MODELS OF DEMOCRACY -BEHIND THE DESIGN AND
USE OF NEW MEDIA JANAGM IN POLITICS VAN DIJK
Introduction
It seems safe to say that there is a relationship between the application of information and communication technology (ICT) and democracy. For decades now the consequences of ICT for social and political life are disputed in an ideological way. Some call them technologies of freedom (de Sola Pool 1983), while for others they are technologies of central control and registration. Visions of the rebirth of Athenian democracy are opposed by nightmares of Orwellian proportions. Others again, like the author of this article (Van Dijk 1991/1994), stress the ambiguity of this technology which is enabling as well as defining and leaves opportunities of choice within certain limits. These are theoretical interpretations of the use of ICT which existed before it was practised on a large scale. In the meantime ICT has massively entered organisations, among them political, government and administrative organisations. The effects of ICT on the practice of these organisations can not be denied any more. The interpretations of the meaning of ICT can be specified now and released from some of the early speculations.
This article is about the positive contribution of the use of ICT to the political system in general and political democracy in particular. It deals with several directions of improvement guided by different conceptions of democracy. This does not mean that negative effects of the use of ICT for the political system and democracy, lightly touched upon in the former paragraph, can be ruled out in advance. People who conceive ICT as a technology of central control and registration will not perceive any beneficial effects for democracy and reject particular views of the renewal of democracy, such as notions of
Jan A.G.M. van Dijk is lecturer in Department of Mass Communication, Faculty of Social Sciences, Utrecht University, PO Box 80140, 3584 CS Utrecht.
KD Ol
cn
on
Ö >
m
direct democracy, as Utopian or dangerous views. They tend to think that the best thing which can happen is a preservation of the current political system and the present quality of democracy. One type of analysis can not be ignored. Several analysts argue that the field of democracy itself is losing ground together with the declining influence of the political system of the nation state. The American computer scientist Mowshowitz (1992) speaks of the advent of so-called virtual feudalism. This is a type of rule by the fragmented global political power of transnational corporations and other organisations who, like feudal lords, surpass the nation states and undermine their power to take or implement decisions. This could even mean the end of democracy, a forecast made by the French political analyst Guéhenno (1994) in his book La fin de la démocratie. Guéhenno observes the crisis of the nation state and the "Libanonization" of politics. All kinds of national and international organisations are filling the vacuum and creating undemocratic or uncontrollable types of political power by means of information and communication networks, informal social networks, corruption and even crime. Guéhenno does not perceive any serious project of a world government. If this dark perspective does not mean the end of democracy, it would surely mean a significant set-back. The democracy within transnational corporations and other global associations is at a far lower level than the democracy of national institutional politics. The power of a European parliament or a United Nations does not match the present power of the nation states with a democratic rule either. The analysis of the decline of the nation state by some other social or political scientists is less dramatic. Ulrich Beck, for instance, in his Der Risikogesellschaft (1986; Risk Society, 1992) observes a displacement of institutional politics, not only on global affairs but on civil society as well. The political primacy of the nation state and institutional politics are questioned, not democracy in its own right. Other places and ways of democracy in civil society can replace or add to the institutional democracy of the nation state. This position at least can be located in one of the positive models of democracy to be dealt with now.
Models of Democracy
From the first introduction of the new information and communication technologies, marked by the design and diffusion of interactive and integrative (multi)media, these technologies appeared to be connected to conceptions of democracy. The new facilities of telepolling, telereferenda and electronic elections immediately spurred visions of the rebirth of the Athenian agora and other means of direct democracy and self-representation (Toffler 1980; Becker 1981; Barber 1984). These visions were rejected with equal vigour by defenders of an (improved) representative system and people who were very sceptic about "push-button democracy" (Laudon 1977; Arterton 1987; Abrahamson, Arterton and Orren 1988). The discussion about the opportunities of the new media for direct or representative democracy has lingered for about twenty years. In these decades two things appeared. First, the conceptions of democracy are much more complicated than a simple dualism between direct and representative democracy. Second, discussions have become less theoretical; the media of ICT are maturing and entering into the daily practice of the political system. One can observe stages of experimentation and beyond.
So we will have to do two things: elaborate the conceptions of democracy connected to the practice of ICT and describe this practice with regard to its introduction
in the political system. Both can only be done here in the typifying and summarising way imposed by the narrow limits of an article. Still we hope to show in a plausible way that some typical views of democracy are connected to particular practices of ICT in politics. This goes as far as the suggestion of a relationship of these views with concrete instruments of ICT in the political system like computerised citizen enquiries, opinion polls, referenda, public information systems, government information and registration systems, electronic town halls and freenets (see Table 1).
TABLE 1: Electronic Instruments of Politics and Democracy Arranged According to Information Traffic Patterns
ALLOCUTION
-	computerised election campaigns
-	computerised information campaigns
-	civic service and information centres
CONSULTATION
-	mass public information systems
-	advanced public information systems (the Internet, etc.)
REGISTRATION
-	registration systems for government and public administration
-	computer-assisted citizen enquiries
-	electronic polls
-	electronic referenda
-	electronic elections
CONVERSATION
-	bulletin board systems
-	discussion lists
-	electronic mail and teleconferencing
-	electronic town halls
-	group decision support systems
We even hope to show that these instruments are disputed by different views of democracy. Sometimes these views are very explicit, as in the conception of direct democracy mentioned above. More often they remain implicit. In this article they are made explicit for the purposes of clarification and explanation. There is not much to prove yet at this stage of introduction of ICT into the political system.
The large number of conceptions of democracy can only be summarised by analytical means. A successful attempt to do this has been made by David Held in his Models of Democracy (1987). Held constructs nine models in the history of democracy. A model is "a theoretical construction designed to reveal and explain the chief elements of democratic form and its underlying structure and relations" (Held 1987,6). He distinguishes four classical and five contemporary models. The classical models are classical democracy (Athens, Rome), protective democracy (liberals against the absolutist state), developmental democracy (stressing the education of the citizenry) and direct democracy (based on a system of councils and/or referenda). The contemporary models are competitive-elitist democracy, pluralism, legal democracy, partici-
patory democracy and developmental autonomy. The first four of these models will be used in this article. We have clearly observed views belonging to them in the design and use of ICT in politics. This is not the case with the last model, developmental autonomy, which seems first of all the favourite model of Held himself. Instead of this model we add a fifth one which is based upon direct democracy. It is clearly present in some designs and uses of ICT in politics. We will call it plebiscitary democracy. Held's analysis can be elaborated and improved by the construction of an analytical space relating the five selected models of democracy. Two dimensions typify the differences in the present views of democracy: what should be the goals and the means of democracy? Should its prime goal be opinion formation or decision making? In other words, is democracy primarily a matter of substantial input or of procedure (an output)? Should these goals be reached first of all by the ways of representative or direct democracy? The selected models of democracy can be located in a two-dimensional analytical space (Table 2).
TABLE 2: Five Models in Two Dimensions of Political Democracy
PRIMARY MEANS	PRIMARY GOAL OPINION DECISION FORMATION MAKING
REPRESENTATIVE DEMOCRACY DIRECT DEMOCRACY	LEGALIST COMPETITIVE PLURALIST PARTICIPATORY PLEBISCITARY
Information and communication technology offers means first of all. So it seems most interesting to find out which means are favoured in the design and use of this technology in the political system. In Table 1 a spectrum is drawn of representative and direct means which will be described in the next sections. Here we will also observe a third dividing line most often corresponding to the representative-direct distinction. ICT can be used to reinforce the present, mainly representative political system to confront its difficulties mentioned in the introduction and to rescue or revive the primacy of (institutional) politics. ICT can also be used to displace politics on to civil society by means of participation, pluralism and direct citizen power, abandoning the attempts to save the present political system attached to the nation state in crisis and removing the political primacy, at least partially, to the associations and individuals of civil society. With the aid of ICT, politics can be moved to the market as well. In this case the freedom of choice of producers and consumers is made equal to the vote of citizens. ICT is treated as a technology of freedom offering ways of decision making which can replace traditional ways of political decision making.
It is very important to realise that these models are theoretical constructions. In fact they are ideal types. In the reality of political systems and views several of them can be combined, often in contradictory ways. For instance, the political system of the United States is a combination of a legalist model (stressing the constitutional separa-— tion of powers and the checks and balances in the system), a competitive model (a presidential state and the election of individual political leaders), a plebiscitary model
(direct elections of officials and a growing number of referenda in the states) and a pluralist model (America as a multicultural society from the very start). Still, these models describe real institutions, conceptions and differences of opinion, as we hope to show in the next five sections.
Legalist Democracy
The first model is based on the classical Western conception of democracy which came into being after the decline of the absolutist state in Western Europe. It is reflected in most contemporary constitutions. The first advocates of the legalist model were Locke and Montesquieu. It is called legalist as it clearly is a procedural conception which takes the constitution and the law as the basis of democracy.
According to most contemporary constitutions state authority is separated into three powers (trias politica) controlled by a system of checks and balances. Another important principle is majority rule. This rule is taken to be universal except for particular basic rights of the individual which are also part of the constitution. In the legalist model, democracy is a means to safeguard the freedom of individuals from authoritarian power. It is not a goal in its own right. A system of representation is proposed. The heart of our political system is the judgement of heterogeneous interests and complex problems by representatives of the people. Direct democracy is rejected. Populism is feared. The power of every political institution and public administration has to be limited by the least possible, but effective rules. The system of politics and public administration has to be small and effective. For this reason the legalist model is very popular among conservatives and liberals (see, e.g., Hayek 1960).
The basic assumption in this model with regard to the meaning of ICT for the political system is that it should solve its basic problem: information shortage. The present crisis of the political system and the nation state (see the Introduction) is viewed as the crisis of organisations which can not sufficiently deal with the increasing complexity of the environment and the system itself, as information is lacking, among others by the obstructions caused by bureaucracy. The so-called gap between governors or administrators and citizens is also conceived as a kind of information shortage on both sides. Finally, all kinds of threats to the separation of powers, and checks and balances in the system, most often caused by the rising power of the executive as compared to the legislative state, are accounted to deficiencies of information as well. It is a matter of sharing the power of information. The problem can be solved by an equal supply of the resources of information to the executive and to parliaments, municipal councils, political parties and other representatives.
So, following the legalist model ICT is designed and used as a means to remove information shortages and reinforce the present political system by more effective and efficient ways of information processing and organisation. ICT is also applied to increase the transparency of the political system. By all these means the system would be capable to confront the problems of complexity. For some proponents of the legalist model, usually not the conservative or liberal ones, this means a revival of the steering ambition of the state, so much discredited in the 1980s.
Which are the favourite media or instruments of ICT following this model of democracy? The chosen instruments should serve two functions. First, they would have to supply more and better information to governors, administrators, representatives and citizens. Second, the interactive communication capacities of the new media might
create a representative government which is more open and responsive to the people, not directly controlled by the people. Both functions can only be fulfilled by instruments of ICT under the control of governors, administrators and representatives. In Table 1 they can by found under the headings of allocution, consultation and registration. The preferred instruments are computerised information campaigns, civic service and information centres, mass public information systems, registration systems for the government and the public administration and computer-assisted citizen enquiries. Registration and conversation media such as electronic polls or referenda and electronic debates between citizens are not adopted at all. They are deeply distrusted.
Competitive Democracy
The second model of democracy is also based on a procedural view of representative democracy. Elections of representatives are considered to be the most important operations of the political system. The advocates of this model strongly reject the possibility of direct democracy. According to the best-known designers of this model, Max Weber (1921) and Joseph Schumpeter (1942), direct democracy is impossible in large, complex and heterogeneous societies. A central role for bureaucracy, political parties and leaders with authority is inevitable. Politics has to be seen as an everlasting competition between parties and their leaders for the support of the voting public. In this way the best leaders and representatives are elected. This is the solution for the problems of complexity and the crisis of the political system. It is also the main difference as compared to the legalist model which is based on a balance of executive and legislative power and on responsive representation. In the competitive model power is entrusted to leaders and experts in the executive power. They rule the apparatus of state, they weigh matters and interests against each other, they solve conflicts with negotiations and they command authority. As leadership is emphasised in this model, it is called competitive-elitist by Held. In one respect this is not a good label for it: populism is one of the best-known electoral strategies in this model.
The competitive model is practised first of all in presidential states and two-party political systems. It is gaining popularity in contemporary politics as the role of persons and personalities in politics grows. This role was reinforced by old media such as television and will be strengthened even further by the audio-visual new media enabling all kinds of techniques in direct mail, marketing, targeting and visual manipulation.
The last-called facilities show the way to the design and use of ICT in politics according to this model. First of all, ICT will be used in election and information campaigns. The voting public will be reached by a combination of television and interactive media which serve as direct channels to target a selective audience of potential voters with differentiated political messages. In the second place, the interested public, the fragmented electoral base of political leaders and parties should have the opportunity to get information about views, stands and voting behaviour of their leaders and representatives. So they need access to mass and advanced public information systems. Finally, the registration systems of the government and the public administration are vital to a strong and efficient state authority. Other means of registration and conversation, such as electronic polls and town halls, are only used for the benefit of the political leadership. Their resemblance to direct democracy is deceptive. For instance, the electronic polls, conferences and interactive television shows in the
campaign of the American presidential candidate Ross Perot in 1992 were meant in the first place to boost the popularity of this leader in his competition with other candidates.
CT»
Plebiscitary Democracy
The design and use of direct channels of communication between the political leaders and the citizenry can be transformed into an altogether different view of politics and democracy. In this case these channels are not used to strengthen the position of governors, politicians and administrators, but to amplify the voice of the citizenry. This is the central tenet of the plebiscitarian model of democracy. It is based on notions of direct democracy as a way of decision making. The essential difference with the competitive and legalist models is the displacement of politics from the government and the administration on to society and the world of the individual citizen in particular. According to the plebiscitarian views the decisions in the political system should be taken by as few as possible representatives and as much as possible by individual citizens by means of plebiscites. For these radical views the supposed democracy of the Athenian agora and the Roman forum, revived in some late-medieval city states, have always been the prime source of inspiration. They were for the Founding Fathers of the American constitution, like Thomas Jefferson. In the eighteenth and nineteenth centuries the United States had several forms of direct democracy. Among them were mass meetings with political speakers and polls among the audience. In the twentieth century these forms have been degraded to the level of caucuses and conventions of the two major political parties controlling them to a large degree, a series of referenda in a number of states and the election of some officials by the citizens.
The advent of ICT and the new interactive media stimulated a renaissance of plebliscitarian views in the United States from the 1960s onwards. The concept of teledemocracy was invented. Many local experiments have been waged (see Arterton 1987). In these experiments old and new media were (re)designed and used to open channels between the local government or administration and individual citizens. Well-known American experimenters were Becker (1981) and Barber (1984). They set their hopes on the technical capacities of the new media. They would be able to remove the age-old practical barriers of direct democracy in a large, complex society. The political primacy of the government and institutional politics, already in a state of crisis, would not have to be saved. A political system based on a continuing registration of the peoples will and, for some advocates, the will of consumers on the market as well, might be able to replace this role and this primacy.
Following the plebiscitarian model the logical preferences in ICT are registration systems of the votes and opinions of citizens. Telepolis, telereferenda and televotes by means of telephone and computer networks, two-way cable television or future information highways are the favourite instruments. As a well-known criticism of this conception of democracy points at its individualisation and atomisation of the citizenry and its simplification of issues, instruments of conversation are added sometimes. This means the design of electronic town halls, teleconferencing and other new discussion channels. Of course, instruments of consultation by citizens themselves, mass and advanced public information systems, can not be discarded either. However, all instruments filled with information by institutional politics are distrusted.
Pluralist Democracy
In the competitive, legalist and plebiscitarian models of democracy nothing seems to exist between the state and the representation on the one hand and the individual citizen on the other. In the pluralist model, to the contrary, attention is called to the role of the intermediary organisations and associations of civil society. Alexis de Tocqueville (1835/1864) observed the conspicuous role of these organisations in the American democracy of the eighteenth century. Robert Dahl (1956) did the same about a century later, depicting a political system based upon a representation of competing and negotiating interest-, pressure-, religious- and ethnic groups or political parties. According to this view the political system should consist of many centres of power and administration. A network conception of politics is favoured as opposed to the centralist views of the legalist model (a pyramid of representation) and the elitist competitive model.
The present decline of the central nation state and the displacement of politics is not deplored. On the contrary, it is acclaimed. In the pluralist model, democracy is not the sovereign power of the majority but an always shifting coalition of minorities. The state should act as an arbiter. If it is supposed to put the different parties in an equal position by some kind of social policy, you have a progressive type of pluralist democracy. If it is supposed to refrain from doing this, you meet a liberal or conservative type of it. So, the pluralist model is a combination of direct and representative democracy. Representation is made continually, not only by professional politicians selected every four or five years, but by all kinds of organisational representatives as well. The constitutional state can be accepted, but its real substance and resources are produced by the intermediary organisations of civil society. In Western Europe the most frequent result was some kind of corporatist state. The words "substance" and "resource" indicate that substantial democracy is preferred to a procedural conception of it. Ultimately, opinion formation in civil society, based on interests, discussions and all views, is more important for democracy than decision making in the central state.
Two characteristics of the new media in ICT are very attractive to this model of democracy. First, the multiplication of channels and stand-alone media supports the potential pluriformity of political information and discussion. Every view and every organisation or association can have its say. They can reach their own and every other interested audience. Second, the advance of interactive communication networks, in contrast to the allocution of broadcasting, perfectly fits to a network conception of politics.
Following these two general preferences, all instruments or systems which can be used to reinforce information and communication inside the organisations of civil society or between them will be favoured. These are most instruments arranged under the patterns of consultation, registration and conversation in Table 1. This time these instruments are used, first of all, by the organisations of civil society to inform and register their membership and external audiences. They are mass and advanced public information systems (during the Internet hype in 1995 only organisations with their own WWW-page seem to count), registration systems, and computerised self-surveys inside organisations. However, the most favourite instruments to a pluralist model of democracy are conversation systems inside or among organisations, associations and individual citizens: electronic mail, discussion lists, teleconferencing and decision support systems for the most complex problems.
Participatory Democracy
The last model of democracy to be described is close to the pluralist model in several aspects. It is a combination of representative and direct democracy. It is based on views of democracy emphasising the substantial aspects and resources of democracy even more than the pluralist model. The big difference is the shift in attention from organisations to citizens. The support of citizenship is the central aim in the model of participatory democracy. Jean Jeacques Rousseau is the first classical advocate of this model. He can be considered a proponent of direct democracy, but not in its plebiscitarian brand. Rousseau's notion of the people's will is not based upon the measurement of the views of individual citizens, but it aims at the development of citizenship by means of collective discussion and education. Educating citizens as active members of the community is the primary aim in this model which clearly originates in the Enlightenment. For Rousseau, the people's will was not a sum of individual wills but some kind of totality revealing the sovereignty of the people as a collective. This totality had to be created in public meetings and legislative assemblies. One of the latter-day interpretations was the council or Soviet type of democracy covering a large part of the Marxist tradition; in its practice this totality was often transformed in totalitarianism.
A necessary condition of this model of democracy is the presence of informed citizens. Present-day proponents of participatory democracy, such as Carole Pateman and C.B. Macpherson, want to stimulate active citizenship. The centres of political power themselves should become more accessible to citizens. They should be responsive to their questions and certainly not only pose questions to them. The individualist bias of the plebiscitarian and competitive views is firmly opposed. Plebiscites, electronic or otherwise, are feared for the isolation of the individual citizen and the possibility of central manipulation. Another threat is a separation of opinion polling and opinion formation. Polling in its own right is considered to be a poor and passive type of political participation directed by simple and prefabricated questions. A complete fragmentation of political practice is expected; collective opinion formation in discussions and educational contexts is preferred.
The logical consequence of this model of democracy is the option of instruments of ICT which are able to inform and activate the citizenry. Computerised information campaigns and mass public information systems have to be designed and supported in such a way that they help to narrow the gap between the "information rich" and the "information poor," otherwise the spontaneous development of ICT will widen it. Therefore the access and the user friendliness of the new media should be improved. According to the participatory view this is the only way to really open or make transparent the political system to the mass of the citizenry.
The electronic instruments of discussion are taken as a second option. They are attractive as they could be means for opinion formation, learning and active participation. Discussion lists on public computer networks, teleconferences and electronic town halls might be very useful. However, a first condition is that not only the social and intellectual elite will participate in them. A second one is their design as suitable instruments of discussion. Both conditions are badly fulfilled at this moment (Van Dijk, forthcoming). The so-called virtual communities created on the Internet and other public networks are extremely overpopulated by male, affluent people with high education. The quality of the discussions in these networks is rather low, as some
communication capacities of these new media do not support fruitful discussion. Not much conclusions and agreements can be reached (Van Dijk, forthcoming).
Disputes About Electronic Instruments of Democracy
The description of the relationship between these five models of democracy and different practices of ICT in politics might be considered as a pure construction of the author. He hopes to show that this relationship is firmly grounded. One way of doing this is to take practical experiments with electronic instruments in politics to reveal which goals the experimenters have and to make it acceptable that these goals are related to different conceptions of democracy. Fortunately (for our argumentation), there are not only preferences for different instruments of democracy, but disputes about the design and use of the some of the same instruments as well. The author of this article has conducted several evaluation researches of these instruments in the Netherlands. He has selected three typical cases of a dispute about the same (kind of) instrument in this country.
Case No. 1: Public Information Systems in the Netherlands
The most general dispute about the uses of ICT for the improvement of the communication between local government, public administration and political parties on the one side and the citizens at the other in the Netherlands is a collision of two proposals for the design and implementation of public information systems informing the citizenry about official and political affairs and giving it means to react and interact as well. Some local authorities, experimenters and service providers propose to develop and to implement mass public information systems. Their aim is to reach the mass of the citizenry with elementary information about all kinds of local affairs. They are using a mix of the present old and new media trying to reach anyone: cable-TV with teletext (ceefax etc.), hybrid networks combining cable-TV, cable-text and a digital telephone (a kind of videotex) and free computer terminals in official buildings. The main public is a television audience.
Others propose to support advanced public information systems. Most often they are young enthusiastic computer hobbyists and activists introducing so-called freenets and entries to the Internet. In the Netherlands they are called Digital Cities, for instance the Digital City of Amsterdam. The goals of these people are much more ambitious. They do not only want to supply elementary information, but advanced political information and real-time discussions as well. They want to offer all the facilities of the Internet. They now can only reach a "modem audience," that is less than 10 percent of the population. However, they suppose that the small elite of present users are the pioneers of the population which will follow very soon. They are not pleased with the limited and one-sided information supply of the mass systems, they want to take every opportunity for discussion and change of the political system now. Their systems have much more to offer (advanced political information, discussion and polling).
Very likely these systems and channels will be technically integrated in the future. However, what is interesting to our argument are the priorities made now. Several models of democracy can be used to explain these differences of design and opinion to some extent. They are not only a matter of technical choices. It is a political choice to inform the mass of the citizenry first of all and to let them react in a limited way only.
This would be the first option in the legalist and competitive models of democracy. It is also a political choice to preselect the high capacities for interaction and discussion in computer networks, even if they can only be accessed and used by a small minority of the population. This option is taken by people who want fundamental changes in the political system, most often more direct ways of democracy as to them official politics is discredited. Within the plebiscitarian model advanced systems like the Internet and the freenets will be used for a multitude of applications, first of all electronic polls and discussions. The participatory and pluralist models are ambivalent on this matter. Their advocates will hail the high quality of participation and the manifold discussion in the advanced systems, but they will complain about the quantity of participation and the low accessibility. For them mass public information systems offer better opportunities.
Case No. 2: Citizen Enquiries of Hoogvliet, Rotterdam
In the first part of 1993, a population of 1,345 citizens of Hoogvliet (a part of Rotterdam) received five different questionnaires sequentially about all kinds of local affairs along the channel of the experimental two-way cable-TV This is a case of a computer-assisted citizen enquiry (see Table 1), which might be a prototype of the electronic polls and votes which are expected to be organised in the future information super highways.
After some discussions by all parties engaged (official authorities of the municipality, local political parties, public information service providers and the designers of the enquiry) three goals were selected for this experiment (see Paans 1993):
1.	to ask for the opinion of the population about local policies and affairs (information demand);
2.	to inform about these policies and affairs (information supply);
3.	to increase the involvement in these policies and affairs (attitude change).
The most interesting thing about this set of goals is that they were heavily disputed. The initiators and designers of the enquiry put forward the first and third goals. However, the official authorities and the political parties insisted that the second goal would be selected. They did not want to accept a one-sided telepoll. Information about policies and affairs should be the primary aim. For our analysis of the models of democracy this really is a splendid example. An attempt is made to put upside down an instrument which is the favourite option in the plebiscitarian model, telepolling, and turn it into an information campaign. This fits very well in the legalist and competitive models. The third goal, to increase involvement, would be a primary option of the participatory model.
The result was a compromise of three goals. The first was not reached, as the response was not representative for the population (293 out of 1,345). The results of the enquiries were barely used in official policies. However, the second and third goals were achieved to a certain extent according to the result of an evaluation questionnaire at the end (see Paans 1993).
Case No. 3: "City Talks" of Amsterdam
Between 1989 and 1994 about a dozen so-called city talks (stadsgesprekken) were organised by the information service of the local administration and local organisations of Amsterdam. These talks were some kind of electronic town hall meetings. Discus-
sions of more than two hours about urgent local matters were broadcast live on local television. For these discussions a panel of politicians and representatives of local organisations and an audience of citizens was gathered in a theatre ("the town hall"). In every city talk three questions were posed to the television audience at home. It was able to respond by means of interactive teletext, videotex and the plain digital telephone by just pressing keys for yes or no. In every programme between 3,000 and 6,000 people among 30,000 to 40,000 viewers responded to one, two or three of these (tele)questions. The responses were processed immediately and presented live in the ongoing discussion influencing the debate. The city talks were organised to reach at least five goals (Van Dijk 1993):
1.	to collect (tele)opinions about a particular local affair among municipal organisations
and individual citizens;
2.	to improve the communication between the city government and the citizenry;
3.	to promote discussion among the organisations in the city;
4.	to continue discussion in the city (city talks should not be only a single meeting);
5.	to activate the citizenry.
These goals were so ambitious and some of the means to realise them so contradictory — for instance, a successful television programme for the masses and a detailed and balanced (every one is able to participate) discussion among a panel and an audience require different directions — that none of these goals appeared to be attained sufficiently in the overall evaluation (Van Dijk 1993). Most interesting to the argument in this article are the tensions between these five goals. The first of it (telepolling) "belongs" to the plebiscitarian model and the second (to remove information shortages on every side) to the legalist one. It was not surprising that this second goal was emphasised by the politicians in the evaluation (Van Dijk 1993). The third goal (to promote discussion among municipal organisations) can be derived from a pluralist model. Again, it was not amazing that this goal was backed first of all by the municipal organisations themselves. They did also want to take the opportunity of a television programme to present themselves to a large audience. Finally, the fourth and fifth goals (continuity and activation) are the first options of a participatory model.
In conclusion, the city talks of Amsterdam were a perfect compromise of all models of democracy, a compromise which did not work as it was too ambitious and the tensions in the realisation of diverging goals remained.
Conclusions
The elaboration of five contemporary models of democracy was made to suggest that the goals and means of designing and using ICT in the political system can be very different. The same goes for the promotion of democracy within this system. It makes a difference whether the primary goal of democracy is opinion or decision making. These goals could be reconciled in practice as some electronic instruments used in politics and for the benefit of democracy are able to serve both aims, for instance computer-assisted citizen enquiries (e.g., the informed choice questionnaire), electronic mail and teleconferencing, electronic town halls, and group decision support systems. To reconcile the different means of democracy is a much more difficult matter. As general models representative and direct democracy are fundamentally opposed. You can have some instruments of direct democracy within a representative system, for instance referenda, but the effects of this combination are contradictory in
many cases. As any one knows, the results of referenda often disturb and contradict the decision-making process in the representative system. This is why these direct instruments should remain subordinate to the workings of the representative system of democracy according to the advocates of this system. On the other hand it is possible, and presumably necessary in practice, to insert some kind of representation in a working system of direct democracy. However, a fundamental problem is that a system of daily telepolls, referenda, tele-elections and electronic debates would undermine the authority of every representative and parliament continually.
In the course of the exposition we have revealed a third dividing line between the views of democracy, often linked to the representative-direct distinction. This line is characterised by contemporary discussions about the steering role of the state. The electronic instruments of democracy and politics mentioned in this article (see Table 1) can be designed and used to reinforce institutional politics and to revive the steering ambitions within states in their present crisis caused by globalisation and increasing complexity. However, they can also be applied to displace politics on to civil society and even on to the non-political sphere of the market, which means a much broader definition of the political system.
The discussion between the advocates of the models of democracy can not be solved. Only compromises can be made. Of course they really are made in daily political practice. However, often these compromises are uneasy and contradictory ones. These kinds of compromises are most likely to increase in the design and application of ICT in politics and democracy. In practice the results often are unclear definitions of the goals of experiments with ICT in politics or they are a compromise of a multitude of very different, even contradictory goals. In the former section we have described some cases which were a partial or even an almost complete failure as a result of fuzzy compromises. So, a practical proposal to be derived from our argument is to take great pains to clarify in advance the goals of every experiment with ICT in politics. If compromises have to be made, they should at least be made in an explicit way. This article was written to assist in this process of clarification.
Finally, three suggestions for compromises will be made. For all models of democracy (1) a more responsive government, (2) a better information supply in both directions — the government and the public administration on the one side and the citizenry at the other — and (3) a more transparent political system would be very important goals. For the legalist and competitive models these compromises would help to integrate some proposals of a plebiscitarian, pluralist and participatory model without having to yield to a system based on telepolling, telereferenda and teledebates from the outset. For the plebiscitarian, pluralist and participatory models they could afford the means to promote plebiscites, pluralist discussion and participation of the citizenry without having to submit to the institutions and workings of official politics.
References:
Abrahamson, Jeffrey B., F. Cristopher Arterton, and Garry Orren. 1988. The Electronic Commonwealth. The Impact of New Technologies upon Democratic Politics. New York: Basic Books.
Arterton, Christopher F. 1987. Teledemocracy. Can Technology Protect Democracy? Newbury F>ark: Sage.
Barber, Benjamin. 1984. Strong Democracy: Participatory Democracy for a New Age. Berkeley: University of California Press.
Beck, Ulrich. 1986. Der Risikogesellschaft. Frankfurt am Main: Suhrkamp. (1992. Risk Society. London: Sage.)
Becker, Theodore. 1981. Teledemocracy: Bringing Power Back to the People. Futurist 15, 6, 6-9. Dahl, Robert. 1956. A Preface to Democratic Theory. Chicago: University of Chicago Press. Guéhenno, Jean-Marie. 1994. La fin de la démocratie. Paris: Editions Flamarions. Hayek, Friedrich. 1960. The Constitution of Liberty. London: Routledge & Kegan Paul. Held, David. 1987. Models of Democracy. Cambridge, UK: Polity Press.
Laudon, K. C. 1977. Communications Technology and Democratic Participation. New York: Praeger.
Mowshowitz, A. 1992. Virtual Feudalism. A Vision of Political Organisation in the Information Age. In P Frissen, A. Koers, I. Snellen (eds.), Orwell of Athene? Democratie en informetiesamenleving. Den Haag: NOTA/SDU.
Paans, Jaap. 1993. Teleparticipatie, electronische participate als vorm van bestuurHjke vernieuwing in de gemeentelijke democratie. Leiden: University of Leiden.
Pool, Ithiel de Sola. 1983. Technologies of Freedom. On Free Speech in an Electronic Age. Cambridge, MA: Harvard University Press.
Schumpeter, Joseph. 1976. Capitalism, Socialism and Democracy. London: Allen & Unwin. (Originally published 1942.)
Tocqueville de, Alexis (1864). De la démocratie en Amérique, Oeuvres Completes, Part 1-3. Paris. Toffler, Alvin. 1980. The Third Wave. New York: Pan.
Van Dijk, J.A.G.M. 1991/1994. De Netwerkmaatschappij, sociale aspecten van nieuwe media, (Network Society, social aspects of new media). Houten, Zaventem: Bohn Stafleu van Loghum (first and second edition).
Van Dijk, Jan. 1992a. Evaluatie-onderzoek Stadsgesprekken Amsterdam, Eindrapport, Utrecht: Universiteit Utrecht.
Van Dijk, Jan. 1992b. Referendum, Onderzoek en Stadsgesprek bij de Bestuurlijke Vernieuwing van Gemeenten. Bestuurswetenschappen 36, 5, 384-405.
Van Dijk, Jan A.G.M. Forthcoming. The Reality of Virtual Community. Trends in Communication i (February).
Weber, Max. 1921. Gesammelte politische Schriften. Tubingen: Mohr.