Earlier today I spoke at the Brookings Institution event “The Future of E-rulemaking: Promoting Public Participation and Efficiency,” which was co-sponsored with the Administrative Conference of the United States. I made two points: we have not yet achieved regulatory transparency, and wiki-government does not overcome Hayek’s knowledge problem. What follows are my remarks.
When we talk about e-rulemaking, we often think about a first generation and a second generation of e-rulemaking.
The first generation is focused on making available online all of the information related to regulation and the rulemaking process, as well as making it simple for citizens to participate electronically in traditional rulemaking. In this way we improve the transparency and accountability of the regulatory process.
The second generation moves beyond the basics to leverage the new social technologies of the internet to increase citizen participation and enhance agency expertise. This is the exciting stuff of using Twitter and Facebook and wikis and collaborative commenting systems to achieve a truly democratic, efficient, and responsive rulemaking process. And while I’m very excited by the prospect of this transformation, I feel I have to suggest some caution.
For one thing, I’m not sure we have successfully graduated from the first generation. Less than two years ago we launched [OpenRegs.com](http://openregs.com) because [Regulations.gov](http://regulations.gov) did not offer something as simple as RSS feeds and had a less than ideal user interface. Since then it has been much improved, but if we look at the recommendations of the [ABA Administrative Law Section’s report on e-rulemaking](http://ceri.law.cornell.edu/erm-comm.php) — in which so many of the folks I see here today participated — or the recommendations of [OMB Watch’s Task Force on e-rulemaking](http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1292911), we can see that we’re a long way from where we should be to say that the first generation is complete.
The data that is made available online is often not standardized or structured in a meaningful way. And the interface for public interaction, in my opinion, could be greatly improved. Technology is not the problem. The technology exists — and freely in many cases — to vastly improve the accessibility and transparency of rulemaking dockets online. What’s missing are institutional reforms to require meaningful transparency. That’s why I’m happy to see Sen. Lieberman’s efforts on the E-Rulemaking Act of 2010, which would address many of these concerns including how e-rulemaking is funded.
I’d also love to see reform of federal contracting rules to allow agencies to entice open source communities and the types of small firms that are at the cutting edge of web innovation to help them develop great tools for e-rulemaking. FDMS and Regulations.gov were originally developed by Lockheed Martin, and they are now operated under contract by Booz Allen Hamilton. I’d love to see the sorts of experiences that firms like 37 Signals or Adaptive Patch would create given the opportunity.
Now, this all is not to say that the first generation of e-rulemaking must be done completely before we can begin experimenting in the second generation. But it does mean that to the extent that there are trade-offs, government should allocate resources to making sure that we have access to all relevant rulemaking data in structured searchable formats. If we’re given the data, third parties will be able to begin the second-generation experimentation by employing social networks to increase awareness, and collaborative tools to distill the wisdom of the crowds.
For example, look at the amazing work that Cynthia Farina has been doing with [Cornell’s e-Rulemaking Initiative](http://ceri.law.cornell.edu/). In partnership with the Department of Transportation it has developed an [experimental platform](http://regulationroom.org/) for citizen outreach through social media, human-moderated discussions, and collaboration on comments. With early access to data from DoT, they have been able to leverage networks, like Facebook and Twitter, and off-the-shelf open source tools, including WordPress and Digress-It, to bring together hundreds of interested citizens to collaborate on two live rulemakings. If more agencies made more of their data available early and often in usable formats, I’d like to think we’d see more experiments like Cornell’s.
Now, in keeping with the cautionary tone of my remarks, I have to also give a bit of a warning about the second generation of e-rulemaking.
Ideally we turn to regulation only when there is a market failure. One reason we prefer markets is that we recognize that regulators can’t possibly have all the information possessed by the myriad individual market actors — information that is communicated by prices. I fear that because new technologies make it easy for regulators to tap into “the wisdom of the crowds,” they may believe that they have solved what Friedrich Hayek called “the knowledge problem.” That’s a conclusion that we must resist.
Another thing the Cornell initiative’s experience has taught us is that it is very difficult to engage ordinary citizens in a rulemaking, much less getting them to make useful contributions, and that the doing so is very labor intensive. Now, I understand that Wikipedia is written and edited by an incredibly small fraction of its users, and yet they’re able to build a remarkable resource. But this small number of wikipedians form a persistent community that has developed over the course of almost a decade, with clear norms and a real culture. While the peer production of knowledge to improve regulation no doubt shows promise, we should understand that we have not solved the knowledge problem, and may never be able to do more than marginally improve regulations.
So let’s focus on finishing the first step toward the promise of e-rulemaking — greater online transparency — so that we can facilitate experimentation toward the next.