Version classiqueVersion mobile
OpenEdition Books


Leen D’Haenens

Part Three: Communities in Cyberspace

Teledemocracy in the Province: An Experiment with Internet-Based Software and Public Debate

Nicholas Jankowski, Cees Leeuwis, Peter Martin, Margreet Noordhof et Jeffrey van Rossum


In November 1996 a month-long experiment was held with teledemocracy in the Dutch province of North Brabant. During that experiment, some 100 residents and representatives of organizations in the province were invited to debate aspects of land use for the province. This public debate took place with the aid of an Internet – based software program which allowed for moderated discussion, periodic polling of participants, and voting. A research report of this experiment was completed in May 1997 and this paper – based on that report and academic theses subsequently prepared – provides an overview and assessment of this experiment with teledemocracy. Data are presented on the participants and their reflections on the experiment. Attention is also paid to the role of the provincial government in this initiative with on-line democracy. These findings are related to two of the central theoretical perspectives which guided the work – Rogers’s and Kincaid’s convergence model of communication and Habermas’s notion of public sphere. The paper ends with a summary of the main findings from the experiment along with consideration of a number of central issues valuable for future research around initiatives with teledemocracy

Texte intégral


  • 1 Rekindled in the sense that there has been interest in teledemocracy at least since the 1960s, when (...)
  • 2 An extensive annotated bibliography has been compiled by London (1994); another more recent annota (...)

1With the popularization of the Internet, concern for forms of teledemocracy has been rekindled and discussions intensified.1 An explosion of conferences, some accompanied by publications, has taken place recently with reference to virtual democracy, electronic democracy and similar expressions.2 Only a small number of the initiatives reported, however, have been monitored by systematic research. And none of these initiatives has included specially designed software to support public debate, opinion polling and voting activities.

2Such an experiment with supportive software took place in November 1996 in the Dutch province of North Brabant, where residents and interest groups were invited to participate in a public debate conducted on an Internet site established for the event. The central issue of the debate was formulated with the question: “Is space running out in North Brabant?”

  • 3 Previous conference papers and presentations (e.g., Jankowski, 1997; Jankowski et al., 1997) about(...)

3The final report of the investigation (Leeuwis et al., 1997) which monitored this experiment served as the basis for this chapter.3 Here, a portion of the data gathered during the investigation of the experiment is presented. First, though, the model for teledemocracy employed in the experiment, called Besliswijzer (“decision assistant”), is outlined. The following section of the paper notes the main research questions and theoretical perspectives which guided the work. The research design is briefly presented thereafter. Next, a profile is given of the participants in the experiment. The major section of the chapter, “participation in the debate,” follows. Involvement in the debate by organizations and the provincial government is subsequently considered. The paper ends with a summary of the main conclusions from the final report on the Besliswijzer experiment and a series of suggestions for further research regarding this and other models for teledemocracy.


4The conceptual model and related computer software employed for this experiment with teledemocracy, known as Besliswijzer, is Internet-based and provides various components of real-life political debates: platform statements by interest groups and political parties, information from media sources (in this case, access to current and archived editions of the regional newspapers), opportunity for contributions to discussions on topics related to the general theme, opinion polling during the course of the debate, and finally, a secret ballot at the end of the experiment related to the central theme of the debate.

5The developers of Besliswijzer made arrangements with the regional government in the province of North Brabant and with a consortium of social and cultural organizations to hold an experiment with the Besliswijzer model and software. One of the salient issues in that province is the increasing scarcity of space, and that topic was selected as the focus for a month-long public debate.

6It was decided to invite 100 persons to participate in the debate. Both private individuals as well as representatives from organizations active in the province were welcome to participate. The debate was divided into phases, each with a main activity. The first week, for example, was intended for general discussion related to land use in the province of North Brabant. During the second and third weeks, discussion focused on various subtopics. In the fourth and final week, concluding statements and proposals for solutions were to be assessed and – in the form of a general secret ballot – participants could vote for their preferred spokespeople. At various moments during the debate, voting options were provided, including a so-called voting advice and a “passion option,” a provision for attaching extra emphasis or weight to a vote.


7The organization which initiated Besliswijzer and funded the research was interested primarily in a series of concerns related to how the software performed, the quality of information provided for participants, the degree to which the editorial group was able to function as an independent organ, whether the moderation of the debate was adequate, the value of the opinion polls and the electronic voting opportunity at the end of the debate, and the degree to which the debate provided added value to political discussion in the province on issues regarding land use.

8The following research questions were formulated to guide the research activities:

  • What were the main characteristics and levels of involvement of the various actors involved in the experiment?
  • What was the nature and quality of the discussion during the experiment?
  • What was the experience and assessment of the various actors regarding Besliswijzer?

9Although the interests of the funding agency were decisive in the formulation of research questions and project design, a number of theoretical perspectives also played an important role for members of the research team, all of whom were trained or employed in fields of communication studies. The main perspectives of interest to the research team are briefly noted below.

3.1 Communication between citizen and government

10The Besliswijzer experiment should be placed within the context of a growing degree of dissatisfaction with the gap between government action and citizen involvement. There is a general feeling that introduction of new forms of communication could improve this situation, but different visions and ideas exist as to what forms of communication are most suitable. Three different visions can be identified. The first emerges from the idea of network management whereby government plays the role of facilitator, allowing diverse interest groups to present their opinions (Koppenjan et al., 1993). The second vision is known as “interactive policy development” (Van Woerkum, 1997). Here, the government retains a central role, but is required to develop better awareness of the context and impact of a particular policy through exchange with those immediately affected. The third vision relates to the ideal of the Athenian agora and the libertarian notions of Thomas Jefferson (1969). Proponents of this last vision are in favor of more direct forms of democracy whereby citizens attain greater possibilities for involvement in political decision making through, for example, referendums.

11The Besliswijzer experiment is particularly interesting with regard to the specific conditions imposed, and how it may relate to the above three visions. The design of this Internet-based software makes it possible for individuals to discuss issues among themselves, but also directly with representatives of government. Also, the planned opinion polling and voting provide opportunity to assess how these options are experienced by those involved in the experiment.

3.2 Impact of technology

12The introduction of new technologies is often accompanied by widespread speculation about the societal consequences of the innovations. On the one hand, these speculations express optimism about the positive transformation of society; on the other hand, doomsday scenarios abound. The relation between the introduction of Internet services and democracy reflects both of these perceptions: a utopia of direct democracy (Becker, 1991; Bullinga, 1995) along with a dictatorship controlled by a new elite (Laudon, 1986). Both visions have a technologically deterministic perception in common. Others (e.g., Van de Donk & Tops, 1992) suggest that the contextual situation of a specific technological innovation may be of importance as to whether an Orwellian or an Athenian development emerges.

3.3 Electronically-mediated versus conventional forms of public debate

13The organization of public debate on a variety of social and political issues has been a relatively common practice in the Netherlands for some time. With the emergence of various network-based media, questions arise regarding the possibilities and limitations of this type of debate as compared to conventional public debates. A number of important differences can be identified, based on social-psychological and communicative insights:

  • the nature and perception of aspects of time (e.g., moment and duration of debate, sense of urgency, nature of exchanges, use of available time for debate);
  • the nature of place and “location”;
  • the nature of the structure of the communication process;
  • group identity processes;
  • the identity of individuals in debate and their contributions to debate.

14On the basis of such general differences, Leeuwis & Voorburg (1997) constructed a number of hypotheses regarding the specific possibilities and limitations of network-based and conventional, media-supported public debates. The most important differences between the two forms of debate seem related to:

  • level of participation;
  • intensity of involvement;
  • quality of information;
  • involvement of particular actors (e.g., government officials);
  • influence of debate on political decision-making.

3.4 Interactivity

15One of the frequently mentioned characteristics of new media is “interactivity.” Many definitions of the term are in vogue (see, e.g., Hanssen et al., 1996), but generally speaking, interactivity is related to the degree of exchange between persons engaged in communicative action. Interactivity is not, in other words, related to the medium of exchange; nor is it the case that interactivity necessarily leads to increased awareness or understanding. Even with these restrictions, it remains valuable to examine the degree of interactive communication which transpires in a system such as Besliswijzer.


16A wide array of data collection and analysis methods was employed to attend to the noted research questions. An Internet-administered questionnaire was distributed among all 87 registered participants at the very beginning of the experimental period. Some 36 persons completed this questionnaire, a response rate of 41 percent. A second Internet-administered questionnaire was distributed at the end of the experimental period. Due to technical problems the response rate was low. It was decided to redistribute the same instrument by conventional post which resulted in 49 usable questionnaires, a response rate of 56 percent. Basic data were extracted from a short registration form all participants were required to complete. This data was also made available to the research team. A similar arrangement was to be made for the so-called Internet site log data recorded during the course of the experiment. Due to technical problems, however, this data proved unreliable.

17Open-ended personal telephone interviews were conducted with individual participants and representatives of organizations. Initial interviews were held with 18 persons; with 10 members of this group, follow-up interviews were conducted. Regarding representatives of organizations, 7 usable interviews – from half of the organizations invited to take part in the experiment – were completed. For an analysis of the contributions to the debate, a selection was made of the discussion lines representing both short and long series of contributions made during the three phases of the debate. Finally, a member of the research team was permitted to listen in on the editorial board meetings held on a weekly basis by telephone. These observations provided insight into the functioning of the board and the periodic assessment of the debate by board members. Detailed information on this and the other research methods may be found in the final report of the project (Leeuwis et al., 1997).


18Given the self-selective manner in which persons chose to participate in the Besliswijzer experiment, the group cannot be considered representative of the general population of the province. Interest in Internet-based discussions undoubtedly played some role in the decision to participate, and it would be expected that at least some of the characteristics of Internet users would be reflected in a profile of Besliswijzer participants. Also, the special nature of the topic of this experiment with teledemocracy – land use within the province – would influence which persons volunteered to take part in the experiment.

19Upon registering to participate, persons were required to complete a brief form requesting contact information and basic demographic data. This source of information provides the most complete overview on all 87 persons who registered to participate. These persons were also requested to complete a short, electronically-delivered questionnaire at the beginning of the month-long experiment. This questionnaire was designed to supplement the information on the registration form.

20No data were available to compare the pattern of responses to the questionnaire with characteristics of the total group of registrants. Consequently, it is not possible to assess the degree to which this data is representative of the entire group. This uncertainty must be taken into account during any extrapolation of findings to the group as whole. From these two sources of information, the registration form and the electronically-delivered questionnaire, the following profile of the Besliswijzer participants has been constructed.

21The age of the participants ranged from 15 to 74 years, with age 40 being the average for the group. Only two of the 87 registrants were female. As for level of education, more than three-quarters had completed a professional training program or university degree. Considering these and other demographic data, the profile of respondents suggests that those who registered to participate in the teledemocracy experiment were well-educated males, employed, who had lived for a long time within the province. The group as a whole had already reached middle age, lived with a partner, had one or more children, and owned their own homes. Apart from the gender bias, this group could be considered a reflection of the established and settled sector of the population with long-term “roots” in the province.

22Most respondents had less than a year’s experience with Internet services. Half spent less than four hours per week logged onto the Internet; almost a quarter of the respondents said they were logged on more than seven hours per week. E-mail facilities were the primary Internet service respondents made use of; half indicated that they used this service “much” or “very much.” The World Wide Web is used to a similar degree; only a few respondents had made use of Internet services like Internet Relay Chat and File Transfer Protocol.

23The interest and involvement of participants in political topics was generally high. Almost 40 percent indicated that they were very interested in political topics, and almost 60 percent indicated interest in land use politics. A majority of the respondents (58 percent) had taken part in discussions and hearings organized by either the regional or municipal government. Some two-thirds had contact with regional or municipal government officials on various matters.

24The expectations of participants in this experiment were generally high. Three-quarters said they expected to make between an “average” to “large” number of contributions during the debate. Not only was a substantial quantity of contributions anticipated, respondents also expected the material to be of high quality. There were, however, differences of opinion regarding how easily the debate would proceed via computer-mediated communication. One third felt the debate would proceed without problems, while another third expected considerable difficulties and cost. Half of the respondents believed that contacts with other participants would remain distant. In summary, the participants seemed quite interested and active in political matters. They were, taken as a whole, optimistic about the value of this experiment with teledemocracy.


  • 4 Although 335 contributions were recorded, 37 were rejected by the monitoring group for the debate (...)

25The debate was organized in three phases: discussion of the problem, discussion of solutions, and the voting phase. Individuals and representatives of organizations were able to contribute statements during all three phases. Of the 87 persons who registered to participate in the experiment, 45 were responsible for the 335 contributions to the debate.4 Table 1 indicates the number of participants, contributions, and discussion lines or subtopics for each of the three phases. Clearly, most activity took place during the second phase of the debate, the solution phase, when almost two-thirds of all contributions were placed in more than half of all the discussion subtopics. The second phase also lasted the longest in terms of time – 15 days, as opposed to 10 days for the problem phase, and 5 days for the final phase when voting took place.

Table 1: Number of participants, contributions and discussion threads during debate




problem phase




solution phase




voting phase








Note: * total includes 37 contributions which were made mainly as preliminary tests during first weeks of debate

26Upon examination of the largest discussion lines during the problem phase, reflect in Table 2, it can be seen that one line far surpasses all others: the line initiated by V; Geel, political representative within the provincial government responsible for land use policy. Nearly three times as many contributions were made to this discussion line as the runner up, and twice as many people participated in that discussion than ti following two lines.

Table 2: Five largest discussion threads during problem phase

initiator contributions


longest chain

Van Geel*












Van Putten**








Note: * representative of regional government, ** individual participant

27Another way to examine participation is the total number of contributions from individuals and the number of reactions each of these individuals received to his or her contributions; see Table 3. The largest number of contributions (28) to the debate across all three phases was made by an individual participant (Bruggink). The participant from the regional government, Van Geel, contributed half that number (15), but received more reactions – an average of 2.4 – for his efforts than any other participant.

Table 3: Average number of reactions for active participants









Van Putten




Van Diepen








Van Wijk








Van Geel*




Verkuylen **




Baeten **




Note: * participant from regional government
** participant from organization in province

28The data in the above tables suggest that the discussion seemed to revolve around contributions made by the representative from the regional government. The debate was, in fact, an unusual opportunity for citizens to engage in discussion with the representative from government directly responsible for the policy question under discussion. There was a clustering of debate activity around the participant most representative of political power. Much less clustering was evident regarding the average number of reactions the representatives of organizations received. The two organizational representatives noted in Table 4 each received an average of one reaction per contribution. There was also a concentration of contributions among a small group of participants. As reflected in Table 1, 45 persons were responsible for all of the contributions during the course of the debate. Most of these persons, however, made only a single contribution during the debate. In fact, 15 persons were responsible for three-quarters of all the contributions to the debate.

29This situation – a small minority being responsible for the bulk of contributions to a debate – is not entirely unusual for public debates, generally speaking, and for electronically-mediated debates specifically. Schneider (1996), for example, investigated a newsgroup for a one-year period and concluded that 10 persons were responsible for 40 percent of the more than 46,000 contributions. Further, often the same persons were found to be reacting to each other during different time periods during the year under investigation – suggesting a limited amount of interaction among different groups of the more than 3,000 persons who were involved in that newsgroup.


30An original intention of the organizers of the debate was to involve organizations in the province in the experiment so as to increase the diversity of opinions presented during the experimental period. A total of 22 organizations directly or indirectly involved with land use activities were invited at the beginning of the debate to submit statements of their perspectives on the central issue of the debate: the scarcity of space in the province. Of those invited, 14 submitted statements but less than half made contributions during the course of the debate. Four organizations made single contributions, and one exceptionally active organization made 11 contributions during the period of the debate.

31The expectation that representatives of interest groups would play a pro-active role in the debate turned out to be highly mistaken. The reason, it turned out, had little to do with lack of motivation; most representatives of organizations interviewed were actually enthusiastic about the experiment. The main problem was, in fact, technical in nature: many organizations either did not have an Internet connection or were insufficiently familiar with the protocols to consult Web pages. Those organizations which were equipped and trained to use Internet complained about the slowness of the network connection or other technical difficulties. And, to complete the list of problems, there was a lack of clarity within some organizations regarding the authority to make contributions to the debate reflective of internal organization policy.

32This last point was also prominent among representatives of the provincial government as an explanation as to why they did not become involved to a greater degree. Governmental bureaucracy is not accustomed to taking part in public debates or issuing public statements without extensive internal considerations of formulation and consequence. Further, there was division between the politically elected figure (Van Geel) who saw value in the debate and less enthusiastic government employees, who were reluctant to have such a public debate interfere with existing policy. The debate, consequently, never achieved much support within the division of regional government concerned with land use. The experiment displayed a rather ad hoc character and ran well outside the conventional channels for policy formulation.

33Most striking was that one of the initial agreements, that a meeting would take place after the debate, with all involved, in order to assess what points had made their way into government policy and thinking generally, proved almost futile. That meeting was postponed twice and finally held in late May, six months after the experiment. Then, at the meeting, reference was made to various statements and ideas generated during the debate, but no indication was made that the experiment would be repeated. Besliswijzer was, as far as the province North Brabant was concerned, a one-time experience with limited impact and value for policy development.


34The experiment with Besliswijzer provided an opportunity to explore the possible contribution that a new communication technology could have on citizen involvement in political debate. The number of participants in the experiment was limited, because of practical considerations, to 100 persons living or working within the province North Brabant. Ultimately, 87 persons registered to take part in the experiment. Of these registrants, only 45 actually made one or more contributions to the debate during the month-long experiment. And of these 45 persons, a relatively small number – 15 percent – were responsible for nearly three-quarters of the 298 contributions made to the public debate.

35The involvement of organizations in the debate was problematic from the very start. A total of 14 organizations were approached to submit initial standpoints on the theme of the debate, and to participate during the course of the experiment. Nearly all of these organizations contributed initial statements but less than half of this group intervened during, or contributed to, the debate itself.

36There were considerable differences in intention between the various actors. Some representatives of the provincial government saw the debate mainly as opportunity for communication between citizens and organizations active within the province. Citizens and organizations, however, expected the provincial government to take an active role in the debate.

37One of the most important intentions of the experiment was that the results of the Besliswijzer debate would contribute in some manner to formulation of regional land use policy. This aspect was one of the most problematic. One reason for this was that the debate was initiated outside the division of the provincial government responsible for land use; the government officials from this department consequently played a very limited role during the debate. Moreover, there was from the very beginning a lack of clarity as to how the debate might enhance policy formation. In fact, land use policy had already reached an advanced stage of completion and it was consequently uncertain what role there might even be for “interesting ideas” emerging from the public debate.

38Most participants had expected a greater level of participation from the provincial government in the debate. Despite their disappointment, the political representative for regional land use was indeed one of the most prominent participants in the debate. However, one of the most important intentions of the experiment was that the results of the debate would find some form of reflection in further policy development. The lack of such reflection was a frequent criticism made by participants at the end of the experimental period.

8.1 Further research

39This experiment, albeit small in scale, suggests a framework of six areas for further study of teledemocracy initiatives. First, the possibilities for acquiring access to the forums of electronically-mediated public debate remain of foremost concern. Access as a concept is complex (see Jankowski, 1995), but two basic areas requiring attention are access to the technology required for involvement (the hard- and software on which teledemocracy depends), and the skills (both technical and communicative) necessary to make use of the technology, once citizens have it at their disposal.

40Second, the control and procedural mechanisms imposed on the debate may be of central importance in how such an event develops and in the degree of citizen involvement. In the case of Besliswijzer, the more or less artificial stages imposed on the discussion may have been determinant. In other situations, the role of a moderator or (as in Besliswijzer) an editorial organ responsible for summarizing and assessing the relevance of contributions, may be critical for the further development and direction of discussion.

41Third, the range and degree of participation in an electronically-mediated public debate is of paramount importance in assessing its value for the democratic process. By design, and because of structural restraints, the Besliswijzer experiment was limited to a small, self-selected minority. It cannot be considered reflective of how such an instrument for teledemocracy would function on a larger scale. It remains a fundamental – as yet unanswered – question whether instruments of teledemocracy such as Besliswijzer can possibly contribute to increasing citizen involvement and improving the nature of the democratic process. This experiment was much too small in scale and narrow in design to suggest a possible answer. The many restrictions of the technology, along with its limited penetration, are formidable obstacles. But even in the best of technological worlds, where all worked perfectly and everyone was “wired,” there would remain serious, nearly insurmountable difficulties, e.g., lack of varied sense data and dependency on text-based communication.

42Fourth, characteristics of contributions to electronically-mediated debates – and how they are similar or different from real-life variants – remain an area full of unknowns. Much effort is ongoing to isolate and study the various facets of real-life versus virtual communication, often with a view towards formulation of prescriptive statements on how the virtual variety can be organized and structured so as to improve “performance.” Leeuwis & Voorburg (1997), for example, have formulated a series of hypotheses based on a review of the literature on computer-mediated communication, regarding the differences between face-to-face and electronic debates, which deserve empirical examination.

43Fifth, the assessment of debate by participants and other involved actors is of central importance in any overall research agenda. Comparing experiences between real-life and virtual debates, along a range of features, is of paramount importance. And sixth, the relation of virtual debates to real-life debates and further political action is uncertain (see Fernback & Thompson, 1995). The same reservation remains for the potential of teledemocracy initiatives: in what manner and to what extent can they influence the real-life political process?


1 Rekindled in the sense that there has been interest in teledemocracy at least since the 1960s, when public discussion emerged around functions and services for cable television systems. More broadly, there has almost always been discussion about the democratic value of whatever communication technology happened to emerge at a particular point in time; see Jankowski & Malina (1996).

2 An extensive annotated bibliography has been compiled by London (1994); another more recent annotated bibliography has been prepared by Flos (1996). Two journals – Media, Culture and Society and Javnost/The Public published theme issues on virtual and electronic democracy in 1996.

3 Previous conference papers and presentations (e.g., Jankowski, 1997; Jankowski et al., 1997) about Besliswijzer were also valuable in the preparation of this chapter.

4 Although 335 contributions were recorded, 37 were rejected by the monitoring group for the debate because of irrelevancy. Most of these 37 contributions were tests submitted by participants during the first weeks of the debate. Further, contributions submitted by the editorial group serving as moderator for the debate have been deducted, to arrive at a total of 278 contributions.


Associate Professor at the Department of Communication, University of Nijmegen, The Netherlands.

Lecturer and Senior Researcher at the Department of Communication and Innovation Studies, University of Wageningen, The Netherlands. E-mail:

Master’s Degree in Communication Science (University of Nijmegen, The Netherlands) and is currently employed as a Customer Support Specialist at Solvay Pharmaceuticals, Weesp, The Netherlands.

Master’s Degree in Communication Science (University of Nijmegen, The Netherlands). E-mail:

Master’s Degree in Communication Science (University of Nijmegen, The Netherlands) and is currently employed as an IT-specialist at IBM Global Services, The Netherlands.

© Les Presses de l’Université d’Ottawa | University of Ottawa Press, 1999

Conditions d’utilisation :