an open data law. It just turned 46 on July 4th, and may be more relevant than ever. Even if we have to keep struggling to maintain its strength, FOI laws’ role in democratic societies should be as certain as a free media.

Let’s take a narrower example. Every time an FDA advisory committee recommends a new drug’s approval, we have the Federal Advisory Committee Act to thank for what we know about the people behind those decisions. While those decisions may range from completely meritorious to rankly corrupt, the response (be it legal, administrative, legislative, or public) is informed by disclosure requirements created by the FACA. While the FACA isn’t perfect, it has unequivocally created accountability. Every ounce of skepticism in every of these news stories is accountability directly created by an open data requirement. FACA is an open data law, and it creates accountability.

Our election laws are another great example. The Federal Election Campaign Act and its amendments are responsible for most of what we know about money in politics in the US. Every time a campaign finance law is enforced — and that means things as commonplace as the reporting requirements for every campaign donor over $200, candidates, PACs, and party committees, the FECA played a role. Every time a voter considers contributions and money in politics in voting, the FECA was involved. And still, even as FECA’s relevance wanes in a post Citizens United / Speech Now world, FECA still prevents corruption. Even the most die hard of frustrated campaign finance advocates’ arguments will be made through information the FECA gives us. FECA is an open data law, and it creates accountability. Is it perfect? No. But it would be idiotic and counterproductive to say that we should throw it all away and return to a world of secret money in politics.

Open legislation provides another great example. House and Senate rules provide for publicly available legislation. Every time someone reads a bill from the US Congress, an open data policy has created public accountability. Legislation should be published better, and legislating often happens beyond public view, but that doesn’t erase the accountability that comes from having public legislation. The House and Senate Rules have (many) open data policies, and they create accountability.

These just represent a few examples. All the vital connections within our democracy are built on open data. The argument that open data is difficult to connect to accountability is demonstrably false. Almost every time you find accountability within our democracy, there’s public knowledge powering it.

Perhaps the counterargument is that this accountability is not real, and that corruption just moves around. This argument, to me, that all shadows are created equal, is also demonstrably false.

Let’s take yesterday’s House Oversight Committee report on corrupt mortgage interest rates. In short, Countrywide bribed legislators by giving them discounted interest rates and other favorable treatment. This should be illegal, and is clearly corrupting. What stands out to me, though, is that this corruption occurred *just past* the line of what is required to be public. Interest rates on mortgages weren’t required to be publicly disclosed until recently, and that’s exactly the place that the corrupt behavior appeared. Skeptics may point and say “See? Let’s give up, corruption will just move again, always beyond our reach.”

But this dismissal misses a key point: not all shadows are created equal. Home ownership is publicly disclosed in the US, and because of that, legislators here don’t generally accept free homes. Huge personal bribes and free mansions are political currency in other countries, and in the US they’re the rare exception. While bad behaviour does follow the shadows, there’s a huge difference between huddling in corners and having a riot in the darkened streets. Our comparatively strong disclosure laws, combined with cultural norms, a free and energetic press, and an effective judiciary, are what protect us from lawless corruption in public service.

Open data can strengthen all of these laws, and that’s part of what Sunlight’s policy work is designed to do. That’s why we’ve tried to broaden and nail down open data policy a bit — so that the power of digital disclosure isn’t reduced to data portals or poorly sourced claims about economic benefits. Open government data is broader, and more powerful than it’s been given credit, and the praise it has received is too one-dimensional. It should be treated with thoughtful respect, since we’re talking about nothing less than human knowledge applied to democracy.

So our treatment of open government data needs to be circumspect, because information is indeed powerful. Advocates need more political savvy, or risk empowering obstruction dressed as transparency. That’s just what a recent proposal from Senator Paul would have done, were it taken seriously, proposing a full day’s delay for every 20 pages of legislation. Similarly, we should be skeptical of political cooption, and remember that bestowing personal recognition on politicians is akin to granting leaders immunity from our future judgment. That’s what advocates did last year in giving President Obama an award, and it empowered the administration to use their previous achievements as a shield against further criticism.

More common than political coopting, though, is the risk that new data policies focus on high profile provisions like contests and data portals, without attempting to address existing needs, or strengthening existing requirements. If you look through the arc of Sunlight’s blogging since 2009, for example, you’ll see critiques of the Open Government Directive shortfalls slowly coalescing into an affirmative agenda for data policy, with a particular focus on rigor and relevance to political power.

That’s not to say other values are inappropriate in transparency advocacy. As we pointed out in #29 of our open data policy guidelines, different policies tend to appeal to different values. As Tom Lee pointed out recently, government transparency advocates who appeal to different values fit quite easily under the same tent. Websites devoted to public services requests are popular precisely because they bring the accountability and efficiency of public online communications to government service requests that are often unaccountably ignored. Service, accountability, and efficiency are all deeply complementary, and suggesting that they are competitive is counterproductive.

Transparency is as complicated as public attention, so we shouldn’t suggest that unintended consequences don’t exist, or that disclosure requirements are like laser surgery. If public access policies ever cause harm to the broader public interest, we should re-examine them (see the end of this post). We haven’t encountered that problem in our work, though. In fact, much of our work explicitly grants access to information by and about government usually only available through expensive commercial services, if at all. In any case, Sunlight’s development priorities are about egalitarian access to democracy, and seek to undermine unfair disparities in access to our democracy.

None of this should be read to suggest that mere disclosure guarantees accountability. Transparency is a necessary but not sufficient condition for empowered knowledge. That’s why we build tools like Scout or Influence Explorer, that’s why we do investigative journalism and train reporters. We’ve seen what happens when disclosure isn’t coupled with public attention: public protections atrophy, and disclosure becomes a fake, rote performance.

We also know that disclosure’s effectiveness needs an informed public in order to succeed. Beyond equitable internet access and freedom of expression, without a literate public with enough free time and journalists to help create sense, accountability will remain just a goal.

Despite everything that open data needs in order to succeed, it is. Open data is as broad as public attention of government. The current round of questioning open data should help us realize that open data empowers all of our knowledge of government, and that without that knowledge, accountability can’t exist.