Using Social Media in Rulemaking

Using Social Media in Rulemaking:
Possibilities and Barriers

Michael Herz
Benjamin N. Cardozo School of Law

Draft Report to the Administrative Conference of the United States
October 21, 2013

This report was prepared for the consideration of the Administrative Conference of the United States. The views expressed are those of the author and do not necessarily reflect those of the members of the Conference or its committees. Using Social Media in Rulemaking: Possibilities and Barriers*

Michael Herz+






A. Social Media

B. The Appeal of Social Media for Rulemaking

C. Federal Agencies Embrace of Social Media

D. Realistic Expectations

1. Built-In Mismatches

2. Costs


A. Outreach

1. Examples of social media outreach

2. Regulation Room

3. The Unified Agenda

4. Beyond the Usual Suspects

B. Gathering Open-Ended Public Input through Social Media

1. U.S. Open Government National Action Plan and E-mail Comments

2. Exchange

3. Blogs

4. On-Line “Suggestion Boxes”

5. Ideation Sites

a. Examples

b. Assessment

C. Education for Effective Commenting

D. Obtaining Informed and Useful Comments from Diverse Stakeholders

1. Blogs

2. Discussions

3. Direct Outreach

4. Ideation Platforms

5. The Crowdsourcing Model

a. Distributed Tasking

b. Information

c. Rulemaking Applications

d. “Situated Knowledge”

E. Collaborative Drafting

F. Retrospective Review of Existing Regulations

G. Direct Final Rulemaking


A. The Administrative Procedure Act

B. Paperwork Reduction Act

1. The PRA Process

2. PRA Compliance as an Impediment to Use of Social Media

3. Legal Issues

C. The First Amendment

D. Federal Advisory Committee Act

E. Ex Parte Contact Policies



“E-topians” believe that technological developments will usher in a brighter future in many domains, not least in how democracies function. New technologies will, it is suggested, enable a robust, meaningfully participatory self-governance, in which an engaged and informed citizenry partners with government officials in a deliberative process and barriers between the governed and the governors are obliterated.

Notice-and-comment rulemaking is the pre-digital government process that most approached the e-topian vision of public participation in deliberative governance. K.C. Davis called notice-and-comment rulemaking the “most democratic of procedures” because all may participate.1 Regulators are required to accept comments from any interested person and consider and respond to them before making a final decision. The mechanism is already in place, all that is necessary is to make it more effectively open to ordinary citizens. If the Internet is to produce a democratic transformation, this is where one might first expect to see it.

In the last decade, the notice-and-comment process for federal agency rulemaking has changed from a paper process to an electronic one. Expectations for this switch were high; many anticipated a “revolution”2 that would make rulemaking not just more efficient, but also more broadly participatory, democratic, and dialogic. In the event, the move online has not produced a fundamental shift in the nature of notice-and-comment rulemaking. The process remains quite recognizable.

At the same time, the online world in general has come to be increasingly characterized by participatory and dialogic activities, with a move from static, text-based websites to dynamic, multi-media platforms with large amounts of user-generated content. At the heart of this move to “Web 2.0” have been social media, blogs, Twitter, Facebook, Youtube, IdeaScale, wikis, Flickr, Tumblr, and the like. Outside the rulemaking setting, federal, state, and local governments have enthusiastically jumped on the social media bandwagon.

If the move online has not produced the hoped-for gains in public participation, democratic legitimacy, and quality, perhaps the problem is not that those goals are unattainable but rather just that agencies have not been using the right technologies. Observers have labeled the existing version of e-rulemakingRulemaking 1.0,” as opposed to a possible “Rulemaking 2.0.”3 Rulemaking 2.0 would share the characteristics commonly associated with “Web 2.0”: interaction, collaboration, non-static web sites, use of social media, and creation of user-generated content.4 For those in the thick of these technological shifts, it seems self-evident that “[i]mproving public input today can’t be done without thinking of the existence and impact of social media.”5

The e-Rulemaking Program Management Office, which houses and embodies Rulemaking 1.0, has itself endorsed the use of these platforms, urging agencies to:

[e]xplore the use of the latest technologies, to the extent feasible and permitted by law, to engage the public in improving federal decision-making and help illustrate the impact of emerging Internet technologies on the federal regulatory process. New tools (such as blogs, wikis, user generated feedback and ratings, social bookmarks, videos, and links to share information in social media networks) serve to promote and facilitate transparency, public engagement, and collaboration. When federal agencies use these tools in the regulatory process, stakeholders have the time to take advantage of information sharing and knowledge transfer. This added form of communication is likely to increase formal and informal stakeholder contributions, thus increasing their participation in the federal decision-making and regulatory process.6

The Administrative Conference has consistently supported full and effective public participation in rulemaking and the use of new technologies to enhance such participation. For example, a recommendation from the dawn of the e-government era, Recommendation 95-3, Review of Existing Agency Regulations,7 includes the following regarding review of existing rules:

Agencies should provide adequate opportunity for public involvement in both the priority-setting and review processes. In addition to reliance on requests for comment or other recognized means such as agency ombudsmen and formally established advisory committees, agencies should also consider other means of soliciting public input. These include issuing press releases and public notices, convening roundtable discussions with interested members of the public, and requesting comments through electronic bulletin boards or other means of electronic communication.8

More recently, in Recommendation 2011-8, Agency Innovations in e-Rulemaking,9 the Conference has endorsed, in general terms, the use of social media in rulemaking:

Agencies should consider, in appropriate rulemakings, using social media tools to raise the visibility of rulemakings. When an agency sponsors a social media discussion of a rulemaking, it should provide clear notice as to whether and how it will use the discussion in the rulemaking proceeding.10

The present study reviews how federal agencies have been using social media to date and considers when, how, and whether agencies should use social media in rulemaking, not just to “raise the visibility of rulemakings,” but as tools in formulating the content of rules. The Conference’s charge is to consider whether and how social media could be used by agencies to improve the rulemaking process and what barriers, especially legal barriers, stand in the way.

This report is in five parts. After this Introduction, Part II summarizes the methodology behind this report. Part III then reviews the accomplishments and shortcomings of electronic rulemaking, noting in particular that, contrary to expectations, e-Rulemaking has not been fundamentally more inclusive and dialogic than the paper process it replaced. Part IIV considers in the abstract how social media and Web 2.0 might be incorporated into the rulemaking process and also why doing so will not be either simple or an unmitigated blessing. Part V reviews ways in which agencies have used social media to date and addresses particular potential applications of social media for outreach, discussion, and input. Part VI considers legal obstacles to such applications. Throughout these sections, recommendations are put forth; these are collected as a group and set out in Part VII.


In preparing this report, I initially reviewed the extensive literature regarding government uses of the Internet and affiliated technologies and studied the relevant statutes and the administrative and judicial materials connected to them. These include not just academic writing but numerous agency reports. I then conducted a number of interviews with agency staff members and leading academics in the area to learn about particular projects that have been attempted or considered and to gather different perspectives on the barriers and possibilities of using social media in rulemaking. (A list of interviewees is attached as an Appendix to this report.) I also did extensive on-line research. This took two forms. One was reading news, blogs, and other content posted by people who are most convinced about the value of social media and enhanced public participation in governance. The other was to explore and experience the various agency uses of social media that are in place.

Two meetings organized by ACUS provided further information and input. First, the Rulemaking Committee met in March 2013 to discuss an early draft of this report. Second, in collaboration with George Washington University, on September 17, 2013, ACUS hosted a workshop on the use of social media in rulemaking. The workshop featured two panel presentations by those with hands-on experience and an open-ended question and answer session with the audience, which consisted primarily of agency staff.11

Given the nature of the project, I did not conduct a survey of federal agencies with regard to their social media practices. I felt that on the one hand, social media use outside of the rulemaking area was too diffuse, fluid, far-reaching—and, on the other, the use of social media explicitly for rulemaking was too rare—to justify the time and effort of such a survey.


Articulating a generally held understanding, the D.C. Circuit has written that the notice and comment process is supposed “(1) to ensure that agency regulations are tested via exposure to diverse public comment, (2) to ensure fairness to affected parties, and (3) to give affected parties an opportunity to develop evidence in the record to support their objections to the rule and thereby enhance the quality of judicial review.”12 In addition, “a chance to comment ... [enables] ‘the agency [to] maintain[ ] a flexible and open-minded attitude towards its own rules.’”13 “To achieve those purposes, ‘there must be an exchange of views, information, and criticism between interested persons and the agency.’”14 President Obama’s Executive Order on Improving Regulation and Regulatory Review takes a similar position: “Our regulatory system . . . must allow for public participation and an open exchange of ideas.”15

The traditional paper-based process, with comments stored in a single docket room in Washington, DC, always and necessarily fell short of this ideal. Barriers to participation reduced the likelihood of “diverse public comment,” limited the opportunity for participation by all affected parties, and meant that some useful information was not reaching the agencies.16 Perhaps more important, the basic structure—a one-shot opportunity to submit comments—prevented any real “exchange . . . between interested persons and the agency.” And note what the D.C. Circuit does not even mention, viz. the possibility of “an exchange of views, information and criticism” among interested persons. Having the agency as a central repository for unilateral comments directed to it, like the hub of a spoked wheel, prevented such dialogue.

Electronic rulemaking was widely anticipated to mitigate these shortcomings. The expectation was that it would produce two basic changes in the way agencies write regulations and, by extension, the substance of the regulations ultimately adopted. First, the Internet massively reduces barriers to public participation in rulemaking. E-rulemaking was thus expected to open to all what had been a largely invisible insiders’ game limited to sophisticated players blessed with access, funds, a Washington, DC presence, and good lawyers. Second, e-rulemaking promised to make the process more dialogic. Instead of a spoked wheel, with the agency at the hub and numerous isolated commenters sending their comments in to the center, all independent of one another, the online process seemed to invite reply periods,17 comments on comments, exchanges through different media, collaborative drafting—in short, a conversation, with genuine give and take.

The expectation was that these two changes would in turn have three significant benefits. Most prosaically, it would be more efficient. Agencies would have less paper to manage, and centralizing the process in would bring economies of scale.

Second, and most grandly, by bringing in a wider range of participants, the process would be more “democratic.” This assertion is often offered as self-evident; the more people participating in a process, the more democratic it is. But this claim requires some unpacking. Broad participation is not actually an end in itself, although agency staffers and commentators often treat it as one. Rather, the democratic value would seem to consist in (at least) three subsidiary values. (a) To the extent that agency rules reflect judgments about values or preferences rather than technical problems involving expertise, they are arguably more legitimate if they reflect popular input. (b) Broader popular participation will produce a more informed citizenry, which in turn will be able to hold political actors accountable through mechanisms other than participation in rulemaking. (c) Broader participation will produce greater buy-in regarding the resulting regulations, which in turn will lead to fuller and less costly compliance.

The third anticipated value of broader and more dialogic participation was that it would, simply, produce better rules. This might happen or several reasons. For one thing, rulemakers would have access to more and better information. As Cary Coglienese has written: “[T]he local sanitation engineer for the City of Milwaukee . . . will probably have useful insights about how new EPA drinking water standards should be implemented that might not be apparent to the American Water Works Association representatives in Washington, DC.”18 E-rulemaking might produce better rules because the process would allow for a fuller vetting of public submissions. Having comments online and readily accessible could result in comments on comments, reply periods, or other exchanges that would test and refine public submissions in a way that does not occur when everyone submits directly, at the last minute, without the opportunity to see what others have submitted.19

E-Rulemaking is in many ways an improvement over the paper-based process it replaced, First, it is easier to submit a comment. This is a plus; it is hardly a transformation. It is easy to send an email or upload a document, but printing out and mailing a document is not that hard either.

Much more important is the ready availability of materials in the rulemaking docket. There is no question that having that material available online improves the ability of commenters to review and respond to it more effectively, and this can only be a good thing. The point is not just that the new regime is more efficient, though it is that.20 It also makes for higher quality comments. No one has proved this, but it is supported by a survey of agency staff by Jeffrey Lubbers21 and informal conversations, and it is what one would expect.

Widely available rulemaking dockets are surely of use to others besides commenters. There’s a lot of good stuff in rulemaking dockets. To be sure, it can be hard to find. One of the things that has made steady and impressive progress on over the years is making it easier to find material on its site. One major breakthrough was creating full-text searching. More recently, in February 2012 the site introduced a set of Application Programming Interfaces (APIs) to enable third parties to search and retrieve material on the site.22 Their ready availability is only one small aspect of sweeping shift in the easy availability of information held by the government. Their value is not an aspect of notice-and-comment rulemaking per se, and for present purposes it suffices just to nod toward, or incorporate by reference, the expansive literature on the utility of make government-held information widely available.23

In addition, an online docket makes it easier for agency staff to do its job. No one has to worry that something has been checked out, more than one person can use a document at a time, people stay out of each other’s way.24 And the docket is available to agency staff who do not work at headquarters.25

These are real improvements, but while the mechanics of notice-and-comment rulemaking have changed, and very much for the better; the nature of the process remains essentially what it was before the move online. E-Rulemaking’s grander anticipated benefits have not yet come to pass.26 With isolated exceptions, there has not been a huge outpouring of lay comments.27 Moreover, though the matter is disputed, lay comments have by and large not been especially helpful or influential. Few people are aware of the opportunity; of those who are, few bother to participate; and few of those who participate manage to submit something useful or persuasive. They generally fail to provide the things that agency staff most need: concrete examples, specific alternatives to the proposal, an awareness of statutory limitations, hard data to back up conclusions, and direct responses to any specific questions the agency may have asked.28 In those rulemakings that have generated extensive lay participation (a distinct minority) the comments have been dominated by duplicative submissions resulting from organized “astro turf” campaigns. Tens or hundreds of thousands of near-identical submissions are a testament to the costlessness of submitting a comment. But such “click-through democracy,” in Stuart Shulman’s phrase, may be a “harbinger[] of a slide into a technological arms race predicated on plebiscite-style governance.”29 Even e-Rulemaking’s greatest enthusiasts acknowledge that “the digitization of citizen participation practices has not worked well. . . . Online participation has thus evolved into ‘notice and spam’ rather than notice and comment.”30

Analyses of individual rulemakings provide further evidence that e-rulemaking has yet to deliver on its ambitious promise of more effectively engaging the public and providing a forum in which lay citizens have an effective voice. For example, Kimberly Krawiec read every nonduplicative comment submitted to the agencies jointly responsible for implementing the so-called Volcker rule.31 In her description, the public comments—some of which are included in the examples given above—were short, lacked specific suggestions, did not grasp the distinction between the statutory provision and the regulation that would implement it, were poorly written, and overflowed with anger. “[T]he contrast with the meticulously drafted, argued, and researched—though far less numerous—letters from industry and trade groups is stark.”32

In short, lay participation has shown haphazard increases in quantity. But that increase has been haphazard, manipulated, uninformed, and largely unhelpful to rulewriters. As for the traditional, sophisticated participants, they are doing what they have always done. Their comments are lengthy, well-researched comments, often prepared by counsel, and generally submitted right at the close of the comment period. (The last-minute submission is generally seen as being in part just a function of human nature, but also the result of wanting to avoid subjecting one’s comments to review and critique by other commenters.33) The fact that the comments are posted on-line or attached to an email is no real change at all.

In addition, e-rulemaking has not proven more dialogic or collaborative than the traditional paper process. other ways, too, the rulemaking process remains completely recognizable. The FCC makes use of reply or rebuttal comment periods as a matter of course.34 But the FCC largely stands alone. Use of reply periods remains quite rare and, strikingly, has not significantly increased with the move of rulemaking on-line.35 Commenters still write their comments in isolation and submit them right before the deadline; the agency still responds in the preamble to the final rule. Instead of providing a shared venue for collaboration and discussion, electronic rulemaking, in Peter Shane’s incisive description, “resembles a global suggestion box, appended to an electronic library.”36

These indicators of e-rulemaking’s mixed success to date in achieving its promised transformation of rulemaking are infrequently acknowledged by those responsible for managing FDMS. For example, trumpets:

Federal regulations have been available for public comment for many years, but people used to have to visit a government reading room to provide comments. Today, the public can share opinions from anywhere on removed the logistical barriers that made it difficult for a citizen to participate in the complex regulatory process, revolutionizing the way the public can participate in and impact Federal rules and regulations.37

It is the last paragraph that is the tricky one. E-rulemaking has undeniably “removed the logistical barriers” to citizen participation—at least many of them. And that has “revolutioniz[ed] the way the public can participate in” federal rulemaking. But the other claim—that the site has revolutionized the way the public can “impact Federal rules and regulations” is more a statement of faith than a statement of fact. This is because (a) the barriers to effective public participation are not only logistical and (b) it is exceedingly difficult to identify, and harder still to measure, the impact of public comments on rules. Another example on is a one-page document provided on the site identifying “Program Impacts and Achievements.”38 To be sure, the achievements listed are real and documentable—numerous awards, undeniable improvements to the site, hundreds of millions of visits, significant cost savings, and the migration of agencies’ “legacy data” onto the site. Notwithstanding its heading, however, the document does not actually identify any “impacts.” Indeed, it would be surprising if that were possible, in part because the impacts of are much harder to quantify than are the site’s other accomplishments. But it is also because the move online has not transformed rulemaking, which remains quite recognizable as a version of the process that has long existed.39


The experience with e-rulemaking to date begs the question ACUS has posed: accepting that the basic goals of e-Rulemaking are desirable, might they be achieved through adoption of new technologies? In particular, could agencies use social media to improve the rulemaking process? In the decade-plus since e-rulemaking began in earnest, the Internet has been transformed by so-called “web 2.0” technologies. That term means different things to different people, but the core concept is that while web 1.0 consisted of static websites and repositories from which users could retrieve information, web 2.0 involves interaction, collaboration, the uploading of user-created content. In short, Web 2.0 is characterized by just the sort of activity that e-rulemaking was anticipated to produce but generally has not.

A. Social Media

The ACUS Request for Proposals did not define “social media,” although its opening paragraph gives a good sense of the project’s anticipated scope:

Social media, including Facebook, Twitter, blogs, and other similar technologies, present new opportunities for agencies to engage the public in rulemaking activities. Such social media tools are uniquely valuable because they facilitate two-way communication. Rather than just pushing information out, social media allows agencies to provide the public with a way to communicate views and information to the agency.40

There is no single, established definition of social media; to the contrary, definitions are abundant and varying.41 EPA has offered this: “any online tool or application that goes beyond simply providing information, instead allowing collaboration, interaction, and sharing.”42 There are numerous such platforms. One possible taxonomy appears on the following page.

Types of Social Media

A Web log (Blog) is a Web-based interactive application that allows one to log journal entries on events, or to express opinions and make commentaries on specific topics. It is a popular content generation tool. Blogs typically consist of text, images, videos, music, and/or audios.

The process of creating a short blog that is primarily achieved through mobile devices to
share information about current events or personal opinions. A well-known example is

A Web-based collaborative editing tool that allows different people to contribute their
knowledge to the content. One author’s content can be modified and enhanced with another author’s contribution. A well-known example of this application tool is Wikipedia.

Social networking
A Web-based tool or model that allows individuals to meet and form a virtual community
through socializing via different relationships, such as friendships and professional relationships, sharing and propagating multi-media information, exchange interests, and communicating.
The rich multi-media contents such as photos, videos, audios are shared through multi- media sharing tools. Typically examples include YouTube, Flickr, Picasa, Vimeo, etc.

An application that uses contents from two or more external data sources, combines and integrates them, and thus creates new value-added information. This is a reuse and repurposing of the source data by retrieving source contents with open APIs (Application Programming Interfaces) and integrating them according to the information needs, instead of navigating them sequentially.

A Web application that can pull the content from sources that are structured in standard
metadata format called RSS (Really Simple Syndication) feeds such that it is easy to syndicate the contents from RSS formatted documents. The RSS feeds or Web feeds can be published and updated by the authors such that the updates can be easily inserted and quickly updated in content aggregation sites. The RSS feeds (also called atoms) are
annotated with metadata such as the author and date information. The RSS based content aggregators include news headlines, weather warnings, blogs, etc. Once the source content is updated, the content aggregator sites will be updated thus always sharing the updated content.

Small applications either on the desktop, a mobile device or the Web. The widgets bring
personalized dedicated content to the user from predefined data sources.

Virtual World
A virtual world is an interactive 3-D computer-simulated world were avatars, controlled
and played by the users, interact with each other as inhabitants.

Social Bookmarking & Tagging
A tagging system that allows the users to describe the content of the Web source with
metadata such as free text, comments, evaluative ratings and votes. This human generated collective and collaborative set of tags forms a folksonomy and helps cluster Web resources.

[Source: Kelsi Klaerrs, Towards Effective Web 2.0 Public Engagement: A Case Study of (May 13, 2013), unpublished manuscript available at, citing Soon Ae Chun et al., Government 2.0: Making connections between citizens, data and government, 15 INFORMATION POLITY 1 (2010).]

The essential features of social media (or, what are generally seen as essentially synonyms, “social technologies” or “social networking”) are usually understood to include:

* the ability to support two-way social interactions in real time;
* the ability to allow creation and exchange of user-generated content (“UGC”); and
* easy and low-cost accessibility by large numbers of people without specialized skills or training.

Surely the best-known social media platform is Facebook, familiar to just about everyone. Facebook has roughly 165 million users (about 130 million daily active users) in the United States and over a billion users worldwide. Users go to the site to communicate with friends, sharing photos, information, and links to websites of interest. Everything of interest on the site is there because a user put it there—the content is user-generated—and the experience is defined by its interactive, communicative nature. Other examples include similar social networking sites, such as Google+; blogs; microblogs, of which Twitter is the dominant example; file or photosharing sites, such as Flickr or Instagram; wikis, which allow unlimited number of individuals to contribute to or edit text; and mechanisms for voting or ranking specific items, such as IdeaScale or Reddit.

Social media is a moving target, constantly evolving and expanding. Any list will be instantly out of date.43 Accordingly, this report for the most part focuses on general types of tools rather than particular platforms.

B. The Appeal of Social Media for Rulemaking

Recall the ways in which e-Rulemaking has fallen short of its original vision: barriers to effective participation remain high because members of the public remain largely unaware and uninformed about the process and particular rulemakings and do not know how to make useful contributions, there is no back-and-forth among commenters or between commenters and the agency, and the process remains largely sealed off from the public at large. Social media tools seem, at least on the surface, to offer a solution to exactly those problems.

First, quite simply, social media sites are the places in the virtual world where the most people can be found. As one leading academic researcher and social media enthusiast writes:

So how do you [i.e. government] expand th[e] pool of participation? How do you collect input from those who may feel marginalized or are simply too busy to invest the time needed to attend a council meeting or other forum? . . . Traditional websites are not the answer . . . [R]elying on an online survey on your agency’s own website is like putting a shining billboard on a backcountry road. It’s pointless! You need to move the message and the debate to a forum where the people are. Enter Government 2.0.44

Second, social media allows significant interaction and dialogue. In contrast to the “one-to-many” nature of traditional media, and the “many-to-one” nature of traditional avenues of public comment and input, social media holds the promise of collaborative discussion among the many.45

Third, a defining characteristic of social media is that the users create most or all of the content. If the goal is seriously to hear, and learn, from pubic submissions, then agencies need to use tools that fully enable and encourage submissions, with low barriers to participation and an openness to varying types and formats.

Given these characteristics, social media has obvious potential value as a way of increasing public participation in rulemaking. It could be an avenue by which the agencies “give interested persons an opportunity to participate in the rulemaking through submission of written data, views, or arguments,”46 i.e. accept comments on a proposed rule. In addition, agencies are in search of input both in the rule development stage and in the post-promulgation stage.

An essential assumption should be made explicit here, namely, that increased public participation is desirable. In much of the discussion about rulemaking, it is taken as a given that more participation is better. If that is the case, then of course use of social media makes sense. If, on the other hand, participation is not an end in itself but rather a means to other ends (including informed rulewriters, better rules, greater public acceptance of rules, or enhanced compliance), then agencies must be more nuanced about when and how they use social media tools.

Recommendation: As part of the rulemaking process, agencies should explore on-line platforms that enable opportunities for public consultation, discussion, and engagement that go beyond merely submitting a written comment to the rulemaking docket.

C. Federal Agencies’ Embrace of Social Media

As the public has gravitated toward social media, government agencies at the local, state, and federal level47 have not been far behind, embracing social media with remarkable enthusiasm in non-rulemaking contexts. Indeed, the enthusiasm and extent of this activity belies agencies’ reputation as risk-averse, slow to change, and nervous about transparency.

The turn to social media has been given a strong push by the Obama Administration’s emphasis on transparency and openness. Clearly drawing on recent interest in “crowdsourcing,” officials from the President down have expressed a desire to tap into the “dispersed knowledge of the American people.” This aspiration is at the heart of President Obama’s much-invoked Open Government Memorandum, issued on the first day of his presidency:

Government should be participatory. Public engagement enhances the Government's effectiveness and improves the quality of its decisions. Knowledge is widely dispersed in society, and public officials benefit from having access to that dispersed knowledge. Executive departments and agencies should offer Americans increased opportunities to participate in policymaking and to provide their Government with the benefits of their collective expertise and information. Executive departments and agencies should also solicit public input on how we can increase and improve opportunities for public participation in Government.48

The memorandum does not mention rulemaking as such, but it is one of the most obvious settings in which agencies can offer citizen “opportunities to participate in policymaking.”

There are currently about 2,000 verified government social media accounts across nearly two dozen different platforms.49 A “Social Media Community of Practice,” which dates to June 2012, brings together more than 200 federal social media managers.50 A 2010 GAO report found that 22 of 24 “major” federal agencies had a presence on YouTube, Facebook, and/or Twitter51; a year later, the number was up to 2352; it is now 24 out of 24.53 Blogs, Flickr pages, and other undertakings are also common.54 So, for example, the Environmental Protection Agency maintains

* 23 Twitter feeds from agency headquarters;
* 12 Twitter feeds from regional offices;
* 16 Facebook pages from headquarters;
* 16 Facebook pages from regional offices;
* 12 blogs;
* Multiple discussion forums;
* A Youtube channel;
* A Flickr photostream;
* A page on Google+;
* A page on Foursquare;
* A page on the government-wide website, which collects all open and recent federal government prize competitions;
* A wiki that gathers information about watershed management;
* An RSS feed for its news releases; and
* A collection of podcasts on environmental topics.55

The EPA is a large agency with a tradition of being relatively open to new technologies. Most agencies do not have quite this array of social media sites. But some do, and many come close.

This has all happened quite rapidly.56 The Federal Web Managers Council has assembled a time line indicating when and how different agencies have embraced social media.57 The first item, somewhat pathetic in retrospect, is from April 1, 2002, when the White House Easter Egg hunt was live-streamed. The next item does not appear for another two years, but the pace picks up pretty dramatically starting in 2008, and the timeline is crammed in the ensuing years. Then it stops as of July 2011, which is in itself an indication that it is no longer news when an agency turns to social media.

How are agencies making use of their extensive social media presence? Such use is dominated by efforts to push information out from the agency to the public, rather than to gather information flowing in the other direction. Agency Youtube channels (many of them combined in the channel) and Twitter feeds, for example, are primarily ways of reaching the public rather than ways of interacting with the public. They provide general information, “tell the agency’s story,” or let people know about available services, benefits, or employment opportunities.58

The Coast Guard’s description of its social media efforts captures this reality.

Our social media program will complement our media relations efforts as part of a comprehensive communications plan to educate and engage our publics. . . . As public affairs professionals, we rely on three basic mediums to tell the Coast Guard story: words, pictures, and video. . . . The Coast Guard will centralize and focus our use of social media tools to complement our media relations program and maximize our impact with unique audiences.59

Many agencies have posted videos on Youtube and/or the agency’s own website. These vary enormously in subject matter. For example, EPA’s 238 videos range from interviews with gay, lesbian, and transgendered EPA employees discussing the struggles they faced growing up60 through a video touting the benefits of hydraulic hybrid vehicles61 to an endorsement of e-waste recycling62 and discussion of mercury emissions from small-scale gold processing.63 TSA’s videos (of which there are 84 as of October 1, 2013) are focused on the particulars of airport screening.

Agency twitter feeds by definition simply alert followers to information available somewhere else, such as, for example, the agency’s videos on Youtube. So, for example, the Centers for Disease Control (310,000 followers) sent out a tweet pointing followers to a Youtube video regarding HIV and African-Americans and promising (over-promising, frankly) that the video will reveal “how you can help stop HIV in your community”).64

Agency Facebook pages provide news about agency initiatives, information about underlying substantive issues, and “tips” of various sorts relevant to the substantive issues with which the agency is dealing, generally aimed at promoting healthier, safer, or more environmentally sound lifestyles. They also encourage cross-media pollination, so to speak, urging viewers to subscribe to the agency’s twitter feed, visit its website, go to its Flickr page and Youtube channel, and so on.

Agency blogs are ubiquitous, and some are widely read. For example, Dipnote, the State Department’s blog, “recently passed 15,000,000 page views and 13,000 comments by the public.”65 The blog’s RSS feed has more than 2500 subscribers.66 Another very active blog is the TSA’s. In its five years of existence it has received approximately 75,000 comments (of which over 20,000 were deleted as inconsistent with the blog’s comment policy).67 The blogs are valuable sources of information about the agency and serve an educational purpose.

In short, federal agencies are deeply engaged and familiar with social media platforms. Most of the technical, bureaucratic, and legal challenges involved in government use of these platforms have been overcome; they are in place and could be put to work in the rulemaking setting.

D. Realistic Expectations

Given the broad enthusiasm for social media and the profound opportunities it creates for public input and participation, one might think that use of these tools in rulemaking is inevitable; indeed, the question might be why it is not already happening. For it has not already happened. Agency use of social media has fallen short of the participatory democratic ideal. First, existing uses are built around the government’s role as a provider of services and information; what is sometimes called a “managerial” model of online state/citizen interaction,68 or “e-government” as opposed to “e-governance”69 (though those terms are slippery). Second, in general, levels of public participation have been disappointingly low. Third, where agencies have sought to gather input or prompt discussion through social media, the quality of participation has been haphazard, with a sizable portion of public contributions consisting of off-topic and unconstructive attacks on the agency or other posters. Fourth, in practice there has been little interchange or dialogue, either among commenters or between the agency and commenters. In other words, so far, social media platforms have shared many of the characteristics of e-rulemaking.

This section reviews certain features of social media that will stand as an obstacle to its achieving the “revolution” that electronic rulemaking has not.

1. Built-In Mismatches

In part, the managerial focus of agency web use reflects a reality about modern government, namely, it does a lot of things besides govern. Most people’s contacts with the government involve the provision of services, not the imposition of rules and regulations. This reality is reinforced by the existence of a perpetual campaign. Politicians want—indeed, need—to be seen as providing direct, tangible benefits to constituents. There’s more payoff in announcing that it will provide services more effectively and efficiently than there is in announcing that it will regulate more effectively and efficiently.

But more is at work here. It is striking that e-government so closely resembles e-everything else. Many theorists of the Internet have envisioned it as a tool for unprecedented political engagement and deliberation, a forum where a thousand, or a million, citizen voices will bloom, a “networked public sphere” to which all have equal access and in which all voices can be heard As Matthew Hindman points out in The Myth of Digital Democracy,70 that just has not happened. Politics is not what the Internet is about. Web users go online to shop, socialize, be entertained, look at pornography, and find information (usually not about politics). “Given the magnitude of traffic flowing to other categories of online content, traffic to political sites is small enough to be a rounding error.”71 The governmental equivalent is that users go to government web sites for services and information that is useful in their daily lives; they do not go online to participate in a great national debate.

Why is this? Presumably the answer lies partly in the nature of the Internet and partly in the nature of the citizenry. The government’s web presence both responds to but also reinforces the non-policy interests of users. And the features of Google and other search engines that Hindman discusses in some detail apply to some extent to government web sites like any other. In addition, most people most of the time just are not that interested. In the words of two political scientists who have staked out the extreme position regarding Americans’ antipathy to political engagement: “The last thing people want is to be more involved in political decision making.”72 Whether they could be, or should be, or could be made to be through e-government,73 are not questions I am going to try to resolve, but they underlie the debate over much in the way of current open government efforts.

Alex Howard, one the most knowledgeable and perceptive open-government journalists, makes the point this way:

Rulemaking and regulatory review are, virtually by their nature, wonky and involve esoteric processes that rely upon knowledge of existing laws and regulations.
While the Internet could involve many more people in the process, improved outcomes will depend upon an digitally literate populace that’s willing to spend some of its civic surplus on public participation.
To put it another way, getting to “Regulations 2.0” will require “Citizen 2.0” — and we’ll need the combined efforts of all our schools, universities, libraries, non-profits and open government advocates to have a hope of successfully making that upgrade.74

There are two points here. One is simple: it is unlikely that large swaths of the citizenry will get caught up in rulemaking, regardless of the technology used. The second is more complex: it is unclear that broad participation by the general public is valuable in rulemaking. Drawing on their work with Regulation Room, the most sophisticated and promising use of Web 2.0 in rulemaking, Farina and Newhart have distinguished four types of potential commenters. The groups and their characteristics are set out in the table on the following page.75 Sophisticated stakeholders are effectively engaged already and always have been; in theory social media might improve the value of their participation, but it is not needed to bring them in to the process. The other three groups are un or underrepresented at present. These are missing stakeholders, such as small businesses or individual consumers, who are generally unaware of rulemakings and not equipped to participate effectively; unaffiliated experts, such as academics, who have historically not been focused on agency rulemaking but who have the capacity to make useful contributions without much help; and members of the general public who are “interested” in a loose sense but generally lack both specialized knowledge and the ability to understand and effectively participate in rulemakings. Farina and Newhart are skeptical about what members of group four have to contribute.

Second, citizen participation in e-rulemaking, IdeaScale, agency Youtube video watching, and the like follows what is the standard distribution curve on the Internet: the power law. A “power law distribution”—in contrast to a “normal distribution,” which shows up as a bell curve—is characterized by a very small number of data points of with high values and a very large number of data points with lower values.76 Whether it is blog readership, website hits, products sold by online retailers, Youtube video viewings in general, or anything else, the Internet

Types of Potential Rulemaking Participants & Their Likely Capabilities

Sophisticated stakeholders

Missing Stakeholders

Unaffiliated Experts

Interested Members of the Public

Who they are

Directly affected by proposed rule (either because their conduct would be regulated or because they would directly benefit); experienced in interacting with the agency in RM and other contexts

Directly affected by proposed rule (either because their conduct would be regulated or because they would directly benefit); do not participate in RM or other agency policy interactions

Scientific, technical or other professionals who are not direct stakeholders, and not employed or retained by a stakeholder in this matter

Individuals who self‐ identify as interested in the proposal, but are not in the previous groups


Trade association of large trucking companies; large mortgage lenders; major airlines

Small trucking company owners; drivers; travelers with disabilities; consumers who went through foreclosure; community bank officials

Researchers on driving fatigue or traffic accident prediction models; accessible designers consumer behavior researchers

Members of the driving public

Awareness of relevant ongoing rulemakings


Typically, low

Typically low, but might vary with field and particular rule

Possibly general awareness in highly politically salient RM; otherwise, low to nonexistent

Understanding of RM process and larger regulatory environment

High; often “repeat players”

May have patchy knowledge of regulations that immediately affect them; unlikely to understand RM process or larger regulatory environment

Hard to predict; likely dependent on field and particular rule

Low to nonexistent

Ability to comprehend meaning and implications of agency’s proposal without help

High; often have staff that specialize in regulation; likely to have in‐house or hired legal and technical experts

Low on deciphering NPRM and supporting cost/benefit projections

High for parts directly relevant to their expertise

Very low on deciphering NRPM and supporting

cost/benefit projections

Ability to produce effective comments without help

High (already have access to the required help)

Low; likely to have relevant situated knowledge but communication is impeded by lack of knowledge of RM process or larger regulatory context

Likely high for parts relevant to their expertise

Very low

Source: Cornell eRulemaking Initiative

produces a handful of hugely popular winners and then a “long tail” of almost completely ignored content.77 Most online rulemakings have only a few public comments for the same reason that most blogs have only a few readers. There are only so many interested citizens to go around, and rulemaking simply does not involve the sort of content that goes viral (which is also always partly a matter of happenstance).

Third, “commenting” on the Internet more often than not devolves into snarky ad hominems, with like-minded folks reinforcing each other’s views78 and “discussion” consisting of little more than polarized and polarizing name-calling. Anyone who has spent time reading comments on popular blogs should be a little nervous about replicating that experience in agency rulemaking. The TSA has disabled comments for all the videos on its YouTube channel, and it is not hard to guess why. The Huffington Post deletes 75% of the comments it receives; in the words of its Managing Editor, this is “either because they are flat-out spam or because they contain unpublishable levels of vitriol.”79

Fourth, the essential thing the Internet does is make it easier to distribute content. It does not make it easier to produce content (except in that, because it takes content, or information, to produce content, the ready availability of more material will make the task of producing more material easier.) One of the reasons for the continued concentration of news sources in the electronic age is that even when distribution is essentially free—with no need to buy paper, use printing presses, hire drivers, etc.—there remain economies of scale in producing the content.80 In submitting rulemaking comments, the hard thing is writing good comments. Distribution was never the problem, since the comment is only sent to a single reader. Obtaining information was part of the problem, and the move online has significantly ameliorated that part. But it was only part of the problem.

Fifth, notice-and-comment rulemaking is often a poor fit with Web 2.0 approaches and assumptions because of the obvious but sometimes overlooked fact that commenting involves words (which also means it involves reading). In contrast, one of the defining characteristics of social media is that it is multi-media and therefore allows communication other than through words. That is breathtaking and wonderful and valuable in many settings. But writing regulations just is not one of them. The Web 2.0 emphasis on photos, videos, mashups, etc. does not have much to offer the rulemaking process. In a presentation to agency staffers, Adam Connor of Facebook offered ten tips for government use of Facebook.81 Number six on his list was a reminder that most rulewriters would not have thought necessary: “Words can have power too” (not do, just can).82 So, admonished Connor, “not everything has to be a picture; not everything has to be a video.” It is interesting that the Facebookers are so taken with visual content that it feels like an insight and a valuable reminder to be told that words “can have power too.” That alone implies that much of what Web 2.0 has to offer may be a poor fit for rulemaking.

In addition, words must be read. Part of what can be demoralizing and overwhelming about comment sites, even ones with well-behaved, moderate, informed participants, is that there are just so many comments.

One of the main problems of user-contributed content on big media sites is often not even that it’s low quality, but that it’s too abundant. The Huffington Post is another such sufferer of the comment-overload affliction. Take a look at its lead story right now: already it has way over 2,000 comments. Who’s supposed to read all those? There may be a few worthwhile comments in there, but how the hell do you find them?83

The Huffington Post gets 200,000 comments a day, more than 70 million a year.84 There are only two ways this volume can be handled. It can be ignored, or it can be read by a computer. HuffPo has tried both methods. It recently rolled out a new platform, “Conversations,” through which the computer reads all the comments and picks the “best.” This option is not legally or practically available to agencies. As the Regulation Room team has written, “orthodox federal Participation 2.0 thinking” holds that more participation is always better precisely because only a tiny percentage of submissions have value, so the only way to get a meaningful number of useful submissions is to have a huge number of total submissions. But “[a] rulemaking agency . . . cannot routinely plan to read 100,000 comments to find 100 that offer some value to the rulemaking. At least until advances in natural language processing research yield nuanced and reliable methods of automated topic categorization, summarization, and content analysis of comments, ‘more’ per se cannot sensibly be the goal of participation system designers.”85

Sixth, social media culture is quite at odds with fundamental characteristics of notice-and-comment rulemaking. Producing useful comments is hard; it requires time, thought, study of the agency proposal and rationale, articulating reasons rather than stating preferences, and constructive engagement.86 In contrast, submitting a blog comment, “liking” a page or photo, rating a movie or book or restaurant, and similar online activities involve virtually effortless, subjective, minimalist, off-the-top-of-one’s-head assertions of a bottom line. Web pages are designed to minimize thought and effort. Farina et al. note that web users tend to scan pages, click on the first available option, and spurn instructions:

Significantly, usability experts study these behaviors in order to design for them, not to change them. The cardinal rule of Web design is “Make it easy”—a principle memorably captured by usability expert Steve Krug in the title of his popular book, Don’t Make Me Think: A Common Sense Approach to Web Usability. Hence, Internet users are now accustomed to websites designed specifically to allow them to engage rapidly and with little effort—the antithesis of the kind of engagement needed for effective rulemaking participation.87

Seventh, and related, a significant piece of social media involves voting of one sort or another. As many have pointed out, rulemaking is not supposed to be a referendum. Indeed, it is rather remarkable how firmly entrenched that understanding of rulemaking is. When e-Rulemaking got underway, a number of people speculated that one consequence would be that rulemaking would become more of a plebiscite, that the technology would push our understanding of the nature of the process. That simply has not happened. That is partly because the deluge of “votes” largely has not occurred, but it also reflects a very firm consensus about the nature of rulemaking. Thus, what social media do best is what rulemaking needs least.

2. Costs

Efforts to engage the public through social media are not costless. First, as just discussed, there are direct costs in equipment and personnel. Handling large volumes of comments over is hard enough. As Bridget Dooling explained in her report on e-Rulemaking, trying to read every comment

in mass comment scenarios forces agencies to sink considerable staff resources into reading or at least skimming comments that are word-for-word identical. For example, if an agency takes this approach with a docket that contains 250,000 comments from an organized mail campaign, even if it takes less than ten seconds to identify and skim each comment, that effort still accounts for almost 700 staff hours or $21,000. This excludes any time needed to summarize the comments for use internally or for the preamble of the final rule. The voluminous influx of comments can drive some agencies to turn to contractors, either to help organize and save public comments in the docket, or to actually review and summarize those comments.88

Those costs will only rise if through social media (a) participation increases and (b) the agency is to engage in actual dialogue, or if it is to moderate or facilitate public postings.

In general, the greatest enthusiasm for using electronic tools in rulemaking has come from academics, followed by public interest or good government organizations, then the White House; the agencies themselves seem most dubious. There are individual counterexamples, of course. But there reside significant doubts about the enterprise among agency staff. Jeff Lubbers’ e-Rulemaking survey revealed this, indicating that agency employees tended to think that the benefits of e-rulemaking flowed to commenters and the burdens fell on them.89 And part of the reason that agencies have flocked to social media for PR and communications and not for rulemaking is that the same people answering the Lubbers survey are nervous that social media will just put them in a deeper hole. They are skeptical about the value of lay comments, and they are very nervous about the extra work involved in reading, moderating, screening, responding to submissions, and they fear the whole thing will be chaotic, off-topic, repetitive, and go way beyond the point of diminishing returns.

A second category of potential costs should not be overlooked. That is the potential backlash resulting from disappointed expectations when promises about the meaningfulness of participation are disappointed. It is not wholly accurate, for example, for to bill itself as “Your Voice in Federal Decision-Making.” While it is impossible to know whether that enticement has led to disappointment, one can see such annoyance, for example, on the White House “We the People” site. This is an open-ended call for suggestions, with the promise that any petition with more than 100,000 signatures (originally 5,000, then 25,000) by a specific deadline will get an official response. As of February 2013, there were 282 pending petitions; 98 petitions had generated responses (it’s not possible to determine how many expired petitions fell short of the signature threshold). In the fall of 2011, the most popular petition was: “Actually take these petitions seriously instead of just using them as an excuse to pretend you are listening.”90 A more recent petition in the same vein requests the White House to “admit that these petitions are just going to be ignored.”91

The White House is working hard to assure people that their individual voices truly are heard and influential. On its “engage and connect” page, for example, one can find a video entitled “Your Voice Matters.” The video shows White House staff, the First Lady, and even the President speaking on the phone to, and even visiting in the homes of, ordinary folks who expressed an opinion. This is followed by a series of individual testimonials: “I really feel that I’m being heard”; “It has been very heartening for me to see how effective I can be”; “It’s clear that your story will be read and will be considered and can have a big impact on the White House and the President and people who are making decisions. It’s worth it.” The final tag line is “Everyone has a part to play.” This particular video’s implicit promises about access and influence are so astounding and unrealistic as to make it a work of fiction and it is hard to believe anyone would take them at face value. However, the general point remains; overpromising can lead to disappointment, backlash, and disengagement.

Recommendation: Before using social media in connection with a particular rulemaking, agencies should carefully consider the potential costs and benefits and identify with specificity what they expect to achieve through the use of social media.

Recommendation: The use of social media may not be appropriate and productive in all rulemakings. When deciding whether to use social media in a particular rulemaking, agencies should keep in mind the following principles:

1. Rulemakings that primarily involve questions of statutory interpretation, technical knowledge, or scientific expertise may be poorly suited to the kinds of responses usually produced by social media.
2. On the other hand, social media may be valuable when an agency seeks to ascertain the perceptions or reactions of regulated parties or the public to the proposed rule.

Recommendation: An agency should use the social media tool(s) that best fit its particular purposes and goals and should carefully consider how to effectively integrate those tools into the rulemaking process it would otherwise use.


A decade into the flowering of e-Rulemaking, four years since the Open Government Memorandum, in the thick of the Web 2.0 explosion, the role(s) of social media in rulemaking remains uncertain. There has yet to be a truly successful demonstration of how social media will or might meaningfully enhance the notice-and-comment process. To date, participation levels have been low and dialogue virtually nonexistent. Sophisticated participants have shunned social media, sticking to traditional notice-and-comment. The quality of contributions from lay participants has not been high, and the impact of their contributions difficult to perceive. On the other hand, the theory is enticing, and the possibility that new technologies could engage stakeholders who have heretofore been on the sidelines, tap into the dispersed knowledge of the public, bring new voices to the table, and democratize the process remains worth pursuing.

However, the preceding discussion, and experience to date, counsel against jumping in with both feet. It would be appropriate and important for agencies to experiment integrating new technologies into the rulemaking process, but this should be done with some caution. To quote Cynthia Farina and Mary Newhart:

Rulemaking 2.0 is very much a work in progress. In our view, it is still an open question whether the time and effort required to achieve meaningful new participation represent the best investment the agency can make in improving its rulemaking processes. This question won’t be answered by the outcomes in one or two rulemakings.92

Before detailing some specific undertakings, a caveat and a clarification are in order. The caveat is that not all rulemakings lend themselves equally to the use of social media. “[T]he local sanitation engineer for the City of Milwaukee” may well have useful insights about the implementation of drinking water standards93; she is less likely to have something useful to say about the adverse health effects of different levels of arsenic exposures.

The clarification is that the question of the value of social media in rulemaking is not the same as the value of lay participation in rulemaking. The two certainly overlap; one goal—indeed, the most visible and salient goal—of social media would be to increase the amount and quality of participation by unsophisticated stakeholders and members of the general public. But that is not the only goal. One might be dubious about the value added by broad public participation but still turn to social media either to increase participation by certain small, targeted groups or to improve the quality of the participation that is already occurring.

A. Outreach

As discussed in Part IV.C, above, existing agency uses of social media have been used almost exclusively for purposes of outreach: providing general information, “telling the agency’s story,” or letting people know about available services, benefits, or employment opportunities. Such uses include informing the public about the rulemaking process generally and individual rulemakings in particular.

1. Examples of social media outreach

Many agencies provide useful information about their rulemakings on their websites. DOT’s regulations page94 and EPA’s “Rulemaking Gateway”95 are prominent examples. But these sites are not really examples of social media nor venues for discussing those rulemakings.96 Agencies have also started to use social media to get the word out about rulemakings. A good example is the Small Business Administration, which uses its blog, “The Small Business Watchdog,” to post announcements of NPRMs of interest to its readership, including those of other agencies.97 It is self-evident that any time an agency issues an ANPR an NPRM, or a final rule, it should notify the public not simply by placing the notice in the Federal Register and posting it to, but also make an announcement, with a link, via its website, blog, Twitter feed, and Facebook account.

Agencies have begun to go beyond a simple notification, for example, by producing videos about particular rulemakings. The actual information in the video could just as easily have been put in a written document, but the hope is that the video format will grab the attention of more people, presenting the same information in a more compelling and salient way. For example, EPA recently produced a lengthy video tied to a planned high-stakes rulemaking. The agency has issued two notices of proposed rulemaking for a regulation that would establish a New Source Performance Standard under the Clean Air Act to limit emissions of greenhouse gasses from fossil-fuel fired power plants. Under the relevant statutory provision, some regulation of existing facilities is also permissible once such a standard is in place. EPA is in the preliminary stages of thinking about what such regulation would look like; because there are many existing coal-fired power plants but no planned ones, the stakes are arguably higher with regard to producing these guidelines than with regard to the regulations applicable to new plants. And not surprisingly it is proceeding with some care. In the summer of 2013 it produced a 33-minute video laying out background information about the Clean Air Act, greenhouse gasses, and electricity generation98 ostensibly aimed at stakeholders with whom it will meet and discuss just how to proceed.

A related tool that some agencies have experimented with is hosting a webinar keyed to the release of a NPRM. For example, in 2009 the Department of Education sought comments on its proposed priorities and criteria for funding innovative educational programs. It used its blog, “Homeroom,” to announce and encourage public comment on the proposal. 99 (It did not permitted web site users to post comments by commenting on the blog entry. Rather, it disabled the commenting function for the post relevant to the rulemaking and instructed the public to comment via or go the traditional off-line route.) Simultaneously with the issuance of the notice, the agency conducted a 37-minute webinar discussing the scope and nature of the proposal. Over 600 diverse stakeholders “attended.” itself has a growing social media presence. It has its own Twitter feed; followers receive a tweet about every new docket created on the site. (Arguably, such feed is too broad and diverse to be useful. also provides for agency-specific RSS feeds, which are more valuable.) As of September 2013, had tweeted over 13,000 tweets and the feed had about 2800 followers. Since October 2011, has also had a Facebook page. The page was intended to serve mix of functions: drawing attention to, teaching about the regulatory process , flagging particular rulemakings, and discussing the regulatory process in general.100 The site has generated just over 3,000 “likes.” The site was fairly active in 2012; however, the last post (as of mid-September 2013) was in September 2012. A scan of the posts and the public comments thereon suggests that the site is not actually that useful; certainly the public comments are confused, diffuse, and unhelpful.

The other crucial government-wide rulemaking site,, is similarly engaged with the primary social media. It maintains a blog, posts some videos and has a YouTube channel, maintains a Twitter feed, and has an elegant subscription tool for learning (by email or RSS feed) about material from particular agencies.

Both and also ties in to social media by allowing visitors to share a link to the page they are viewing via Twitter, Facebook, or email.

2. Regulation Room

A useful example of outreach efforts via social media is provided by “Regulation Room.” Housed at the Cornell E-Rulemaking Initiative and led by Professor Cynthia Farina of the Cornell Law School, Regulation Room is a website that uses Web 2.0 approaches and tools to facilitate public discussion and feedback in connection with federal agency rulemakings.101 The site is conceived and operated by researchers from computing and information science, communications, conflict resolution, law, and psychology. Its basic goals are to improve the amount and quality of public participation in rulemaking.

To date, Regulation Room has participated in five rulemakings—four with the Department of Transportation and one with the Consumer Financial Protection Bureau. In each instance, most of the participants (and in two cases virtually all of the participants) had no prior experience with rulemaking. The number of comments has ranged from 32 to 931; the number of registered users from 53 to 1189.

The Regulation Room team identifies three barriers to fuller public participation.102 The first is ignorance. Citizens know very little about agencies and next to nothing about the role, nature, or importance of rulemaking. Second, unawareness—even individuals who have a more sophisticated understanding of rulemaking will often just not know about a particular rulemaking even though it is of direct relevance to them. The third barrier is information overload from the sheer mass and technical complexity of the materials found in rulemaking dockets and the preambles to proposed rules.103 The project has made strenuous efforts to overcome these obstacles. The first two are addressed through extensive outreach, primarily but not exclusively through Twitter, Facebook, and other social media. These efforts have increased participation somewhat, though the absolute numbers remain modest, hovering in the hundreds rather than the thousands. Note that outreach is in some respects only a small part of the issue; while it is of course a challenge to get people to come to the site, it is also a challenge to get those who do come to stay and participate. Serious participation is burdensome and challenging; most people will not do it. Thus, in Regulation Room’s most recent rulemaking, which involved CFPB regulation of mandatory disclosures to consumer borrowers, 8908 unique visitors came to the site, making 12,665 total visits. However, the average visit was only about 3 minutes long, and of the almost 9000 visitors only 144 registered and of those only 67 posted a comment.104

On the basis of its experience, the Regulation Room team offers a set of useful suggestions regarding outreach to potential rulemaking participants. Agencies should:

* Develop a communications plan specifically tailored to the rule and to the types of missing stakeholders or other potential new participants the agency is trying to engage.
* Not underestimate the power of conventional media or overestimate the ease or likely impact of using social media.
* In outreach messages, give information about participation as much emphasis as information about substance, emphasizing that the agency wants their participation and will take it into account.
* To motivate action, be clear and specific about how the proposed rule would affect the targeted participants, positively or negatively.
* When asking organizations to pass on the participation message to members or followers, recognize that they may need persuasion about why such individual participation will not hurt organizational interests.105

3. The Unified Agenda

In the pre-Internet era, the important innovation in increasing awareness of agency rulemaking was the Unified Regulatory Agenda. President Reagan’s E.O. 12,291 required “[e]ach agency” to publish its regulatory agenda in April and October of each year. This requirement was partially codified by the Regulatory Flexibility Act.106 President Clinton’s E.O. 12,866, which remains in effect, elaborated on these requirements by establishing the current version of the Unified Regulatory Agenda. The Agenda is published twice a year, in April and October, and provides uniform reporting of regulatory and deregulatory activities under development throughout the federal government. Agencies identify all ANPRs, NPRMs, and final rules expected to be issued in the next year, and may also list “Long-Term Actions.” As the ABA’s Section of Administrative Law noted back in 2000, the Agenda “provides important information to agency heads, centralized reviewers, and the public at large, thereby serving the values of open government.”107

Since 2007, the agenda has been posted to the Internet rather than distributed in printed form.108 The agenda is officially compiled by the GSA’s Regulatory Information Service Center on the basis of agency submissions, but it is OIRA that oversees the process. The agenda is accessed through the website, which maintained by GSA but is about OIRA. With the move on-line, the justification for only updating the Agenda twice a year largely evaporates. The Agenda would be more valuable, timely, and useful if it was kept “ever green,” updated by each agency as soon as it decides to proceed with a relevant action.109

In addition, the Unified Agenda can provide a trigger for agency notification regarding individual rulemakings. That is, a natural moment for a Tweet, RSS notification, Facebook posting, or similar outreach would be when a status item for a rulemaking is first included, or later modified, in the Unified Agenda.

4. Beyond the Usual Suspects

All of the foregoing efforts are valuable. They do, however, reflect a basic paradox: the people who are easiest to reach are those who least require reaching. For example, someone who goes to an agency website’s rulemaking page or to, or who knows about the unified agenda, or who subscribes to an RSS feed, already is relatively well-informed. The greater challenge is reaching people who are simply unaware of the rulemaking process generally.

Consider one effort to reach that group which had rather mixed results. In 2010, the White House held a competition to produce a short video about rulemaking. The idea was to educate the public about rulemaking generally (i.e., to respond to the first of the problems identified by the Cornell team). This effort, called “Rulemaking Matters!,” was a classic example of contemporary principles: (a) using a competition, which is a form of crowdsourcing, (b) to produce a video rather than a text, (c) which would be made available through multiple on-line platforms, in order to (d) make government more transparent and participatory.110 Several of the submissions, including the winning video,111 can be found on Youtube. It is difficult to quantify the value of such an undertaking. However, as of October 2013, the winning video has a mere 686 views on Youtube and does not seem to be up on at all (though it is posted to the YouTube channel). Other entrants also have Youtube views in the hundreds. The messages are broad, abstract, and unfocused; the basic pitch of all of them is that regulations are really important and that provides a channel for public input. None is especially helpful or impressive. Between their so-so quality and their low viewership, these cannot be considered a particular success.

Interestingly, ReasonTV, a conservative YouTube channel with a virulent anti-regulatory stance, has posted two modified versions of EPA’s own video announcing the contest. One includes subtitles which criticize and subvert the ostensible message of EPA’s video112; the other changes the speakers’ voices into robots.113 Each has been viewed tens of thousands of times. These attack videos are in fact classic Web 2.0 projects, involving creative adaptation or appropriation, citizen initiative, and user-produced content. These particular efforts do not, of course, advance the rulemaking project; to the contrary, they undermine it. But they actually stand as an example of the possibility that engaged and creative individuals might build on the agency’s own efforts to highlight or publicize a rulemaking—and as a reminder that when that happens, the agency no longer controls the message.

The gap between the popularity of the Rulemaking Matters! submissions and the ReasonTV videos illustrates one other important point regarding outreach. Highly visible content becomes highly visible because it goes viral—that is, individuals who see it share it with others and so total views grow exponentially. Simply being available to be found online, or mentioned in one Twitter feed that has a few thousand followers, is a necessary but not sufficient step to reach a large number of people. Material only really becomes visible when it develops a following and is widely shared. For something to catch someone’s eye in the first place, and then generate a sufficiently enthusiastic response to prompt numerous viewers to share it, it has to be really interesting, funny, or edgy.114 Unfortunately, material that is “official” and produced by the government, and/or material pertaining to agency rulemaking will almost certainly lack some or all of those features. Thus, it is very hard for the government to compete for attention with privately produced material.115

Recommendation: Agencies should use social media to inform the public about agency activities, the rulemaking process in general, and specific rulemakings. Agencies should take an all-of-the-above approach to alerting potential participants to upcoming rulemakings, posting to its website and blog and sending notifications through multiple channels. Social media provide a more effective means to reach interested persons that have traditionally been under-represented in the rulemaking process.

Recommendation: Agency rulemaking notifications should be directed at both individuals and organizations, be clear and specific about how the proposed rule could affect targeted participants, and include details about the mechanisms and value of public participation.

Recommendation: At a minimum, agency notifications regarding rulemakings should be tied to inclusion in the Unified Regulatory Agenda; the addition or change of any item in the Unified Regulatory Agenda should trigger notification via social media.

Recommendation: The Regulatory Agenda should be updated continuously, or at least monthly.

B. Gathering Open-Ended Public Input through Social Media

Rulemaking” involves much more than “notice and comment.” Social media have much, perhaps most, to offer not during the actual comment period, but prior to issuance of the NPRM and after promulgation of the final rule. This section will review various of these efforts and consider their application to rulemaking during the initial rule-development, prior to the official comment period. This is a period in which the agency’s scope of inquiry is extremely broad, the questions more open-ended, and the interchange less formal.

Negotiated rulemaking (“re neg”) provides a ready analogy. Under the Negotiated Rulemaking Act, the entire reg neg process is a mechanism for developing a proposed rule. The proposed rule is then published in the Federal Register and the ordinary notice-and-comment process takes place. Use of social media differs from regulatory negotiation in important respects. There are, ideally, many more participants and the idea is not all to reach a mutually acceptable compromise. But the two share important elements. Both open up the traditional rulemaking process, create a more dialogic exchange, and have a slightly awkward fit with the traditional process. Agencies resolved that awkwardness for reg neg by having the whole process take place before the NPRM, and Congress took the same tack.116 Similarly, agencies would be unconstrained by rulemaking requirements when gathering input via social media prior to the NPRM.

Indeed, social media might be especially useful and appropriate in this setting. Social media provide an avenue for lay participation in the process (and, again, that is not the sole point). As a generalization, it is probably fair to say that the lay public is better at identifying problems than at identifying solutions. Such input is especially relevant at the early stages of the rulemaking process, when the agency needs to understand the existing state of affairs, what’s working and what isn’t, where improvements must be made, and so on: in short, what’s the problem? Furthermore, for all interested persons, a looser, more dialogic exchange may be especially useful in at an early, problem-identifying stage.

In this setting the tools highlighted in the previous section—Facebook, Twitter, videos, and blogs—are not so central (though blogs may still be valuable). These tools are very good for one-way communication; they allow the agency to communicate with outsiders. They are much less valuable as a platform for gathering responses. To be sure, agency postings do generate comments. A basic feature of Facebook, for example, is that viewers can comment on anything that is posted. The page owner can turn off comments, but agencies generally have not done so, and the folks at Facebook encourage them not to.117 Accordingly, almost every post prompts a handful of both “likes” and comments. Facebook does not have a “don’t like” option, so the number of “likes” that a post accumulates is only the roughest sort of guide to its actual popularity. Comments are a mixed bag, many are brief expressions of support118 or derision,119 some wholly off-topic,120 some substantive. By and large, however, they are not especially helpful and are ignored by the agency. Similarly, Youtube allows for comments, but the reality is that government Youtube videos tend not to accumulate that many comments, the comments they get tend toward the short, irrelevant, and obnoxious, and (as a result) the comments that are made seem to go unread. Indeed, TSA has simply disabled the comments feature for most or all of its videos.121

It is not uncommon for agencies to tweet general requests for the public to submit ideas. However, these are pretty unsuccessful. To pick a random example, on February 1, 2013, EPA tweeted: “It’s time for #EPAtips again! What are some unexpected ways you’ve found to save energy this winter?” Responses could be tweeted or posted on Facebook. The same day, it tried again: “Tell us some unexpected ways you’ve found to save energy this winter. Use hashtag #EPAtips. We’ll retweet our favorites.” The next day: “Last chance to share your #EPAtips with us! What are some unexpected ways you’ve found to save energy this winter?” And then two days later: “Thanks to everyone who shared their #EPAtips with us!” The exuberant (or desperate) exclamation marks notwithstanding, it appears that not a single “unexpected way to save energy” was submitted.

In short, these tools can be important in alerting members of the public to opportunities to submit information, ideas, or comments. But they will not themselves be the channel for doing so. Instead, agencies might establish some sort of mechanism or platform through which members of the public can make comments or suggestions and engage in a discussion with each other and the agency. Agencies have made some attempts to gather public input regarding policy questions through social media.122 Some examples follow.

1. U.S. Open Government National Action Plan and E-mail Comments

The most bare-bones, least “social media-y” mechanism for receiving public input is email. This is essentially no different than using; it is a Web 1.0 tool. Consider the following example.

The United States has spearheaded an international collaboration called the Open Government Partnership (“OGP”). Each participating country is to establish a National Action Plan that lays out the open government initiatives it plans to undertake. The U.S. Plan was issued in September 2011 and “laid out 26 concrete steps the United States would take to promote public participation in government, increase transparency, and manage public resources more effectively.”123 Among these, interestingly, are measures to increase public participation in the development of regulations—not by relying on social media but rather by improving In March 2013, the White House announced it was beginning work on a second National Action Plan.125 In preparing both the first and the second Action Plan, the White House has, not surprisingly, sought public input. What is notable is that in doing so the people in the administration among those ostensibly most committed to public participation and open government chose the mode of comment that is the most basic and least visible, dialogic, or open: email. In August 2011, OIRA Administrator Cass Sunstein and Chief Technology Officer Aneesh Chopra solicited input on the National Action Plan, asking three questions about improving,, and the federal web policy in particular. Readers were asked to submit their thoughts by email.126 That approach was repeated two years later with regard to the second National Action Plan.127 Gathering comments by email is less transparent than using, since no one other than the government official sees the comments. (Chopra and Sunstein said they would post a summary of comments; I have not been able to find one.) The choice of this relatively traditional and hidden tool for public input, which did not go unnoticed in the blogosphere,128 suggests that at least some in the White House are themselves unconvinced of the value of social media as a tool for public participation in policymaking.

2. Exchange

Until recently, maintained an open discussion site—officially a “partner site,” distinct from itself—called Exchange. The idea behind Exchange was to establish a general discussion forum for feedback either about itself or about particular agency initiatives. Agencies were invited to seek public input on a planned rulemaking before it begins. Very few did so. EPA sought feedback regarding two changes to the Toxic Release Inventory129 and with regard to a planned rulemaking to require electronic reporting of discharge monitoring results from holders of Clean Water Act discharge permits.130 And that’s about it. The site was “retired” in 2013.131

With regard to the handful of instances in which agencies attempted to use Exchange, the results were disappointing. For example, with regard to the Clean Water Act rulemaking, EPA asked seven specific questions.132 Most of these had about 3,000 views; one, regarding Internet access, had 25,000. The Internet question prompted 15 comments, all of which were simply very general expressions of enthusiasm for the Internet; the other six questions produce between zero and two responses each, none of any real value.133 There is no dialogue among commenters and no response from the agency. Not surprisingly, this effort was seen as a failure within the agency.134

3. Blogs

As mentioned above, agency blogs, while ubiquitous, are used primarily to distribute information about the agency. That is not inherent in the platform. In principle, a blog could support the sort of back-and-forth dialogue among readers that email submissions cannot. To date, however, they have not been useful tools for the sort of dialogue and input that might be transferable to the rulemaking setting.

For example, reading the TSA blog is discouraging. The posts from TSA vary between the quasi-promotional and the informational. All in all, the agency is attempting to: (a) get the public on its side; and (b) make the screening process go more smoothly by providing information about what cannot be brought onto a plane and other travel tips. The comments—and bear in mind, more than a quarter of comments never see the light of day because they violate the TSA comment policy—are primarily furious attacks on the TSA’s competence, integrity, and value, leavened by the occasional “atta boy” (which the attackers assume are posted by TSA employees trying to make it look as if there is more support for their efforts than there is). It is hard to believe that anything of value gets communicated to the agency through this medium. The official responses from the agency are polite and restrained, but it appears that individual TSA employees occasionally voice their frustration with the attacks in the comments. For example, TSORon wrote on 1/31/13: “Very few actually understand what security is or what it takes to achieve it. And then there are those who don’t want to know, and are very happy in their ignorance. That alone accounts for the vast majority of the posters here.”135 Which, of course, only prompted aggravated, nonsubstantive responses.136

In short, the TSA Blog reads a good deal like the comments section of non-government blogs, which is discouraging. However, TSA is an agency that is uniquely visible to and uniquely and directly adversarial with members of the public. Not every agency blog will suffer from the same pathologies. Agencies have been somewhat more successful in using blogs to solicit comments and reactions with regard to specific proposals or problems. For example, in developing its National Broadband Plan, the FCC established a dedicated blog, The Broadband Blog, also known as “Blogband.”137 The blog was designated an official part of the record for the proceeding.138 The Commission issued an official statement making that designation and admonishing that “interested persons are advised to review not only ECFS, but also Blogband to ensure that they are aware of all relevant views expressed to the Commission concerning the National Broadband Plan.”139 Despite some fanfare, comments were neither especially numerous nor especially substantive.

The CFPB’s blog140 got underway in February 2011; the agency posts on average every 2 or 3 days. The page states that “The CFPB blog aims to facilitate conversations about our work. We want your comments to drive this conversation.” That has not quite come to pass; many of the posts have generated a couple of comments, but it is clear that the comments are not driving anything and that no real “conversations” are taking place.

In short, the blogs have yet to become a really effective venue for discussion on rulemaking topics. It is hard to say exactly why this is. However, the same recurrent issues seem to be present: not enough people are reading the blogs or participating, they are not convinced that anyone is listening, those who do participate are in an angry rather than a collaborative frame of mind, the questions are too open-ended and unfocused, and (related to each of the foregoing) the agency itself is silent.

4. On-Line “Suggestion Boxes”

A similar effort is the general discussion page on the FCC’s web site,141 which hosts ongoing discussions on various topics. This promising idea is undermined, as is so often the case, by a lack of participation, and particularly informed participation. For example, the FCC asks: “How can we improve our APIs?” The top-ranked comment in response is “what is API?” It is one of only two comments, and it has received only ten votes (plus one comment, which answers the question). Similarly, the FCC asks: “How can the FCC better engage the public on rulemakings?” (a question of great interest to this writer, at least).142 That question has produced 25 suggestions, of which the most popular has 17 votes. Of the 25, only five are actually about the rulemaking process as opposed to this or that substantive concern. These five suggestions endorse Regulation Room, endorse the use of social media (the entirety of this comments says, “use social media outlet”), urge the FCC to create an index of rulemakings, and suggest alerting the world to possible regulations via social media before they have been proposed. None of those are foolish suggestions, but neither are they helpful or novel.

Many agencies now have general pages where members of the public can submit ideas or suggestions and vote on those submitted by others. These tend to be quite open-ended. Examples include the FCC’s “reboot” site,143 HUD’s “Ideas in Action” site,144 and the CFPB’s “tell your story” page145 and now-completed “Open for Suggestions” campaign,146 which also solicited public input through multiple outlets, including Twitter, e-mail, and YouTube videos. The CFPB received hundreds of suggestions and posted video responses to many on its YouTube channel.147 The Bureau also launched a blog and social media outposts on Twitter, Facebook, Flickr, and YouTube. These channels have been providing a steady stream of information from the public about problems with consumer financial products and suggestions for how to address those problems. The Bureau asserts that it is analyzing this information so that it can inform its priorities and policymaking,148 though concrete affects or results are hard to identify.

5. Ideation Sites

The most prominent real-world examples of using social media for policy input involve “ideation” tools. Sometimes referred to as brainstorming or social voting tools, these involve open calls for suggestions on a particular topic. Anyone can submit an idea or suggestion, the submissions are visible to all other visitors to the site, and other visitors comment on the suggestions vote thumbs up or thumbs down on them. The idea is that the most popular ideas, defined as those with the highest net number of thumbs up votes, will percolate to the top, and the least popular ideas will sink to the bottom, of the user-generated list. GSA has promoted these tools as mechanisms “that make it possible for agencies to engage with many more people and help analyze, absorb, and use the public’s ideas and suggestions.”149 GSA has also negotiated Terms of Service agreements that make agency adoption of one leading platform, IdeaScale, quite straightforward.

a. Examples

Most major agencies have tried IdeaScale. In particular, some three-dozen agencies, with prompting and support from the General Services Administration, used IdeaScale sites in developing their Open Government Plans.150 Several have used the software in other settings as well.151 Results have been mixed. A few examples follow.

The Federal Emergency Management Agency maintains an ongoing IdeaScale site.152 As of mid-October, 2013, almost 4000 users have posted 1347 ideas (541 are visible), made over 5000 comments, and cast 27,000 votes. More than most such sites, the FEMA site seems to have attracted a set of participants with relatively high levels of expertise. Many of the ideas attracted comments in the mid two figures, and the discussions are substantive. The least popular ideas were also voted against decisively, again suggesting the level of expertise among the users.

“EPA Conversations,” the most popular of EPA’s several IdeaScale sites, began in mid-2012 and generally invites ideas about environmental protection. Its tagline is “Look beneath the surface, Address the Issues, Expand the Conversation.”153 A Youtube video (with very high production values and 72,000 views) sends viewers to the site. As of mid-October 2013, it has 475 users who have come up with 147 ideas. The most popular idea boasts 131 votes, while the least popular has garnered a net vote total of -16. (This least-popular idea was: “Global warming is NOT happening.”) Only 27 ideas have more than 10 votes. Most of the ideas are constructive, but none of them convey information that would be news to environmental professionals. The number one idea, for example, is to ban single-use plastic bags. This approach has been part of the solid waste conversation for decades, with various jurisdictions having adopted such measures and others having explicitly rejected them. Novel ideas might be valuable; old ideas that showed massive support might be valuable. But 121 “votes” in favor of banning plastic bags does not advance the conversation. To date, EPA itself has not been a visible presence on the site; it has not responded to or acted on any of the posts.154

The FCC has a wide variety of ideation sites.155 Among the most interesting and highly touted is the site it hosted in developing its national broadband plan.156 The plan is not a regulation, so its development was not required to proceed through rulemaking. The FCC nonetheless elected to solicit public input during the plan development process. In particular, the FCC established a docket and took public comments, in addition to holding a number of legislative-type public hearings. The FCC’s approach thus provides an example of the use of a separate IdeaScale site to supplement an otherwise standard notice-and-comment process. During the two years or so the IdeaScale site was open, it drew 279 ideas, 536 comments, and 4,685 votes from 934 registered users. IdeaScale touts this project as a success in its own promotional efforts.157 As with EPA Conversations, the suggestions the site generated were fairly vague and generic. The two most popular were “Bring the United States mobile broadband pricing in line with the rest of the world” (131 votes) and “Net Neutrality is Vitally needed, even in Cities” (117 votes).158 As with the EPA site, these propositions may be true, but they are neither novel nor helpful, the propositions are utterly familiar to those inside the agency, and the fact that 100+ people endorsed them does not really mean much.

Interestingly, the Commission treated the IdeaScale site and a dedicated blog as of equal status with its “official” Electronic Comment Filing System (ECFS). It took all the ideas and comments and placed them in the official docket for the broadband plan proceeding.159 That does not mean that the ideas actually affected the substance of the plan. The “Civic Engagement” section of the plan (Chapter 15) does mention the existence of the IdeaScale site and notes that many citizens had given ideas, votes, and comments. It does not indicate that those ideas, votes, and comments were incorporated into the plan and does not specifically discuss any of them.160 Similarly, IdeaScale boasts about the FCC site in its promotional materials, asserting that “those who participated in the public dialogue directly affected changes that were implemented by the FCC in the National Broadband Plan.” But it gives no concrete examples to back up this assertion.161

b. Assessment

Ideation sites have an enormous intuitive appeal. The wide-open request for ideas might generate a brilliant suggestion no one has ever thought of, and the opportunity to vote (a) saves the agency the time and trouble of reading duplicative comments, (b) saves commentators the time and trouble of writing such comments, and (c) taps in to the wisdom of crowds. For these reasons, and despite the lackluster performance of the real-world examples, ideation sites have a role to play at the rule development stage. However, as the above examples show, there is nothing magic about this approach. In particular, low participation (which undermines so many of these sorts of undertakings) can doom the effort from the get go. In addition, if the participants do not know anything that those in the agency do not know, the exercise is pointless. Third, there is some tension between the technocratic, expertise-based, legally constrained nature of most rulemaking and the referendum-like nature of ranking submissions by popularity. Fourth, the most popular submissions may be such not because they are the highest quality or even genuinely most accurately reflect the independently held views of those who are voting. It may just be that voters are swayed by the identity of the submitter, or the very fact that a submission seems to be possible, or simply the fact that the comment was the first they read and seemed basically right. Fifth, someone might heartily approve of one part of a comment and dislike another (though that is a problem with a technological solution, allowing voting on the whole comment or just parts).

Finally, the process is potentially subject to easy manipulation through orchestrated campaigns and get out the vote drives. There is no much evidence of such efforts to date, though isolated examples exist.162 But if in fact agencies were known to be taking rankings seriously, such efforts would surely follow. Suppose an NGO submits a comment and then urges all its members to go online and vote for it. Now the most popular comment has 200,000 more votes than the second most popular. At that point, the purpose of ranking comments has been defeated. Moreover, there is nothing the agency can do to prevent it—any bar on such orchestrated endorsements would likely violate the First Amendment and would in any event be unenforceable.

It is worth noting that several agencies have set up internal ideation sites to solicit suggestions for improving agency operations from staffmembers. These uses are analyzed and endorsed in a 2009 report from the White House Innovation and Information Policy Group.163 The report identified six “lessons learned” about what has to be done for an ideation site to be successful:

* Senior and mid-level leadership support and participation are essential.
* Significant human capital is needed to successfully manage an idea generation program.
* Long-term success is dependent on acknowledgement of the innovators, i.e. there must be a strong “reward and recognition component.”
* A good communications and growth strategy is essential for the roll-out & continued community engagement (support has to be built up—it doesn’t happen the day of the launch).
* A key message we heard from nearly all the program managers we interviewed is that a “build it and they will come” philosophy does not apply to idea generation tools. The importance of developing a culture in which idea generation is promoted, celebrated, and rewarded cannot be understated.
* Site moderation and rules of engagement are critical.164

These lessons are applicable, mutatis mutandis, to public ideation sites. Notably, each is more difficult to accomplish when the participants are not all employees of the same agency but are instead the broad and diffuse public. These sites will not success without significant hands-on involvement by the agency; they are not machines that run of themselves.

Recommendation: Agencies should consider using social media during pre-rulemaking or policy development proceedings where the goal is to understand the current state of affairs, collect dispersed knowledge, or identify problems. To enhance the number and value of public contributions, an agency seeking to engage the public for these purposes should, to the maximum extent possible, make clear the sort of information it is seeking, clarify the role of public suggestions, and directly engage with participants by acknowledging submissions, asking follow-up questions, providing substantive responses, and giving credit and recognition to the most valuable submissions.

C. Education for Effective Commenting

As discussed in Part II, the quality of comments submitted by the lay public is generally poor. One response to these shortcomings has been efforts to tutor lay commentators so as to increase the quality of their submissions. For example, the Department of Transportation has on its website a description of the rulemaking process that includes a section entitled “How Do I Prepare Effective Comments?,” which lays out exactly the sort of things that are helpful to the agency and generally absent from public comments.165 There is no indication that these have done much good.

The most concerted effort to guide and tutor lay commenters has been Regulation Room. The site provides materials about the rulemaking process and the nature of comments.166 Concrete advice, similar to what can be found at and some individual agency websites, about how to write effective comments is also provided. In addition, moderators will flag comments that are of especially high quality, identifying them as examples of the sorts of characteristics that make a comment effective.

All of this is useful, but it only goes so far and it does not take advantage of the full potential of social media. Agencies, the GSA, and/or the e-Rulemaking Program should consider more robust educational programs, including such possibilities as producing videos about the commenting process, hosting webinars with actual agency personnel reviewing actual examples of useful and unhelpful comments, maintaining an on-line database of exemplary (good and bad) comments, conducting an on-line “master class” in which officials go over and seek to improve a draft comment with its author. None of these would be transformative, and the audience will be limited. But a list of tips is abstract and can be difficult to apply; the real learning would come from a more hands-on, multi-media instruction that directly engages with actual comments.

Recommendation: The General Services Administration, the e-Rulemaking Program Management Office, and other federal agencies, should consider using social media to create and distribute educational programs about rulemaking. These efforts could include: producing videos about the commenting process and posting on an agency website or video-sharing website; hosting webinars in which agency personnel discuss how to draft useful and helpful comments; maintaining an online database of exemplary rulemaking comments; or conducting an online class in which officials review a draft comment and suggest ways to improve it.

D. Obtaining Informed and Useful Comments from Diverse Stakeholders

Perhaps the consummation most devoutly to be wished would be that the use of social media would enable agencies to obtain better comments on proposed rules. “Better” might mean many things, but the potential of social media is (a) to reach a broader range of potential participants, (b) to facilitate the submission of more useful and informative comments from those participants, and, related, (c) provide a forum in which there can be an actual discussion or dialogue among interested persons. Tentative experiments in this direction have occurred, and there is great potential for more.

One technical development is relevant to the discussion below in several places, so it is worth noting here. As of this writing, the eRulemaking Program has developed, and is in the process of rolling out, a new “commenting API.”167 The API allows third parties—which could be individual federal agencies or private organizations—feed comments directly into the Federal Docket Management System from their own websites. As of this writing, the API is not yet up and running, but the eRulemaking program has provided it to several agencies and expects that they will soon have commenting widgets on their websites that will utilize the API.

1. Blogs

From 2009 to 2012, the U.S. Forest Service developed a new rule establishing an overall planning process governing the management of national forests. The process was standard in many respects, beginning with a Notice of Intent (NOI), which produced a set of comments, followed by a Notice of Proposed Rulemaking,168 which produced another set of comments, leading to a final rule.169 The Service, however, said that the rule “was developed through the most collaborative rulemaking effort in Agency history.”170 After issuing the NOI, the Service held two-dozen “regional roundtables” to discuss the rule.171 During the comment period, the Forest Service held some 75 public forums, spread through all the forest service regions.172 Although the number of public meetings was unusually high, these aspects of the process are not unusual. More unique was that the Forest Service established at the outset a blog dedicated to the rulemaking, the Planning Rule Blog.173 The blog began in 2009 with release of the NOI and concluded in 2012 with publication of the final rule. It was modestly active, with just a few dozen posts. These posts produced approximately 300 comments from readers, most of which did not generate responses from other commenters or the Service. Note that the actual rulemaking produced 300,000 comments, a thousand times more than the blog. Thus, the blog can hardly be seen as a central aspect of the overall process.

Notably, the Forest Service kept the blog and its comments carefully separate from the official notice-and-comment process. At the top of the blog home page there appears this notice: “Welcome to the Planning Rule blog! While we hope you will engage with us on the Planning Rule Blog, please remember these are not official comments.” The message is repeated and expanded at the bottom of the page: “NOTE: Blog posts do not constitute formal comments. Comments submitted to this blog do not constitute formal comments such as those submitted during Federal Register comment periods. Official formal comments must be submitted during formal comment periods.”174 This approach is consistent with ACUS Recommendation 2011-8: “When an agency sponsors a social media discussion of a rulemaking, it should provide clear notice as to whether and how it will use the discussion in the rulemaking proceeding.”175 The development of the commenting API simplifies this separation while also making it less problematic. It should be straightforward for an agency to attach the commenting widget to the blog, so if a blog reader wishes to make an official comment there is no need to leave the blog and go to; with one click, the comment form appears.

Although the Forest Service rulemaking did not go very far down the path to conducting the rulemaking via blog, it points in that direction. A standard blog presentation could be effectively used for generating and receiving public comment. The format and presentation are familiar; discussions will be threaded, which makes them easier to follow and review; it is convenient to have the key material (a portion of a preamble, for example) at the top of the page for ready reference while scrolling through comments below.

The format is still less than ideal, however. Farina and Newhart explain:

[T]he standard blog structure has disadvantages in the rulemaking context[, in which] . . . there will likely be a large amount of explanatory text compared with the typical blog post. Scrolling back to recheck a point thus becomes more tedious and difficult. As users must expend more effort to find and return to the relevant portion of the explanation, their comments are likely to become less focused and detailed. Moreover, because all comments are made below the text regardless of which part of the text they address, isolating discussion about a specific topic becomes difficult, especially as the comment stream grows. This is a disadvantage for participants while discussion is going on, and for rule makers after the comment period closes.176

(What is true of blogs is even more the case for Facebook pages and Twitter feeds. They are very good tools for communicating specific bits of information or clarifying minor points; they are not strong tools for extensive and productive discussions.) Furthermore, as discussed above,177 the culture of blogs is often not conducive to serious and constructive commenting.

While a blog for a specific rulemaking is not the preferred venue for extensive discussions or the submission of comments, it can serve useful functions. It can flag important issues and documents; be a forum for informal discussion; provide an opportunity for the agency to answer specific, minor questions about the rulemaking; alert interested persons to any public hearings or other relevant proceedings and link to recordings thereof; and general keep the rulemaking visible. It seems likely that a blog will lead to more comments; the jury is out on that question, but the 300,000+ comments submitted on the Forest Service Planning Rule are certainly notable.

Recommendation: For each rulemaking, agencies should consider maintaining a blog dedicated to that rulemaking for purposes of providing information, updates, and clarifications regarding the scope and progress of the rulemaking. The blog should include a widget for submission of official comments to the rulemaking docket. In general, the blog should not, however, be used as a tool for extended discussions of substantive questions at issue in the rulemaking.

2. Discussions

If there is a single largest gap between expectation and reality with regard to e-rulemaking, it is the fact that the move on-line has not at all increased the amount of discussion among commenters or between commenters and the agency. The signal opportunity of social media platforms is to provide a forum for such discussions. The exact structure of such a forum is beyond the scope of this report and would be refined over time. Regulation Room is the leading existing example and the obvious starting point.

As we have seen, agencies have been reluctant to engage in actual conversations with the public on social media sites. This reluctance is easy to understand. Such engagement takes time and effort. It could easily backfire if someone says something without authorization or that is ill-considered and comes back to haunt the agency. There might be some concern that, as in negotiated rulemaking, agency “facilitation” may become agency domination or direction of the discussion.178 And some agency officials have concerns whether direct engagement or moderation of the discussion is permitted by the First Amendment.179

Nonetheless, this is a forgone opportunity to increase the quantity and quality of participation. In part, the simple fact of having a facilitator encourages participation, for it indicates that someone is actually listening, paying attention, and interested in what you have to say. In addition, a talented facilitator can nudge, press, and seek clarification and substantiation, correct statements that are flat-out wrong or misunderstand the proposal, keep the discussion on topic, ask others for responses, all of which enormously increases the likelihood of true dialogue and thoughtful, useful input.

Indeed, this is one the main lessons of Regulation Room.180 The Regulation Room team is large, talented, and hard-working and boasts a wide range of expertise. The site has struggled to achieve its original vision of engaged citizen participation in a dialogic, collaborative process that produces information and insight not otherwise available to the agency. Nonetheless, it has had some success in promoting responsive and useful comments. The Regulation Room team concludes, plausibly, that this success has resulted from labor-intensive, hands-on facilitation and discussion-leading, not by simply creating a forum, listing a few tips, and leaving the public to it.

Recall that Regulation Room seeks to overcome three barriers to effective public participation: ignorance, unawareness, and information overload. The first two have been discussed above. The third involves the way in which mammoth preambles can overwhelm and defeat anyone without technical and/or legal training. “To participate effectively—that is, to do more than simply express support or opposition in general terms—participants must master lengthy, intricate proposals embedded in a mass of linguistically, technically and legally sophisticated material. This intimidates and overwhelms most people, even those who have some relevant working knowledge of, or experience with, the substantive issues being addressed.”181

To address information overload, a group of students and faculty disaggregates, “translates,” and summarizes the material in the preamble, presenting a series of more bite-sized “issue posts” and specific aspects of the proposal. Commenters can address their comments to specific items within an issue post,182 and the comments are threaded. In addition, trained facilitators actively moderate the discussion. “The moderators police inappropriate content and help with site use questions but, far more important, they help lower the barriers of both information overload and ignorance of the rulemaking process by mentoring effective commenting. They point users to relevant information, prompt them to provide more details, and encourage them to react to different positions.”183 According to the Regulation Room team, this facilitation has been essential to productive on-line discussions. 184

Cary Coglianese’s ACUS report on agency use of electronic media185 endorsed this idea. On his view, the facilitator would be an independent contract, not an agency official:

Although the facilitator would not be speaking on behalf of the agency, the facilitator’s objective would be to steer the conversation in a fashion that could be more helpful to the agency’s decision makers. This could mean that agency managers would stay in contact with the facilitator, perhaps conveying their desire to follow up on a particular line of comments or perhaps to raise questions that would be helpful if they were answered by participants in the online conversation. The facilitator could pose questions, float ideas, and even offer his or her own explanations for features, issues, or decisions that the agency has made—but all without binding the agency.186

Recommendation: When soliciting input through a social media platform, agencies should provide a version of the NPRM that is “friendly” to lay users. This involves such steps as breaking preambles into smaller components by subject, summarizing those components in plain writing, layering more complete versions of the preamble below the summaries, and providing hyperlinked definitions of key terms.

Recommendation: Agencies should consider, in appropriate rulemakings, retaining facilitator services to manage rulemaking discussions conducted through social media.

3. Direct Outreach

The Consumer Financial Protection Bureau is arguably the federal agency currently most committed to and engaged in the use of social media and other cutting edge technologies. Its most significant use of innovative approaches has been in its rulemaking concerning mortgage disclosures, for which the NPRM appeared in August 2012.187 The proposed regulation would impose new disclosure requirements in connection with consumer mortgages. Central to the rulemaking was designing a mandatory form that lenders would use to provide information about a loan to a prospective borrower. The Bureau developed two prototypes for a Loan Estimate form, and each went through several iterations. It obtained feedback on the forms in two ways. One was in a series of “qualitative testing” in which individuals reviewed the forms, attempted to use them, and gave their reactions and suggestions. Over time, the Bureau performed qualitative testing with 114 persons, of whom 92 were consumers and 22 were members of the loan industry.188

Separately and in addition, beginning in May 2011, the Bureau sought feedback from the general public by: (a) posting two versions of the proposed Loan Estimate form to its blog, from which visitors could email feedback to the Bureau; and (b) posting the forms to its web site, where visitors could use an interactive tool to review and give feedback. Announcements of the availability of the prototype forms were made on the Bureau website and through its Twitter feed. Over the following ten months, it received 27,000 comments. It would seem that the old-fashioned, in-person qualitative testing proved to be the critical basis for changes to the forms through this process. However, the general public feedback through the online interactive tool also had some influence, particularly in identifying problem areas.189 It is worth noting, too, that this is a setting in which simply asking individuals what they prefer—in essence, inviting them to cast a vote—could be valuable. In general, agencies and commentators have been appropriately wary of turning rulemaking into a referendum. But sometimes the ultimate question really is simply “what do people prefer,” and, if a broad segment of the population can be reached, and interactive tool can provide helpful information about that question.

Interestingly, part of the information gleaned by the Bureau did not come from direct comments or reactions, but rather from the Bureau’s observing the behavior of web site visitors. For example, staff aggregated the results into “heat maps” that indicated which parts of the forms people paid attention to most.190 The Bureau insists that it found this useful.191 It is not clear, at least to this author, exactly what useful information the heat maps convey. Still, this experience hints at something potentially quite important. The information provided through citizen use of social media may be indirect, or meta. The agency may learn something not from what is directly communicated to it, but what is indirectly revealed by what members of the public do online and in interacting with the agency. GSA’s Center for Excellence in Digital Government has stressed the insights to be gleaned from social media metrics and assembled guidance and statements of best practices to assist agencies in this regard.192

Recommendation: The information provided through citizen use of social media may be indirect or amount to metadata. Agencies should consider whether the information they need can also be learned not from what is directly communicated by the public, but from what is indirectly revealed by how members of the public interact with the agency online.

4. Ideation Platforms

I discussed above the use of ideation sites for open-ended solicitations of ideas from the public. Agencies might also use ideation platforms for commenting on proposed rules. To date, only one agency of which I am aware has actually done so. In November 2009, the FCC issued a Notice of Proposed Rulemaking concerning Preserving the Open Internet.193 Strikingly, it identified seven different methods for submitting comments: through, on the FCC web site, by email, by U.S. mail, in person, through special arrangements for those with disabilities, and . . . by posting comments on either the blog194 or the IdeaScale site195 dedicated to the rulemaking. The blog seems not to have attracted any comments. The IdeaScale site drew over 5,000 users, who submitted 422 ideas, 2,387 comments on those ideas, and 37,000 votes. The most popular idea had 467 net “agree” votes; the least popular 132 net “disagree” votes. As is typical of IdeaScale ideas and comments, all of these were quite brief, generally not more than a few sentences in length. I have not read all the comments, but it would seem that Commission staff did not themselves participate on the IdeaScale site. The Commission then transferred all of the IdeaScale posts into the official docket. Total filings in the docket exceeded 100,000.196

In the Report and Order (R&O) issuing the final rule, the FCC did not discuss this unusual procedure or mention the IdeaScale comments in particular. The R&O does list “major commenters,” which include those whose submissions are discussed in the R&O. It further acknowledges that “[t]he Commission also received tens of thousands of brief comments in this proceeding, which are not listed here but which were considered.”197 Presumably, the IdeaScale submissions were among these “brief” comments, and the Commission read but did not directly respond to them. Indeed, almost every comment actually referenced in the preamble was filed by a sophisticated corporate, industry, or NGO stakeholder.

The rulemaking is striking in that it is one of the few, and perhaps the only, example of a rulemaking proceeding in which an agency provided for commenting via social media. On the one hand the process seems to have gone entirely smoothly. On the other, it seems to have added little, if anything, to the primary mechanism, which was submitting comments through the agency’s docket management system.

Section V.B, above, discussed certain drawbacks of ideation sites. These include low participation, lack of useful suggestions, the fact that the top vote-getters may be popular for reasons other than merit, the tendency toward the referendum model, and the possibility of manipulation. These concerns are likely to be most salient during the comment period after a notice of proposed rulemaking has been issued. In particular, concerns about manipulation would be especially strong. Although this seems not to have occurred in the FCC rulemaking, it seems inevitable that if interest groups received the signal that the agency actually cared about how many votes different comments got, they would rally the troops and organize astroturf voting campaigns. Accordingly, despite its appeal, having people vote on comments to achieve a rank would be at best a waste of effort, and possibly counterproductive.198 Rankings do have a potentially valuable role during rule-development, however, which involves open-ended brainstorming, a wide-ranging consideration of possible alternatives rather than focused assessment one in particular, and in retrospective review of existing regulations (see section V.F, below).

Recommendation: Agencies should not employ tools through which users can vote on or rank comments submitted in response to a Notice of Proposed Rulemaking.

5. The Crowdsourcing Model

Crowdsourcing involves distributing production and problem-solving, capitalizing on dispersed knowledge and the ancient principle that “many hands make light the work.” A big project can be broken into a large number of discrete, manageable tasks that are distributed to a large number of participants, often but not necessarily volunteers. Prominent examples have involved the development of the open-source software Linux and Wikipedia. There are two basic sorts of crowdsourcing. In one, exemplified by Linux, large numbers of people perform discrete tasks as part of a larger project. Amazon’s Mechanical Turk platform199 is another prominent example of this sort of distributed tasking. Mechanical Turk is an on-line marketplace through which “requestors” can hire “providers” to perform “human intelligence tasks” (or “HITs”), generally the sort of things that people are good at and computers bad at. Unlike the development of Linux, Mechanical Turk involves actually paying the participants. The other category of crowdsourcing, exemplified by Wikipedia, large numbers of people provide small bits of information that add up to a usefully comprehensive body of knowledge.200 Neither is a new idea—law enforcement has crowdsourced surveillance forever, hence anonymous phone tips lines and “if you see something, say something” campaigns. However, the capacity for and ease of crowdsourcing are hugely enhanced by the Internet.

a. Distributed Tasking

In some circumstances, agencies may be able to crowdsource direct, voluntary participation in agency tasks. For example, the National Archives and Records Administration (NARA) recruited volunteers through Wikisource to transcribe scanned images of materials in NARA’s collection. Over the course of a year, volunteers transcribed and verified over 400 documents. Similarly, NARA has uploaded over 13,000 images to Wikimedia commons and asked volunteers to edit the “tags” (the key word(s) associated with an image through which searchers can find it). 201 While this is only a small fraction of NARA’s overall digitization effort, almost all of which is done in house, it at least serves as a “proof of concept.”

The federal agency that has most successfully and extensively relied on distributed tasking is the National Aeronautics and Space Administration. NASA’s crowdsourcing efforts date back to at least 2001, when it relied on thousands of ordinary citizen “clickworkers” to study pictures of Mars and identify craters.202

Though enormously valuable in certain settings, distributed tasking will rarely have an application to rulemaking.

b. Information

Such use of private individuals for particular tasks has little, if any, potential application to rulemaking. The sort of crowdsourcing that in concept could be valuable to rulewriters involves turning to the crowd for its wisdom, for information. This is what President Obama has in mind when he invokes “dispersed knowledge in the American people.” It is an accepted truth among open government enthusiasts around the world that, to quote (from many possible sources) the New Zealand’s Digital Strategy: “The collective wisdom of the public is valuable—after all, the people who are affected by government policy are in the best position to suggest how to make it better.”203

Some agency information collection activities have effectively drawn on dispersed knowledge. The best-known example is the Peer to Patent (P2P) project developed by Beth Noveck.204 P2P allows patent examiners to access general knowledge with regard to whether a supposedly new (and therefore patentable) invention in fact is such. It is generally seen as having been a notable success.205

Similarly, the federal government has increasingly turned to challenges and prizes as a way of stimulating new ideas and inventions. The idea is both old and enormously simple: offer money and fame to whoever manages first to solve a particular problem or invent a particular device. (Note that the sponsor need not be the government; the best-known non-governmental examples are the numerous competitions sponsored by the X Prize Foundation.) The 2010 America COMPETES Reauthorization Act provides general authority to agencies to conduct prize competitions to spur innovation, identify new ideas, encourage participants to develop new skills, or otherwise advance the agency’s mission.206 In the three years between September 2010 and September 2013, 58 federal agencies ran 288 challenge competitions.207 However, overall participating agencies have been enthusiastic and these competitions have generated useful ideas and innovations.208

The most obvious and successful crowdsourcing efforts that enlist numerous people outside the government to pitch in to complete a task have involved using individuals as “eyes and ears” for various sorts of reporting. Many of these are at the local government level; perhaps the best-known is the SeeClickFix system for reporting potholes, broken traffic lights, graffiti, and the like.209 A conceptually similar use of reports from scattered individuals is the U.S. Geological Survey’s “Did You Feel It?” site.210 After an earthquake, individuals who felt it so indicate on the website, allowing the USGS to precisely map the intensity of the quake.

A promising crowdsourcing effort is the Consumer Product Safety Commission’s new site for reporting unsafe products, This online database allows consumers to register concerns or complaints about safety issues with purchased products. Complaints are not published directly to the database upon receipt. Instead, manufacturers are first given the opportunity to respond and dispute the complaint. The CSPC posts the complaint only if it is satisfied of its genuineness.211 The National Highway Traffic Safety Administration has a similar site,, that allows individuals to submit complaints about their vehicles and search others’ submissions.212

c. Rulemaking Applications

None of the foregoing involve rulemaking, of course. The question is whether agencies might more effectively tap into the wisdom of crowds, really reaching “the crowd” rather than just a relative handful of self-selected stakeholders, through more effective use of new technologies. Notice and comment can be conceptualized as a sort of crowdsourcing, and e-Rulemaking in particular is often justified in exactly these terms. Indeed, the site hosts a document entitled “Public Comments Make a Difference” that directly invokes crowdsourcing concepts in explaining why public participation in rulemaking matters:

Distributed expertise. No agency has perfect knowledge. Some of the information needed to conduct risk assessments for health and safety, for example, comes from those regulated. This information can alert agencies to unforeseen options or consequences of proposed rules. Also, the more comments, the greater likelihood of collecting the most accurate, useful, and current information for the development of rules [regulations].213

However, as discussed in Part II, so conceived it has largely failed. Broad lay participation has generally not been forthcoming, and when it has been it has not magically produced the sort of wisdom that crowdsourcing enthusiasts promise. As the Regulation Room project has done much to demonstrate, the sort of informational challenges involved in rulemaking will in most cases (not all) not lend themselves to elucidation by broad lay commentary. The issues are often technical, scientific, or complex; the relevant facts are defined by particular legal requirements and settings; the process is far down the path toward a final decision; laypersons will simply not possess the information that rulewriters most need.

There is an additional obstacle. Crowdsourcing in particular, and deliberative decisionmaking in general, works best for what are sometimes called “eureka problems,” i.e. problems with answers that are self-affirming or clearly correct once stated.214 Rulemakings almost never involve eureka problems. There just is no clearly correct answer to the trade-offs, scientific disputes, and questions of degree that are generally in play in agency rulemaking.

Perhaps it is no surprise, then, that in a recent report from the IBM Center for the Business of Government devoted to, and entitled, Using Crowdsourcing in Government, the word “rulemaking” simply does not appear.215

For certain sorts of questions, however, the crowdsourcing model is compelling. There will be particular rulemakings where an agency might benefit, in a crowd-sourced, Mechanical Turk sort of way, in getting a whole bunch of volunteers to try things. The CFPB’s Know Before You Owe process216 is an example. The agency sought to determine which of two disclosure forms was more helpful and comprehensible. It did the obvious thing: it had a bunch of people look at the forms and give their reactions. There are other settings where that sort of direct feedback would be helpful. One could imagine, for example, giving different groups different versions of a warning label, letting each look at it and then take a little test about what they noticed, retained, and understood.

Agencies should not assume that all rulemakings will be enhanced by a crowdsourcing approach. However, where public or user response is precisely the question to be determined, direct submission to the public at large will provide useful information and should be pursued.

d. “Situated Knowledge”

While the dispersed knowledge/crowdsourcing idea is easily oversold, a narrower version is robust. We return again to the findings of the Regulation Room project. An important conclusion is that the most useful lay comments will come not from members of the general public but from individuals who possess “situated knowledge.” “This knowledge is based on their on-the-ground experiences with the kinds of problems, circumstances, or solutions involved in the proposed regulation.”217 Such knowledge might reveal levels of complexity of which the agency was unaware, hidden contributions to existing problems, possible unintended consequences of particular proposed solutions, or ways of thinking about a problem that just had not occurred to policymakers without day-to-day, on-the-ground experience.218

An example has been the “President’s SAVE Award.” Created in 2009, this award invites suggestions from federal employees as to how the government can save money. Each year, a set of finalists and one winner is chosen from all the submissions, although other suggestions can of course also be implemented. In the five years of the awards existence, there have been 85,000 submissions. According to Steven VanRoekel, U.S. Chief Information Officer, “more than 80 SAVE Award ideas have been incorporated into the President’s Budgets, saving hundreds of millions of dollars.”219 The SAVE Award is not magic; only one in a thousand suggestions has made it into the President’s budget and the savings are a drop in the bucket. But it counts as a success, and it draws on the fact that federal employees who see the nuts and bolts of federal spending on the ground will have useful ideas about where savings can be made. This is an example of situated knowledge, albeit outside the rulemaking setting.

Recommendation: Agencies should realize that not all rulemakings will be enhanced by a crowdsourcing approach. However, where the public or user response is the question to be determined, direct submission to the public at large may provide useful information. In addition, agencies should seek to encourage, and be receptive to, comments from lay stakeholders with “situated knowledge” arising out of their real-world experience.

E. Collaborative Drafting

A more ambitious form of public engagement in rulemaking would be not simply to solicit more valuable comments and discussion, but to allow persons outside the agency to participate in the actual drafting of the text of a regulation or preamble. Numerous collaborative drafting platforms exist; well-known examples include Google Docs, Mediawiki (the software used for Wikipedia), and, more recently, GitHub. All are web-based tools that allow multiple users to edit a single document. The software keeps track of who made what change when and also provides pages for discussions. For simplicity’s sake, I will simply refer to these various tools generically as wikis.

Wikis can be seen as a particular instantiation of crowdsourcing, drawing on both the dispersed knowledge and the distributed tasks components. Wikipedia stands as the wildly successful and truly astonishing model of what is possible through a wiki. Several federal agencies have had significant success with intra-agency “pedias” used to create and share knowledge within the agency. Two leading examples are the State Department’s “Diplopedia,” which was launched in 2006 and gathers information from and for State Department foreign affairs specialists,220 and “Intellipedia,” launched the same year and used by the intelligence community.221 EPA’s “Watershed Central Wiki”222 gathers broad information regarding watershed management best practices. It is open to participation by people outside the agency, but only those who actually do business with the agency.

These success stories involve compiling information. That is much easier than actual drafting of texts. Wikipedia works partly because it involves essentially factual material. A Wikipedia entry on “the best music” or “the right religion” would not be stable. Drafting regulations involves contested issues with enough disagreement about first principles, especially on issues that the general public cares about, to make an open wiki rather challenging in the rulemaking setting. Nonetheless, might wikis be used for public drafting of government documents, including statutes and regulations?

Some private attempts to draft law—generally, proposed legislation—via a wiki have been made. One example was, which was created in 2011223 and now seems inactive (although the web site is still up). Lexpop features the rather optimistic tag line “More voices. Solve everything.”224 It hoped to engage large numbers of citizens in collaborative drafting and problem solving, including through drafting legislation. A Massachusetts state legislator agreed to introduce a net neutrality bill drafted through the lexpop wiki. However, nothing seems to have come of it as the project never got past the threshold problem of finding participants. In the somewhat plaintive words of its founder, “[a]s we’ve learned over the past week, it’s tough to crowdsource without a crowd.”225

A more promising but as yet still insignificant effort is the “Madison Project,” sponsored by the Open Government Foundation and created with the involvement and encouragement of Representative Darrel Issa. Representative Issa used an early version of the software in drafting the proposed OPEN Act.226 Any entity, person, or government can use the software, which is open source.227 A Madison Project site posts draft legislation and enables users to modify, annotate, and comment on the bill’s text. Highlighting a passage opens a comment bubble; the comment shows up in the right-hand margin. Others can then “like,” “dislike,” or comment; the most popular suggestions are automatically moved to the top of the list. The software has had some modest use, with fairly meager results.228

In short, the early results are intriguing but there has yet to be a compelling example of effective public legislative drafting by wiki. (Though there are reports of greater successes in other countries, notably Russia.229)

Experience with regulation-writing has been even skimpier. The private site did use Madison to solicit public input on the FCC’s Net Neutrality rule.230 However, this was not a serious effort. They posted the final regulation, after the rulemaking process had concluded, inviting users to “[a]dd your expertise to the regulation to see if we can't find a better way that we all agree on.” A grand total of one person made a comment, and that comment read, in its entirety, “Amen sister!”

To my knowledge, no federal agency has yet posted proposed text for a regulation for public editing via a Wiki. But the idea has its supporters. A well-known “TED Talk” by Clay Shirky promotes the idea, specifically endorsing the use of GitHub as the platform of choice.231 In 2012 GitHub was used to correct a typo in a Consumer Financial Protection Bureau document;232 an event which serves as an extremely modest proof of concept. Ben Balter of GitHub states that he would “love it” if agencies used GitHub in that way. In his view, “we are there legally and technically. The problem is cultural. Culturally we are not going to be there for a while.”233

In short, thus far efforts at legal drafting by wiki have come to grief at the threshold: wikis depend on large numbers of knowledgeable people acting in good faith being willing to share the labor for free. It can happen; Wikipedia is the endlessly invoked example.234 But it will not happen always or automatically.

The fact that participation has been low does not mean the idea is doomed. Higher rates of participation seem possible; moreover, useful collaborative drafting can occur even without thousands of participants—indeed it may go more smoothly. So the open source software model is theoretically applicable to regulatory drafting. For several reasons, however, it will be applicable only to a small set of rulemakings. There are two basic reasons why, even with broad participation, the open software model cannot simply be transplanted to mass rulewriting.

First, open-source software projects by their nature largely draw participants who know what they are doing; the projects are sufficiently technical that there are high barriers to entry for incompetent programmers. An open-to-all regulation-drafting wiki would lack that built-in, self-enforcing screening mechanism.

Second, and much more important, drafting by wiki works best when the participants share common motivations and respect for common constraints—for example, when they are all members of the same organization or work for the same company—who are committed to an ongoing project.235 It is called “collaborative drafting” for a reason; if the participants are not “engaged in a mutually beneficial relationship to meet pre-defined goals,”236 it just does not work. But most rule-writing involves, at least to a degree, a zero sum game in which directly opposing interests are at odds. In that setting, drafting by wiki is doomed; it will become a constant process of each side undoing the other’s changes in the hopes of being the last version standing.

Thus, there is every reason for agencies to hesitate with regard to an open, public drafting of actual regulatory text. For exactly the same reasons, however, wikis may be useful for (a) drafting within the agency or (b) drafting of comments by like-minded stakeholders. The former is easily achieved through an intra-agency wiki. One could imagine the agency setting up a platform through which the latter could occur as well. However, it is not clear that that is really the agency’s job and there are potential legal and other complications with the agency hosting one interest groups comment process.

Recommendation: Agencies should not set up public collaborative drafting sites for the production of regulatory or preambular text or for the production of comments. Agencies should experiment with collaborative drafting platforms internally within the agency for purposes of producing regulatory documents.

F. Retrospective Review of Existing Regulations

Since 1978, every U.S. president has directed agencies to evaluate or reconsider existing regulations. Most recently, President Obama’s Executive Order 13563, “Improving Regulations and Regulatory Review,” issued in January 2011, instructs agencies to consider how best to promote retrospective reviews of regulations that may be outmoded, ineffective, insufficient, or excessively burdensome, and to modify, streamline, expand, or repeal them in accordance with what has been learned.237 A variety of statutory requirements also require retrospective review of particular sorts of regulations. Most significantly, the Regulatory Flexibility Act requires that agencies review, every ten years, existing regulations that have a significant economic impact upon a substantial number of small entities.238 The CFPB’s statute requires that it undertake a subsequent review of every “significant” rule or order within at least five years of its issuance.239 Wholly apart from these legal requirements, there is general consensus that retrospective review is, simply, a good idea and something agencies should do whether required to or not.240

Despite this paper commitment and general consensus, prior to the current round, retrospective reviews have fallen short of expectations.241 Current undertakings seem to be more rigorous and to reflect a greater commitment than their predecessors,242 but the jury remains out on their ultimate success. No doubt many factors have contributed to these mixed results, including a lack of commitment on the part of the agencies, the fact that the satisfaction, political payoff, and sense of progress from eliminating an old regulation will generally be less than for the adoption of a new one, and a lack of resources.243 However, many observers have concluded that historically one significant shortcoming of retrospective review has been a lack of meaningful public participation. 244 In 2007, GAO reported:

Agencies stated that despite extensive outreach efforts to solicit public input, they receive very little participation from the public in the review process, which hinders the quality of the reviews. Almost all of the agencies in our review reported actively soliciting public input into their formal and informal review processes. They reported using public forums and industry meetings, among other things for soliciting input into their discretionary reviews, and primarily using the Federal Register and Unified Agenda for soliciting public input for their mandatory reviews. For example, USDA officials reported conducting referenda of growers to establish or amend AMS marketing orders, and CPSC officials reported regularly meeting with standard-setting consensus bodies, consumer groups, and regulated entities to obtain feedback on their regulations. Other agencies reported holding regular conferences, a forum, or other public meetings. However, most agencies reported primarily using the Unified Agenda and Federal Register to solicit public comments on mandatory reviews, such as Section 610 reviews. Despite these efforts, agency officials reported receiving very little public input on their mandatory reviews.245

GAO also spoke to observers outside the agencies. They were less impressed by the intensity of agencies’ outreach, but there was general consensus that: (a) there had been little public input; and (b) more public input would have been useful.

The current round of retrospective review provides an opportunity for agencies to rely on Web 2.0 for meaningful public input. Several are doing so. Some efforts are modest, though not unimportant. For example, EPA’s plan states that it will seek nominations of regulations in need of review from the public, other agencies, and EPA staff. It will do so “via the Semiannual Regulatory Agenda, a press release, and related outreach tools.”246 The “related outreach tools” are unidentified but if they are “related” to the semiannual agenda and press releases they are not exactly at the technological cutting edge. In fact, in early 2011 EPA did solicit comments not just by press release, but also on its open government website, Twitter feed, and Facebook page. The actual comments, however, are to be submitted to

The Department of Labor established an IdeaScale site for submission both of ideas about the agency’s retrospective review plan and to identify regulations appropriate for review.247 The site garnered 20 ideas, 1314 users, and 420 votes, of which the overwhelming majority were in support of a particular suggestion regarding electronic disclosure of certain information.248 It would be a happy story if this idea was uncovered by this public process; in fact, the agency had issued a Request for Information on electronic disclosure two months before the IdeaScale went live.249

Similarly, the Department of Transportation supplemented its public meetings and use of with an IdeaScale site “to allow the public to more easily and effectively submit comments on our process and suggestions for specific rules that should be reviewed.”250 The theory was that IdeaScale “permit participants to discuss ideas with others and agree or disagree with them, perhaps making it particularly useful for individuals and small entities, including local and tribal governments. We also suggested it might help participants refine their suggestions and gather additional information or data to support those suggestions.”251 The site produced 53 suggestions from 47 participants; 11 comments were submitted about others’ suggestions.252 It is pretty clear, however, that none of these suggestions proved particularly valuable. The agency prepared a 164-page summary of the comments and the actions it was taking in light thereof. The responses to the IdeaScale suggestions vary, but almost all indicate that the suggestion is not feasible, not legal, not within the agency’s jurisdiction, or something the agency was already aware of and working on.253

It is not clear why the DOT experience was not more successful. It is possible that more strenuous outreach would have produce more and more useful submissions. The value and extent of public input will depend hugely on the way in which it is sought. This is common sense. It is also borne out by experience. For example, in the early years of the Bush Administration, OMB solicited nominations of regulations that could or should be modified in order to increase net benefits to the public. In 2001, OMB made its solicitation in the draft of its annual report to Congress on the costs and benefits of federal regulations. It received suggestions for modifying a grand total of 71 regulations. The following year, its process was “much more ambitious” and it received almost 1700 nominations.254
In any event, the lackluster DOT IdeaScale experiment is not a reason for abandoning the use of Web 2.0 approaches as part of the retrospective review process. To the contrary, this seems a particularly promising opportunity for the use of social media. First, retrospective review does not trigger the APA; the rulemaking that produced the rule is complete, and the rulemaking to modify or withdraw the rule being reviewed, if any, has not yet begun. Second, regulated entities and the general public are in a particularly good position to provide the information the agency needs, which involves how a regulation is actually operating in practice. Third, as noted, members of the public, as a generalization, are better at identifying problems than solutions. But that is exactly what retrospective review is trying to uncover. Fourth, the goal of the process is to learn about actual experiences under the regulation; it is more likely that laypersons will have relevant “situated knowledge” in these settings. Finally, the opportunities social media provide for multiple participants to ask questions of each other, compare notes, and engage will be especially useful when the inquiry concerns actual experience under a particular regulatory requirement.

Recommendation: Agencies should consider using social media in support of retrospective review of existing regulations, particularly to learn what actual experience has been under the relevant regulation(s).

G. Direct Final Rulemaking

Direct Final Rulemaking (DFRM), endorsed by the Administrative Conference of the United States in 1995,255 is a technique for bypassing the notice and comment process for uncontroversial legislative rules. Under this procedure, the agency skips notice and comment and simply publishes a final rule and a statement of its basis and purpose. The notice also indicates that the agency expects the rule to be noncontroversial and provides that the rule will be effective within a relatively short time, such as 60 or 90 days, unless significant adverse comments are received. If no one objects, the rule stands; some agencies publish a “confirmation notice” to indicate that there has been no significant adverse comment. If adverse comments are received, the agency withdraws the rule; it may then reissue it as a proposed rule, thus commencing full-fledged notice-and-comment procedures. 256

DFRM is a standard tool for many agencies, led by EPA and the Department of Transportation. Over one four-year period, federal agencies issued over 1000 direct final rules, of which 62 elicited significant adverse comment.257 In 2005, 27% of the rules EPA published in the Federal Register were direct final rules.258 However, it has its detractors and leaves many uneasy, suspicious that notice and comment is forgone inappropriately.259

Social media could play an important role in ensuring that DFRM is not abused. When considering or pursuing DFRM, an agency needs to know, quickly, whether there is meaningful objection. Social media platforms offer an effective way of getting word out while also providing a forum for preliminary discussion.

Recommendation: Agencies should consider using social media in connection with direct final rulemaking to quickly identify whether there are significant or meaningful objections not initially apparent.


A. The Administrative Procedure Act

The conventional wisdom is that the Administrative Procedure Act stands as a barrier to the use of social media in rulemaking. In 2008, the Federal Web Managers Council (FWMC) produced a document entitled Social Media and the Federal Government: Perceived and Real Barriers and Potential Solutions.260 It identified ten barriers to greater use of social media by government agencies and possible ways to overcome them. Number ten on the list was the “hesitation and confusion as to how to incorporate [social media] during the rulemaking process,” given that the APA was not written with these tools in mind.261

The FWMC suggested that “[t]he National CTO or OMB should issue guidance to help agencies use collaborative social media tools to enhance the rulemaking process, while still complying with the APA.”262 Such guidance has not been forthcoming, and discussions with lawyers and others within agencies indicates a continued concern about how it would be possible to engage in the sort of fluid, dialogic give and take that characterizes social media while still complying with the APA.263

As described above, to date agencies have generally separated social media discussion from “official” comments as part of a rulemaking proceeding. Agencies have been quicker to gather input from ideation sites and blogs, for example, when the ultimate agency product is a report, or a policy document, or a strategy—i.e. something other than a regulation subject to APA procedures. Where such sites were tied to an actual rulemaking, such as the Forest Service’s blog on its planning rule,264 the agency has made clear that blog comments are not rulemaking comments. Similarly, the FCC’s general comment policy, which covers all “thoughts, ideas, and feedback you submit to the FCC online,”265 includes the following in a section headed “Relation to Proceedings”:

Unless explicitly stated otherwise, commenting on these platforms is not a substitute for submitting a formal comment in the record of a specific Commission proceeding. The Commission will not rely on anonymous comments, or comments posted elsewhere due to the difficulty of verifying the accuracy of information. Accordingly, any persons interested in examining the record should review the traditional Electronic Comment Filing System.266

The careful separation of social media interactions from official commenting rests primarily on concerns that public comments on a blog could constitute submissions the agency would then have to docket and respond to, and which the agency therefore ignores, or reasonably overlooks, at its peril.

The essential features of notice-and-comment rulemaking are embodied in the term itself: adequate public notice, a meaningful opportunity to comment, agency consideration and response to those comments, and (to a somewhat uncertain extent) inclusion of all relevant background materials and all public submissions in a docket, available to all (including a reviewing court). A large body of caselaw fleshes out these requirements.

Section 553 requires that an agency give all “interested persons”—in short, everyone—“an opportunity to participate in the rulemaking through submission of written data, views, or arguments.”267 It is implicit in the statute, and courts have consistently held, that this “opportunity to participate” must be meaningful. Accordingly, the notice of proposed rulemaking must indicate what the agency proposes to do, why, and on the basis of what studies or information.268 More important for present purposes, the agency must take the comments seriously and not ignore them.269 This is explicit in §553, which provides that the agency shall issue a final rule only “[a]fter consideration of the relevant matter presented.”270 Irrelevant submissions can be ignored, but “relevant matter presented” must be considered.

This requirement of consideration is enforced by and reflected in a requirement, which is not explicit in the text of the APA, to discuss and respond to significant comments.271 Any agency is not required Although the FCA is not required “to discuss every item of fact or opinion included” in public submissions,272 but, it must respond to those “comments which, if true, ... would require a change in [the] proposed rule.”273

This obligations has four interconnected sources. First, the APA requires the agency to include in a final rule “a concise general statement of [the rule’s] basis and purpose.”274 As elaborated by the courts of appeals, this requirement has become much more burdensome than its text suggests; as many have said, the required statement is neither concise nor general and must go beyond simply laying out the rule’s basis and purpose.275 Two decades ago, in Recommendation 93-4, Improving the Environment for Agency Rulemaking, ACUS recommended amending §553 to bring its text in line with the judicial gloss thereon by eliminating the words “concise” and “general” and “codifying existing doctrine that a rule must be supported by a ‘reasoned statement,’ and that such statement respond to the significant issues raised in public comments.”276 The key early decision was Automotive Parts & Accessories Ass’n v. Boyd, which stated that while “[w]e do not expect the agency to discuss every item or opinion included in the submissions made to it[,] . . . [w]e do expect that . . . [it] will enable us to see what major issues of policy were ventilated by the informal proceedings and why the agency reacted to them as it did.”277 That idea has hardened into a firm requirement to respond to significant comments. The D.C. Circuit has noted that the “purpose [of the statement of basis and purpose] is, at least in part, to respond in a reasoned manner to the comments received, to explain how the agency resolved any significant problems raised by the comments, and to show how that resolution led the agency to the ultimate rule.”278

The duty to respond to significant comments is also grounded on the APA’s reference to the agency’s “consideration” of the comments,279 which is textually linked to the issuance of the statement of basis and purpose and also implies actual discussion of the comments. In addition, courts have concluded that the failure to discuss relevant comments would be in itself arbitrary and capricious.280 The courts require “reasoned decisionmaking,” consideration of all relevant factors, review of available alternatives. Agencies cannot show that they have met these requirements without discussing significant issues raised by commentators. Indeed, some courts have grounded the obligation to respond solely on these requirements,281 though the better view is that they stem from §553 as well. Some scholars have argued that much of this caselaw is invalid under the Supreme Court’s Vermont Yankee decision, which rejected judicial expansion of agency procedures beyond the requirements directly imposed by statute.282 Whether or not that critique is correct, the scholars who make it do not dispute that agencies must consider and respond to all relevant and significant comments. The requirement seems firmly entrenched.

Finally, courts have required discussion of relevant comments so as to enable effective judicial review.283

The obligation to respond does not require the agency to discuss every comment, no matter how off-topic or wrong-headed. “The APA requirement of agency responsiveness to comments is subject to the common-sense rule that a response be necessary.”284 Accordingly, “comments must be significant enough to step over a threshold requirement of materiality before any lack of agency response or consideration becomes of concern.” Courts have generally been sympathetic to agencies faced with a large volume of comments, not requiring individualized responses to every submission.285 Individual commenters are not entitled to a response. Rather, the courts’ key concern is that agencies consider all relevant issues.286 “‘The failure to respond to comments is significant only insofar as it demonstrates that the agency's decision was not based on a consideration of the relevant factors.’”287 Moreover, it can safely ignore comments that are trivial, duplicative, or irrelevant. A standard formulation is that it must address all comments that are “relevant and significant.”288

Consistent with this principle, agencies are permitted to respond minimally to comments that are conclusory, setting out a bottom line supported by little or no documentation or argument. For example, in Brown v. Secretary of HHS,289 the agency had issued regulations that limited to $1,500 the equity value of an automobile that would be ignored in considering a family’s resources for purposes of determining eligibility for certain welfare benefits. The proposed rule had included the $1,500 figure, and a dozen or so comments objected that it was too low. The agency’s response was minimal: “We stand by our original position. The choice of $1,500 as the maximum equity value for an automobile was based on the data from a Spring 1979 survey of food stamp recipients. We regard the limit of $1,500 equity value in an automobile as reasonable and supportable.” But this satisfied the court:

Only a dozen comments were submitted on the automobile resource exemption, of which ten took issue with the $1,500 amount. Each of these comments was fairly brief, criticizing the figure as generally too low. Only one of them suggested an alternative to the $1,500 figure. None of them suggested any alternative data upon which to base the figure. Given the nature of the comments, we do not find the Secretary's brief response so inadequate as to violate [the duty to respond].290

Finally, some individual agency statutes contain express requirements to consider and respond to comment. For example, the Clean Air Act requires that EPA accompany any final rule issued thereunder with “a response to each of the significant comments, criticism, and new data submitted in written or oral presentations during the comment period.”291 The Regulatory Flexibility Act requires the final regulatory flexibility analysis (RFA) to review public comments on the draft RFA and highlight changes made in light of those comments.292 Under the Unfunded Mandates Reform Act, an agency issuing a “significant” regulation must prepare a written statement that summarizes its “evaluation of . . . comments and concerns” presented by State, local, or tribal governments regarding the proposed rule.293

Requiring agencies to consider and respond to comments rests on three assumptions about how notice and comment operates. First, the agency will be able to tell that a particular submission is a “comment” relevant to a specific NPRM. Second, each “comment” will be a stand-alone submission that could be considered in isolation. Third, the overall burden of comments will be manageable. These assumptions underlie the caselaw in this area, and are also reflected in agency regulations related to the commenting process. Consider, for example, the Department of Transportation imposes the following requirements for written comments:

Your comments must be in English and must contain the following:
(a) The docket number of the rulemaking document you are commenting on, clearly set out at the beginning of your comments.
(b) Information, views, or arguments that follow the instructions for participation that appear in the rulemaking document on which you are commenting.
(c) All material that is relevant to any statement of fact in your comments.
(d) The document title and page number of any material that you reference in your comments.294

Requirements such as these are designed to separate “comments” from general background chatter, so that the agency and any reviewing court will know what is in the record, what the agency must consider, and what it must respond to. It also ensures that the agency does not overlook any submission that was intended to be a comment. The legality of such requirements has never been in question. But these requirements are not easily translated to informal, ongoing, multi-forum discussions on social media. In this new setting, material that belongs in the rulemaking docket blends into the general background chatter. Indeed, that is exactly the point.

Given current premises regarding the rulemaking record and judicial review, agencies cannot rely on input outside the record through online discussions. Ironically, they could have at the time the APA was adopted, when informal rulemaking bore no resemblance to on-the-record decisionmaking.295 The process has moved increasingly toward an on-the-record model, however. As a result, agencies have to clearly identify what is in the record, have to be inclusive in making those decisions, and have to base their decision on such material.296 Anything that the agency actually relies on has to be in the rulemaking record.

What options, then, do agencies have to read or participate in social media discussions regarding rulemakings separate from the submission of comments denominated as such to a rulemaking docket?

1. Option One: Include everything. One approach to adapting existing rulemaking requirements to the use of social media would be to throw everything into the record—any on-line comment or discussion regarding a proposed rule would be placed in the record. The FCC took this approach in its Open Internet rulemaking.

There seems little question about the legality of this approach. Indeed, it would reflect lawyerly caution; no one can complain that they were excluded. The problem is that at some point the “record” comes to more less overlap with “the Internet.” That is, of course, an exaggeration. Still, in an era of limitless information, the value of the rulemaking process is not only in gathering up useful information, but also in winnowing out distracting, unhelpful, or duplicative material.297 We may in the foreseeable future reach a point where technology renders the idea of a docket containing all relevant material a quaint anachronism. There is too much relevant material, and it can all be dumped into the docket with a click of the mouse. It is possible that the time has come to reconceptualize the rulemaking docket as something more akin to the Joint Appendix in an appeal; a winnowed down subset of the whole, limited to the stuff that really matters. That would, of course, be a huge shift, and probably outside the scope of this project.

Option 2: Require participants to designate comments as such and submit them to the rulemaking docket. The alternative is to update requirements such as those from DOT to take into account the fact that a good deal of discussion of a proposed rule might take place on social media. The essential idea would be to do exactly what the Forest Service did with its Planning Rule blog. The agency hosts, facilitates, and participates in a discussion through social media, but makes clear that nothing said in that discussion is actually a “comment,” and, accordingly, does not require an agency response and will not be included in the rulemaking record. Any such restriction should include instructions on how to submit a comment and a link to or the agency web site to do so. This could be done most efficiently through the eRulemaking programs new commenting API, which would enable submission of a comment to the FDMS without leaving the social media site.

This approach raises more serious legal questions. After all, there have to be some limits on the agency’s ability to define what a “comment” is. Arbitrary or pointlessly onerous requirements (font size, indicating the RIN on each page, only filing between 3:00 and 5:00 p.m. on Thursdays, having the comments signed by a member of the bar) would violate § 553’s requirement that the agency “give interested persons an opportunity to participate in the rulemaking through submission of” comments.298 However, the sort of limitations just mentioned seem appropriate and legally permissible, and there could be no question about an agency’s authority to require use of the Commenting API. They are not more onerous or restrictive than the limitations currently in place. The agency’s obligation is to give “an opportunity.” It is not legally obligated to give every opportunity or to actively round up comments.299 And its obligation is to permit “submission[s].” That term implies an active, self-conscious decision on the part of the “interested person” to provide the agency with relevant material.

Option 3: Forgo use of social media during the comment period. Whatever constraints the APA imposes on the use of social media in rulemaking apply only after the agency has issued a notice of proposed rulemaking, before issuance of the final rule, and in rulemakings that are not exempt from the requirements of §553. Accordingly, opportunities for experimentation, free of concerns about the APA, exist in the early stages of a rulemaking (for example in developing or gathering reaction to an ANPR), in undertaking retrospective review of regulations, in developing interpretive rules and statements of policy, and when issuing direct final or interim final rules. An ultra-cautious agency could simply forgo use of social media related to a regulation during the comment period. That would be carrying lawyerly caution to an unnecessary extreme.

Recommendation: The APA does not restrain agency use of social media before an NPRM is issued or after a final rule has been promulgated.

Recommendation: When an agency uses social media in connection with a notice-and-comment rulemaking, it has two options for determining how the discussion will be treated under the APA:

The agency may decide to include all comments submitted via social media in the rulemaking record. Agencies should consider using an application programming interface (API) or other appropriate technological tool to efficiently transfer content from social media to the rulemaking record.

The agency may decide that no part of the social media discussion will be included in the rulemaking record, be considered in developing the rule, or be responded to in the final rule. An agency that selects this option should communicate the restriction clearly to the public, provide instructions on how to submit an official comment, and provide a link to the rulemaking docket on or other agency rulemaking portal.

Recommendation: Regardless of which approach an agency pursues pursuant to the previous recommendation, it should communicate clearly to the public how the social media discussion will be used in the rulemaking.

B. Paperwork Reduction Act

The Paperwork Reduction Act requires agencies and OMB to scrutinize any planned information collection requests (ICRs). Concerns that requests for feedback might qualify as an ICR and trigger PRA requirements have blocked or discouraged agency reliance on social media. Fortunately, OIRA has clarified that the PRA does not apply to most agency requests for public feedback, although some additional clarity might be helpful.

1. The PRA Process

The PRA has the noble though largely unrealized goal of eliminating unnecessary paperwork imposed by the federal government. It is aimed in particular at government agency collections of information from the public. Such collections can take many forms, a reality evident in the statute’s broad definition of “information collection”:

the obtaining, causing to be obtained, soliciting, or requiring the disclosure to third parties or the public, of facts or opinions by or for an agency, regardless of format, calling for either answers to identical questions posed to or identical reporting or record keeping requirements on ten or more persons, other than agencies, instrumentalities, or employees of the United States.300

The definition contained in OMB’s regulations is even broader:

the use of report forms; application forms; schedules; questionnaires; surveys; reporting or recordkeeping requirements; contracts; agreements; policy statements; plans; rules or regulations; planning requirements; circulars; directives; instructions; bulletins; requests for proposal or other procurement requirements; interview guides; oral communications; posting, notification, labeling, or similar disclosure requirements; telegraphic or telephonic requests; automated, electronic, mechanical, or other technological collection techniques; standard questionnaires used to monitor compliance with agency requirements; or any other techniques or technological methods used to monitor compliance with agency requirements. A “collection of information” may implicitly or explicitly include related collection of information requirements.301

Operating somewhat at cross-purposes, the PRA establishes an extensive process of pre-clearance for new Collections of Information (CIs) or extension of existing CIs. Agencies must establish an internal review process, and must consult with the public and affected agencies, seeking comments on the need for the information, its practical utility, the accuracy of the agency's burden estimate, and ways to minimize that burden. The PRA requires the agency to publish a notice in the Federal Register and open a 60-day comment period. After all that, the agency must submit a clearance package—an Information Collection Request (ICR) —for approval by OIRA, during which review another 30-day comment period is opened.302

The ICR must include a supporting record, including anything submitted in response to the Federal Register notice, and a formal certification that the proposed CI is needed, not necessarily duplicative, reduces the burden on respondents, uses “unambiguous terminology,” is consistent with the existing record-keeping practices of respondents, and indicates how long respondents must keep relevant documents.303

Agency submission of an ICR to OMB entails another Federal Register notice. In this notice, the agency must summarize and describe the need for the proposed CI, describe likely respondents, estimate the annual burden, and give notice that the comments may be submitted to OMB and the agency. OIRA has sixty days to review the submission; with notice to the agency it can extend that period by another 30 days.304 OIRA may not approve the CI for a period of longer than three years. If OIRA fails to act within the sixty-day period, the CI is deemed approved, though only for one year. Executive agencies are bound by an OIRA disapproval; independent agencies can override OIRA’s disapproval.305

2. PRA Compliance as an Impediment to Use of Social Media

Submitting ICRs to OIRA is not an impossible nightmare, but it is a significant burden and involves a lengthy delay. The whole process significantly, and understandably, discourages agencies from pursuing actions that might trigger that obligation.

The implications for agency use of social media are evident. The whole point of using social media in rulemaking is to gather feedback, input, and information. Thus, a request for input or feedback in a social media forum might be deemed a “collection” (see the definitions above) of “information,” i.e. “any statement or estimate of fact or opinion, regardless of form or format, whether in numerical, graphic, or narrative form and whether oral or maintained on paper, electronic or other media.”306 And pretty much by definition it will be aimed at more than ten people. Historically, concern over triggering PRA requirements seems to have been a meaningful constraint on agency use of social media, discouraging agencies from sending out surveys in conjunction with retrospective review of existing regulations,307 prompting the CFPB to forgo some solicitation of feedback in formulating its Know Before You Owe proposed rule,308 leading agencies to forgo use of blogs,309 and lurking as a general disincentive to use of web 2.0 materials that involve soliciting feedback.310

OIRA has been sensitive to these concerns. In 2009, as part of a general request for comments on improving the administration of the PRA,311 it asked: “What practices could OMB implement under the PRA to facilitate the use of new technologies, such as social media, as well as future technologies, while supporting the Federal Government's responsibilities for Information Resource Management?”312 The following year, OIRA Administrator Cass Sunstein released a memorandum, discussed in detail below, that clarified the scope of the PRA and read it not to apply to a wide range of social media activity.313 A month later Sunstein issued a memorandum on generic clearances, reminding agencies, and clarifying, that certain web-based activities, such as getting feedback on the agency’s website, are not exempt, but could be approved through a generic clearance, meaning that the agency would not have to obtain approval each time it undertook the request.314

The OIRA memos significantly reduced the concern and uncertainty about the PRA and social media, but did not eliminate them. The following section addresses key legal issues.

3. Legal Issues

The PRA applies to requests for information “regardless of form or format” 315 and OMB regulations are sweeping and extend to “automated, electronic, mechanical, or other technological collection techniques.”316 Accordingly, there is no tidy, blanket exemption for requests made via social media. However, the Act does not define the term “information.” OIRA has defined the term with care in light of the overall purposes of the Act.317 The Social Media Memorandum identifies three sorts of submissions that, while undoubtedly “information” under the lay or dictionary definition, are not “information” within the meaning of the PRA:

General Solicitations. 5 C.F.R. 1320.3(h)(4) excludes “facts or opinions submitted in response to general solicitations of comments from the public, published in the Federal Register or other publications, regardless of the form or format thereof, provided that no person is required to supply specific information pertaining to the commenter, other than that necessary for self-identification, as a condition of the agency’s full consideration of the comment.”

Public Meetings. 5 C.F.R. 1320.3(h)(8) excludes certain “facts or opinions obtained or solicited at or in connection with public hearings or meetings.”

Like Items. 5 C.F.R. 1320.3(h)(10) reserves general authority for OMB to identify other “like items” that are not “information.”

Between these three exceptions, most public participation in agency social media sites is exempt from the PRA. It would certainly seem that requests for comments on possible or proposed regulations would qualify as a “general solicitation.” In addition, the “public meeting” exception is read broadly to include on-line interactions that replicate public meetings. “For purposes of the PRA, OMB considers interactive meeting tools—including but not limited to public conference calls, webinars, blogs, discussion boards, forums, message boards, chat sessions, social networks, and online communities—to be equivalent to in-person public meetings.”318 And “like items” include such things as rankings and ratings.319

This construction of the PRA is sensible. Of course, if we assume that the PRA serves a valuable function, cutting back on it comes at a cost. But there is something peculiar about seeing Web 2.0 discussions as involving paperwork “burdens.” First, they are voluntary. Some have argued that all voluntary CI’s are or should be outside the PRA.320 However, the statutory text is not limited to mandatory reporting. A collection is “the obtaining, causing to be obtained, soliciting, or requiring the disclosure to third parties or the public, of facts or opinions.” While “obtaining” could possibly be stretched to imply compelled disclosure of some sort, “soliciting” just cannot. This reading is confirmed by the provisions regarding what notice must be provided to private parties responding to information requests. That notice must, among other things, indicate “whether responses . . . are voluntary, required to obtain a benefit, or mandatory.”321 Accordingly, the application of the PRA to voluntary disclosures is settled.322 Moreover, the line between “voluntary” and “mandatory” collections is not easy to draw. Just like offers can be too good to refuse, requests for information, though nominally voluntary, might amount to de facto requirements.323 Accordingly, a blanket rule that voluntary CI’s are not covered might well be a mistake and would be inconsistent with the existing statute.

However, much in the way of voluntary requests do not implicate any PRA concerns. (ACUS’s 2012 recommendation regarding the PRA reflects this fact by urging agencies and OMB to make robust use of generic clearances and fast-track approvals for voluntary collections.324) General requests for informal public input via social media are a prime example of a voluntary request that simply does not involve the sort of thing that should require PRA clearance. Participation is truly voluntary, non-participants can free-ride, and the requests are identical in nature to notices that have always been understood to fall outside the PRA, and little or nothing is to be gained by going through the process.

To provide additional clarity, OIRA should amend its “general solicitations” definition to eliminate or expand the reference to “the Federal Register or other publications.” After all, given the (sensible) 2010 Guidance, many general solicitations that fall within the exemption are not found in “publications.”

Recommendation: To provide clarity regarding the application of the Paperwork Reduction Act (PRA) to agency use of social media, the Office of Information and Regulatory Affairs (OIRA) should amend its “general solicitations” definition to eliminate or expand the reference to “the Federal Register or other publications.”

C. The First Amendment

Almost any use of social media to solicit comments and conduct discussion of rulemaking proposals would involve some limitations on public submissions. For example, most agencies using social media at present have comment policies, as does Agencies typically use these policies to prohibit, or at least reserve the right not to post, submissions that:

* contain obscene, indecent, or profane language;
* contain threats, or defamatory statements;
* contain hate speech directed at race, color, sex, sexual orientation, national origin, ethnicity, age, religion, or disability;
* reveal your own or others' sensitive/personal information (e.g., Social Security numbers);
* contain information posted in violation of law, including libel, condoning or encouraging illegal activity, and revealing classified information, or posts that might affect the outcome of ongoing legal proceedings; or
* promote or endorse services or products, including links to external commercial sites.325

Not surprisingly, agencies also require that comments be on topic, though they often still post off-topic submissions, albeit in a separate space often labeled “off-topic.”326 Similarly, one could imagine an agency barring a particularly obstreperous or problematic participant from its Facebook page or an uncooperative participant from a wiki, etc.

Such limitations are indispensable to the effective functioning of rulemaking discussions. However, they raise nontrivial concerns under the First Amendment. Obviously, the government cannot impose a general restriction on speech that it finds “offensive . . . or otherwise objectionable,” as the DHS comment policy does. And in general First Amendment protections on the Internet are robust; it is not a forum such as broadcast television or radio where the inherent characteristics of the medium justify greater government regulation.327 The question is whether such restrictions are permissible in these particular settings.

No court has yet specifically considered the constitutionality of such restrictions, although litigation is pending in federal district court challenging the Honolulu Police Department’s banning one problematic commenter from its Facebook page.328 Two bodies of doctrine might apply in this setting.

First, a good deal of government space on the Internet is devoted primarily if not exclusively to government speech. Agency web sites, for example, are almost entirely a forum for agency speech. Similarly, a Twitter feed is, by definition, simply government speech. When the government is speaking it is not bound by the rules of viewpoint neutrality that apply when it is regulating private speech. Government speakers can take sides; they can present a single perspective; they can be one-sided.329 When the government is “engaging in [its] own expressive conduct, . . . the Free Speech Clause has no application.”330 This “government speech doctrine” would seem to relieve agencies of any obligation to provide a forum for opposing views on their own web sites, and several lower courts have so held.331

Tidy though the government speech approach is, it is not very helpful as applied to social media. Social media platforms of any value to rulemaking cannot be portrayed as instances of government speech. The whole point is that rather than simply being an instance of the agency pushing out information, the site is a space for dialogue, exchange, and the receipt of public input. Social media applications in rulemaking have to involve government listening, not just government speaking. So they present the much harder question of whether an agency, having opened up a space for public debate and exchange—a space whose very purpose is to allow such debate and exchange—can then shut out certain voices. The answer to that question turns on the niceties of public forum doctrine.

Public forum doctrine governs what principles apply to government limitations of private speech on public property. Traditional public forums include parks and sidewalks, the sorts of public spaces that have “immemorially been held in trust for the use of the public and, time out of mind, have been used for purposes of assembly, communicating thoughts between citizens, and discussing public questions.”332 In traditional public fora, First Amendment principles apply with full force: the government can regulate the content of speech only insofar as is necessary to achieve a compelling governmental interest; reasonable time, place, and manner restrictions are permissible if they serve significant interests and leave open ample alternative channels.333

Under prevailing caselaw, social media sites do not qualify as traditional public fora. Given the Supreme Court’s reluctance to analogize new kinds of public spaces such as airports to streets, sidewalks, and parks,334 the Internet seems too recent and unfamiliar for any of its spaces to receive that designation. Several lower courts have rejected arguments that government websites are public fora335 (though, again, what holds for a web site does not necessarily hold for a blog, Facebook page, or other more collaborative online space).

The government has no choice but to treat traditional public fora as such. In contrast, a “designated public forum” “consists of public property which the state has opened for use by the public as a place for expressive activity.”336 Typical examples include theaters, stadiums, university meeting spaces, and other government property that has been opened up for expressive activities. Government limitations of speech in designated public fora are subject to all the same constraints as in traditional public fora, with the major and important exception that because the forum has been voluntarily created by the government, the government is free to shut it down.337

So far so good. It is possible that an agency-sponsored social media site could be found to be a designated public forum. But for that to occur the agency would have had to have opened it up to all comers, to write on all subjects. Agencies would be ill-advised to do that, since it could end up significantly tying their hands. And indeed, as we have seen, few if any have taken that approach. Comment policies limit submissions to items that are on a particular topic, consistent with certain standards of decorum, and consistent with concerns and legal constraints regarding privacy, defamation, and other considerations. In an example of a peculiarity of public forum doctrine, restraints on speech are self-justifying; that is, by restricting permissible speech, the agency ensures that it is creating, if anything, only a “limited public forum.” And within a limited public forum, content-based (though not viewpoint-based) restrictions are permissible.338

The “limited” public forum is a subspecies of the designated public forum that is set up “for a limited purpose such as use by certain groups, . . . or for the discussion of certain subjects.”339 Accordingly, in a limited public forum the state has a much freer hand in limiting speech. The black-letter rule is that restrictions need only be reasonable in light of the purposes served by the forum and viewpoint (as opposed to content) neutral.340 Thus, a school district can limit a public forum to the discussion of a particular topic, such as school board business,341 and a school or university can limit a public forum it establishes to use by student groups, but cannot pick and choose among student groups based on the viewpoint they espouse.342

Finally, some governmental controlled spaces where private speech may occur are not public fora of any sort. A public forum, even if limited, sure sounds like something very different from a nonpublic forum. However, as one commentator has written, the “line between the designated ‘limited’ public forum and the nonpublic forum is maddeningly slippery, and some would even say non-existent, notwithstanding their linguistically opposed labels.”343 This is true along two axes. First, because the whole question of first amendment restrictions would not arise if there were not some private speech taking place, “nonpublic forums” tend to be government spaces with some though limited opportunities for private speech, just like limited public forums. Second, even in a nonpublic forum the government cannot restrict speech as it sees fit; the limitations are essentially those that apply in a limited public forum: restrictions must be reasonable and cannot be based on disagreement with the viewpoint expressed.344

In determining whether an area is a limited public forum or a non-public forum, the two critical factors are historical practice and government intent.

Professor Robert Post has suggested that an important distinction, normatively and descriptively, is between management and governance. The state’s authority over a resource is a matter of management where the resource is internal to the governmental organization and used for institutional ends: an agency conference room, its internal email system, a court-room, a legislative hearing. The state’s authority over a resource is a matter of governance when the resource is external and used by individuals of varying roles and statuses for public discussion and exchange.345 Post’s fancy theorizing has not been explicitly adopted by the courts, but it is grounded on certain intuitions that judges seem to have and can help explain the cases. It also recalls the discussion above about the distinction between citizens and customers. To the extent social media are used as part of the rulemaking process, they are an aspect of governance, not management. As such, they are venues where First Amendment values are very much in play. In addition, precisely because the whole point of these sites (in contrast to a standard agency web site, with its clear and tightly controlled message) is to hear voices other than the agency’s, the government speech doctrine is inapplicable.

The handful of commentators who have investigated this question have concluded that agency social media sites would be classified as limited public forums.346 They are areas that the government has opened up for discussion, allowing—indeed, inviting—all comers to participate and share information and views. These venues are not, however, wholly open-ended; they are set up to discuss specific and limited topics. Accordingly, agencies can restrict the topics, can impose reasonable limits of decorum, decency, and mutual respect, and can shut the whole thing down if it wants. But it cannot deny access to participants because of the particular message or viewpoint they express. Once the forum is created, participation is not a matter of grace.

Recommendation: Agencies should consider First Amendment principles when facilitating or hosting social media discussions. Agencies may define or restrict the topics of discussion, impose reasonable limitations to preserve decorum, decency, and mutual respect, or decide to terminate a social media discussion. Agencies may not, however, deny access to participants based on the viewpoint or message they express.

D. Federal Advisory Committee Act

The Federal Advisory Committee Act (FACA)347 governs the formation and operation of advisory committees by federal agencies. The overall goal is to avoid over-reliance on such committees, block special interest influence via advisory committees, and ensure that committees are balanced, expert, and transparent.348 Agencies can create new advisory committees only after public notice and a determination that doing so is in the public interest. An advisory committee must have a “clearly defined purpose” and a balanced membership and open its meetings to public observation.

Some agency staff have expressed concern that online discussion or consultation with a (self-appointed) group might amount to the use of an advisory committee in violation of the Act.349 The Act defines an advisory committee as “any committee, board, commission, council, conference, panel, task force, or other similar group” established by statute, “established or utilized by the President,” or “established or utilized by one or more agencies, in the interest of obtaining advice or recommendations for the President or one or more agencies or officers of the Federal Government.”350 Read broadly, this definition could conceivably reach social media consultees, who are a “group” that is “utilized” by an agency in order to obtain “advice or recommendations.” (Note that this concern is present in all social media contexts in which public input is sought, not just rulemaking.) The concern seems misplaced, however, for at least three reasons.

First, social media participants are not a “committee.” Commenters in an informal rulemaking have never been understood to be an “advisory committee, and receiving input from a number of individuals in the form of electronic posts or comments rather than written comments does not change the nature of the participants or the process. Commenters would not become an advisory committee simply because there is more back and forth and there may be some de facto collaboration (for example, through a wiki) over time. This is particularly the case because consultation through social media involves input from a shifting, informal, ad hoc group that is quite different from the cohesive group with a defined purpose at which the Act is aimed. Outside the social media context courts have frequently held that amorphous, ad hoc assemblages are not advisory committees.351 Reeve Bull’s study on FACA makes this point, concluding that

agencies can likely exploit many of the recent advances in social media with little concern of running afoul of FACA. For instance, an agency’s receiving comments on its Facebook page or posing a question to the general public via Twitter or a blog and receiving responses thereto is unlikely to trigger FACA, since the agency has not established any formal committee from which it is seeking group advice but instead is simply receiving individual inputs from an amorphous, unorganized assemblage of individuals.352

Second, FACA applies only if the advisory committee is either “established” or “utilized” by an agency. To “establish” a committee, the agency must select its members.353 But when an agency participates in or hosts a social media discussion, it is not selecting the members (to the contrary, it is inviting all comers and engaging with a self-selected group), funding the group, or limiting its participants. In an ordinary-meaning sense, the agency would be “utilizing” the committee (if it is a committee), but the Supreme Court has rejected this “literal,” “straightforward,” “dictionary” reading of the term.354 Instead, an agency “utilizes” a committee it has not created only where “a group [is] organized by a nongovernmental entity but nonetheless so ‘closely tied’ to an agency as to be amendable to ‘strict management by agency officials.’”355 Plainly, voluntary participants in a Web 2.0 forum are not under the agency’s “strict management.”

Third, one of the two basic purposes of FACA was to “ensure that committees operated objectively and were not improperly captured by special interests.”356 Combatting differential access and ensuring a voice for the public is exactly what Web 2.0 platforms do. They do not do it perfectly, of course, but it would be at least ironic, and arguably perverse, if using these tools triggered application of FACA when traditional notice and comment does not. The key point is that a FACA committee is a closed group; some people are one it and some are not, and the agency decides who is who and, thus, who has influence. Because a blog, Facebook, a wiki, or a general discussion forum are open to all, the underlying concerns to which FACA is addressed simply do not arise.

All of this is not to say that it would be impossible for an agency advisory committee to operate through social media, for example by holding an asynchronous virtual meeting..357 But doing so would require steps going far beyond participating in the ordinary discussions that take place in these settings.

E. Ex Parte Communications Policies

Suppose an agency official is reading a blog—it could be her own agency’s blog or something wholly unrelated—on which there is discussion relevant to an ongoing rulemaking. Is that an impermissible ex part contact? Should it be?

Suppose, first, that the decisionmaker in an on-the-record proceeding consults blog posts by a party to the proceeding. That would be the strongest case for finding an impermissible ex parte contact. A recent student note has suggested that blogs discussing pending Supreme Court cases might constitute ex parte communications because of the potential to influence the Court's decision.358 On the other hand, Justice Kennedy has acknowledged, with approval, that his clerks read blogs discussing pending cases.359

The general rule as a matter of the APA and other statutory constraints is that there are restrictions on ex parte contacts in a notice-and-comment rulemaking. The APA’s ex parte provision applies only in formal proceedings.360 The D.C. Circuit flirted with imposing such limitations in the 1970s in the Home Box Office decision.361 However, that case was quickly limited largely to its facts362 and no more recent decision has imposed such limits. However, agencies have adopted their own ex parte contact policies, some of which are quite restrictive. An example is the Department of Transportation’s “Policies for Public Contacts in Rulemaking,” adopted in 1970. In brief, significant influential contacts occurring before issuance of an NPRM should be noted in the NPRM or in a memorandum in the docket; any contacts occurring during the comment period should be noted in the docket (i.e. a summary of the discussion or a copy of anything in writing); and later contacts that might influence the agency must be noted in the docket and, if the comment is significant, the docket may have to reopened to permit a reply. It would seem that these rules would prohibit our hypothetical blog-reader from engaging in such activities.

DOT’s policy can be contrasted with the CFPB’s,363 which is more accepting of some contacts. The CFPB also requires the oral contacts be summarized and included in the docket and written contacts be directly placed in the docket. However, the prohibition kicks in only after the NPRM is published. Second, covered communications are defined as “any written or oral communication by any person outside the CFPB that imparts information or argument directed to the merits or outcome of a rulemaking proceeding.” Some contacts might involve something other than “information or argument” or not go to the merits and so not be covered. Third, the policy expressly excludes “[s]tatements by any person made in a public meeting, hearing, conference, or similar event, or public medium such as a newspaper, magazine, or blog.”364

The Administrative Conference has commenced a research project on ex parte communications in informal rulemaking.365 One aspect of that project is likely to be consideration of agency use of social media, which, as the foregoing suggests, has made the longstanding concerns about such contacts more salient than ever. It is clear that agencies should address social media platforms directly in their ex parte contact policies. Since ACUS is investigating these issues directly, it would be premature to consider just what those policies should consist of in this report.

Development of social media resources for rulemakings would involve agencies in an ongoing conversation, or at least the monitoring of ongoing private conversations, that would seem to violate the DOT policy but be permitted by the CFPB policy.

Recommendation: Agencies should develop appropriate ex parte contact policies that explicitly address the use of social media in informal rulemaking.


This section gathers the recommendations made throughout this report in one place and indicates where the recommendation appears and the pages of supporting discussion.
Recommendation: As part of the rulemaking process, agencies should explore on-line platforms that enable opportunities for public consultation, discussion, and engagement that go beyond merely submitting a written comment to the rulemaking docket.
Read the relevant part of the Consultant's Report: The Promise of Web 2.0; Social Media and The Appeal of Social Media for Rulemaking.

Recommendation: Before using social media in connection with a particular rulemaking, agencies should carefully consider the potential costs and benefits and identify with specificity what they expect to achieve through the use of social media.
Read the relevant part of the Consultant's Report: Realistic Expectations.

Recommendation: The use of social media may not be appropriate and productive in all rulemakings. When deciding whether to use social media in a particular rulemaking, agencies should keep in mind the following principles:
1. Rulemakings that primarily involve questions of statutory interpretation, technical knowledge, or scientific expertise may be poorly suited to the kinds of responses usually produced by social media.
2. On the other hand, social media may be valuable when an agency seeks to ascertain the perceptions or reactions of regulated parties or the public to the proposed rule.
Read the relevant part of the Consultant's Report: Realistic Expectations.

Recommendation: An agency should use the social media tool(s) that best fit its particular purposes and goals and should carefully consider how to effectively integrate those tools into the rulemaking process it would otherwise use.
Read the relevant part of the Consultant's Report: Realistic Expectations.

Recommendation: Agencies should use social media to inform the public about agency activities, the rulemaking process in general, and specific rulemakings. Agencies should take an all-of-the-above approach to alerting potential participants to upcoming rulemakings, posting to its website and sending notifications through multiple channels. Social media provide a more effective means to reach interested persons that have traditionally been under-represented in the rulemaking process.
Read the relevant part of the Consultant's Report: Current and Potential Uses of Social Media in Rulemaking and Outreach.

Recommendation: Agency rulemaking notifications should be directed at both individuals and organizations, be clear and specific about how the proposed rule could affect targeted participants, and include details about the mechanisms and value of public participation.
Read the relevant part of the Consultant's Report: Outreach.

Recommendation: At a minimum, agency notifications regarding rulemakings should be tied to inclusion in the Unified Regulatory Agenda; the addition or change of any item in the Unified Regulatory Agenda should trigger notification via social media.
Read the relevant part of the Consultant's Report: Outreach.

Recommendation: The Regulatory Agenda should be updated continuously, or at least monthly.
Read the relevant part of the Consultant's Report: Outreach.

Recommendation: Agencies should consider using social media during pre-rulemaking or policy development proceedings where the goal is to understand the current state of affairs, collect dispersed knowledge, or identify problems. To enhance the number and value of public contributions, an agency seeking to engage the public for these purposes should, to the maximum extent possible, make clear the sort of information it is seeking, clarify the role of public suggestions, and directly engage with participants by acknowledging submissions, asking follow-up questions, providing substantive responses, and giving credit and recognition to the most valuable submissions.
Read the relevant part of the Consultant's Report: Gathering Open-Ended Public Input Through Social Media.

Recommendation: The General Services Administration, the e-Rulemaking Program Management Office, and other federal agencies, should consider using social media to create and distribute educational programs about rulemaking. These efforts could include: producing videos about the commenting process and posting on an agency website or video-sharing website; hosting webinars in which agency personnel discuss how to draft useful and helpful comments; maintaining an online database of exemplary rulemaking comments; or conducting an online class in which officials review a draft comment and suggest ways to improve it.
Read the relevant part of the Consultant's Report: Education for Effective Commenting.

Recommendation: For each rulemaking, agencies should consider maintaining a blog dedicated to that rulemaking for purposes of providing information, updates, and clarifications regarding the scope and progress of the rulemaking. The blog should include a widget for submission of official comments to the rulemaking docket. In general, the blog should not, however, be used as a tool for extended discussions of substantive questions at issue in the rulemaking,
Read the relevant part of the Consultant's Report: Obtaining Informed and Useful Comments from Diverse Stakeholders.

Recommendation: When soliciting input through a social media platform, agencies should provide a version of the NPRM that is “friendly” to lay users. This involves such steps as breaking preambles into smaller components by subject, summarizing those components in plain writing, layering more complete versions of the preamble below the summaries, and providing hyperlinked definitions of key terms.
Read the relevant part of the Consultant's Report: Discussions.

Recommendation: Agencies should consider, in appropriate rulemakings, retaining facilitator services to manage rulemaking discussions conducted through social media.
Read the relevant part of the Consultant's Report: Discussions.

Recommendation: The information provided through citizen use of social media may be indirect or amount to metadata. Agencies should consider whether the information they need can also be learned not from what is directly communicated by the public, but from what is indirectly revealed by how members of the public interact with the agency online.
Read the relevant part of the Consultant's Report: Direct Outreach.

Recommendation: Agencies should not employ tools through which users can vote on or rank comments submitted in response to a Notice of Proposed Rulemaking.
Read the relevant part of the Consultant's Report: Assessment and Ideation Platforms.

Recommendation: Agencies should realize that not all rulemakings will be enhanced by a crowdsourcing approach. However, where the public or user response is the question to be determined, direct submission to the public at large may provide useful information. In addition, agencies should seek to encourage, and be receptive to, comments from lay stakeholders with “situated knowledge” arising out of their real-world experience.
Read the relevant part of the Consultant's Report: The Crowdsourcing Model.

Recommendation: Agencies should not set up public collaborative drafting sites for the production of regulatory or preambular text or for the production of comments. Agencies should experiment with collaborative drafting platforms internally within the agency for purposes of producing regulatory documents.
Read the relevant part of the Consultant's Report: Collaborative Drafting.

Recommendation: Agencies should consider using social media in support of retrospective review of existing regulations, particularly to learn what actual experience has been under the relevant regulation(s).
Read the relevant part of the Consultant's Report: Retrospective Review of Existing Regulations.

Recommendation: Agencies should consider using social media in connection with direct final rulemaking to quickly identify whether there are significant or meaningful objections not initially apparent.
Read the relevant part of the Consultant's Report: Direct Final Rulemaking.

Recommendation: The APA does not constrain agency use of social media before an NPRM is issued or after a final rule has been promulgated.
Read the relevant part of the Consultant's Report: The Administrative Procedure Act.

Recommendation: When an agency uses social media in connection with a notice-and-comment rulemaking, it has two options for determining how the discussion will be treated under the APA:
The agency may decide to include all comments submitted via social media in the rulemaking record. Agencies should consider using an application programming interface (API) or other appropriate technological tool to efficiently transfer content from social media to the rulemaking record.
The agency may decide that no part of the social media discussion will be included in the rulemaking record, be considered in developing the rule, or be responded to in the final rule. An agency that selects this option should communicate the restriction clearly to the public, provide instructions on how to submit an official comment, and provide a link to the rulemaking docket on or other agency rulemaking portal.
Read the relevant part of the Consultant's Report: The Administrative Procedure Act.

Recommendation: Regardless of which approach an agency pursues pursuant to the previous recommendation, it should communicate clearly to the public how the social media discussion will be used in the rulemaking.
Read the relevant part of the Consultant's Report: The Administrative Procedure Act.

Recommendation: To provide clarity regarding the application of the Paperwork Reduction Act (PRA) to agency use of social media, the Office of Information and Regulatory Affairs (OIRA) should amend its “general solicitations” definition to eliminate or expand the reference to “the Federal Register or other publications.”
Read the relevant part of the Consultant's Report: Paperwork Reduction Act.

Recommendation: Agencies should consider First Amendment principles when facilitating or hosting social media discussions. Agencies may define or restrict the topics of discussion, impose reasonable limitations to preserve decorum, decency, and mutual respect, or decide to terminate a social media discussion. Agencies may not, however, deny access to participants based on the viewpoint or message they express.
Read the relevant part of the Consultant's Report: The First Amendment.

Recommendation: Agencies should develop appropriate ex parte contact policies that explicitly address the use of social media in informal rulemaking.
Read the relevant part of the Consultant's Report: Ex Parte Communications Policies.


Scott Albright, Jessica Bell, Marilyn Kuray, Carl Ann Siciliano, EPA, 1/9/13
Alissa Ardito and Stephen Shaw, Department of Housing and Urban Development, 1/24/13
Ben Balter, GitHub, 5/7/13
Bryant Crowe,, 11/19/12, 10/18/13 (and follow-up email exchanges)
Bridget Dooling, Office of Information and Regulatory Affairs, 11/2/12
Neil Eisner, Department of Transportation, 10/23/12
Cynthia Farina, Cornell Law School and Regulation Room, 10/26/12
James Hupp, Daniel Munz, Lea Mosena, Consumer Financial Protection Bureau, 1/15/13
Joel Kaufman, Federal Communications Commission, 1/9/13
Beth Noveck, New York Law School, 12/5/12
Rebecca Orban, U.S. Coast Guard, 2/19/13


* This report was prepared for the consideration of the Administrative Conference of the United States. The views expressed are those of the author and do not necessarily reflect those of the members of the Conference or its committees.
+ Arthur Kaplan Professor of Law, Benjamin N. Cardozo School of Law, Yeshiva University. I gratefully acknowledge enormously helpful comments on prior drafts by the staff of the Administrative Conference of the United States (especially Emily Bremer, who is the staff counsel for this project), Cynthia Farina, and Jeffrey Lubbers. I also benefitted from research assistance by Cardozo students Timothy White and Candace Chung.
2 The website trumpets: “ removed the logistical barriers that made it difficult for a citizen to participate in the complex regulatory process, revolutionizing the way the public can participate in and impact Federal rules and regulations.”, “About Us,”!about
Program. See also Beth S. Noveck, The Electronic Revolution in Rulemaking, 53 EMORY L.J. 433 (2004); Stephen M. Johnson, The Internet Changes Everything: Revolutionizing Public Participation and Access to Government Information through the Internet, 50 ADMIN. L. REV. 277 (1998).
3 See, e.g., Cynthia R. Farina et al., Rulemaking 2.0, 65 U. MIAMI L. REV. 395 (2011) [hereinafter Rulemaking 2.0]; Cynthia R. Farina et al., Rulemaking in 140 Characters or Less: Social Networking and Public Participation in Rulemaking, 31 PACE L. REV. 382 (2011) [hereinafter Rulemaking in 140 Characters]; Stephen M. Johnson, Beyond the Usual Suspects: ACUS, Rulemaking 2.0 and a Vision for Broader, More Informed and More Transparent Rulemaking, 65 ADMIN. L. REV. 77 (2013).
4 See, e.g., Beth Noveck, Turning Rule Writers Into Problem Solvers: Creating a 21st Century Government That’s Open and Competent by Improving Regulation and Regulatory Review, Cairns Blog (Jan. 26, 2011),; BETH SIMONE NOVECK, WIKIGOVERNMENT: HOW TECHNOLOGY CAN MAKE GOVERNMENT BETTER (2009).
5 Clay Johnson, A More Social Open Government, Expert Labs Blog (Aug. 12, 2011),
7 See Adoption of Recommendations, 60 Fed. Reg. 43,108, 43,109 (Aug. 18, 1995).
8 Id. at 43,110 (¶ IV.A) (emphasis added) (footnote omitted).
9 See Adoption of Recommendations, 77 Fed. Reg. 2257, 2264 (Jan. 17, 2012).
10 Id. at 2265 (¶ 3).
11 Details of the event as well as copies of three of the speakers’ power point presentations are available at
12 Int’l Union, United Mine Workers of Am. v. Mine Safety & Health Admin., 407 F.3d 1250, 1259 (D.C. Cir. 2005).
13 McLouth Steel Prods. Corp. v. Thomas, 838 F.2d 1317, 1325 (D.C. Cir. 1988).
14 Home Box Office, Inc. v. FCC, 567 F.2d 9, 35-36 (D.C. Cir. 1977), quoted in Prometheus Radio Project v. FCC, 652 F.3d 431, 449 (3d Cir. 2011).
15 E.O. 13,563, § 1(a), 76 Fed. Reg. 3821, 3821 (2011).
16 See, e.g., Marissa Martino Golden, Interest Groups in the Rule-Making Process: Who Participates? Whose Voices Get Heard?, 8 J. PUB. ADMIN. RES. & THEORY 245, 245-67 (1998).
17 See, e.g., Neil Eisner, “Policy Direction & Management” (Center for the Study of Rulemaking, Mar. 16, 2005), available at (Department of Transportation official endorsing reply periods and anticipating that they “will be tremendously increased as more agencies have electronic, internet-accessible dockets”).
18 Cary Coglianese, Weak Democracy, Strong Information: The Role of Information Technology in the Rulemaking Process, in GOVERNANCE AND INFORMATION TECHNOLOGY: FROM ELECTRONIC GOVERNMENT TO INFORMATION GOVERNMENT 101, 117 (Viktor Mayer-Schonberger & David Laze eds., 2007).
19 Other enumerations of expected benefits of more open and inclusive policymaking are possible. For an overlapping but slightly different list, see ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT, DIRECTORATE FOR PUBLIC GOVERNANCE AND TERRITORIAL DEVELOPMENT, FOCUS ON CITIZENS: PUBLIC ENGAGEMENT FOR BETTER POLICY AND SERVICES 23-24 (2009). This volume identifies the following anticipated gains:
* Greater trust in government.
* Better outcomes at less cost.
* Higher compliance.
* Ensuring equity of access to public policy making and services.
* Leveraging knowledge and resources.
* Production of more innovative solutions.
20 The Federal Docket Management System is reported to have saved the government $30 million over five years when compared to paper-based docketing. Office of Mgmt. & Budget, Exec. Office of the President, Report to Congress on the Benefits of the E-Government Initiatives 10 (2010), available at
21 See Jeffrey S. Lubbers, A Survey of Federal Agency Rulemakers’ Attitudes About E-Rulemaking, 62 ADMIN. L. REV. 451 (2010). Lubbers asked agency staff about sixteen activities that e-Rulemaking might have made easier or harder than in a paper-based process. Strikingly, respondents reported that each of the sixteen tasks had become easier. The one that scored second highest was: “disseminate information relevant to the agency’s proposed rulemaking (e.g., studies, economic analyses, legal analyses), so as to generate more informed commenters.” Id. at 461.
22 The APIs enable software programs and websites to automatically get data from the portal. They expand the reach of data and allow organizations to use and “repackage” it in various ways. Users of the data APIs include the Sunlight Foundation’s Docket Wrench,, Bloomberg Government,, and the Federal Register’s “Federal Register 2.0” website,, all of which are linked from,!developers;page=showcase.
23 See, e.g., Jerry Brito, Hack, Mash, and Peer: Crowdsourcing Government Transparency, 9 COLUM. SCI. TECH. L. REV. 119 (2008); David Robinson et al., Government Data and the Invisible Hand, 11 YALE J.L. & TECH. 160 (2009); Richard Thaler, This Data Isn’t Dull. It Improves Lives, N.Y. TIMES, March 13, 2011, at B5.
24 Indeed, the task that scored highest in the Lubbers survey—i.e., generated the highest level of agreement—was “Coordinate the rulemaking internally by allowing many people to look at the same rulemaking docket without getting in each others’ way.” Lubbers, supra note 21, at 461.
25 A Department of Transportation staffer reports that in the bad old days “one DOT organization found it necessary to fly a staff member from Boston to Washington, D.C., several days each week just to locate and review docketed material housed throughout the nine separate docket offices.” Christine Meers, Taking Government to the People (unpublished manuscript), quoted in Thomas C. Bierle, Discussing the Rules: Electronic Rulemaking and Democratic Deliberation 14 (April 2003) (Resources for the Future Discussion Paper 03-22), available at
26 Useful overviews include Cary Coglianese, Enhancing Public Access to Online Rulemaking Information, 2 MICH. J. ENVTL. & ADMIN. L. 1 (2012) [hereinafter Coglianese Report]; Rulemaking 2.0, supra note 3, at 417-19.
27 Cary Coglianese, Citizen Participation in Rulemaking: Past, Present, and Future, 55 Duke L.J. 943, 952-58 (2006).
28 See, e.g., Mariano-Florentino Cuéllar, Rethinking Regulatory Democracy, 57 ADMIN. L. REV. 411, 443 (2005) (noting that “individual commenters came across as being angry and exasperated,” “failed to understand the distinction between the regulation and the statute,” and rarely offered “anything remotely resembling a concrete proposal”). Cuellar identified five criteria for what makes rulewriters take comments seriously:
(a) Did the commenter distinguish the regulation from the statutory requirements?; (b) Did the commenter include at least a paragraph of text providing a particular interpretation of, and indicating an understanding of, the statutory requirement?; (c) Did the commenter propose an explicit change in the regulation provided in the notice of proposed rulemaking (NPRM)?; (d) Did the commenter provide at least one example or discrete logical argument for why the commenter's concern should be addressed?; and (e) Did the commenter provide any legal, policy, or empirical background information to place the suggestions in context?
Id. at 431. Not surprisingly, lay commenters generally compare poorly with ones with professional training on these criteria.
29 Stuart W. Shulman, Click-Through Democracy, 20 USA SERVS. INTERGOVERNMENTAL NEWSLETTER 42, 42 (2007).
30 NOVECK, supra note 4, at 138.
31 See Kimberly D. Krawiec, Don’t “Screw Joe the Plummer”: The Sausage-Making of Financial Reform, 55 ARIZ. L. REV. 53 (2013). The Volcker rule is a provision of the Dodd-Frank Act that prohibits banks to engage in proprietary trading or to acquire or obtain an interest in a hedge fund or private equity fund. The complex details were left to several agencies, including the Department of the Treasury, the Federal Deposit Insurance Corporation, the Securities and Exchange Commission, the Office of the Comptroller of the Currency, and the Board of Governors of the Federal Reserve System, to work out through rulemaking. In October 2011, the agencies issued a joint Notice of Proposed Rulemaking. See Prohibitions and Restrictions on Proprietary Trading and Certain Interests in, and Relationships with, Hedge Funds and Private Equity Funds, 76 Fed. Reg. 68,846 (Oct. 11, 2011) (to be codified at 12 C.F.R. pts. 44, 248, 351, and 17 C.F.R. pt. 255). The agencies received over 8,000 comments. See Krawiec, supra note 31, at 72. Of these, about 6,550 were more or less identical, based on a form letter provided by a consortium of public interest groups to their members. Id. The remaining 1,450 comments were submitted by 1,374 distinct commenters; 1,281 (93%) were private individuals. Id. Half of those, in turn, had submitted the same form letter as the other 6,550, with just a sentence or two added or changed. Id. at 21. So that left 515 individual comments by private individuals. Id. at 73. The remaining comments were from industry trade groups (26), asset management firms (16), academics (14), public interest, research, advocacy, and labor groups (12), insurance companies (10), financial institutions (8), and Congress (7). Id. at 85, Table 1.
32 Krawiec, supra note 31, at 58. A similar conclusion is reached by Thomas A. Bryer, Public Participation in Regulatory Decision-Making: Cases from, PUB. PERFORMANCE & MGMT. REV. (forthcoming), available at Bryer reviewed three rulemakings and, like Krawiec, came away distinctly underwhelmed by the quality of lay comments, concluding that “if costs are not accepted to better prepare citizens to be effective participants in the regulatory decision making process, then the democratization experiment might best be called for the facade it is and terminated.” Id. at 24.
33 Steven J. Balla, Public Commenting on Federal Agency Regulations: Research on Current practices and Recommendations to the Administrative Conference of the United States 30-33 (March 15, 2011), available at
34 FCC Rules of Practice, 47 C.F.R. § 1.415(c) (“A reasonable time will be provided for filing comments in reply to the original comments, and the time provided will be specified in the notice of proposed rulemaking.”).
35 Steven J. Balla, Public Commenting on Federal Agency Regulations: Research on Current Practices and Recommendations to the Administrative Conference of the United States 9-10 (2011).
36 Peter M. Shane, Turning GOLD into EPG: Lessons from Low-Tech Democratic Experimentalism for Electronic Rulemaking and Other Ventures in Cyberdemocracy, in Online Deliberation: Design, Research, and Practice 149, 154 (Todd Davies and Seeta Peña Gangadharan eds. 2009).
37 See!aboutProgram.
39 See generally Bryer, supra note 32; Jennifer Shkabatur, Transparency With(out) Accountability: Open Government in the United States, 31 YALE L. & POL’Y REV. 79 (2012); Cynthia R. Farina, Achieving the Potential: The Future of Federal E-Rulemaking, 62 ADMIN. L. REV. 279 (2010) (executive summary of the report of the Committee on the Status and Future of Federal e-Rulemaking).
40 See Administrative Conference of the United States, Request for Proposals: Social Media in Rulemaking 1 (June 8, 2012), available at
41 A list of 50 different definitions can be found at The list compiler’s own definition is: “user-generated content that is shared over the Internet via technologies that promote engagement, sharing and collaboration.”
42 EPA Chief Information Officer, Social Media Policy 5 (June 20, 2011),
43 Three years ago, the National Archives and Records Administration set out one useful typology of social media, with examples (some of which are already incomplete or out of date):
Web Publishing: Platforms used to create, publish, and reuse content.
Microblogging (Twitter, Plurk)
Blogs (WordPress, Blogger)
Wikis (Wikispaces, PBWiki)
Mashups (Google Maps, popurls)

Social Networking: Platforms used to provide interactions and collaboration.
Social Networking tools (Facebook, LinkedIn)
Social Bookmarks (Delicious, Digg)
Virtual Worlds (Second Life, OpenSim)
Crowdsourcing/Social Voting (IdeaScale, Chaordix)
File Sharing/Storage: Platforms used to share files and host content storage.
Photo Libraries (Flickr, Picasa)
Video Sharing (YouTube, Vimeo)
Storage (Google Docs,
Content Management (SharePoint, Drupal)
NARA Bulletin 2011-2, Guidance on Managing Records in Web 2.0/Social Media Platforms (Oct. 20, 2010), available at
45 Id. at 4.
46 5 U.S.C. § 553(c).
47 This report is limited to federal agencies, but the adoption of social media by state and local governments has been if anything more robust than its adoption by federal agencies. For an overview, see MERGEL & GREEVES, supra note 44. This divergence reflects in part considerations of scale. Social media promises that every individual can participate and be heard. Well, there are over 300 million people in the country. The reality is that it would be both an overinvestment of citizen resources and a logistical nightmare for the agencies if all citizens, or a significant portion, sought to engage in a dialogue with federal agencies. There is just no way for a single agency to engage with and listen to each of them that has a beef or a suggestion.
48 Transparency and Open Government: Memorandum for the Heads of Executive Departments and Agencies , 74 Fed. Reg. 4685 (Jan. 26, 2009), available at OMB Director Peter Orszag’s Open Government Directive, issued in December 2009 in response to the Memorandum, sounded the same theme: “Participation allows members of the public to contribute ideas and expertise so that their government can make policies with the benefit of information that is widely dispersed in society.” Office of Mgmt. & Budget, Exec. Office of the President, Memorandum for the Heads of Executive Departments and Agencies: Open Government Directive (Dec. 8, 2009), available at See also Open Gov. Partnership, National Action Plan for the United States of America 1, 9 (Sep. 20, 2011).
49 Joseph Marks, GSA Releases Guidelines to Measure Social Media Impact, Nextgov (Feb. 19, 2013),
50 See
51 U.S. GOV’T ACCOUNTABILITY OFFICE, GAO-10-872T, CHALLENGES IN FEDERAL AGENCIES’ USE OF WEB 2.0 TECHNOLOGIES (2010). A contemporaneous but less thorough report from the National Archives and Records Administration also describes extensive social media by six agencies. NARA, Nat’l Records Mgmt. Program, A Report on Federal Web 2.0 Use and Record Value (2010).
53 The NRC still does not have a Facebook page; it does, however, maintain a Twitter feed, a blog, and a Youtube channel.
54 Cary Coglienese’s December 2011 report to the Administrative Conference, Federal Agency Use of Electronic Media in the Rulemaking Process, found that of the 90 agency websites reviewed, 55 had an RSS feed option, 43 linked to Youtube, 24 to Flickr, 39 to Facebook, and 14 to other social media applications. See Coglianese Report, supra note 26, at 30. Those numbers have surely risen since.
55 All of these are linked from the agency’s social media web page,
56 The “Government and Social Media Wiki” keeps track of government uses of social media. As of September 15, 2013, it lists 157 separate federal administrative entities with some sort of social media presence and provides links to all of them. For general overviews, see INES MERGEL, SOCIAL MEDIA IN THE PUBLIC SECTOR 243-63 (2013) (Appendix listing federal agencies’ social media accounts and tabulating usage in 2010 and 2011); PARTNERSHIP FOR PUBLIC SERVICE, #CONNECTED.GOV: ENGAGING STAKEHOLDERS IN THE DIGITAL AGE (2013); Soon Ae Chun et al., Government 2.0: Making Connections Between Citizens, Data and Government, 15 INFO. POLITY: THE INT'L J. OF GOV'T & DEMOCRACY IN THE INFO. AGE 1 (2010).
57 See
58 The Department of Energy reportedly has established an island in Second Life in order to reach potential employees.
59 U.S. Coast Guard, Social Media and the U.S. Coast Guard: Right Tool … Right Level … Right Audience 1 (2011).
60 See
61 See
62 See
63 See
65 Dep’t of State, Open Government Plan (Version 2.0) 19 (Apr. 9, 2012), available at
documents/organization/188085.pdf. These numbers are somewhat suspect, in that the 2010 version of the plan and the 2012 version both state that the blog “recently passed 15,000,000 page views.”
66 Id.
67 These figures appear on the blog’s “comment policy” page.
68 The term “managerial” is from Andrew Chadwick and Christopher May, Interaction Between States and Citizens in the Age of Internet: “e-Government” in the United States, Britain, and the European Union, 16 GOVERNANCE 271 (2003). The authors distinguish “managerial,” “consultative,” and “participatory” approaches to online interactions between the state and its citizens. See id.
69 See Accenture Institute for Health and Public Service Value, From e-Government to e-Governance: Using new technologies to strengthen relationships with citizens (2009).
71 Id. at 63. See also id. at 81 (“Only about three of every hundred site visits is to a news and media Web site. Slightly more than one site visits in a thousand is to a political Web site. Pornographic content is two orders of magnitude more popular than political content.”)
73 See Peter Muhlberger, Should E-Government Design for Citizen Participation? Stealth Democracy and Deliberation, Proceedings of the 2006 international conference on Digital government research (arguing that online deliberation will itself stimulate citizen engagement and “ameliorate” “stealth democracy beliefs”).
74 Alex Howard, relaunches with APIs, integrates social media, hopes for public participation, gov20.govfresh (Feb. 21, 2012),
75 The table is taken from Cynthia R. Farina & Mary J. Newhart, Rulemaking 2.0: Understanding and Getting Better Public Participation 14 (IBM Center for The Business of Government 2013) [hereinafter CERI 2013 Report]. It can also be found in Professor Farina’s power point presentation from the September 17 social media workshop, which is available at
76 Strictly speaking, a power law distribution is present if frequency—for example, how often a particular event occurs—varies as a power of some attribute of that event—for example, its size.
77 See generally CLAY SHIRKEY, HERE COMES EVERYBODY 122-30 (2009).
78 See generally CASS R. SUNSTEIN, REPUBLIC.COM 2.0 (2009).
79 Andrew Beaujon, Huffington Post deletes 75 percent of incoming comments, POYNTER, (Aug. 27, 2013).
80 HINDMAN, supra note 70, at 133.
81 See Adam Connor, Ten Quick Tips for using Facebook, available at
82 Id. at 6:30.
83 Hamish McKenzie, The Verge and the Huffington Post Attempt the Impossible: Making Comments Smarter, Pandodaily (Feb. 1, 2013),
84 Jeff Sonderman, How the Huffington Post handles 70+ million comments a year, POYNTER, (Oct. 23, 2012).
85 Cynthia R. Farina et al., Rulemaking vs. Democracy: Judging and Nudging Public Participation that Counts, 2 MICH. J. ENVTL. & ADMIN. L. 123, 146-47 (2012).
87 CERI 2013 Report, supra note 75, at 35.
88 Bridget C.E. Dooling, Legal Issues in E-Rulemaking, 63 ADMIN. L. REV. 893, 901 (2011).
89 See Lubbers, supra note 21, at 473.
90 Old petitions are not maintained on the site, so this is no longer visible.
91 The full text reads: “We petition the President of the United States to finally admit what we all know; That this platform is utterly useless, and that the responses the President provide are only given to trick people into making them think that the President actually cares about them in any way.”
92 CERI 2013 Report, supra note 75, at 40.
93 Coglianese Report, supra note 26, at 117.
95 See See generally Coglianese Report, supra note 26.
96 As then Associate Administrator Lisa Heinzerling put it in a blog post (expressing her own views and not those of the agency, as all EPA posters do on the Greenversations blog) when the gateway first went up, “This is a new web site that makes EPA’s rulemaking process more transparent and easier to follow. It gives you the tools to understand how you can get involved in EPA’s priority rulemakings, how a rulemaking might affect you, and where each rule falls in our rulemaking process.” See Heinzerling’s post mentions a discussion forum on the Gateway; that seems no longer to exist (if it ever did).
97 See, e.g., Small Business Regulatory Alert: Deadline Extended for Comments on Passenger Vessel Accessibility Guidelines, The Small Business Watchdog (Aug. 14, 2013), (blog post); Regulatory Alerts: FDA Issues Two Draft Rules on Safety of Imported Food, The Small Business Watchdog (Aug. 8, 2013), (blog post).
98 The video can be viewed at
99 See
100 The maiden post can be read at
101 The project’s useful self-description is available at See also Rulemaking in 140 Characters, supra note 3, at 388-93.
102 Rulemaking 2.0, supra note 3, at 417-18.
103 Id. at 417-19.
104 Regulation Room, Home Mortgage Consumer Protection, Final Summary § 2,
105 CERI 2013 Report, supra note 75, at 39.
106 5 U.S.C. § 602(a).
107 ABA Section of Administrative Law & Regulatory Practice, A Report for the President-Elect of the United States, 52 ADMIN. L. REV. 1099, 1104-05 (2000).
108 The portion of the agenda mandated by the Regulatory Flexibility Act must, under that statute, be published in the Federal Register. 5 U.S.C. §602(a). Accordingly, that portion continues to be produced in hard copy.
109 The ABA’s Administrative Law Section has endorsed, and has lobbied for, for an evergreen Unified Agenda. Former EPA official Rick Otis also championed this idea at ACUS’s September 17, 2013 Social Media Workshop. Related legislation is pending in Congress. See H.R. 2804, All Economic Regulations are Transparent Act of 2013 (113th Cong.) (requiring, among other things, monthly Unified Agenda submissions by agencies to OIRA, and monthly Internet posting thereof by OIRA).
110 Here is EPA’s press release about the competition:
The U.S. Environmental Protection Agency (EPA) and the eRulemaking program have partnered to sponsor the Rulemaking Matters! video contest. The contest will highlight the significance of federal regulations and help the public understand the rulemaking process.
Federal agencies develop and issue hundreds of rules and regulations every year to implement statutes written by Congress. Almost every aspect of an individual’s life is touched by federal regulations, but many do not understand how rules are made or how they can get involved in the process.
This video contest is an opportunity for everyone to learn more and participate in an open government. With a short 60 to 90 second video, citizens should capture public imagination and use creativity, artistic expression and innovation to explain why regulations are important to everyone, and motivate others to participate in the rulemaking process.
Individuals and groups of all ages may participate. Entries must be received by May17, 2010. The winner will be awarded $2,500, and their video posted on the and EPA Web sites.
Press Release, EPA Announces Video Contest on Rulemaking (April 15, 2010),
111 The Rulemaking Matters! Mosaic,
114 See Kevin Allocca, Why videos go viral (TED Talk, Nov. 2011), (identifying three factors that result in a Youtube video going viral: endorsement by “tastemakers” who share with a large community, chance for others to participate by doing parodies or different versions, and content or presentation that is unique and unexpected and humorous).
115 Another example: as of October 2013, the TSA has posted 84 videos to Youtube; for the most part, these explain and justify its airport screening procedures. See As government videos go, these have an enormous number of views; three have been seen more than 100,000 times; 18 more than 10,000 times. But dwarfed by all the anti-TSA videos posted by news outlets or individuals who are furious at the agency. These are more numerous by orders of magnitude, and many have orders of magnitude more views. Inescapably, more people will watch “TSA Agent Touches my Vagina at San Diego International Airport,” or 100% foolproof solution to stop TSA from stealing your valuables out of your carry-on bag,” or “TSA Agent Found With ABC IPad,” than will watch “Why Shoes on the Belt?”or “TSA-Choose Your Lane.”
116 See Negotiated Rulemaking Act of 1990, Pub. L. No. 101-648, 104 Stat. 4969, codified at 5 U.S.C. §§ 561-70.
117 Connor, supra note 81, at approx. 4:30, slides available at
118 “Well done!”
119 “Ban the EPA!” “Let’s not appreciate the commie EPA propaganda!” “Another waste of taxpayer money, folks. Pathetic.”
120 “I think I would be an asset to the EPA. I'll be seeking employment soon. I'm graduating with an MPA this spring.” (This was posted as a comment to an EPA video about LGBT employees at the agency, which some commenters attacked as itself pretty off topic.) In response to a post about grocers donating unsold food to food banks: “No GMO's. stop fracking now! Leftover food is the least of your worries. Our country is poisoned by Monsanto daily!”
121 The TSA YouTube channel is at
122 I limit this section to input about policy because that is what rulemaking involves. Social media can also be useful in gathering other sorts of information from the public. Most obviously, it is a mechanism for citizens seeking assistance to alert government officials to something requiring attention. There are numerous examples at the state and local level. For example, many municipalities have had success with “clickfix,” a platform that allows residents to alert local officials of conditions such as potholes, uncollected garbage, nonworking traffic lights, stray wildlife, etc. See, e.g., (Burlington, Vermont clickfix site); see generally (clickfix home page, with general information and links to city-specific sites).
123 Nick Sinai, U.S. Deputy Chief Technology Officer, We Want Your Input on Building a More Open Government, White House Open Government Initiative Blog (Sep. 3, 2013),
124 The Open Government Partnership, National Action Plan for the United States of America 7 (Sep. 20, 2011).
125 Lisa Ellman and Todd Park, Open Government: A Time for Self-Assessment, White House Open Government Initiative Blog (March 29, 2013),
126 Aneesh Chopra and Cass Sunstein, Open Government and the National Plan, White House Blog (Aug. 8, 2011),
127 See Sinai, supra note 123.
128 See, e.g., Alex Howard, White House Asks for Feedback on Second National Action Plan for Open Government, E Pluribus Unum Blog (Sep. 6, 2013),
One 2011 observer, drawing on Sunstein’s own academic writing, offers this analysis:
[T]he problem is that most of the tools available (e.g. ranking, ideation) do not meet reasonable standards of good “choice architecture”, to use Sunstein’s terms. One might imagine that as Sunstein went through the different options available, he foresaw all the effects that could be generated by the tools and their design: reputational cascades, polarization, herding… In the end, the only remaining alternative, although unexciting, was e-mail. In this case at least, preferences are independently aggregated, and the risks of informational and social influence are mitigated.
Tiago Peixoto, Why Cass Sunstein Will Take E-Mails Only (Aug. 9, 2011),
129 See (noting that the agency was considering adding two chemicals to the TRI list; no comments as to either).
130 See
131 Because the site has been retired, the URLs given in the footnotes in this section no longer work. It does not appear that the material from Exchange is available at a current on-line location. When in operation, Exchange could be found at
132 See
133 These include the following, which is the sort of thing that might bring an agency official to tears and highlights just how badly this sort of thing can work. EPA asked: “If EPA published a rule requiring mandated electronic reporting, what effect would this have on your state or local government, on the public or on the federal government?” That question received one response: “I am very sure that the effects will pronounced more on both sides but i guess it is debatable. It will be interesting to see what others' view point is on the electronic reporting effects on the public and the government.”
134 EPA Interview, 1/9/13.
135 See
136 Anonymous responded: “Why is TSA employee ‘TSORon’ allowed to insult American citizens on this gov't website?” And frequent poster “Wintermute” wrote: “Ahh... there’s the insults we’ve come to expect. I’ll put my knowledge of security up against any TSO’s knowledge any day, even though I’d hardly call myself an expert (just a geek with some interesting hobbies).” Id.
137 See
138 FCC, FCC Explains Relationship of Blogband to the Record in the National Broadband Plan Proceeding, 29 FCC Rcd. 11999 (Sep. 22, 2009).
139 Id.
140 See
141 See
142 See
143 See
144 See
145 See
146 See (archived page).
147 See
148 See Leonard J. Kennedy, Patricia A. McCoy, and Ethan Bernstein, The Consumer Financial Protection Bureau: Financial Regulation for the Twenty-First Century, 97 CORNELL L. REV. 1141, 1158-59 (2012).
149 Gov’t Servs. Admin., OpenGov Citizen Engagement Tool: Frequently Asked Questions for Federal Agencies, No. 2.a,
150 See the list at
151 In addition to the examples that follow, see, e.g., U.S. Dep’t of State, Share Your Ideas, OPENSTATE,; U.S. Dep’t of State, Opinion Space: A Universe of Viewpoints and Ideas,; U.S. Nuclear Regulatory Comm’n, Share Your Ideas, OPENNRC,
153 See
154 A separate EPA ideation effort is U.S. Envtl. Prot. Agency, Share Your Ideas, OPENEPA,, which seems to have generated less interest, participation, and valuable submissions.
The failure to generate any new ideas is typical. For example, the Department of Veterans’ Affairs set up an IdeaScale site tied to producing its Open Government Plan. The site produced 186 ideas and 3084 votes from 618 users. After the site closed, the Department produced a document setting out and responding to the ten most popular ideas. Dep’t of Veterans Affairs, VA’s Open Government Plan: The Top 10 Ideas Received from the Public (n.d.), available at Although the rhetoric of this document celebrates the enormous value of the suggestions, the reality of the responses indicates that the agency was already familiar with the complaints (e.g. long hold times on telephone calls) and proposed remedies; the touted efforts to improve were evidently in the works already, wholly independent of the IdeaScale process.
155 Links are collected at the commission’s “discuss” page. Note that the agency has shifted from using IdeaScale to a similar ideation platform, Uservoice. See, e.g.,
forums/105541-a-new-fcc-gov-feedback (general feedback on and suggestions for the FCC web site).
156 See
157 See
158 See
159 FCC, FCC Explains Relationship of IdeaScale Postings to the Record in the National Broadband Plan Proceeding, DA 10-42 (Jan. 11, 2010), available at
160 See
161 IdeaScale, FCC: National Broadband Plan Government Case Study (2011). The study also asserts that the Broadband IdeaScale site logged 60,000 responses from the public, a number that appears to be just wildly off.
162 For example, a Minnesota GIS Manager used his LinkedIn page to urge people to vote for a particular idea on FEMA’s IdeaScale site. The follow-up posts suggest the effort paid off. Two days after the initial post he noted “Our votes keep going up. We are ranked 9th so far.” Ten days after that: “We are only three votes away from being the number one ranked idea.” And two weeks later: “Ranked number 1 for over a week. Vote count is now at 93 and continues to climb! Thanks for your support. If you have not voted, please do so.”
163 Innovation Tools Subgroup, White House Innovation and Information Policy Group, Electronic Management Tools for New Idea Generation in the Federal Workplace (Sep. 2009).
164 Id. at 18-22.
165 See do I prepare effective comments?. provides a similar guide. See, Tips for Submitting Effective Comments,; see also FCC, Rulemaking Process at the FCC, FCC Encyclopedia,; FDA, The Importance of Public Comment to the FDA,; NOAA Fisheries Service, Alaska Regional Office, Tips for Submitting Effective Public Comments,
167 Crowe interview, 10/17/13; Crowe presentation, ACUS Social Media Workshop, 9/17/13; The Open Government Partnership: Government Self-Assessment Report for the United States of America 31 (March 29, 2013).
168 See National Forest System Land Management Planning, 76 Fed. Reg. 8480 (Feb. 14, 2011) (to be codified at 36 C.F.R. pt. 219).
169 See National Forest System Land Management Planning, 77 Fed. Reg. 21,162 (Apr. 9, 2012) (to be codified at 36 C.F.R. pt. 219).
170 See
171 See
172 See (schedule of 2011 public meetings on the planning rule).
173 See
174 See
175 Adoption of Recommendations, 77 Fed. Reg. 2257, 2265 (¶ 3) (Jan. 17, 2012). The legal questions are addressed in Part VI, infra.
176 CERI 2013 Report, supra note 75, at 28.
177 See supra Part V.B.3.
178 Under the Negotiated Rulemaking Act, a “person designated to represent the agency in substantive issues may not serve as facilitator or otherwise chair the committee.” 5 U.S.C. § 566(c).
179 See, e.g., Help a Govie - Dealing with Legal Issues about Online Moderation Policy, GovLoop Blog (Nov. 12, 2012), (blog post from an official at a local government agency lamenting that her legal team had advised that “that if a government entity allows the public to post content, any resulting public content or commentary may not, in any way, be moderated”). The First Amendment issues are addressed in Part VI.C., infra.
180 See supra at Part V.A.2.
181 Rulemaking 2.0, supra note 3, at 418.
182 Regulation Room is built on WordPress 3+, an open-source blogging tool. The ability to attach comments to specific sections of a post or document results from adding, which the site builders consider “the most important” of the open-source plugins they have added to WordPress. Rulemaking 2.0, supra note 3, at 412.
183 Id. at 391.
184 See, e.g., CERI 2013 Report, supra note 75, at 33-35.
185 See Coglianese Report, supra note 26.
186 Id. at 43.
187 Integrated Mortgage Disclosures Under the Real Estate Settlement Procedures Act (Regulation X) and the Truth in Lending Act (Regulation Z), 77 Fed. Reg. 51,116 (2012) (to be codified at 12 C.F.R. pts. 1024 & 1026).
188 The whole process and its results are detailed in an exhaustive report from the outside consultant who assisted the Bureau. See Kleimann Communication Group, Inc., Know Before You Owe: Evolution of the Integrated TILA-RESPA Disclosures (July 9, 2012), available at
189 CFPB Interview, 1/15/13.
190 An image of such a heat map can be seen at
191 Mortgage Disclosure is Heating Up, CFPB Blog (June 24, 2011),
mortgage-disclosure-is-heating-up/. The post observes:
Our feedback tool recorded where users clicked as they reviewed the draft disclosures. A heatmap is a way of displaying those clicks as a graphic that shows which areas were clicked on most. Simply put, it’s a way for us to see, at a glance, what areas of our draft disclosures attracted the most and least attention. . . .
So, what can this image tell us? Here are a few highlights:
• Respondents were interested in the bottom line. The full loan amount at the top of the page, the projected payments section at the bottom of the page, and the estimated closing payment on the second page all received a lot of clicks.
• People had a great deal to say about the “Key Loan Terms” and “Cautions” sections.
• People commented on the first page of the draft form much more than on the second. This is a pretty common occurrence, and on its own, it serves as helpful advice for our designers about where to put certain important information. But the information on the second page (like closing costs, for example) is also an essential part of mortgage disclosure. That’s why the next round of testing will focus on the second page.
192 See Justin Herman, Social Media Metrics for Federal Agencies, Howto.Gov (April 19, 2013),; see also Justin Herman, Presentation at the Social Media Workshop (Sep. 17, 2013).
193 Preserving the Open Internet, Broadband Industry Practices, 74 Fed. Reg. 62,638 (Nov. 30, 2009) (to be codified at 47 C.F.R. pt. 8).
194 See
195 See
196 Preserving the Open Internet, 76 Fed. Reg. 59,192, 59,192 (Sep. 23, 2011) (to be codified at 47 C.F.R. pts. 0 & 8).
197 In the Matter of Preserving the Open Internet, Broadband Industry Practices, FCC 10-201, Report And Order, 25 FCC Rcd. 17905, 18001 n. 501 (rel. Dec. 23, 2010).
198 Farina and Newhart reach the same conclusion. CERI 2013 Report, supra note 75, at 27, 39.
200 More refined taxonomies are possible. Daren Brabham identifies four categories:
(1) Knowledge Discovery and Management,
(2) Distributed Human Intelligence Tasking,
(3) Broadcast Search (i.e. seeking solutions to “eureka” problems with empirically provable solutions, as through challenges), and
(4) Peer-Vetted Creative Production, as, for example, in having a group design a structure or art project.
DAREN C. BRABHAM, USING CROWDSOURCING IN GOVERNMENT 11 (2013). This is not inconsistent with the two-category account: items 1 and 3 are two types of information-gathering; items 2 and 4 are two types of task-sharing.
201 These efforts described in Partnership for Public Service, Engaging Stakeholders in the Digital Age 16-17 (2013).
202 See Pamela Licalzi O'Connell, Mining the Minds of the Masses, N.Y. TIMES, March 8, 2001, at G1.
203 New Zealand Ministry of Economic Development, The Digital Strategy 2.0 11 (2008).
204 See NOVECK, supra note 4, at 12-14, 48-49;
205 In addition to Professor Noveck’s own writing, see BRABHAM, supra note 201, at 13-14.
206 15 U.S.C. § 3719.
207 The website gathers in one place challenges from all federal agencies.
208 See generally Office of Science and Technology Policy, Implementation of Federal Prize Authority: Progress Report (March 2012).
209 Tens of thousands of municipalities use seeclickfix or equivalent platforms that allow individuals easily to go online and report non-emergency conditions needing attention. These are essentially web-based versions of 311 systems. See generally
211 See generally 16 C.F.R. pt. 1102 (establishing the rules governing the CPSC’s publicly available consumer product safety information database).
212 See
213, Public Comments Make a Difference, available at
215 BRABHAM, supra note 201.
216 See supra note 189 and accompanying text.
217 Rulemaking 2.0, supra note 3.
218 Id.; CERI 2013 Report, supra note 75, at 16.
219 Steven VanRoekel, Acting OMB Deputy Director for Management and U.S. Chief Information Officer, Guidance for 2013 SAVE Award Program, Memorandum for the Heads of Departments and Agencies (July 26, 2013).
220 See INES MERGEL, USING WIKIS IN GOVERNMENT: A GUIDE FOR PUBLIC MANAGERS 13-14 (IBM Center for the Business of Government 2011); Chris Bronk & Tiffany Smith, Diplopedia Imagined: Building State’s Diplomacy Wiki, 2010 INT’L SYMP. ON COLLABORATIVE TECHNOLOGIES AND SYSTEMS 53.
221 For an enthusiastic description, see Jason Miller, 4 Studies in Collaboration: Case 1: Intellipedia, FED. COMPUTER WEEK (Feb. 29, 2008)
223 Lexpop was a successor to the Write the Bill Wiki, In late 2010 and early 2011, 15 different legislative topics were posted to the wiki. Topics included climate change, tax reform, net neutrality, and several education policy topics. The hope was that participants would discuss and actually draft legislation on these topics. In actuality, only a handful of people participated, the postings were skimpy and general, and the project really never got off the ground.
225 Have a few minutes for a gov 2.0 experiment? (March 14, 2011),
226 Online Protection and Enforcement of Digital Trade Act, H.R. 3782, 112th Cong. (2012). The Madison Project site is at
228 See, e.g., (site with two proposed bills that have generated some comments).
229 See, e.g., Vasiliy Burov, Evgeny Patarakin & Boris Yarmakhov, A Crowdsourcing Model for Public Consultations on Draft Laws, PROCEEDINGS 6TH INT’L CONF. THEORY & PRACTICE OF ELECTRONIC GOVERNANCE 450 (2012) (reporting on successful uses of wikivote software to draft Russian laws on education and on fisheries).
231 Clay Shirky, How the Internet will (one day) transform government (2012), The talk has approximately 790,000 views as of this writing.
232 Ben Balter, CFPB Accepts First Citizen-Submitted Code on Behalf of Federal Government (April 15, 2012),; Robert McMillan, How GitHub Helps You Hack the Government, Wired (Jan. 9, 2013), (“For the first time, the Consumer Protection Bureau was accepting a direct change to one of its internal documents not from someone inside the agency but from an average citizen somewhere across the country.”).
233 Balter interview.
234 Shirky, supra note 232, at 130-35.
235 Mariake Guy, Wiki or Won’t He?: A Tale of Public Sector Wikis, ARIADNE Issue 49 (Oct. 30, 2006), (“Before you even begin to develop a wiki you need to know your audience and think about whether this is what they want. Are they a group who can and will converse? The likelihood is that if they cannot converse offline then they probably will not online.”).
236 Michael Vallance, Phillip A. Towndrow, & Charles Wiz, Conditions for Successful Online Document Collaboration, 54 TECHTRENDS 20, 20 (2010).
237 Exec. Order No. 13,563, 76 Fed. Reg. 3821 (Jan. 18, 2011); see also Exec. Order No. 12,866; Exec. Order No. 12,044, § 4; Exec. Order No. 12,291.
238 5 U.S.C. § 610.
239 12 U.S.C. § 5512(d).
240 Eisner & Kaleta, supra note 308, at 147-48.
241 The (Not So) New Executive Order on Regulatory Review, and What to Expect, 41 Envtl. L. Rep. 10505, 10509 (2011) (remarks of Susan Dudley) [hereinafter Dudley Remarks].
242 For a rosy account of the first year or so of the effort, see Executive Office of the President, Council of Economic Advisers, Smarter Regulations Through Retrospective Review (May 10, 2012).
243 Dudley Remarks, supra note 242, at 10509 (stating that to do retrospective review well, “we really need to change the incentives regulatory agencies face”); Eisner & Kaleta, supra note 308, at 148, 155.
245 Id.
246 EPA Open Government Plan at 52.
247 See The site is no longer up.
248 U.S. Dep’t of Labor, Plan for Retrospective Analysis of Existing Rules 5 (August 2011).
249 Id. at 9.
250 Dep’t of Transportation, Plan for Implementation of Executive Order 13563: Retrospective Review and Analysis of Existing Rules 11 (August 2011).
251 Id. at 4.
252 Id. at 5.
253 See id. at 25-189 passim (Attachment I). The Department of Labor has also shown some enthusiasm for such an approach.
254 Office of Mgmt. & Budget, Exec. Office of The President, 2006 Report to Congress on the Costs and Benefits of Federal Regulations and Unfunded Mandates on State, Local, and Tribal Entities 89 (2006), available at The report does not indicate exactly what steps went into this “more ambitious” outreach.
255 Administrative Conference of the United States, Recommendation 95-4, Procedures for Noncontroversial and Expedited Rulemaking ¶ I(A) (1995).
256 See, e.g., 14 C.F.R. § 11.31 (FAA DFRM procedures). See generally Ronald M. Levin, Direct Final Rulemaking, 64 Geo. Wash. L. Rev. 1 (1995).
257 Lars Noah, Doubts About Direct Final Rulemaking, 51 ADMIN. L. REV. 401, 410 (1999).
258 Stephen Johnson, Ossification’s Demise? An Empirical Study of EPA Rulemaking from 2001-2005, 38 ENVTL. L. 767 (2008).
259 See, e.g., Michael Kolber, Rulemaking Without Rules: An Empirical Study of Direct Final Rulemaking, 72 ALB. L. REV. 79 (2009) (finding that in a ten-year period the FDA withdrew fully forty percent of its direct final rules when they generated significant adverse comments).
260 Federal Web Managers Council, Social Media and the Federal Government: Perceived and Real Barriers and Potential Solutions (Dec. 2008), available at
Fed%20Govt_BarriersPotentialSolutions.pdf [hereinafter FWMC].
261 Id. at 4. See also Ethan Klapper, Social Media in Rulemaking is a No-No Because of 1946 Law, (Feb. 23, 2009) (blog post).
262 FWMC, supra note 261, at 4.
263 Eisner interview; EPA interview; HUD interview; Coast Guard interview. See also Alex Moll,, Facebook Post (Oct. 3, 2011), (describing Facebook page as “a place to discuss with fellow participants (and possibly government officials if we can work with policy leaders to address constraints within the Administrative Procedure Act policies and process) the substance of Federal regulations or to submit questions for response”) (emphasis added).
264 See supra notes 173-174 and accompanying text.
265 FCC, Comment Policy,
266 Id. The CFPB comment policy is similar, stating that for “Public comment periods for proposed regulatory actions [visitors should] Use our Notice and Comment page. Blog comments will help drive our conversation, but they cannot be considered formal public comments on proposed actions.”
267 5 U.S.C. § 553(c).
268 See Portland Cement Ass’n v. Ruckelshaus, 486 F.2d 375 (D.C. Cir. 1973); RICHARD J. PIERCE, SIDNEY A. SHAPIRO & PAUL R. VERKUIL, ADMINISTRATIVE LAW AND PROCESS 333-38 (5th ed. 2008).
269 See, e.g., Alabama Power Co. v. Costle, 636 F.2d 323, 384 (1979) (“[T]he opportunity to comment is meaningless unless the agency responds to specific points raised by the public.”).
270 5 U.S.C. § 553(c).
159 Louisiana Federal Land Bank Ass'n, FLCA v. Farm Credit Admin., 336 F.3d 1075, 1080-81 (D.C. Cir. 2003).
272 Public Citizen, Inc. v. Fed. Aviation Admin., 988 F.2d 186, 197 (D.C. Cir.1993).
273 Am. Mining Cong. v. United States EPA, 907 F.2d 1179, 1188 (D.C. Cir.1990).
274 5 U.S.C. § 553(c).
275 David L. Franklin, Two Cheers for Procedural Review of Guidance Documents, 90 TEX. L. REV. 111, 122 (2012).
276 Adoption of Recommendations, 59 Fed. Reg. (¶ IV.D) (Feb. 1, 1994), available at
277 407 F.2d 330, 338 (D.C. Cir. 1968).
278 Indep. U. S. Tanker Owners Comm. v. Lewis, 690 F.2d 908, 919 (D.C. Cir. 1982).
279 5 U.S.C. § 553(c).
280 See, e.g., Public Citizen, Inc. v. FAA, 988 F.2d 186, 197 (D.C.Cir.1993) (“The requirement that agency action not be arbitrary or capricious includes a requirement that the agency ... respond to ‘relevant’ and ‘significant’ public comments.”).
281 See, e.g., Thompson v. Clark, 741 F.2d 401, 409 (D.C. Cir. 1984) (“The failure to respond to comments is significant only insofar as it demonstrates that the agency's decision was not based on a consideration of the relevant factors.”).
282 The leading example is Jack M. Beermann and Gary Lawson, Reprocessing Vermont Yankee, 75 GEO. WASH. L. REV. 856 (2007).
283 Indep. U.S. Tankers, 690 F.2d at 919 (“The statement of basis and purpose ‘should enable us [i.e. the reviewing court] to see what major issues of policy were ventilated . . . and why the agency reacted to them as it did.’”) (quoting Automotive Parts & Accessories Ass’n v. Boyd, 407 F.2d 330 (1968)).
284 NRDC v. U.S. EPA, 859 F.2d 156, 188 (D.C.Cir.1988) (per curiam).
285 See, e.g., Citizens for Health v. Leavitt, 428 F.3d 167, 186-87 (3d Cir. 2005) (rejecting argument that agency failed adequately to respond to comments, finding that the agency’s overall explanation was adequate and noting, though in passing, “the large volume of public comments”).
286 See City of Waukesha v. EPA, 320 F.3d 228, 257-58 (D.C. Cir. 2003).
287Thompson v. Clark, 741 F.2d 401, 409 (D.C. Cir. 1984), quoted in Texas Mun. Power Agency v. EPA, 89 F.3d 858, 876 (D.C. Cir. 1996); accord City of Waukesha v. EPA, 320 F.3d 228 (D.C. Cir. 2003); Am. Iron & Steel Inst. v. EPA, 115 F.3d 979, 1005 (D.C. Cir. 1997) (finding comment response sufficient if it “demonstrates that the agency at least considered whether it should adopt [an alternative] model”).
288 Grand Canyon Air Tour Coal. v. FAA, 154 F.3d 455, 468 (D.C. Cir. 1998). Of course, every comment must be read, if only to determine which ones are “relevant and significant.” In a rulemaking with thousands or tens of thousands of comments, this is not a trivial burden. Some agencies have relied on software to screen comments, at least to separate out repeats from unique submissions.
In Recommendation 2011-1, Legal Considerations in e-Rulemaking, the Administrative Conference endorsed use of reliable comment analysis software. See Adoption of Recommendations, 76 Fed. Reg. 48,789, 48,790 (Aug. 9, 2011). As Bridget Dooling’s report to the conference explained:
Software that uses natural language processing is one promising technology, because it could help staff identify duplicate comments, providing confidence that all unique comments and personalized portions of partially duplicative comments are considered efficiently. This time-saving approach does not diminish agency consideration because it would still give agencies access to the number and content of all comments received.
Dooling, supra note 88, at 902.
289 46 F.3d 102 (1st Cir. 1995).
290 Id. at 110.
291 42 U.S.C. §7607(d)(6)(B).
292 5 U.S.C. §604(a)(2).
293 2 U.S.C. § 1532(a)(5)(C).
294 49 C.F.R. §106.65.
295 The 1947 Attorney General’s Manual on the APA explained:
Each agency is affirmatively required to consider “all relevant matter presented” in the proceeding; it is recommended that all rules issued after such informal proceedings be accompanied by an express recital that such material has been considered. It is entirely clear, however, that section [553(c)] does not require the formulation of rules upon the exclusive basis of any “record” made in informal rule making proceedings. Senate Hearings (1941) p. 444. Accordingly, . . . an agency is free to formulate rules upon the basis of materials in its files and the knowledge and experience of the agency, in addition to the materials adduced in public rule making proceedings.
296 See generally Administrative Conference Recommendation 2013-4, The Administrative Record in Informal Rulemaking (June 14, 2013), available at
297 Integration of rulemaking with the web and social media might ultimately blow apart the model of the all-inclusive record that has dominated since Portland Cement and Nova Scotia. That is still some time away, but it seems conceivable.
298 5 U.S.C. § 553(c).
299 See Dooling, supra note 88, at 924-25.
300 4 U.S.C. § 3502(3)(A). This language reflects the 1995 amendments. Originally, the Act definition might have been read to refer literally to paperwork: “the obtaining or soliciting of facts or opinions by an agency through the use of written report forms, application forms, schedules, questionnaires, reporting or recordkeeping requirements, or other similar methods.” By replacing the specific examples with the phrase “regardless of format,” the 1995 amendment broadened, or clarified the breadth of, the definition, which inescapably applies to electronic information collections.
301 5 C.F.R. § 1320.3(c)(1).
302 See 44 U.S.C. § 3506(c)(2)(A).
303 See 44 U.S.C. § 3506 (c)(3).
304 44 U.S.C. § 3507(b).
305 44 U.S.C. § 3507(c).
306 5 C.F.R. § 1320.3(h) (emphases added).
307 Neil R. Eisner & Judith S. Kaleta, Federal Agency Reviews of Existing Regulations, 48 ADMIN. L. REV. 139, 152 (1996).
308 See Panel at ABA Annual Meeting, August 2012.
309 Rulemaking 2.0, supra note 3, at 397 & n.10.
310 Vivek Kundra & Michael Fitzpatrick, Enhancing Online Citizen Participation through Policy, Open Gov’t Blog, June 16, 2009, (quoting a federal employee as observing: “[The Paperwork Reduction Act] imposes a burden to obtain any user-generated input … The result is that we often don’t go to the trouble.”). See also OMB Watch Comments on PRA request for comments at 3; Sternstein, Government Seeks to Update Paperwork Rule, available at http://www. The FCC has urged OIRA and Congress to make clear that the PRA does not apply when agencies seek feedback or input from the general public about the services they provide or the regulations they are considering. See Federal Communications Commission, National Broadband Plan, Rec. 14.19 (2009) (“The Executive Branch’s review of the Paperwork Reduction Act should aim to enable government to solicit input to improve government services.”).
311 Office of Management and Budget, Improving Implementation of the Paperwork Reduction Act, 74 Fed. Reg. 55269 (Oct. 27, 2009).
312 Id. at 55272. Few of the resulting comments addressed this question. One that did came from OMB Watch, which urged OMB to take a strong and clear position that “blogs, wikis, and other web-based social media tools agencies use to gather voluntary public comment will not be considered information collection requests in need of OIRA review and approval.” OMB Watch comments at 3.
313 Memorandum from Cass R. Sunstein, OIRA Administrator, to the Heads of Executive Departments and Agencies and Independent Regulatory Agencies regarding Social Media, Web-based Interactive Technologies, and the Paperwork Reduction Act (April 7, 2010), available at
assets/inforeg/SocialMediaGuidance_04072010.pdf (hereinafter Sunstein Social Media Memorandum).
314 Memorandum from Cass R. Sunstein, OIRA Administrator, to the Heads of Executive Departments and Agencies and Independent Regulatory Agencies regarding Paperwork Reduction Act—General Clearances (May 28, 2010), available at
315 44 U.S.C. §3502(3)(A).
316 5 C.F.R. § 1320.8(a)(5).
317 Id. § 1320.3(h).
318 Sunstein Social Media Memorandum, supra note 314, at 5.
319 Id. at 6.
320 A useful discussion can be found in Stuart Shapiro, The Paperwork Reduction Act: Research on Current Practices and Recommendations for Reform: Report to the Administrative Conference of the United States 29-31 (2012), available at (describing the arguments for and against such an exemption and noting how deeply the issue has divided PRA experts).
321 44 U.S.C. § 3506(c)(1)(B)(IV).
322 See, e.g., Lubbers, supra note 21, at 119 (“Purely voluntary surveys, even if used to determine whether regulated parties have problems with existing regulations, are covered, as are focus groups used to determine whether a regulation is clear or is burdensome, if more than ten persons are involved in the group.”).
323 See generally Cary Coglianese, Richard Zeckhauser & Edward Parson, Seeking Truth for Power: Informational Strategy and Regulatory Policymaking, 89 MINN. L. REV. 277 (2004).
324 Administrative Conference Recommendation 2012-4, The Paperwork Reduction Act, par. 3, available at
325 EPA Comment Policy for Exchange. Or consider the Department of Homeland Security’s Facebook Comment Policy, which reads in part:
DHS reserves the right to review all comments and remove any that contain profanity, personal attacks of any kind, spam, refer to Federal Civil Service employees by name, contain offensive terms that target specific ethnic or racial groups, promote commercial products, are geared toward the success or failure of a partisan political party, group, or candidate, incite hate, or are subject to a claim of infringement or deemed to be an infringement of intellectual property, or that is otherwise objectionable.
Dep’t of Homeland Security, Facebook Comment Policy,
326 E.g., TSABlog. Or EPA: “We encourage you to share your thoughts as they relate to the topic being discussed.”
327 See generally Reno v. American Civil Liberties Union, 521 U.S. 844 (1997).
328 See Hawaii Def. Found. v. City & Cnty. of Honolulu, No. CV 12-00469-JSM-RLP (D. Hawaii).
329 Summum v. Pleasant Grove City, Utah, 129 S. Ct. 1125 (2009) (holding that town, having erected a monument of the Ten Commandments, was not obligated to erect a monument of the “Seven Aphorisms” of the Summum religion in the same park).
330 Id. at 113; see also Johanns v. Livestock Marketing Ass’n, 544 U.S. 550, 553 (2005) (“[T]he Government’s own speech . . . is exempt from First Amendment scrutiny.”); see generally David S. Ardia, Government Speech and Online Forums: First Amendment Limitations on Moderating Public Discourse on Government Websites, 2010 B.Y.U. L. REV. 1981, 2010-29 (2010).
331 Sutliffe v. Epping Sch. Dist., 584 F.3d 314 (1st Cir. 2009); Page v. Lexington Cnty. Sch. Dist. One, 531 F.3d 275 (4th Cir. 2008).
332 Hague v. CIO, 307 U.S. 496, 515 (1939).
333 See Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37, 45 (1983).
334 Int’l Soc’y for Krishna Consciousness, Inc. v. Lee, 505 U.S. 672 (1992) (finding that airport terminal is not a traditional public forum).
335 See Hogan v. Twp. of Haddon, 278 Fed. Appx. 98 (3rd Cir. 2008) (holding that township website was not a public forum or limited public forum, and local resident “had no constitutional right to . . . post on the Township's website”); Putnam Pit, Inc. v. City of Cookeville, 221 F.3d 834, 842-45 (6th Cir. 2000); Vargas v. City of Salinas, 205 P.3d 207 (Cal. 2009) (holding that city website was a non-public forum).
336 Perry Educ. Ass’n, 460 U.S. at 45.
337 Id. at 46.
338 See Matthew D. McGill, Unleashing the Limited Public Forum: A Modest Revision to a Dysfunctional Doctrine, 52 STAN. L. REV. 929, 931 (2000) (lamenting that “within a limited public forum it is impossible for one to differentiate between a presumptively invalid content-based restriction on speech and a legitimate adjustment of the content parameters that define the forum”).
339 Perry Educ. Ass’n, 460 U.S. at 46 n.7.
340 Pleasant Grove City, 555 U.S. at 470; Good News Club v. Milford Cent. Sch., 533 U.S. 98, 106-07 (2001).
341 Id. (citing Widmar v. Vincent, 454 U.S. 263, 276 (1981) and City of Madison Joint Sch. Dist. v. Wis. Pub. Employee Relations Comm’n, 429 U.S. 167, 175 n.8 (1976)).
342 Good News Club, 533 U.S. 98 (setting aside school’s policy of not allowing a children’s Christian club from meeting after at the school after the school-day was finished when it did provide meeting space for secular groups).
343 Lyrissa Lidsky, Public Forum 2.0, 91 B.U. L. REV. 1975, 1990 (2011).
344 Perry Educ. Ass’n, 460 U.S. at 46.
345 Robert C. Post, Between Governance and Management: The History and Theory of the Public Forum, 34 UCLA L. REV. 1713 (1987).
346 Patricia E. Salkin and Julie A. Tappendorf, Social Media and Local Governments: Navigating the New Public Square 52-53 (2013) (though suggesting that social media might be either a designated or a limited public forum, depending largely on the privacy and sharing settings established by the user); Ardia, supra note 331, at 1998-99; Lidsky, supra note 344, at 1998-99.
347 Pub. L. No. 92-463, codified at 5 U.S.C. App. 2.
348 Reeve T. Bull, The Federal Advisory Committee Act: Issues and Proposed Reforms 37 (Sept. 12, 2011), available at
349 See ACUS Staff, Ongoing Web Forum Meetings of Federal Advisory Committees: A Proposed Use of “New Media” under the Federal Advisory Committee Act 2 (Mar. 17, 2011), available at
350 FACA § 3(2).
351 See, e.g., Ass’n of Am. Physicians & Surgeons v. Clinton, 997 F.2d 898, 913 (D.C. Cir. 1993) (“The point, it seems to us, is that a group is a FACA advisory committee when it is asked to render advice or recommendations, as a group, and not as a collection of individuals.” (emphasis in original); Citizens for Responsibility and Ethics in Washington v. Leavitt, 577 F. Supp. 2d 427 (D.D.C. 2008) (groups not covered by FACA because they had no formal organizational structure, met only twice, and involved participants conveying their opinions regarding their individual areas of expertise with no attempt to achieve consensus); Grigsby Brandford & Co., Inc. v. U.S., 869 F. Supp. 984 (D.D.C. 1994) (FACA did not apply to single briefing by unstructured, ad hoc group); Nader v. Baroody, 396 F. Supp. 1231 (D.D.C. 1975).
352 Bull, supra note 349, at 37.
353 Judicial Watch, Inc. v. United States Dep’t of Commerce, 736 F. Supp. 2d 24, 33 (D.D.C. 2010); People for Ethical Treatment of Animals, Inc. v. Barshefsky, 925 F.Supp. 844, 847-48 (D.D.C.1996).
354 Public Citizen v. Dep’t of Justice, 491 U.S. 440, 453 & n.8, 454 (1989).
355 Food Chemical News v. Young, 900 F.2d 328, 332-33 (D.C.Cir.1990) (quoting Public Citizen, 491 U.S. 440), cert. denied, 498 U.S. 846 (1990).
356 Bull, supra note 349, at 3.
357 See Bull, supra note 349, at 45-47. See also Administrative Conference Recommendation 2011-7, The Federal Advisory Committee Act – Issues and Proposed Reforms, para. 6 (point out that “agencies also may host virtual meetings that can occur electronically in writing over the course of days, weeks or months on a moderated, publicly accessible web forum”).
358 Rachel Lee, Note, Ex Parte Blogging: The Legal Ethics of Supreme Court Advocacy in the Internet Era, 61 STAN. L. REV. 1535 (Apr. 2009).
359 Jess Brvain, Justice Kennedy On Law School, Blogging, And Popular Culture, WALL ST. J. LAWBLOG (Oct. 20, 2013, ; Orin Kerr, Justice Kennedy on Blogs, VOLOKH CONSPIRACY (Sep. 3, 2010),
360 5 U.S.C. 557(d).
361 Home Box Office, Inc. v. FCC, 567 F.2d 9 (D.C. Cir. 1977), cert. denied, 434 U.S. 829, reh’g denied, 434 U.S. 988 (1977).
362 See Iowa State Commerce Comm’n v. Office of Federal Inspector, 730 F.2d 1566 (D.C. Cir. 1984); Action for Children’s Television v. FCC, 564 F.2d 458 (D.C. Cir. 1977) (holding that bars on ex parte contacts in rulemaking apply only “where the rulemaking proceedings involve ‘competing claims to a valuable privilege’”).
363 CFPB Bulletin 11-3, “Policy on Ex Parte Presentations in Rulemaking Proceedings” (Aug. 16, 2011), available at
364 Id. § B(1)(b)(i) (emphasis added).
365 See




Draft of October 21, 2013.