Democracy and E-Rulemaking: Web-Based Technologies, Participation, and the Potential for

University of Massachusetts AmherstScholarWorks@UMass Amherst

eRulemaking Research Group Science, Technology and Society Initiative

2007

Democracy and E-Rulemaking: Web-BasedTechnologies, Participation, and the Potential forDeliberationDavid SchlosbergNorthern Arizona University, [email protected]

Stephen ZavestoskiUniversity of San Francisco

Stuart W. ShulmanUniversity of Massachusetts - Amherst, [email protected]

Follow this and additional works at: https://scholarworks.umass.edu/erulemaking

Part of the Science and Technology Studies Commons

This Research, creative, or professional activities is brought to you for free and open access by the Science, Technology and Society Initiative atScholarWorks@UMass Amherst. It has been accepted for inclusion in eRulemaking Research Group by an authorized administrator ofScholarWorks@UMass Amherst. For more information, please contact [email protected].

Schlosberg, David; Zavestoski, Stephen; and Shulman, Stuart W., "Democracy and E-Rulemaking: Web-Based Technologies,Participation, and the Potential for Deliberation" (2007). eRulemaking Research Group. 1.Retrieved from https://scholarworks.umass.edu/erulemaking/1

https://scholarworks.umass.edu?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarworks.umass.edu/erulemaking?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarworks.umass.edu/sts?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarworks.umass.edu/erulemaking?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/435?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarworks.umass.edu/erulemaking/1?utm_source=scholarworks.umass.edu%2Ferulemaking%2F1&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

Democracy and E-Rulemaking:Web-Based Technologies, Participation,

and the Potential for Deliberation

David SchlosbergStephen ZavestoskiStuart W. Shulman

ABSTRACT. Deliberative democratic theorists and public participation scholars have become in-creasingly interested in institutionalized forms of citizen discourse with the state, including thosefacilitated by information technology. However, there have been very few empirical studies of theclaims that the Internet will make public participation more inclusive and deliberative. We reportthe results of an exploratory survey of 1,556 citizen participants in regulatory public comment pro-cesses in the United States. Our analysis focuses on the differences in deliberative indicators be-tween those who submitted their comments using newly available electronic tools and those whopostal mailed or faxed letters on paper. We also examine differences between those who submittedan original letter and those who submitted a version of a mass-mailed form letter. Overall, the datafound modest evidence of the presence of deliberative democratic practices. More interesting arethe apparently fundamental differences between citizens who submit original comments and thosewho submit form letters. We discuss the implications of these findings as they relate to the use ofinformation technology to increase government-citizen deliberation. doi:10.1300/J516v04n01_04

David Schlosberg is Professor and Chair of the Department of Political Science at Northern Arizona University,where he teaches political theory and environmental politics. He is also affiliated with the Environmental Scienceand Policy, Environmental Studies, and Grand Canyon Semester programs. Schlosberg has had recent work sup-ported by the National Science Foundation, has been a Fulbright Senior Scholar in the Social and Political TheoryProgram at Australian National University, and has taught at the London School of Economics and Keele Universityin the UK. He has published widely on environmental political thought, environmental movements, and theories ofpluralism. His latest book is Defining Environmental Justice (Oxford 2007).

Stephen Zavestoski is Chair of the Department of Sociology and the Environmental Studies Program at the Uni-versity of San Francisco. His research areas include environmental sociology, social movements, and sociology ofhealth and illness. In addition to exploring the use of Internet technology as a means of increasing public participa-tion in environmental decision-making, Dr. Zavestoski is currently engaged in research examining the mobilizationof transnational social movement networks in response to the increased risk of industrial hazards brought on byglobalization.

Dr. Stuart W. Shulman is Director of the Sara Fine Institute in the School of Information Sciences at the Univer-sity of Pittsburgh. He is also the founder and Director of the Qualitative Data Analysis Program (QDAP) at Pitt’sUniversity Center for Social and Urban Research, which is a fee-for-service coding lab working on projects fundedby the National Science Foundation, the National Institutes of Health, DARPA, and other funding agencies. He hasbeen Principal Investigator and Project Director on related National Science Foundation-funded research projectsfocusing on electronic rulemaking, human language technologies, coding across the disciplines, digital citizenship,and service-learning efforts in the United States. Dr. Shulman is the Editor-in-Chief of the Journal of InformationTechnology & Politics.

Address correspondence to: David Schlosberg, Professor and Chair, Department of Political Science, NorthernArizona University, Box 15036, Flagstaff, AZ 86011-5036 (E-mail: [email protected]).

Journal of Information Technology & Politics, Vol. 4(1) 2007Available online at http://jitp.haworthpress.com

© 2007 by The Haworth Press. All rights reserved.doi:10.1300/J516v04n01_04 37

[Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mailaddress: <[email protected]> Website: <http://www.HaworthPress.com> © 2007 by TheHaworth Press. All rights reserved.]

KEYWORDS. Deliberative democracy, e-rulemaking, Internet, public participation

INTRODUCTION:CITIZEN PARTICIPATION

AND THE WEB

Public participation and citizen deliberationare hallmarks of democratic theory. Over thepast two decades, there has been expanded in-terest in participation, engagement, and delib-eration as crucial aspects of democraticpractice. As Dryzek (2000, p. 1) notes, “the es-sence of democracy itself is now widely takento be deliberation.” While some deliberativedemocrats argue that deliberation already oc-curs incurrent liberaldemocraticgovernments,legislatures,and/orcourts (e.g.,Bessette,1994;Rawls, 1996), most in the field call for expand-ing public participation and deliberation onpolicy issues in various ways: from citizen ju-ries, to a national expansion of town hall-stylemeetings, and further to transnational dis-course in civil society (Barber, 1984; Bohman,1996; Dryzek, 2006; Habermas, 1996; Young,2000).

Given this call for an extension of delibera-tive opportunities, a number of democratictheorists and practitioners have begun to ex-amine the implementation of democratic en-gagement. One area ripe with potential toincrease citizen participation is the use of theInternet (Grossman, 1995; Hill & Hughes,1998). Web-based participation and delibera-tion range from online consultations with MPs(Coleman, 2004), to online policy dialoguesand deliberative polling2 at the national and in-ternational levels (Beierle, 2004; Janssen &Kies, 2004; Fishkin, 2000), to coordinatedWeb-based networking among groups in civilsociety (Dryzek, 2000). Some authors havegone so far as to argue that these types ofWeb-basedparticipationcouldbe theanswer tothe decline in social capital and in general inter-est in citizenship (Blumler & Coleman, 2001;Coleman & Gotze, 2001; numerous essays inShane, 2004). Others, however, see the Internet

asaplace that limitsdemocraticengagementbyfostering the practice of communicating onlywith like-minded citizens (Sunstein, 2001,2002). It seems there is agreement on only onething: The Web has potential. Whether that po-tential is for improving or diminishing deliber-ation is an open question.

As a result of these trends, and of the openquestion of the potential of online democraticengagement, a growing research community islooking closely at the possibilities for onlinepoliticaldeliberation.This new scholarshiphasbegun tomoresystematicallyarticulateand testtheories about the role of deliberation (Beierle,2004; Berkman Center, 2005; Schlosberg &Dryzek, 2002; Sunstein, 2001), information(Bimber, 2000, 2003), communications tech-nology (Coleman & Gøtze, 2001; Froomkin,2003), design (Noveck, 2004), as well as otherfactors linked to theories of democratic gover-nance. This research community, however, isgenerally long on theory, hopes, and predic-tions, while too often short on empirical data.

Responding to the interest inonlinedelibera-tion, and to the lack of data in online delibera-tive studies, we examine how deliberative oneform of institutionalized discourse actually is.Contrary to much research and development inthis field, in this study we did not seek to de-velop new forms of online interaction that opti-mize deliberativebehavior; rather, we set out toevaluate the deliberative nature of one newform of citizen participation in the US: that ofparticipation in online rulemaking, commonlyknown as e-rulemaking.3 Our goal here is to ex-amine the longstanding, but as yet unsubstanti-ated, optimistic claims in the literature that theconvenience of use, the non-hierarchical na-ture, the egalitarian potential, and the interac-tivecharacterof the Internetcould lead to richerand deeper communication. Our central aim isto evaluate the move to Web-based public par-ticipationinrulemakingagainstvariouscriteriaestablished by theorists of deliberative democ-

38 JOURNAL OF INFORMATION TECHNOLOGY & POLITICS

racy. The key question is whether or not anonline forum produces higher indicators ofdeliberation than more traditional forums ofpublic input on rulemaking.

E-RULEMAKING AND PARTICIPATIONIN THE US

There is a growing body of literature focus-ing on the importance of public participation inpublic administration (Bingham, Nabatchi, &O’Leary,2005;Goodsell,2006;Roberts,2004;Thomas&Strieb,2003;Yang,2005), aswell asin rulemaking in particular (Furlong, 2004;Golden, 1998; Langbein & Freeman, 2000;Lubbers, 2006; Shulman, 2005; West, 2005).Issues of inclusion, deliberation, trust, and in-terest group influence are central to thesestudies.

Scholars who have begun to research the po-tential of e-rulemaking in particular (Coglia-nese, 2004; Lubbers, 2002; Noveck, 2004;Shulman, 2003, 2004a, 2004b; Zavestoski,Shulman, & Schlosberg, 2006) are fundamen-tally concerned with the aspects and quality ofpublic participation, as well as its impact on theprocess and outcomes. As Lubbers (2006,p. 221) observes, the “main touted benefitsfrome-rulemaking,ofcourse, are increasedop-portunity for informationdissemination,publicparticipation, and governmental transparency,along with better outcomes and greater trust ingovernment.”Suchclaims,however,haveverylittle empirical support (see Benjamin, 2006).Our focus is specifically on the potential ofelectronic avenues for comment to expand de-liberation in the traditional rulemaking set-ting. In addition to contributing to thediscussions of the potential of the Internet inexpanding democratic deliberation generally,this study’s survey of actual citizen partici-pants in the e-rulemaking process offers aninitial, if exploratory, empirical contributionto these discussions of citizen participation ingovernmental decisions.

Our case begins with the fact that the UnitedStates federal government is, by design, facili-tating the electronic submission of citizencom-ments during federal regulatory rulemakingcomment periods.4 Rules are available forviewing on the Web, and many agency docket

systems have been designed so citizens cansearch for broad topics or specific rules andthen simply type a comment that goes directlyinto the rulemaking comment docket. In re-sponse, citizens and interest groups are takingadvantage of newly developed, Web-basedtools and services for generating large numbersof public comments. The confluence of thesetwo trends–the pull of an email or Web accessi-ble federal system for collecting public com-ments and the push of advocacy coalitions andtheir newly ubiquitous electronic tools–has al-tered the rulemaking environment. As Lubbers(2006, p. 218) notes, the “age of e-rulemakingis upon us.”

E-rulemaking is a unique addition to the par-ticipatory and deliberative realm, and so it is ofinterest to us for this study for that and a varietyof additional reasons. First, the development ofnew rulemaking technology has embodied ademocratic direction (Carlitz & Gunn, 2002).Many agencies now use open electronic dock-ets,whichallowcitizens toseeandcommentonthe rules proposed by agencies, the supportingdocumentation, and the comments of other citi-zens.5 Second, electronic rulemaking systemsare highly structured, hence quite differentfrom other Web-based discourse that isone-way or homogenous. Sunstein (2001,2002) argues that the Web enables people topay attention to other, like-minded people, andignore those who are unlike them or who dis-agree with their positions on issues. The Web,according to this view, diminishes exposure toheterogeneity and is far from the ideal of a realpublic forum. Yet the argument here is that thestructure of e-rulemaking may enable citizensto engage the positions of others, includingthose with whom they disagree. The opendocket architecture of e-rulemaking has at leastthe architectural potential to mitigate some oftheanti-deliberativedangers lurkingelsewhereon the Web.

Third, rulemaking goes somewhere: simplyput, the process frequently leads to actualchanges of agency-enforced rules, though itremains a matter of some debate whether pub-lic comments actually play a role, either di-rectly or indirectly through Congress. In thisarticle, a focus on e-rulemaking differs fromother examinations of Web-based discourse,suchasonlinedeliberativepolling,cyberjuries,

Schlosberg, Zavestoski, and Shulman 39

or Web-based policy discussions. Rulemakingrequires agencies to respond to, and to incorpo-rate, substantive public comments. It may bethe only form of online deliberation that regu-larly ends in some form of actual implementa-tion by the state. Finally, the design of the 1946Administrative Procedure Act, which is the ba-sis for notice and comment rulemaking, was toincrease the gathering of substantive informa-tion; hence it is a valid area of study for deliber-ative democratic practice.6

OVERVIEW

Weofferbaselinedataonindicatorsofdeliber-ation in e-rulemaking collected through a surveyof 1,556 participants in regulatory public com-ment processes. Our analysis focuses on the dif-ferencesbetweenthosewhousednewlyavailableelectronic tools and those who mailed or faxedletters on paper. We also examine differences be-tween those who submitted an original letter andthose who submitted a mass-mailed form letter.7The point of examining the type of submissionwas to get information not only about the use ofelectronic comment opportunities, but also aboutthe way that citizens conceive of their role in thepublic participation process.

In what follows, we first outline our ap-proach to choosing the particular rulemakingsfrom which our sample of survey respondentswasdrawn,andwe layoutourapproach tomea-suring deliberative indicators. After describingthe survey methodology, we focus on threefindings: (a) the presence of high levels of de-liberative engagement across all survey re-spondents, (b) the absence of a significantdifference in self-reported practices betweenelectronic and paper commenters, and (c) thesignificant differences between respondentswho submitted original comments and thosewho submitted form letters. Finally, we con-clude with the implications of our findings forthose interested in public participation inrulemaking, citizen deliberation, and thepotential of the Web for increasing both.

THE APPROACH AND CASES

The choice to focus the study on large com-ment-receiving regulatory actions was based

onseveral factors.Oneof thecentralchallengesfor research in this field is that most cases areexceptional (Golden, 1998; Yackee, 2005).The rulemaking process varies widely withinand across agencies, and the architecture ofelectronic interfaces is often novel, experi-mental, or entirely idiosyncratic; we under-stand the limitations of the case-basedapproach. However, we selected three regula-tory actions where we predicted deliberationby citizens should be more likely to occur. Welook only at those cases where the architectureof the online notice and comment process per-mittedcommenters toviewothercommentsbe-fore writing their own, and where the totalnumber of public comments received num-bered in the tens or hundreds of thousands. Oursurvey respondents are therefore consciouslyandstrategicallydrawnfromexceptionalratherthan ordinary rulemakings.8 If the deliberationenhancement predictions of the cyber-opti-mists are to be borne out anywhere, we ex-pected that this would have been more likelyaround controversial policy issues with opendockets. That we did not find evidence ofgreater deliberation by participants in the on-line comment process is more striking in lightof this case selection bias. However, as our se-lectionof cases was not random, and instead fo-cused on controversial rules in agencies withopen docket systems, we cannot claim broadapplication of our findings. This is an explor-atorystudyfocusedspecificallyonquestionsofdeliberation in the chosen atypical rules andsystems. Nonetheless it is interesting baselinedata that provides an empirical and theoreticalbasis for talkingabout the roleofdeliberationinthe context of mass public comment periods.

Given our interest in controversial regula-tions that elicited large numbers of public com-ments, we settled on the three following cases.First was the EPA’s advanced notice of pro-posed rulemaking (ANPR) on the Clean WaterAct regulatory definition of the Waters of theUnited States (Waters).9 The ANPR invitinginput was published in the Federal Register onJanuary 15, 2003 with a deadline for commentsof March 3, 2003. The central question waswhether or not the EPA would issue a new rulechanging the extent of the federal jurisdictionover so-called isolated wetlands. Whereas de-velopment lobbies saw the prospect of a Bush


administration rulemaking as an opportunity tofree up land that had been protected for 30years, environmentalists feared the changewould undermine core principles of thelandmark 1972 Clean Water Act.

The second rule selected was the EPA’sproposed National Emissions Standards forHazardous Air Pollutants (Mercury).10 Theproposed mercury rule was published January30, 2004, with a deadline for public commentsof March 30, 2004. Our dataset contains com-ments submitted as late as the end of June 2004.While an EPA press release claimed the pro-posedactions represented“the largestairpollu-tion reductions of any kind not specificallymandated by Congress,”11 environmental andpublic interest groups countered that the rulewould undermine the intentions of the CleanAir Act and increase mercury in the environ-ment.

Finally,wechose toexamineaDepartmentofTransportation (DOT) ANPR on the CorporateAverage Fuel Economy Standards (CAFE).12

The CAFE ANPR inviting public input waspublished in the Federal Register on December29,2003,withadeadlineforcommentsofApril27, 2004. This proposal sought public com-ments on revising the CAFE program’s struc-ture to address continuing criticisms of theprogram related to energy security, trafficsafety, economic practicability, and the defini-tion of the separate category for light trucks.

SURVEYING FOR E-DELIBERATION

Much research on deliberative democracyaims to identify deliberative attributes, such asautonomyfrompower, reflexivity,heterogene-ity, inclusion, equality, etc., that are conduciveto better decisions and democratic legitimacy(Dahlberg, 2001; Froomkin, 2004; Janssen &Kies, 2004; Witschge, 2004). Yet major differ-ences exist across theories of deliberative de-mocracy: research ranges from the specificaspects of speech to the effect of discursiveprocesses on the public sphere.

In this study, we focus on a just few key at-tributes of deliberative democracy (from, forexample, theorists such as Bohman, 1996;Dryzek, 2000, 2006; Young, 2000). One of thebasic concepts in the field is that deliberation is

reflective rather than simply reactive. We as-sume reflection is based on collecting diverseinformation and forming an understanding ofvarious positions on an issue. A second centralconcept is that such engagement with other po-sitions will bring recognition of others in theprocess. Participants in democratic delibera-tion ideally listen to others, treat them with re-spect, and make an effort to understand them.Third, deliberative theory examines the rela-tion between discourse and the transformationof individual preferences. The ideal of deliber-ation is that of communication that actuallychanges the preferences of participants as theyengage the positions of others. The perceivedauthenticity of the process and of citizen effi-cacyarealso central todeliberativedemocracy,as deliberation is offered as a more authenticform of political participation.

Our questionnaire included a bank of itemsintended to measure these dimensions of delib-eration. While we do not claim to cover the fullrange of concerns of every deliberative theorist,takentogetherourmeasurescaptureakeysubsetof theconceptscentral to recentdevelopments indemocratic theory, and they give a reasonableproxy indication of the level of deliberative ac-tivity present in the rulemaking process.

We posit thatonestraightforwardway tomea-sure the optimistic expectations of improved de-liberation is to compare the practices of thosesubmitted in a traditional manner, through paperor fax submission of comments, with the prac-tices of those who participated using email orWeb-based opportunities. Examining the widelycited potential of the technology and the actuallandscape of public comment a bit further, wealso posited that those who take the time and ef-fort to write original comments will displayhigher indicators of deliberation than those whosend copies or variants of mass emails. Our aimwas to acknowledge and explore claims thattwo-wayelectroniccommunicationchannelsandmass submission campaigns can either enhanceor decrease deliberation.

SAMPLING FRAME CONSTRUCTION

We constructed a sampling frame that wouldbe used to complete the telephone survey. Sub-


mitted comments become part of the public re-cord, so we were able to rely on relatively openaccess to the comment sets on each rule.13 Thegoal was to complete 375 surveys for each of thefollowing four types of commenters: (a) elec-tronic submission of form letters, (b) electronicsubmission of originals, (c) paper submissionof form letters, and (d) paper submission oforiginals.Table1 lists thenumberofcompletedsurveys for each of the four types of comment-ers in the threedifferent rules.Table2describesthe total number of comments on each rule, thenumber of comments to which we had access,the limitations with respect to the way in whichthe accessible comments had been selected bythe agencies, and the approach we took to sam-pling for each rule.

As Table 2 illustrates, we had to employ anumber of different approaches to reach oursample size goals.14 In each case, graduate re-search assistants trained as sample collectorslocated the comments on the Federal agencyWeb-baseddocket systems(EPA’sEDOCKETor DOT’s Docket Management System).15

Comments were available from these WebsitesaseitherAdobeAcrobat (.pdf)or text (.txt)files. In the case of the mercury rule, EPA alsoprovided us with a large number of .txt filescontaining just over 536,000 emailed com-ments.16 Determinationofsubmission typewasbased on the content and/or appearance of thesubmitted comment. Paper submissions to theDOT (traditional letters and faxes, for exam-

ple), were stamped with receipt dates beforethey were scanned into the docket; electronicsubmissions often have telltale informationheaders, and lack such a date stamp. Form let-ters include identical content and were submit-ted by multiple participants. Determination ofan original comment was based on whether theletter contained text that differed from identi-fied form letters or petitions. Once the mainform letter variations were identified and cod-ers became familiar with their rhetoric, originalletters were easily identified. Occasionally aform letter had been modified enough by acommenter so that it blurred the difference be-tween original and form. In these instances, weused as a standard the inclusion of at least onesubstantiveargumentor viewpointnot found inthe baseline version of the form letter.17

As sampling progressed, it became apparentthat we lacked access to a sufficient number ofform comments on the EPA rules to ensure abalanceofcomment typesacrossall three rules.This was due to the EPA’s practice of puttingone example of each form letter, rather than ev-ery single submission, into the EDOCKET sys-tem. In response, we relied on access to agreater number of form submissions in theCAFE comments to complete the sample.

Since potential respondents were to be con-tacted by telephone, we obtained telephonenumbers either from the actual comment or bylooking them up using a Web-based phonenumber database.18 The use of a systematic,


TABLE 1. Summary of Completed Surveys

random sampling method meant that when wecould not locate a phone number, we moved tothe next nth comment. Due to the range of diffi-culties faced–from agencies failing to provideaccess to the entire set of submitted comments,to over-sampling one case, to obtaining phonenumbers for individuals–the results of the sur-vey are not generalizable to the whole popula-tion of citizen commenters on these regulatoryactions and should be understood as the resultsof an exploratory project. While these issuesshouldbeaddressed inany future survey of citi-zen commenters and in the management byagenciesof futurepubliccommentdatasets, the

data that followrepresent theonlymajorsurveyof the practices of citizen commenters that wehave seen, and they offer important insights onthe ways that citizens participate in therulemaking process.19

THE SURVEY

A telephone survey was administered usinga computer-assisted telephone interviewing(CATI) system. It was completed by 1556 re-spondents between the dates of August 30 andNovember 24, 2004: a cooperation rate of 48%,


TABLE 2. Case Characteristics and Data Access

with a margin of error of plus or minus 2.5%.20

The survey asked questions regarding the re-spondents’ general commenting practices,such as the number of times that they had com-mented, how much information they obtainedbefore commenting, how they typically submitacomment,whether theyrefer toothercitizens’comments and, if so, the effect this has on theircomments, and the reasons that they com-mented. Respondents were also asked whetherthey thought their comments were reviewed bya government employee, whether they heardabout the final agency decision, and if so,whether they were satisfied with the final deci-sion. In addition, respondents were asked ques-tions about agency Web sites, including thefrequencyofvisits, the typeof information theyaccessed, whether they used these Web sites tosubmit a comment, and if they would be likelytosubmitacommentonanagencyrule in thefu-ture.

Finally, respondents were asked if they be-lieve submitting comments has the ability tochange the outcome of the final rule. Demo-graphic informationcollected includeage,gen-der, education, income, political ideology,voting behavior, race, ethnicity, and Internetuse; we note differences along demographiclines below only when statistically signifi-cant.21

SURVEY FINDINGS

We organize the discussion of our findingsaround three important discoveries. First, weobserved high levels of self-reported delibera-tive activity across all types of commenters.Second, we found that electronic commentersdo not appear to be any more deliberatively en-gaged than paper commenters. Third, ratherthan finding significant differences betweenelectronic and paper commenters, the differ-ences we found were between individuals whosubmitted original comments and those whoposted form letters.

The Prevalence of Deliberative Indicators

There are indicators that all types ofcommenters practice certain types of delibera-tive activity. In this section we report on four

indicators of deliberative discourse: thefrequency with which commenters seek outinformation, the tendency to review other cit-izens’ comments, the propensity to gain anunderstanding of the positions of others, andthe predilection for changing one’s own posi-tion after being exposed to the arguments ofothers. The findings are summarized in Tables3 and 4.

Commenters Are Information-Seekers

The use of information is high for all types ofcommenters. When asked how much informa-tion they receive on rules before submitting acomment, 45.2% said they get a lot of informa-tion, and a full 90% say theyget a lotor some in-formation. Those that write original papercomments claim the most: nearly 51% say theyget a lot of information before submitting acomment.Over71%of thosesurveyedsaid thatthey referred to the arguments, studies, state-ments, or positions of agencies or independentorganizations before submitting a comment;men were slightly more likely than women torefer to outside arguments. Again, those thatsubmitted original paper comments were at thetop with 76.7%. Agency Web sites are impor-tant sources of information for commenters: afull 50% surveyed said they used these sites indeveloping theircomment.Women,alongwithpeople with lower incomes and people aged 60andover,were significantly less likely toget in-formation on a proposed rule from a federalagency’s Web site. Overall, a large majority ofcommenters claim they are seeking out infor-mation, even those who submit form letters.Few commenters, at least from what they re-port, simply submit comments without tryingto understand the issue.

Commenters Review Others’ Comments

Over 70% of those surveyed said that theyhad read the comments of others at some point.As these comments are only available either inthe agency docket rooms in DC or on the newlydeveloped agency Web sites, it may be that alltypes of commenters are using the agency Websites to examine the docket, when such com-ments are available.22 Demographic differ-ences are insignificant on this general question


of reading others’ comments. For those thatspecifically reported using the agency Websites in the comment process, 69.4% said thatthesitehelpedthemreviewothercitizens’com-ments. Here, men, those with higher incomes

(over $70K/yr), and people under age 60 weresignificantly more likely to use agencywebsites than women, those with incomes un-der $70K, and people over age 60.23 Counter toour original hypotheses, such Web access was


TABLE 3. Summary of Paper vs. Electronic Deliberation Measures

reported highest (75.5%) by those who submit-ted original paper comments. Still, overall re-porting of the review of others’ comments ishigh regardless of submission type or demo-graphic, illustratingattention to thepositionsofothers in the rulemaking process.

Commenters Gain an Understandingof Other Positions

Reading of other citizens’ comments is notjust for information: Commenters report thatthey gain an understanding of the positions of


TABLE 4. Summary of Form vs. Original Differences in Deliberation Measures

others as well. Overall, nearly three quarters(73.2%) say they get a better understanding ofthe positions of other citizens by reading theircomments; people with higher incomes aremore likely to report better understanding thanthose with lower incomes. In addition, 41.5%of the respondents report that they found othercitizens’comments tobepersuasive.As thedif-ferenceacross typesofcommenters is insignifi-cant, these findings suggest that commenters ingeneral are gaining an understanding of thepositions of other citizens commenting on arule.

Commenters Change Their Own Positions

Finally, over one third (36.3%) of those sur-veyed report that theirpositionon an issue actu-ally changed after reading others’ comments.That is less than the 47% who report no changein their position, but the percentage that ac-knowledges such change is significantand sug-gests that the limited discourse made possibleby access to others’ comments is having an im-pact on the reasoning of citizen commenters.

A great disparity exists between what re-spondents identified as their deliberative prac-tices in the survey and what we observed interms of deliberative practices in a separateanalysis of the actual comments submitted.Though for purposes of maintaining confiden-tiality we could not match up specific surveyrespondents with the respondent’s actual com-ment, we were able to review a random sampleof 501 comments submitted on the EPA’s Wa-ters rule. All 501 commenters became part ofour sampling frame for the survey, which sug-gests that the comments reviewed are represen-tative of comments made by all participants inthe survey.

In performing our content analysis of thecomments, we employed three different measuresof deliberativeness. First we looked to see ifcommenters were referencing other com-ments in their submissions, which would sug-gest that they had read other comments.Second,wemeasured thefrequencywithwhichcommenters referenced viewpoints other thantheir own. While this may not be evidence thatthey read other comments, it does suggest a cer-tain degree of deliberativeness. Finally, wewondered whether commenters were making

specific reference to the documents in thedocket (e.g., the Federal Register notice orother documents EPA included in the docket).We found that a mere 1.4% of the comments(7/501) made specific reference to anothercomment. A scant 5.6% of comments (28/501)made reference to viewpoints other than theirown. Roughly the same, 5.8% (29/501), madespecific reference to official documents in thedocket.

We suspect that the disparity between self-reports of deliberativebehavior and observableindicators of deliberation in the commentsthemselves has to do with two factors. First,there is a known bias toward socially desirablesurvey responses indicating greater levels ofdeliberation. Second is the likelihood thatcommentersactuallydidengage indeliberativeacts like reviewing other comments or docu-ments in the docket, but that such actions aredifficult to observe in the written text of publiccomments.

Differences Between Paperand Electronic Commenters

One main goal of the survey was to look fordifferences between those who submitted com-mentsonpaper,either throughpostalmailorfax,and those who used agency Web-based forms,interest group Web sites, or email to comment.The survey suggests that those differences sim-ply do not exist. There was a significant differ-ence between electronic and paper commentersononly twoquestionsandinoppositedirections.Paper commenters, by 74.6% to 67.1% (df = 2;p < .01) over Web-based commenters, weremore likely to refer to the “arguments, studies,statements, or positions made by agencies orindividual organizations.” Since paper sub-mitters are more likely to say that they refer-ence other people’s work, an essential practicefor creating quality discourse, our hypothesisthat electronic commenters would demon-strate greater deliberative activity than papercommenters is not supported.

We suspect this finding is due to the fact thatmany submitters of original paper commentsalso use the Internet and Web-based agencydockets incollecting informationfor their com-


ments. While there is a distinction between pa-per and electronic submission of comments, alltypes of commenters used electronic means togather information.For example, the survey in-dicates that of those who submit comments inany format, nearly half (49.1%) gathered in-formation from agency Web sites. But there isonly a modestly significant (p < .05; df = 2)greater number of electronic commenters(51.9%) than paper commenters (48.1%) whoreport having used an agency Web site. As forthe overall lack of discursive indicators by elec-tronic commenters, it may be that the technol-ogy, which makes commenting quite simple,encourages the rapid submission of comments,which is antithetical to more thoughtful and care-fully reasoned arguments.24 Also, many formcomments are generated via Web services thatoffer commentersno chance to review the com-ments already submitted by others.

Differences Between Originaland Form Commenters

By far, the most significant differences inthis study are between those who submit origi-nal comments and those who submit form com-ments (see Tables 4 and 5). These differencesstart with demographics. Men, commenterswith higher levels of education, and those withhigher incomes are significantly more likely tosubmit an original comment, while women,commenters with less education, and thosewith lower incomes are significantly morelikely to submit a form comment. Contrast thatwith the fact that we found no significant de-mographic differences between those whosubmit paper comments and those who useelectronic means. As with all political partici-pation, sex, income,andeducationseemtoplayan important part in the overall composition ofe-rulemaking input. This finding supports thatof past examiners of rulemaking participants,such as Golden (1998).

Form versus original: Differences in infor-mation-seeking. The differences between orig-inal and form commenters move beyonddemographics and include the basic use of in-formation. Over half (54.2%) of originalcommenters report having used an agency Web

site to read informationonaproposed rule.Thiscompares to only 44.2% of the form comment-ers,asignificantdifference(df=2;p<.01).Bothform and original submitters, however, claimthey gather informationon rules before submit-ting a comment: 48.0% of original submittersclaim to receive “a lot” of information, com-pared to 42.4% of form submitters (df = 4; ns).Similarly, there is not a great difference in therate at which the two types of commenters re-port referring to other arguments in their com-ments. Nevertheless, the nature of a comment–original or form–is a better predictor of the useof information before commenting than is themethod of submission.

Form versus original: Differences in view-ing of others’ comments. While there is no sig-nificant difference between original and formcommenters on their reading of others’ com-ments, their perceptions of others’ commentsas persuasive, or their having changed theirmind as a result of reading another comment,original commenters are significantly morelikely to report (76.7% vs. 69.7%) gaining “agreater understanding of the positions or argu-ments of other citizens by reading their com-ments” (df = 2; p < .05). While both sets ofcommenters read the positions of others, origi-nal submittersaremore likely to reporthavingabetter understanding of those positions. Thedifferences, as well as some similarities, aresummarized in Table 4.

Form versus original: Differences in trust.In addition to the modest differences betweenoriginal and form commenters on the delibera-tion indicators described above, there are sig-nificant differences between the two on anumber of indicators of trust in the process andof the agency involved. For example, 56.6% oforiginal commenters (both paper and elec-tronic) believe their comments were actuallyread by a government employee, compared toonly 43.4% of form commenters (df = 2; p <.01). This is one of the most significant differ-ences we found between form and originalcommenters. Electronic form commenters alsoappear to be the most cynical in terms of feelingthat their participation will have an impact ontheir satisfaction with the final rule. Con-versely, those that sent paper original com-ments are the most satisfied with theirparticipation and the outcome. Form submitters


arealsomore likely to say that theirparticipationled to a negative view of the agency running therulemaking (56.9% for form commenters, vs.43.1% of original commenters). Originalcommenters are almost 17% more likely(58.4% to 41.6%) to report a positive view ofthe agency (df = 2; p < .01). Originalcommenters report being slightly more satis-fied than form commenters with agency deci-sions on issues they have commented on(54.3% of originals are satisfied vs. 45.7% ofform submitters) (df = 2; p < .05).

Finally, users of form letters are simplymore negative about the government in gen-eral. By 61.6% to 38.4% compared to originalcommenters they “rarely” or “never” trust thegovernment to do what is right (df = 2; p < .01).Simply put, original submitters have signifi-cantly higher levels of trust in the govern-ment.25 These differences are reported above inTable 5.26

On the Value of Electronic Commentand Mass Email Campaigns

There is one other key finding regardingthe difference between form and originalcommenters. Though it contradicts the lackof trust in government noted above, formcommenters are more likely than originalcommenters to think that groups that organizemass mail campaigns have the ability to changeproposed rules (86.7% to 81.7%). This maypartly explain why form commenters are muchmore likely to submit comments more oftenthanoriginalcommenters.Sixty-twopercentofform commenters report submitting commentsmore than ten times, while only 44% of originalcommenters report that level of participation.Thisdifference,however,canalsobeexplainedby the expertise and time involved in manyoriginal comments.

This faith thatmassemailcampaignshaveanimpact may explain the increase in the popular-


TABLE 5. Original vs. Form Differences in Trust and Satisfaction

ity of the tactic. Nearly 50% of those surveyedsaid they submitted their last comment throughan interest group Web site, and almost 40% re-ported that this method will also be how theycomment next time. Only those that had sub-mitted paper original comments said that theywouldcontinue that routeoverallothers.Whileagencies such as the EPA and DOT haveworked to improve the information on theirWeb-based docket systems, and although thefederalgovernmentcontinues todevelopaFed-eral Docket Management System with a singleWeb-based public comment portal (Regula-tions.Gov), very few commenters plan to usesuch systems: only 12% versus the nearly 40%who plan to use interest group Web sites. Theseresults indicate that thepracticeofmass-mailedform letter comments originating from variousinterestgroupswillcontinuefor thenear future.

Yet electronic form commenters show thelowest scores on many deliberative indicators.Commenters who submitted using form emailsvia interest group Web sites were the leastlikely to look at other information and the leastlikely to report that their positions havechanged as a result of reading others’ com-ments. A better understanding of these differ-ences may impact how agencies respond topublic comment and how interest groups refinetheir campaigns. Many interest groups, in addi-tion to drawing on their legaland scientific staffto draft detailed comments, respond to therulemaking process with an aggregative ap-proach, soliciting mass numbers of identical ornear-duplicate comments from their membersand other interested citizens. A key question iswhether or not this technology improves or de-grades the overall efficacy of citizen discourse(Shulman, 2006).

CONCLUSION:THE POTENTIAL

OF E-PARTICIPATION

The distinction between paper and elec-tronic commenters, which was the basis of ouroriginalsetofhypotheses, simplydoesnotexistas we imagined. A majority of commenters, re-gardless of the medium of submission, are us-ing electronic means of researching an issue,with paper commenters reporting a greater use

of Web-based agency docket systems. Com-paring paper and electronic commenters on re-cent rules does not help us understand whetherthe new electronic systems are more delibera-tive than past paper-based notice and commentprocesses.27 Still, it is crucial to note the highlevels of deliberative indicators across theboard; nearly three quarters of the respondentssaid they had read the comments of others andgot a better understanding of them, and over athird changed their own positions after readingthose comments. This indicates that taking therulemaking process online probably did makethe process more deliberative, though it be-came more deliberative in equal measure forthose that submitted comments online andthose that did so through traditional means.

However, the difference between originaland form-based participation is central to un-derstanding potential deliberative activity inthe rulemaking process. Original commentersembody significantly more of the deliberativequalities we hypothesized given the move to anaccessible open-docket system. The range ofsignificant differences between original letterwriters and form letter submitters might be par-tially explained by the ease with which interestgroups can spread information to constituentsabout proposed rules open for public commentand the sophistication of email action alert sys-tems that allow individuals to participate bydoing little more than clicking the send buttonon an interest group’s Web site. Though manyof these participants, even electronic form sub-mitters, reported to us that they seek out infor-mation before sending in their comments, formsubmitters are nevertheless much more cynicalabout the process and much less deliberative intheir engagement. One way of understandingthis is that simply making the comment processavailableon theWebdoesnotnecessarilymakeit more deliberative. Current systems enablemassemailcampaignsandsowthepotential formuch less deliberative input. This offers evi-dence for the cyber-pessimists, who believethat the Internet may actually decrease deliber-ation in rulemaking.28

So why don’t interest groups solicit moreoriginal, substantive, and deliberative com-ments? The weakness could be in movementstrategy, rather than citizen lack of interest orcapability.The environmentalgroups involved


in these rulemakings engage their membershipwith an aggregative approach, soliciting massnumbers of identical or near-duplicate com-ments which the agencies then ignore or de-lete.29 This aggregative approach has beenused for years with regard to issues within thelegislative process, yet a more substantiveand deliberative approach would be more fit-ting for rulemaking under the APA process.And according to the survey, citizencomment-ers have shown an interest in more deliberativeparticipation–reading others’ comments, learn-ing, and participating in something more sub-stantive than mass emails.

If interest groups seek to expand citizen de-liberation in rulemakings, theyneed touseWebtechnology to solicit more substantive com-ments.For example, theycouldask members toenter postal codes, and then prompt them to re-port something about a local stream, mercuryemitting industry, or health problems. Groupscould also distribute parts of a proposed ruleand ask constituents to comment substantivelyon a specific section of interest. As West (2005,p. 661) has argued, stakeholders need to beaware that “effective public comment alsoentails reasoned argumentation.”

If agencies seek to use the Internet to in-crease deliberation over rulemaking, they ap-parently need to do more than simply allowcitizens to submit comments by email. The re-sults here illustrate that, in order for e-rule-making to be more deliberative, agencies needto develop ways to encourage and support orig-inal comment. They also need to address thepessimism and distrust linked to form comment-ers. Some have suggested better informationdelivery, the availability of related studies andanalyses, and word-searchable notices anddockets (e.g., Lubbers, 2006, p. 222). Perhapsanother way to improve the process would be todevelopa betteruser interface in theopen dock-ets: Noveck (2004) advocates such a de-sign-oriented approach. Agencies could alsorandomly respond to comments online dur-ing the rulemaking process, or supplement theformal comment process with online dialogs inorder to illustrate attention paid to citizen com-ments (see, e.g., Stanley & Weare, 2004). Lub-bers (2006,p.236-7)argues that the“flipsideofincreased public participation, of course, is in-creased responsibilities of agencies to digest

andreact toahighervolumeofcomments.”Ourpoint here is that agencies might be moreproactive so that what they digest and react tomight be more constructive.

The point is that the potential to increasedeliberation–something valued by citizens,agencies, and the scholarship on public par-ticipation–exists in the practices of both agen-cies and interest groups. Certainly, we see thatsome citizens are interested in rules, in infor-mation surrounding various issues, and in whatother citizens have to say in the comment pro-cess;manycitizensarealsowillingtohavetheirown positions challenged and possibly trans-formed in the engagement with others.

Yet we also see that technology exists notonly to enhance the deliberative process (theopen dockets and access to information onagency Web sites), but also to degrade dis-course (the easy click-to-send Web pages oninterest group Web sites). One could arguethat the first generation of Internet participa-tion in rulemaking may have actually de-creased the proportionate level of deliberationin rulemaking. But we could also take thestance that the technology has allowed the rawnumber of substantive and deliberative com-ments to increase. In other words, the potentialof the technology may increase both types ofcomment–deliberative and non-deliberative,substantive and non-substantive. It may be thatin addition to the masses of non-deliberativeformcomments, theopen-dockete-rulemakingforum produces a larger number of commentsthat reveal high indicators of deliberation. Evi-dence exists for both the cyber-pessimists andoptimists here. More information on this issueneeds tobegathered,but it is an importantques-tion for e-rulemaking–and for democratic par-ticipation in governance more generally.

So we conclude, in a way, where we started,bynoting thepotentialof the Internetgenerally,and of electronic rulemaking specifically, toenhance democratic deliberation in citizen par-ticipation. Our central goal was to examine theoptimistic claim that online participationwould lead to richer and deeper communica-tion between the public and government.While we found no evidence that electronicparticipation,per se, isanymoredeliberativeorsubstantive than traditional forms of participa-tion, we did find that citizen participants in gen-


eral exhibit numerous deliberative attributes,that those that engaged the process enough tocontribute original comments embodied thehighest measures of deliberative activity, andthat participants expressed a desire for increasedavenues for participation and influence.

Our research suggests that information tech-nology needs to be more proactively developedand applied in order to overcomeexistingbarri-ers in government-citizen interaction and de-liberation. Obviously, the technology will notstand still; we only hope that research like thiswill push agencies and interest groups alike todevelop systems that increase the amount of in-formation, expand the exchange of views, andimprove the democraticprocess in the develop-ment of better policy.

NOTES

1. This project was funded by a grant (SES-0322622)from the National Science Foundation, Social and Eco-nomic Sciences, Program on Social Dimensions of En-gineering, Science and Technology (SDEST), Ethicsand Values Studies. Any opinions, findings, conclu-sions, or recommendations expressed in this materialare those of the authors and do not necessarily reflectthose of the National Science Foundation. The authorswould like to thank Cary Coglianese, David Levi-Faur,Vincent Price, and the anonymous reviewers for theirinsightful comments on previous versions of this paper,and Fred Solop, Jamie Bowie, Kristi Hagen, AnneMottek-Lucas, and Sara Rinfret for their help with boththe survey and the analysis of the responses. KarenMossberger served as Editor Pro Tem for this manu-script. Dr. Mossberger selected the reviewers and madethe editorial decision to accept the manuscript, pendingsatisfactory response to recommended changes. Thedata presented here are available online at: http://erulemaking.ucsur.pitt.edu/data/sdestsurvey.zip

2. See: http://cdd.stanford.edu/polls/docs/ 2004/onlinedp-release.pdf

3. The authors are members of the larger, multi-in-stitution eRulemaking Research Group, which has itshome page at http://www.erulemaking.ucsur.pitt.edu

4. The federal eRulemaking Initiative (http://www.regulations.gov/eRuleMaking.cfm) is one of 24 E-Gov-ernment efforts at the federal level (http://www.whitehouse.gov/omb/egov/). On the progress of thePresident’s Management Agenda to date, see the GAOreport “Electronic Government: Initiatives Sponsoredby the Office of Management and Budget Have MadeMixed Progress” (GAO-04-561T) available at: http://www.gao.gov/new.items/d04561t.pdf

5. Open docket systems exist at the EPA and DOT,examined here, as well as the FCC and some sub-agen-cies in USDA and Commerce.

6. Again, our focus is not on developing new formsof deliberative democracy; it is on measuring delibera-tive indicators in comment systems that are now no lon-ger simply one-way but potentially interactive.

7. Interest group-initiated mass mailed postcards,familiar from past activism, have been modestly en-hanced as customizable e-form letters, often by expen-sive for-profit intermediaries. See, for example, thesuite of services available at http://www.convio.com/site/PageServer

8. This goes against the methodology of others ex-amining rulemaking, such as Golden (1998), who ex-plicitly states the argument to choose “typical” rules to“avoid any bias that might be introduced by examiningonly high-profile rulemakings” (p. 251). de Figueiredo(2006) examines all electronic filings at the FederalCommunications Commission in order to examine theevents that lead to spikes in participation. Yet our pointis to explicitly identify cases where deliberation wouldmost likely be present in order to test the hypotheses ofthe cyber-optimists.

9. See Federal Register, 68, 1991-1998 (availableat: http://snipurl.com/1fl6x).

10. See Federal Register, 69, 4652-4752 (availableat: http://snipurl.com/1fl8e).

11. See http://snipurl.com/dabh12. See Federal Register, 68, 74908-74931 (avail-

able at:http://snipurl.com/1fl8a).13. We would like to acknowledge the assistance of

technicians and administrators in the EPA and the DOTwho made it possible to receive bulk downloads of com-ments in the respective dockets for the purpose of re-search. The complete collections are available at theeRulemaking Testbed hosted by Carnegie Mellon Uni-versity (http://erulemaking.cs.cmu.edu/data.php). OnNovember 11, 2004, a group of 55 scholars submitted aletter on behalf of the Section of Administrative Lawand Regulatory Practice of the American Bar Associa-tion calling for more uniform standards in the preserva-tion of rulemaking data. See the full letter online athttp://www.abanet.org/adminlaw/OMBcomments.pdf

14. The original goal was for a sample size of 1,500,with even distributions of 125 in each of the four com-ment categories (electronic/form, paper/form, elec-tronic/original, paper/original) for all three rules. Due todifficulties in meeting the targets in each category, espe-cially with respect to form letter for the Waters rule (seeTable 1), we over sampled on the CAFE rule in order toreach the overall target of 1,500.

15. Thanks go to Michael Aquino, Tina Eyraud, MegInokumu, Jonathan Nez, Suzuki Susumu, Paul Vaughn,and Baohua Yen.

16. After supplying the 536,000 + text files, the EPAdetermined that nearly 50,000 of the emails were exactduplicates, triplicates, spam, or submissions for otherrulemakings, hence there is a discrepancy between the


estimated total number of comments received and thenumber of comments in our sample frame.

17. This practice has been confirmed in numerousinterviews and focus groups with agency personnel;such a standard has been practiced because agencies arelegally required by the APA to respond to substantivecomments.

18. We used www.whitepages.com and found thatwe were able to obtain phone numbers for slightly morethan 60% of the names and addresses we entered.

19. Furlong (2004) uses a survey to examine interestgroup participation in rulemaking but limited the surveyto groups, rather than to individual commenters. Golden(1998) also used a survey, but it focused on how citizensbecame informed about a rule and how they knew whenand where to comment. Again, we know of no other sur-vey that examines the practices of citizen commenterson rulemakings.

20. A cooperation rate differs slightly from a re-sponse rate; cooperation rate is the proportion of peoplecontacted on the phone who actually completed the sur-vey. As we did not start with a fixed set of subjects,rather we sampled from a larger group until we had thenumber of surveys desired, it is not appropriate to assigna response rate to the project. Given the sampling issues,although we cannot generalize from this sample to thefull population of commenters, we have included esti-mates of error and tests of significance to provide read-ers with benchmarks that compare the likelihood ofgenerating our results if the sampling frame were thetrue universe of commenters. There are also other prob-lems with operationalizing our questions within themethodology of survey research. Participants may un-derstand the questions in ways different than we in-tended, self-reporting may exaggerate discursiveindicators, and citizens may simply be mistaken aboutwhat they actually did during the rulemaking process.Still, we think it is central in an examination of these is-sues to get direct input from a large number of citizenparticipants in the rulemaking process, and we are confi-dent that our methods meet the standards of survey re-search.

21. While we are discussing “citizen” commenters,we should make clear that a small percentage of our re-spondents were involved in the rulemaking process inroles other than as a private citizen. Of those surveyed,86.4% reported that they generally commented as a pri-vate citizen, 7.1% as a paid employee, 3.4% as an unpaidvolunteer, and 3.2% as something else (though mostlyas a representative of an interest group). As we were in-terested in the e-rulemaking process as a whole, we didnot separate out any part of the population from thisstudy.

22. Then again, as only 50% say they visited agencyWeb sites, and as it seems unlikely that 20% physicallyvisited a docket room, this number needs further expla-nation. Most likely, comments were viewed online. Itmay be that some who report reading others’ comments

saw samples on interest group, as well as agency, Websites.

23. This follows many studies, including Thomasand Streib (2003), that have found visitors to govern-mental Web sites more likely to be white, with higher in-come and education.

24. We did not collect data regarding the time citizencommenters took to prepare their comments. While itseems intuitive that original commenters would takemore time, future empirical research should includesuch a question.

25. For a very interesting discussion of the flipsideof this issue–public administrators’ trust in citizens–seeYang (2005).

26. We do not report the results of comparisons be-tween paper and electronic commenters on measures oftrust and satisfaction because we found no significantdifferences on any of the measures reported in Table 5.

27. One could try to explore differences betweencurrent rulemaking processes and past, pre-Internet pro-cesses, but given the weakness of the human memory, asurvey would be an inappropriate method.

28. See, for example, the discussions in Thomas andStreib (2003), Lubbers (2006), and Shulman (2006).

29. It is important to note that environmental groupsdo submit substantive comments on rules, developedwith legal and /or scientific staff, at the same time thatthey solicit mass comment from their membership.From the standpoint of the group, this may be a rationalstrategy: they get to frame the substantive critiques theway they like at the same time they maintain the activityof members (and, often identify potential new membersthrough outreach on the issue). Still, on the goal of im-pacting the substance of a rule, interest groups seem toignore the potential substantive and deliberative inputtheir members could bring to a rulemaking.

REFERENCES

Barber, B. (1984). Strong democracy: Participatorypolitics for a new age. Berkeley: University of Cali-fornia Press.

Beierle, T. C. (2004). Digital deliberation: Engaging thepublic through online policy dialogues. In P. Shane(Ed.), Democracy online: The prospects for politicalrenewal through the Internet (pp. 155-166). NewYork: Routledge.

Benjamin, S. M. (2006). Evaluating e-rulemaking: Pub-lic participation and political institutions. Duke LawReview, 55, 893-941.

Berkman Center for Internet & Society. (2005). Onlinedeliberative discourse droject. Retrieved February28, 2005 from http://cyber.law.harvard.edu/projects/deliberation/

Bessette, J. M. (1994). The mild voice of reason: Delib-erative democracy and American national govern-ment. Chicago: University of Chicago Press.


Bingham, L. B.,Nabatchi, T., & O’Leary, R. (2005). Thenew governance: Practices and processes for stake-holder and citizen participation in the work of govern-ment. Public Administration Review, 65, 547-558.

Bimber, B. (2000). The study of information technologyand civic engagement. Political Communication, 17,329-333.

Bimber, B. (2003). Information and American democ-racy: Technology in the evolution of political power.New York: Cambridge University Press.

Bohman, J. (1996). Public deliberation: Pluralism,complexity and democracy. Cambridge, MA: MITPress.

Carlitz, R. D. & Gunn, R. W. (2002). Online rule-making: A step toward e-governance. GovernmentInformation Quarterly, 19, 389-405.

Coglianese, C. (2004). E-rulemaking: Information tech-nology and regulatory policy. Regulatory PolicyProgram Report (No. RPP-05). Cambridge, MA:Harvard University, John F. Kennedy School ofGovernment, Center for Business and Government,Regulatory Policy Program.

Coleman, S. (2004). Connecting parliament to the publicvia the Internet: Two case studies of online consulta-tions. Information, Communication, and Society,7(1), 1-22.

Coleman, S., & Gøtze, J. (2001). Bowling together: Onlinepublic engagement in policy deliberation. RetrievedJanuary 31, 2003 from http://www.bowlingtogether.net/

Dahlbergh, L. (2001). Computer-mediated communica-tion and the public sphere: A critical analysis. Jour-nal of Computer-Mediated Communication, 7 (1).Retrieved March 1, 2005 from http://www.ascusc.org/jcmc/vol7/issue1/dahlberg.html

Dahlberg, L. (2001, December). The Internet and demo-cratic discourse: Exploring the prospects of onlinedeliberative forums extending the public sphere. In-formation, Communication & Society, 4(4), 615-633.

de Figueiredo, J. M. (2006). When do interest groupsuse electronic rulemaking? In eRulemaking at thecrossroads: The 7th annual international conferenceon digital government research (pp. 19-20). Re-trieved April 30, 2007 from http://erulemaking.ucsur.pitt.edu/doc/Crossroads.pdf

Dryzek, J. S. (2000). Deliberative democracy and be-yond: Liberals, critics, contestations. New York:Oxford University Press.

Dryzek, J. S. (2006). Deliberative global politics. Cam-bridge, UK: Polity Press.

Fishkin, J. S. (2000, November). Virtual democraticpossibilities: Prospects for Internet democracy. Pa-per presented at the conference on Internet, Democ-racy, and Public Goods, Belo Horizonte, Brazil.Retrieved August 13, 2006 from http://cdd.stanford.edu/research/papers/2000/brazil_paper.pdf

Froomkin, A. M. (2003). [email protected]: To-ward a critical theory of cyberspace. Harvard LawReview, 116, 751-873.

Froomkin, A. M. (2004). Technologies for democracy.In P. Shane (Ed.), Democracy online: The prospectsfor political renewal through the Internet (pp. 3-20).New York: Routledge.

Furlong, S. R. (2004). Interest group participation inrule making: A decade of change. Journal of PublicAdministration Research and Theory, 15, 335-370.

Golden, M. M. (1998). Interest groups in the rule-mak-ing process: Who participates? Whose voices getheard? Journal of Public Administration Researchand Theory, 8, 245-270.

Goodsell, C. T. (2006). A new vision for public adminis-tration. Public Administration Review, 66, 623-635.

Habermas, J. (1996). Between facts and norms: Contri-butions to a discourse theory of Law and democracy.Cambridge, MA: MIT Press.

Janssen, D. & Kies, R. (2004, May). Online forums anddeliberative democracy: Hypotheses, variables andmethodologies. Paper presented at the Conference onEmpirical Approaches to Deliberative Politics, Euro-pean University Institute, Florence, Italy. RetrievedApril 30, 2007 from http://edc.unige.ch/edcadmin/images/onlineforums.pdf

Kerwin, C. (2003). Rulemaking: How governmentagencies write law and make policy (3rd ed.). Wash-ington DC: CQ Press.

Langbein, L. I. & Freeman, J. (2000). Regulatory nego-tiation and the legitimacy benefit. N.Y.U. Environ-mental Law Journal 9, 60-151.

Lubbers, J. S. (2002). The future of electronicrulemaking: A research agenda. Administrative andRegulatory Law News, 27, 6-7, 22-23.

Lubbers, J. S. (2006). A guide to federal agencyrulemaking (4th ed.). Chicago: American Bar Asso-ciation.

Noveck, B. S. (2004). The electronic revolution inrulemaking. Retrieved February 28, 2005 fromhttp://snipurl.com/88jf

Rawls, J. (1996). Political liberalism. New York: Co-lumbia University Press.

Roberts, N. (2004). Public deliberation in an age of di-rect citizen participation. American Review of PublicAdministration, 34, 315-353.

Schlosberg, D. & Dryzek, J. S. (2002). Digital democ-racy: Authentic or virtual? Organization & Environ-ment, 15, 332-335.

Shane, P. (2004). Introduction: The prospects for elec-tronic democracy. In P. Shane (Ed.), Democracy on-line: The prospects for political renewal through theInternet (pp. xi-xx). New York: Routledge.

Shulman, S. W. (2003). An experiment in digital gov-ernment at the United States National Organic Pro-gram. Agriculture and Human Values, 20, 253-265.

Shulman, S. W. (2004). The Internet still might (butprobably won’t) change everything: Stakeholderviews on the future of electronic rulemaking. Re-trieved April 11, 2005 from http://erulemaking.ucsur.pitt.edu/doc/reports/e-rulemaking_final.pdf


Shulman, S. W. (2005). eRulemaking: Issues in currentresearch and practice. International Journal of Pub-lic Administration, 28, 621-641.

Shulman, S. W. (2006). Whither deliberation? Massemail campaigns and U.S. regulatory rulemaking.Journal of E-Government, 3(3), 41-64.

Shulman, S. W., Schlosberg, D., Zavestoski, S., &Courard-Hauri, D. (2003). Electronic rulemaking:New frontiers in public participation. Social ScienceComputer Review, 21, 162-178.

Stanley, J. W. & Weare, C. (2004). The effects ofInternet use on political participation. Administra-tion and Society, 36, 503-527.

Sunstein, C. R. (2001). Republic.com. Princeton, N.J.:Princeton University Press.

Sunstein, C. R. (2002). On a danger of deliberative de-mocracy. Daedalus, 131(4), 120-125.

Thomas, J. C. & Streib, G. (2003). The new face of gov-ernment: Citizen-initiated contacts in the era ofe-government. Journal of Public Administration Re-search and Theory, 13(1), 83-102.

West, W. (2005). Administrative rulemaking: An oldand emerging literature. Public Administration Re-view, 65, 655-68.

Witschge, T. (2004). Online deliberation: Possibilitiesof the Internet for deliberative democracy. In P.Shane (Ed.), Democracy online: The prospects forpolitical renewal through the Internet (pp. 109-122).New York: Routledge.

Yackee, S. W. (2005). Sweet-talking the fourth branch:The influence of interest group comments on federalagency rulemaking. Journal of Public Administra-tion Research and Theory, 16, 103-124.

Yang, K. (2005). Public administrators’ trust in citizens:A missing link in citizen involvement efforts. PublicAdministration Review, 65, 273-285.

Young, I. M. (2000). Inclusion and democracy. Oxford,UK: Oxford University Press.

Zavestoski, S., Shulman S., & Schlosberg D. (2006).Democracy and the environment on the Internet:Electronic citizen participation in regulatoryrulemaking. Science, Technology & Human Values,31, 383-408.

Received: 02/15/2007Revised : 04/16/2007

Accepted: 04/24/2007

doi:10.1300/J516v04n01_04


Documents

Democracy and E-Rulemaking: Web-Based Technologies, Participation, and the Potential for