Formal Agenda Setting (National and Local Level)

Korthagen et al. describe and analyse five digital democratic tools which serve or had served to support formal agenda setting at the national and local level (crowdsourcing for a constitution in Iceland, Future Melbourne Wiki, the Slovenian Predlagam.vladi.si, participatory budgeting in Berlin Lichtenberg and the Dutch internetconsultation.nl). The authors place a strong focus on the participatory process and practical experiences. For a better understanding of these tools and how they are used in practice, interviews were conducted with administrators and researchers familiar with the respective tools. Strengths and weaknesses are identified and possibilities for improvements explored. Most of the tools have had their impact on a political or policy agenda, although the extent of the impact differs and cannot always be quantified. This success is partly due to the embeddedness of most of the tools in formal policy or political processes. An important finding of this chapter is that although e-participation can have an impact on the political agenda setting, it does not always imply that there is an impact on the final decision-making process.


Introduction
This case is known internationally because it was the first time 'crowdsourcing' was used to draft a new constitution. It is considered a classic example of how digital tools-next to offline participatory events-can contribute to democratic processes.
Internet and social media were used to involve citizens and experts from Iceland but also from abroad in the process of writing the actual text of the constitution. The idea behind crowdsourcing is that when ordinary citizens and experts combine their judgements, the results are qualitatively better than when based on the experts' judgement alone (Surowiecki 2004).
In November 2009, Prime Minister Johanna Sigurdardottir of Iceland sent a bill to the parliament to start a constitutional revision process. There was a clear motive.
Since 2008 the country suffered from a national economic meltdown which caused the country's stock exchange, currency and banks to crash. The public asked for fundamental reforms, like the separation of legislative and executive powers, more direct public participation, electoral reforms supervision of the financial sector, public ownership of natural resources (Kok 2011) and less influence for the president, who was severely criticised because of his close connections with the banking world (Meuwese 2012). In that same month, different grass-roots organisations calling themselves 'The Anthill' organised a national forum to discuss the future of Iceland. With 1500 Icelanders (0.5% of the population) joining the organisers, they hoped to obtain some sort of mandate representative enough to make the Althingi (the Iceland Parliament) listen (Landemore 2015). The Althingi decided that citizens should be involved in drafting a new constitution replacing the first one, dating 1944 after independence from Denmark. During the four phases of the constitutional revision process, citizens have had quite some influence on the decision-making process in different ways (Meuwese 2012): 1. The initiative to revise the constitution (2009) (a) Public protest resulted in a bill to start the revision of the constitution (agenda setting). (b) The first National Assembly was organised bottom up and lead to a decision by parliament that citizens should be involved in the drafting process (agenda setting).
2. The choice for a constitutional assembly and working method (2010).
(a) The Constitutional Assembly of 25 people were elected by popular vote (decide). (b) The assembly organised a second National Forum with 950 randomly selected citizens who laid down key notions to be included in the new constitution (co-decide).
(a) Citizens were invited to make suggestions and give comments (co-decide).
(b) The assembly discussed the last amendments and voted on the final draft (decide).

The approval (2012).
(a) In a non-binding referendum, the draft constitution was approved by a majority of the Iceland population (co-decide). (b) The new government, however, did not adopt the crowdsourced constitution (contest).
In the end, the parliament never took up the proposed constitution. It was never brought to vote, so it never went into effect. The new government that took office following the 2013 elections has established a committee to prepare further decisionmaking about the new constitution. The committee published a provisional report in the spring of 2014 identifying the Constitutional Council's draft as one of several possible alternatives for a new constitution (Edwards and De Kool 2016). Since then the constitution remained on ice. Bjarni Benediktsson became Prime Minister on 11 January 2017, but he has never been a great supporter of the whole crowdsourcing process. Since the end of November, however, Katrín Jakobsdóttir became president of Iceland. Under her guidance, the current government has been preparing a new constitutional process aiming for a revision of the constitution in a period over 7 years. The draft constitution of the Constitutional Council will be taken into account, but will not be the foundational text in this new process. Most important is that extensive public consultation will be again part of the process.
The costs of the first constitutional reform process from 2009 to 2011 were paid by the Icelandic Parliament. The total budget is, however, unknown (Kok 2011).

Participants
Participants were involved in different ways and at different stages of the constitutional reform process. In general, the target group of the whole process were not only citizens and experts from Iceland but also from abroad.
At the start, the 25 members of the Constitutional Council were selected from a group of 522 citizens who nominated themselves. They had to gather signatures to place themselves on the ballot. Of the 25 members, 15 got less than a thousand votes, which has been considered as a lack of authority and legitimacy on behalf of the assembly (Kok 2011). Besides this, the members were not representative for the Iceland population and turned out to be highly educated. A disproportional amount of them were professors and students of political studies. The 950 people who participated in the National Forum were randomly selected from the national population register. Quota sampling was used to ensure representativeness in terms of age, gender and geographical distribution, which was only successful for gender (Landemore 2015).
The phase of writing the constitution was open to everybody, as mentioned, even to non-nationals. Translating techniques were used and Council members communicated often in English on their Facebook or Twitter accounts, making the process very accessible for foreigners. The only thing required on the website of the Council was an email address (Kok 2011). The 5% Icelandic citizens without Internet access could participate through letters ("hundreds of letters", interview with Council member) and telephone calls to the Council members. There were also public deliberations with the Council that all people could attend but it was not possible to participate in the discussions.
No reliable data has been published to indicate the level of online engagement. However, the fact that just over 4000 people have "liked" the Council's Facebook page during the constitutional drafting seems to indicate that it was not very intensely used. It is therefore unclear whether the revision process has succeeded in truly engaging the Icelandic population (Kok 2011). On the other hand, when it comes to the contributions of the participants, Meuwese (2012) reports that the Council received about 3600 comments, as well as 370 "formal suggestions" on the official website. Tushnet (2015: 9) thinks the amount of contributions was quite substantive: "Not a trivial number in a nation with a population of under 400,000." The majority of the participants of the online crowdsourcing process seem to have been mostly ordinary Iceland citizens. Many professionals disregarded the whole process or gave their comments only after the draft constitution was finished, according to the interviewed Council member. The interviewed researcher says: "Maybe boycotting is not the right word, but I have the impression that the viewpoint at the Faculty of Law of Reykjavik was that: 'Writing a constitution is for experts.' To some extent there been done some preparatory work done by experts and in the end the draft version of the constitution was also edited by experts. So my impression is that professionals were not involved in the digital participation process but did intervene in everything around it"(interview with researcher).
Although there was a clear urgency that helped to mobilise the population, a communication strategy was needed. The Ministry of Justice and Human Rights introduced the candidates for the Constitutional Assembly on a special website and published a brochure with information about the candidates and the elections which was distributed to all homes in the country. The national Icelandic broadcasting service broadcasted over 50 radio shows. In the end, 36% of the total population voted, which is a low turnout for Iceland's standards (Kok 2011). Reasons mentioned were amongst others: low media coverage, lack of debate between the candidates about basic issues, disagreement about the importance of a Constitutional Assembly and most probably also a general election fatigue. Looking at the first two reasons mentioned, it seems that the communication strategy failed in that respect.
It is difficult to discover how the crowdsourcing process worked and how different groups of citizens were actively approached to get involved. One interviewee (researcher) confirms that she did not hear of any communication strategy: "It wouldn't fit with the overall ad hoc character of the process." Tushnet (2015: 9) adds: "It is also not clear whether specific attempts were made to include different groups of citizens-beyond the so-called participation elite-like minorities in the crowdsourcing process." Also, not much can be found in the literature on what Althingi did to get as high a turnout as possible for the non-binding referendum on the final draft of the constitution. The turnout was 47%, which is comparable with the turnout at similar referenda in other countries (Edwards and De Kool 2016).

Participatory Process
During the 4 months the drafting process lasted, the Council published preliminary drafts online for public review in 12 rounds and in three different working groups consisting of the Council members. Participants could comment on these drafts. Some of these suggestions which were approved for consideration by the Council were posted on the website. The suggestions could be debated online by participants. The Council would then consider the arguments for and against the suggestions and decide whether to include them in the next draft. During the last round the Council discussed the draft per article which was followed by voting by the participants (Kok 2011).
Participants could follow the activities of the Council closely. During the drafting process short interviews with delegates were uploaded on Youtube and Facebook daily. On Thursdays at 13:00 there was live broadcast from the Constitutional Council meetings on the webpage and on Facebook. There were schedules for all meetings, all minutes from meetings of groups, the Board and the Council as well as the Council's work procedures. News from the Council's work was posted on the webpage as well as a weekly newsletter. This made the process very transparent and reflected a high degree of accountability towards the public. It was quite innovative for the Council to use digital means-especially social media to actively solicit input from citizens (Gylfason and Meuwese 2016). According to the interviewed Council member, privacy or safety issues were a "complete non-issue" because "there was no confidential information at stake or anything".
However, the drafting process seems to have been organised ad hoc with little time available and under a lot of pressure, without a method to structure all the input from the participants (interview with researcher). The interviewee saw this confirmed at a conference in Iceland on the whole constitutional process. As an example of the unstructured process she mentions: "If some comment came in on Facebook, then I believe that an intern or someone else working at the Council and started working on it. [. . .] There was not a protocol or something, like: how do we deal with the comment coming in and how do we send this information to the meetings of the Council?" (interview with researcher). Meuwese (2012) mentions that most of the posts of participants on social media were generic and a fair number of suggested substantive provisions for inclusion in the draft constitution. Kok (2011) argues that many of the suggestions were more policy recommendations-for example, the prohibition of livestock maltreatmentrather than constitutional rights. On the other hand, the contributions to the official Council website were, according to the interviewed Council member, quite substantive. The concern about the crowdsourcing method that Council members had at the start turned out to be unfounded. "We were concerned that maybe this would be abused by too many people; that our website might sort of be filled by people writing rubbish as you see on many newspaper websites. But that did not happen" (interview with Council member).
During the drafting process there were forms of deliberative democracyamongst participants and between the Council members and the participants (mostly through Facebook and the website)-as well as aggregated forms of democracy (presenting preliminary texts and voting). The process of writing by the Council members and then presenting to the public and receiving comments made it a predictable process (Van Hulst et al. 2016). In other phases of the process there was a similar kind of balance between deliberation and aggregation.
However, the comments themselves were not aggregated. According to the interviewed Council member, neither aggregation nor a systematic approach was needed. The reason was that it was possible for the Council to take a look at every contribution and comment, because of the small size of the population of Iceland. "But if this was the US for example, you had to multiply everything by a thousand and then it would be impossible for them to read everything, so they would simply make a random selection" (interview with Council member).
In the end, the offline process might have been more important than the online process, according to the interviewed researcher: "I think the deliberation within that group of 25 people [the Council] was more important than the processing of all the input from participants. The Council members were elected for a certain reason, for their ideas, so they were mandated. And there was of course the pre-work of the National Forum [in which 1,000 citizens participated] which they were using. So, the input from citizens did not only come from the digital participation process but also through the Forum. Cause 1,000 citizens, in Iceland that is 0.3% of the population".
The transparency of the process seems to have created a lot of public appreciation and a sense of co-ownership with the participants (Van Hulst et al. 2016). The fact that people could monitor the writing of the constitution, that they were regularly informed, that they got personal emails from the members of the Council in response to their suggestions and comments, could have contributed to the perceived legitimacy of the design. It created a sense of responsibility for the document in the entire population, including those who have not even tried actively to take part in the experiment, but, crucially, they knew that they could participate if they would have liked. The public appreciation is partly reflected in the result of the referendum where the yes camp won 67% of the vote and on every issue. Tushnet (2015) has some doubts in relation to the co-ownership. He thinks that constitutions drafted by laypersons, who have no continuing interest in actually implementing it, may be defective because in the end they have no responsibility for the operation of the government or society they have been creating.

Results
During the first three phases, the people of Iceland have had quite some influence on the process, but in the last decision-making phase, citizen involvement was strongly reduced. The strategy of writing a draft constitution through crowdsourcing could lay claim to legitimacy. Still, it is not possible to track back the input of participants to the final draft. "We can only find out because members of the Council tell us so, but there is no archive of any sort. It is all anecdotal. The only thing we know for sure is that there has been a deliberative process in the Council itself" (interview with researcher).
The Constitutional Council's strategy was to let the parliament vote on the proposal before they had to resign in March 2013. This was important because the Iceland constitution can only be changed by a majority of two successive parliaments. After the election, another second vote by the new parliament could take place. However, contrary to the parliamentary procedures, the draft constitution was never put to a vote, so it never went into effect. The final outcome ran up against considerable resistance from institutionalised political circles, for example, on the provision on public control of the natural resources. Gylfason (2014: 29), member of the Constitutional Council, notes in his article: "Even so, in a direct affront against democracy, Parliament hijacked the bill as if no referendum had taken place. It is one thing not to hold a promised referendum on a parliamentary bill [. . .]. It is quite another thing to disrespect the overwhelming result of a constitutional referendum by putting democracy on ice as is now being attempted in Iceland by putting a new constitution already accepted by the voters into the hands of a parliamentary committee chaired by a sworn enemy of constitutional reform as if no referendum had taken place. Parliament is playing with fire." It is only until recently that a new constitutional reform is announced by the new government. Possibly, elements of the earlier crowdfunded draft constitution will be used in this new process. However, since the link with the formal political process has never been made in the process from 2009 till 2012, Edwards and De Kool (2016: 52) are still right, when they conclude: "This makes the Iceland case study a significant example of the tension that can arise between representative democracy and participative democracy." 8.2 Future Melbourne Wiki: A Strategic City Vision by the Community

Introduction
In 2007, the City Council of Melbourne, Australia, decided to replace its strategic vision for the city with a plan drawn up by the community: The Future Melbourne project. Starting March 2007, six visions for Melbourne-a city for people, a creative city, a prosperous city, a knowledge city, an eco-city and a connected city-were drafted in cooperation with several stakeholders such as community groups, academics and civil society organisations. This draft was then published as a wiki webpage. Consultations were organised: First, 2 weeks in which changes in the document were made by specific stakeholders, followed by a month of public consultation in which the wiki was open for anyone to edit. Various meetings and events were organised to gather input for the document, making the project a combination of on-and offline community activity, taking place over the course of 18 months.
The City Council, consisting of nine members supported by an urban planning management team, initiated the plan (Kang 2012). The aim was to enhance community collaboration on the future of Melbourne, to create a plan that had "a political buy-in, not only from an electoral point of view but also from an institutional point of view" (interview with initiator), by engaging the whole community: citizens, institutions, businesses and organisations. The Council was the first in Australia to incorporate e-democracy practices in policymaking. All activities, both on-and offline, were funded from the Council's budget (Melbourne Planning Committee 2008).
The Future Melbourne Reference Group was established to manage and critically monitor the process, consisting of prominent ambassadors from Melbourne. The city's urban planning team's role was to service this leading group (interview with initiator). Various groups of people were actively invited to provide input via a range of events, such as public meetings, roundtables, Internet forums and exhibitions sponsored or organised by the Melbourne City Council. The final method of participation, to publish and edit the outlines of the plan on a wiki, can be seen as a form of crowdsourcing for policymaking (Van Hulst et al. 2016).
The wiki was made using the open source application TWiki. It consisted of two components: pages on each vision of the draft plan and a chat page for each wiki page. This made it possible to edit the content and to discuss the content and the edits made. For editing the document, registration with an email address was necessary. All contributions throughout the process and outcomes of offline activities were fed back into this wiki by City of Melbourne officers (Kang 2012;van Hulst et al. 2016). "The wiki played a very particular role at the end of the project as a vehicle to transparently do a final edit and publish the plan in a way that it was accessible" (interview with initiator). A dedicated Future Melbourne Wiki team answered questions by participants, corrected errors of facts made in edits, linked citizens to relevant documents and updated participants on events and developments concerning the project. An evidence library was added so that contributors could reference their edits and a "Do-it-yourself community meeting kit" (interview with organiser) could be downloaded from the library. It explained the process of organising a community meeting, and it included forms for collecting and filing input into the library. These options helped to integrate on-and offline processes and bridged larger geographic distances in collaborative processes.
The use of a wiki sparked some controversy within the City of Melbourne administration, as it was perceived as a threat to some employees within the administration and a risky mode of communication for the city. The initiators' response was to suggest it would be just another layer to the process, not a substitute to any part of the normal urban planning process (interview, organiser). "What we said to council was, 'you don't have to agree to this plan. [. . .] All you have to do is acknowledge that this is a plan by the community. And in future when the council starts to take actions that have a strategic component, it will use this as a resource.' So, we split those processes" (interview with initiator).

Participants
The City Council sought contributions of the citizens of Melbourne, but also welcomed input by a broader community (Kang 2012). During public consultation, the website received around 30,000 page views. The draft was visited by over 7000 unique visitors in 1 month. And 131 members of the public registered to edit the contents of the plan and collectively made several hundred contributions to the draft (Collabforge 2009). The Post Implementation Review of the project reveals that of the 129 registrants of whom demographic data was recorded, 59% were men (Collabforge 2009). Of participants that made changes to the document, an even larger share was male: 88% (Van Hulst et al. 2016). Over one-third of the users were aged between 27 and 35, but most changes to the wiki were made by participants between 36 and 45 years old. There is no data on the socio-economic status of participants.
Most edits were not made by the public though, but by City of Melbourne officers. They were responsible for around 60,000 views and made 11,500 edits (Collabforge 2009, van Hulst et al. 2016). This is a substantially larger share than the several hundred edits made by citizens during 1 month of public consultation. For the officers, the wiki was open for editing for a longer period of time. On the other hand, the officers acted as messengers, transferring input from meetings and events with stakeholders and citizens into the wiki (Van Hulst et al. 2016). Using a directediting tool such as a wiki works as a filter, the interviewed organiser notes: "If you have an idea and you have enough energy and interest and understanding to demonstrate your idea in context by going in and editing a strategy document or a policy document, that's pretty high on the ladder of engagement. And so, out of your total base of engagement, there is only a small subset who's going to ultimately take up that opportunity. [. . .] It acts as a filter and it filters out a lot of the noise in many respects." The Future Melbourne project approached issues of access and engagement using several strategies. Local library staff was trained to offer support and the wiki was enhanced with translation options. Content could immediately be translated into seven out of ten most spoken languages in the Melbourne area (Collabforge 2009). There was also an online help function or personal assistance by City of Melbourne employees during business hours. Community engagement sessions were organised for community representatives and leaders, helping them to engage in the plan and to edit it.
The project was advertised through the city's regular communication channels, such as the website and newsletters. The main newspaper ran a series for 6 weeks, monitoring the process (interview with initiator). Officers of different departments were also asked to engage "their" stakeholders in the city, to ensure all the various sectors of the community would get to know the process and the opportunities it offered. Mark Elliott, owner of Collabforge, the organisation involved in designing the wiki, indicates that perhaps the process was not promoted enough: "A key problem was coordination between internal units at the City of Melbourne; key units whose role it was to promote the opportunities to participate weren't fully supporting or understanding of the online wiki component" (Van Hulst et al. 2016).
An analysis of the wiki views shows spikes of activity towards the end of both the public consultation and the stakeholder engagement periods (Collabforge 2009). As the interviewed initiator states, "What we didn't wantand we did end up getting some of this unfortunately-is officers [. . .] writing down what they thought should happen. And so it was a mixture. It wasn't a pure result in the end." Better targeting and inclusion of "normal citizens" would have strengthened the process, according to the interviewed researcher, as would better training in how to use the tool.

Participatory Process
The wiki was the principal means of online participation. Here, the draft plan based on (offline) stakeholder engagement was published. Citizens had several ways of participating: reading and monitoring the document online, making changes to the text in the wiki, discussing the draft through the wiki's discussion pages and face-toface at the various events, and co-writing on envisioned scenario's for Melbourne's future, a process assisted by author Steve Bright. The only condition to participate online was a working email address (Van Hulst et al. 2016). The interviewed organiser states that the main strength of the tool is the shift it creates in participants' mindsets, "from a critical outlook-do I agree with this?-to a constructive frame of mind-how can I improve this?".
The wiki was one part of a larger participatory process. The strategy of the process was to build the collaborative community organically, incrementally increasing the number of participants and opening the tool to larger groups. But initially only a few stakeholders took the chance to participate in the consultations. "I think that was because, at that time, people did not understand what the opportunity was. [. . .] Most of them just used the traditional channels" (interview with organiser). This was partially solved by taking a proactive approach. "They [City officers] would take a laptop around. And they might go to one of the major universities and speak with the vice chancellors and talk through where the plan was at, at that point. And as they were getting feedback they would just start writing and either feed that directly into the wiki, or as a comment in the wiki to come back to. And then they were able to just tell them: 'This is how we're writing it. Everybody has access to it. You can have access to this later if you want to go and correct what I'm saying, but I'm just putting it in on your behalf right now.' And just right there, that was a big breakthrough. . . It's almost like the community sentiment started to shift right there with that. [. . .] And just that opportunity changed the way people thought about the organization that was running the consultation" (interview with organiser). Van Hulst et al. (2016) also indicate that a strength of this project is the possibility of directly contributing to texts.
The subsequent phase was the public consultation of 4 weeks in which the wiki was opened up to the public for editing. This was supported with offline meetings and events. This combination between on-and offline participation made this project most successful (interview with researcher). The wiki provided the means for largescale offline consultation to work. "Face-to-face engagement is [. . .] much richer. But it is just not scalable, and that is the key issue" (interview with organiser).
During the public consultation period there was no clear ownership of the document and no clear assessment of the proposals put forward (Collabforge 2009). "Last one in gets final say [. . .] isn't a very constructive collaborative process", stated one respondent (Collabforge 2009: 7). One city department "literally waited until ten minutes before midnight on the deadline to go in and make their contributions that they'd already planned, so that nobody could change what they'd put in there, basically" (interview with organiser). No rule was established on deleting input either. One respondent stated: "It was hard to put so much time and work into contributions, just to see them deleted without knowing why" (Collabforge 2009: 9). This also impeded negotiation on contributions. Another issue was the lack of transparency on who was editing. It was unclear whether individuals were editing on behalf of a larger organisation or not. And as only an email address was needed for verification "in theory, all changes could have been made by the same person" (van Hulst et al. 2016: 31). A "neutral" point of view or a jury would have been helpful to mediate the different vested interests (Collabforge 2009). However, "the role of the moderator then becomes quite powerful. Who moderates the moderator?" (interview with researcher).
After the public consultation, the draft was shaped into a presentable form by city officers. Many of these edits were about making sure the final document read as if it had been written by a single author. Issues of privacy and security were hardly addressed, perhaps because around the time of the process (2007)(2008), online safety and privacy was not as big an issue as it is these days (interview with researcher). No abuse was reported (interview with organiser).

Results
Including a wiki in the development of a new strategic city plan was innovative. The Future Melbourne project won the 2008 Victorian Planning Award and the 2009 President's Award, bestowed by the Planning Institute of Australia (Kang 2012).
The draft plan was published in the run up to the council elections, and a number of candidates and the new mayor then adopted several of the plan's principles for their electoral programme. When the new council was selected, it was already partially in favour of the Future Melbourne project, and it reformed their committee structure to echo the plan. All committees now fell under the header of "Future Melbourne committee". Councillor portfolios were to represent the different parts of the plan (such as "eco-city", "connected city" and "knowledge city"). This way, a large part of the plan was incorporated in the Council's strategic plan. When the council was re-elected after 4 years, they again incorporated many of the plan's ideas into their 4-year strategy. This meant the implementation of the plan effectively ran 8 years (interview with initiator).
In fact, the wiki component was essential to the political uptake of the plan, states the interviewed initiator: "I think the wiki ended up being a very key component of the buy-in as much as anything because of the symbolic value that it brought to the process. The, if you like, slightly outrageous and outlandish transparency and empowerment that we were offering through that." Respondents in the analysis by Collabforge (2009) state that as a collaborative tool the wiki was helpful in bridging organisational silos, to let the city's officers connect and communicate ideas.
Van Hulst et al. (2016) point out that the final plan was only assessed and accorded by the Council, but not by the citizens. There was no opportunity for them to indicate whether they still agreed with the final outcome of the plan. One respondent in the analysis by Collabforge (2009: 11) states: "My big questions were, the work that I did, where is it going, how will it be considered and used?" The City Council received many ideas for the development of the city through the public consultation process, but after reporting on some of these ideas in a news article on its website, it states in between brackets: "Please note: There is no guarantee that all suggestions can be incorporated into the Future Melbourne draft plan. A number of the recommendations fall outside the City of Melbourne's areas of responsibility" (Melbourne Planning Committee 2008). Expectations were also managed during the various offline meetings with the public, by being clear about the parameters of the process, for what the wiki was used and that the ultimate decisions on political issues was the responsibility of the Council (interview with initiator).
In many cases of online participatory processes, trust in government tends to actually decrease as a result of the lack of embedding the participation process and results in the decision-making process, according to the interviewed researcher. In other words, participants have their say, but don't see their opinions considered in the further process. Consequently, their trust in these processes diminishes. This wiki case seems to be an exception to the rule, as the online participation was well anchored in the rest of the decision-making, including monitoring and accounting for actions taken (interviews with organiser and researcher).
What helps, states the interviewed organiser, is if the document is "digital by default": The content created and edited must be the final work product. Which does not mean that the organisers do not have final say over the content, but that this must happen transparently and by way of dialogue. The wiki tool manages revisions, showing participants what has happened with their contributions. "Participants readily accept the role that government must play when it is done openly and responsively", according to the interviewed organiser. Guidelines for participation ensure clarity on where responsibilities lie. "So before somebody even is allowed to start, they need to tick a box which says: I understand the terms here" (interview with organiser).
Despite its success, the wiki was not made a standard component in the urban planning process of the city. Drafting a new plan for the city in 2014 included public consultation, but this time without the use of a wiki (State of Victoria 2014; Van Hulst et al. 2016). The interviewed initiator states that the council did not want to start from scratch but only needed to refresh the plan. They wanted to do this in a much shorter timeframe. Instead of a wiki, a citizen's jury was formed to revise the plan into Future Melbourne 2026.

Introduction
Predlagam vladi is a Slovenian government-initiated e-participation platform which was launched on 11 November 2009. It is an Internet-based interface for petitiontype proposals by citizens which seek to amend the current regulation in certain areas or matters. It is described as "[. . .] the first practical attempts to institutionally democratise the link between the citizens and the Slovenian government through use of the Internet's emancipatory potential" (Oblak-Črnič and Prodnik 2015: 100-101). The greatest potential of the tool is perceived to lie not in the area of political decision-making, but in the possibility of introducing socially relevant issues to the public sphere and contributing to rational opinion-formation. Seeing as the platform allows an exchange of opinions between Slovenia's citizens and policymakers at the state government level, it can be described as falling into the agenda-setting phase of the policy cycle (Predlagam vladi 2016a). During one of the interviews conducted within our study, an administrator of Predlagam vladi stated: "[. . .] it is kind of a notification system on what issues are out there and what things do people think that we should change or what are the problems, so that bureaucrats and politicians are informed on even those smaller issues that don't get media attention [. . .]." While there is no option to submit suggestions offline within the Predlagam vladi structure, Slovenian citizens have the fundamental right to petition and are given the option to address their suggestions to the competent ministries or Prime Minister offline. The website is currently being managed by the Office for Communication (UKOM), with various ministries formulating responses to proposals. Governmental departments and agencies are obliged to respond to proposals submitted to them within 20 days by the order of the Government 38200-11/2009/9 dated 23.7.2009 (Predlagam vladi 2016b). It is currently entirely within the judgement of the government officials in which way input is incorporated, or whether it is taken into consideration at all (interview with researcher).

Participants
Within the first year of the website launch, a total of 2897 users registered. For the following 5 years the Government Communication Office reported 12,891 registered users (Offerman 2014). According to the interviewed administrator, the number of registered users is steadily increasing over the years: "I would say that we see an increase in citizen participation when we have a change of government. I believe because the people are more hopeful that now maybe the new government will be more susceptive to their ideas [. . .]." In their analysis, Oblak-Črnič and Prodnik (2015) deem Predlagam vladi to have a demographically open and inclusive architecture, not technically limiting the participation of individuals based on age, gender or ethnicity. There is, however, a language bias, since the platform operates in Slovenian. During an analysis of Oblak-Črnič et al. in 2011, a sample of 218 users showed that almost a quarter were between 35 and 44 years of age, with 34% having a high degree of education, 30% a secondary school education and 10% a higher education (Oblak-Črnič et al. 2011). Data on representativeness concerning minority groups or the share of professionals participating is not available, seeing as personal data from the users is not collected.
Proposals are published primarily by individual actors. Only registered users can submit their own proposals, participate in discussions and vote on proposals. The data requirements for registration are quite low, with the system merely requiring a name, surname and email address. Seeing as the validity of names is not verified, there is no obligation to use real names. "The only thing that we require is that they use a valid email [. . .]. At the beginning [we] were going with the idea that everybody should use their own name. [But then] how would we ensure that? Well, then they would have to use a government certificate that is issued by the government [. . .], but we didn't want to [introduce] any obstacles that would hinder participation [. . .]" (interview with administrator). According to Oblak-Črnič, the portal could also be accessed through OpenID or Facebook accounts, describing the usage as "plain and simple" and "suitable even for people with low computer literacy" (Oblak-Črnič 2013: 416). Nicknames are used to identify users within the comment section and the vote casting, which contributes to a sense of anonymity (Oblak-Črnič 2013).
During the launch of the website, there were numerous press releases, press conferences and media reports made on the possibilities of participation. Currently no communication campaign exists specifically for the promotion of the tool. At one point there was a weekly 10-minute segment on national television in Slovenia, where a particular issue brought up on the website was discussed by a representative of the government or a citizen who drove the proposal. During the interview, the administrator explained: "We felt that the media coverage that we got at the beginning was quite big and we were under the impression, that even if we spent ten thousand or twenty thousand euro for advertising, we wouldn't get the numbers that one media report gets in the evening news." In the past, a banner and advertisement for the tool could also be found on the government website. It was reported that on average journalists will inquire about the statistics of the site about two to three times a year, in order to formulate a media report. According to an interviewed researcher, however, the general public is not sufficiently aware of the existence of this tool: "I don't think the general public knows about the [tool]. I don't think it has been systematically promoted [. . .]." They speculate this lack of systematic promotion may be tied to there not being enough staff to tackle an increase in workload.

Participatory Process
All registered users can post proposals for new government policies. The proposal can then be voted on during the course of 14 days, with other Predlagam vladi users being able to cast their vote in favour of or against the proposal. During this phase public deliberation takes place within the comments section and it is possible for the author to modify their original proposal based on this deliberation process. Governmental agencies can theoretically also enter this deliberative process (Oblak-Črnič 2013).
The ability to amend the initial proposal based on the forum deliberation was not initially provided, but is based on user feedback, as explained to us by an administrator: "At the beginning we had two stages. One was the deliberation, or the discussion stage and the other was the voting stage. The difference was that once the deliberation stage was over, the user who proposed the suggestion could not change it anymore and so the editing of the proposal was then locked, and it was put to the vote. But we got the suggestion from the users that the voting could be started immediately, so now when the proposal is published by the moderator the users can comment on it and also vote on it immediately. The user who proposed the suggestion can still edit it, but if he or she does, then all the votes are deleted and all the users who voted on the proposal are informed by email that the proposal has changed, and they should vote again. The change, that was implemented on user request has simplified the voting procedure and resulted in a higher number of proposals being voted through" (interview with administrator).
Moderators help initiators improve their proposals, for example, by applying keywords. They are in close contact with the relevant authorities for each proposal and get in touch with them when the proposal is launched with a request for active involvement. The moderators also judge the adequacy of the response of the competent authority (Oblak-Črnič 2013). The process of submitting proposals is described by the interviewed administrator as follows: "When users send forward a suggestion it is not published immediately. We read it, and what we check is if we had a similar suggestion in the past. [. . .] There are two options then. If a similar suggestion is still under consideration and it is in debate by the users, we reject the second suggestion and reply that they should participate in the debate on this issue which is already going on, and we send them the link where they can participate. In the other case, if the suggestion was already answered by the Ministry under the current government, we [. . .] reject their suggestion and send the link to the [response] and invite them to read the [response]. If they still think their suggestion is different or if they still think that their suggestion should be put forward, [. . .] they should take into account the [response] and maybe amend their suggestion a little and we [will] publish it." The comments are also moderated. Seeing as they are posted in real time, moderation consists of checking whether the comments comply with the terms of use.
The proposal is submitted to the competent authority of the government of the Republic of Slovenia for an official response if at least 3% of users active in the last 30 days vote in favour of the proposal and if there are more votes in favour of than opposed. In the case of a positive response, follow-ups are posted annually or every 6 months on the implementation status of the proposal. In the case of a negative response, the users are informed that the procedure is finished, with an explanation as to why there will be no action on the suggestion made within the proposal (interview with administrator).

Results
Regarding transparency, the website itself provides data on the number of open topics and responses from the first year after launch, dating from 2009 to 2010. During our interview with a researcher familiar with the tool, we were informed that there are statistics on the number of open topics, which are sent to everyone recorded on an email list. Information on how many proposals ended up being effective, however, is harder to come by. The responses of the governmental bodies are all published online. When inappropriate comments are hidden, they are replaced by a message from the moderators explaining the reason for its removal.
Predlagam vladi offers fairly detailed statistics on the first year of proposals handled on the website, giving us the following insights: In less than 24 hours of the website going online, more than 80 proposals and draft proposals had been created. Between November 2009 and November 2010, a total of 1.201 proposals had been generated, 7.021 comments published and a total of 11.521 votes cast. Of these 1.201 proposals, 251 went on to be submitted for consideration to the competent government authorities, with 458 being rejected for not receiving sufficient votes. Other reasons for exclusion were, for example lacking concrete solutions to the matter at hand or being too similar to a proposal which already existed (Predlagam vladi 2016b). Within the first year there were 235 responses issued by the competent governmental authorities. On average these responses were sent within 23.9 days. A total of 11 proposals of the first year were ultimately successful in subsequent measures being taken.
One such successful example is the proposal for a reduction in the rate of valueadded tax for baby diapers: On 9 March 2010, the Ministry of Finance supported the proposal and stated that it would be included in the following amendment of the Law on Value Added Tax. On 2 April 2010 this was implemented. Further examples of proposals which have been successfully implemented in some way are the proposal to upgrade the software of state administration computers to allow the viewing of . odf formats, a motion for a clear position on the pandemic flu vaccination, a proposal to increase the number of parking spaces for motorcyclists, a motion to be allowed anonymous votes on the Predlagam vladi website and a proposal to extend the voting time of Predlagam vladi from 7 to 14 days (Predlagam vladi 2016b). When looking at the successful cases from 2010, one can observe that many of these issues didn't have to do with policy changes but were related to software and the Predlagam vladi tool itself. It would be of interest whether the more recent successful proposals are of a similar nature. Additionally, some issues tackled in successful proposals were already present in the public sphere and part of discussions prior to being submitted to Predlagam vladi. It is therefore unclear whether it was the e-tool which led to a successful implementation or if the issues would have been implemented regardless (interview with researcher).
According to the website of Radiotelevizija Slovenija (www.rtvslo.si), Slovenia's national public broadcasting organisation, only 1.5% of 1505 responses from government authorities were positive by 2015. By the end of January 2015, a total of 13,088 users were registered with 5185 proposals having been submitted, 1748 of which garnered enough votes to be submitted to the competent authorities. Of these 1748 proposals, 1505 received a response. According to the report, half of the proposals were rejected on the basis of the proposed solution already being (or having been) in the process of implementation. Of the 1.5% positive responses, many were related to the accessibility of e-services (Cerar 2015). Many of the proposals are difficult to address, such as defining nuisance dog barking, the prohibition of church bells, the castration of rapists, a decrease of television commercials and the reimbursement of schooling costs by physicians who go abroad. This can help explain the low number of proposals receiving positive feedback from government authorities.
The interviewed administrator reported a total of around 35% of proposals being submitted to competent authorities, while the number of consequent actions taken by the government is very low, an estimated 40 out of 2000. Even if the response is positive, there are still further obstacles to overcome regarding implementation. An example of this was a case in 2009 where the decision was taken that the suggestion should be implemented, but due to the amendment being relatively small, it was decided to wait for a different matter to initiate the process of the law being opened and revised. In March of 2016, 7 years later, this finally happened and the suggestion from 2009 could be implemented. Bureaucratic procedures often hinder the realisation of positive responses in this way.
Something which managed to successfully increase positive responses from government authorities was the decision to endorse five or six proposals on a monthly basis: "We at one point wanted to give these suggestions, before we send them to the competent ministries, a bit more weight. [. . .] What we did was, we, with the cooperation from the General Secretary of the government and the chairman of the Committee for State Order and Public Affairs, pick 5 or 6 suggestions monthly that we think are really good and that should be implemented, and then we put them forward to the committee and they discuss the suggestions. The committee consists of representatives of all the ministries" (interview with administrator). These proposals are then submitted to the competent authorities, the idea being that they now carry more weight and are more likely to succeed. Interestingly, not only the success rate of endorsed proposals was raised through this method, but that of all proposals in general, indicating that the importance of the entire tool was raised in the eyes of the ministry representatives. Within a year the amount of positive responses received from government tripled compared to the six previous years (interview with administrator).
When there is a change in government, some users attempt to pass proposals which were previously rejected, a strategy which has proven successful in the past with a new government positively responding to a suggestion which had received a negative response before (interview with administrator). The same interviewee speculated that despite the high amount of negative responses, users appreciate the feedback the ministry provides, as it shows the ministry is giving adequate consideration to their suggestions. They point out that there is, however, no data available as hard evidence. The whole system is currently being upgraded, with the idea of presenting positive examples of successful proposals on the front page. This way it can be communicated that good suggestions can lead to implementation. The interviewed researcher was less optimistic concerning the government feedback: rather than letting the participants feel heard, they claim the responses are often perceived as standardised.
This interviewee also raised interesting criticism with the observation that the format of the tool was too open and the recommendation that it should be more structured with more information given on what kind of input the government wants from citizens. This of course goes hand in hand with limiting the scope of participation and bureaucratising the manner in which a proposal must be made. The interviewed researcher added that the open structure of the Predlagam vladi tool would not be an issue if there was enough staff which could process the ideas.

Introduction
After a large corruption scandal in 2005, the plea of civil society activists for participatory budgeting in Berlin was granted by the borough council that structurally implemented a participatory budget (PB) process in Berlin Lichtenberg. The instrument became part of a citywide administrative reform project to increase the effectiveness and efficiency of public spending (Röcke 2014: 133). The council formulated the following purposes for the PB: "mutual agreement in policy decisions; effective and fair budgeting; transparency; and educating citizens about financial matters" (Shkabatur and Fletcher 2016).
Increasingly PBs are employed in Germany, starting from 1 in 1999 to 14 in 2007 and on to 96 in 2013 (Ruesch and Wagner 2014). Most of them have the objectives to modernise local governments by participation and to become more responsive to the needs and wants of citizens. But they are no instruments for direct democracy. "Citizens do not have the right to make a final decision, this is a consultative procedure" (interview with administrator). The PB process in Berlin Lichtenberg thus concerns citizen involvement in agenda setting and providing concrete input for policy options.
The PB model was created in a workshop with policymakers, party foundations, civil society members and experts and was organised by the Federal Agency for Political Education. The basic structure of German Participatory Budgets has three phases (Ruesch and Wagner 2014): 1. Information provision about the budget and the participatory budget procedure 2. Consultation and participation of citizens, who contribute by making proposals, providing feedback on proposals and/or by making a planning of the budget 3. Decision-makers and civil servants explaining the outcomes of the process and justifying their decisions The rules of the process in Berlin Lichtenberg are decided upon by the administration and the district executive. Some public meetings were organised to discuss the process and a group of organised citizens can put forward suggestions for changes in the process (Röcke 2014: 148-149). As in all German PB processes, the outcomes of the participatory process do not have a binding status and largely depend on the willingness of the borough council or citizen jury to adopt them (Van Hulst et al. 2016, interview with researcher). In annual reports effects of the PB and choices for or against citizens' proposals are justified (Van Hulst et al. 2016). Since 2012, the software applied by Berlin Lichtenberg is based on an open source software and adapted to the needs of the tool (interview with administrator).
The maintenance costs of the participatory budgeting are estimated by the former mayor: "Each year Berlin's borough of Lichtenberg spends 60,000 euro on participatory budgeting. This is used for the Internet presentation, brochures and rentals of event space" (Emmrich 2010: 69). The municipal budget that is opened up for citizen participation and consultation in an annual cycle is about 10% of the total borough budget. "The budget in Berlin-Lichtenberg comprises 576 million euro, of which almost 90% are obligatory payments. This leaves 32 million euro as the subject of participatory budgeting" (Emmrich 2010: 69). These discretionary expenses comprise: support of public health, business counselling, planning parks and free space, public libraries, general support for children and adolescents, cultural services of municipal institutions, the music school, voluntary services for elderly people, sports, maintenance of green spaces and playgrounds, planning of green spaces, support of the local economy and an adult education centre (Van Hulst et al. 2016;Röcke 2014: 198). Within this budget personnel costs are included, which makes the actual budget for PB much less.

Participants
The target group of the participatory budgeting in Berlin Lichtenberg which is described on the website is "everyone who lives or works in Berlin Lichtenberg". However, it doesn't seem to be verified if participants really live or work in Lichtenberg unless they want to cast a vote (Van Hulst et al. 2016). The requirements to participate can be found on the website as the "Rules of the game". In the online discussions participants are encouraged to use their own name and surname, but it is not obliged. When participants propose a plan, they are obliged to give their contact details (interview with administrator).
From 2008 onwards a randomly selected group of citizens (10% of the population) receives a personal invitation letter for the public neighbourhood meeting signed by the mayor. In 2013, flyers with information about the PB in German, Russian and Vietnamese language are offered to inhabitants, which also included an invitation to make a proposal. Other flyers are designed especially for youngsters (interview with administrator). The researcher we interviewed states that next to these flyers, decentralised meetings in particular were an important way for community workers to reach people with a wide range of interests, backgrounds and experiences.
Different channels for communication and interaction were used for the whole participatory budget process, which enlarges the possibilities for different groups of inhabitants and employees in the area: • The website (in German), including a blog, a section for praises and criticism, an agenda with events, an online voting tool (and its results) and a link to the total budget overview in xls. • Offline information by a newspaper and possibility to write letters and proposals.
• Community centres that organise public meetings and can be reached by phone.
In addition, childcare is provided during every public meeting and a sign language interpreter is present. • Local borough coordinators spread information on participatory budgeting proposals.
This multichannel approach and the timeline with short information about the decision-making process around the proposal are used to increase the userfriendliness and accessibility of the PB process.
Despite these measures and the positive reviews, there is still room for improvement: "I think here the problem is that it's a bit too formal, too technically sophisticated, but not really anchored in a sort of political activism or community activism of really going out, getting to the people, getting them involved. [. . .] I think this kind of political communication and inclination of why this is important [and] what change it makes [. . .] is probably missing a bit" (interview with researcher).
Röcke (2014) reports not a very large mobilisation of citizens participating in the whole process (online and offline), but the "digital turn" increased the number of participants every year. In addition, Lichtenberg profits from a growing participatory culture: "We have the luck and the good conditions in Lichtenberg of a high willingness of citizens to participate. Participatory budgeting is just one component of citizens' participation" (interview with administrator). However, participation always depends on time restraints and personal interests (interview with administrator). Although the participation process only attracts a small part of the population, the main goal is to collect good proposals for the neighbourhood. Assessing preferences among a high number of participants is therefore not the most important requirement for a successful process.
Not much research has been done about characteristics of participants and whether PB in Berlin Lichtenberg is inclusive. "The fact that participation is based on self-selection often excludes the less well-off residents of Lichtenberg. The under-represented groups include immigrants, elderly residents, uneducated residents, and young families," suggests Van Hulst et al. (2016: 42). Shkabatur and Fletcher (2016) report on Participedia.net that "Participants were mostly young and middle-aged citizens of up to 50 years old, with a level of education higher than in the general population."

Participatory Process
Participants can make proposals to spend money, to save money or to make proposals that do not entail costs for the discretionary budget. Participants can formulate proposals worth a maximum of 1000 € for activities or facilities in the neighbourhood. Since 2008 participants can also make proposals regarding construction investments in the district (of around four million euros) and neighbourhood projects that are provided by voluntary associations but financed by the district (Röcke 2014). The proposals can be put forward in written form, via the Internet and at neighbourhood assemblies (Röcke 2014: 145, interview with administrator). The process runs both on-and offline. "There should always be parallel means. Because not everyone is comfortable with just one way [of participating" (interview with administrator). Most proposals on the Internet are formulated in only one to three lines, without links, further justification or elaboration (Röcke 2014). Participants can also make comments on proposals of other participants, vote on a proposal and write blog posts.
To aggregate and weigh the proposals several strategies are combined. Participants make a list of the project proposals in an order of priority within public meetings organised in the 13 neighbourhoods. The next step is that the lists with the five priorities of the 13 neighbourhood meetings and the top ten from the Internet "vote" is sent as a survey to a representative sample of the population (50.000 randomly selected households) (Röcke 2014: 145-146). The respondents are asked to make their own list of five projects that they perceive to have most priority. Successively, the district council and the Parliamentary Committees make the final decisions on the allocation of the budget, which they explain in a public meeting (Röcke 2014). The citizen jury decides upon the allocation of the neighbourhood budget proposals (Van Hulst et al. 2016). Proposals that do not fit the established budget can be rejected by the district council or it can be decided that people may vote on it (online or offline). If there are sufficient votes in favour of the proposal, the district council considers whether the proposal is financially feasible, and it will be implemented (Van Hulst et al., 2016). Feedback on the implementation of the proposal or an explanation of the rejection is provided per proposal.
Participants can monitor the decision-making process online. The proposals are shown on the website and their status is indicated with the help of a three-colour system. Green indicates that a proposal has been carried out, red for a proposal that has been rejected, and yellow for a proposal that is still under discussion. This contributes to the user-friendliness, but sometimes proposals have the status of "under discussion" for quite a long time: in that case it's not clear for participants what will happen to these proposals (Van Hulst et al. 2016). In addition, "the whole process is monitored by a committee (Begleitgremium) composed of around 15 people (civil society representatives, civil servants and politicians)" (Röcke 2014: 145). The understanding of the participatory and political processes by Lichtenberg citizens has matured over time.
The tool gives the possibility of interaction and exchange of views. Van Hulst et al. (2016) only found a maximum of 15 comments on a proposal, which limits the diversity of views exchanged. Also, in the public meetings participants mainly focus on their individual wishes and an overarching perspective and assessment of the budgets is not attained (Röcke 2014). "A rather weak point is this facilitative discussion aspect-to really have an informed discussion about why do we take this decision and not another one" (interview with researcher). The interviewed administrator also affirms that participants not often use the possibility to interact or discuss proposals, and the organisation wished that would happen more often. This functionality will receive more emphasis in the near future (interview with administrator).
The website displays information on data protection: What data is saved, what it is used for and how it can be protected. It is an issue with the organisation, but no abuse has been reported so far (interview with administrator). Sometimes they check whether proposals are put forward by "real" people. But no hacks or influencing of voting outcomes have occurred (interview with administrator).

Results
In June 2016, the website counted 821 proposals since 2005, from which 426 proposals have been implemented. Although the participatory budgeting is consultative, it is quite influential on the decisions made (Ruesch and Wagner 2014: 12). For each proposal a brief account of the government is available online that explains why a project is or is not yet achieved or pending, which becomes visible by just clicking on the title of a proposal. Decisions are explained and evaluated in public meetings as well (Röcke 2014, interview with administrator). The participatory budgeting committee also has the task to discuss the evaluation of proposals with the participants (interview with administrator).
In addition, an annual accountability document is published on the website. However, Röcke (2014: 148) is quite critical on this form of accountability, because she saw many examples of only very short, simple rejections. This is an important concern, as most procedures in Germany are consultative and citizens do not make decisions concerning implementation; accountability is crucial for German PB. Without sufficient feedback on the final use of their input in budget planning, citizens are not motivated to invest their time (once again) in participating. This process of accountability can be further developed in the design of the online tool for participatory budgeting. Ruesch and Wagner (2014: 10) point to another example of PB in the city of Bonn, whereby means of visualisation of the overall structure of the municipal budget is presented in an understandable way for the public. Röcke (2014: 134) concludes that the selected approach is "technically sophisticated, but citizens have hardly any procedural and political powers in the process; in addition, the process of accountability is not very well developed". On the other hand, she concludes that the procedure has led to a greater dialogue between citizens, officers and politicians, although the deliberative quality of the procedure could be better (Röcke 2014: 149f.). At least this process works in a way that it increases politicians' trust in citizens: "[The administration] now also sees that citizens have expert knowledge, and that decision-making is not just based on files produced behind a desk. And that the triangle of politics, administration and citizens has already worked in other contexts" (interview with administrator). The tool did receive the European Public Sector Award and the Theodor Heuss Medal (Ruesch and Wagner 2014: 12).
The costs of the maintenance of the participatory budgeting are 60.000 euros; however, high costs for personnel are not included here. We cannot make a harsh judgement on the cost-effectiveness. The (former) Mayor Christina Emmrich argued PB leads to a win-win situation: "The citizens gain more transparency, a say on the budget and requirement-suited priorities and they see they are taken seriously. Political bodies gain objectivity and higher quality of budget-political discussions, more legitimacy and more identification of citizens with the community. The administration gains more information and proximity to the people, as well as more transparency in the setting of priorities" (Emmrich 2010: 69).
As stated earlier, Lichtenberg seem to have a participation culture. Throughout the years, it has been shown to citizens that they can participate in various decisionmaking processes. "This has a positive effect: it motivates [to participate]" (interview with administrator). Increasing trust in government by obtaining legitimacy for decisions and making the decision-making process more transparent was the motivation from the start, she further elaborates. But whether PB has in fact had this effect on local citizens is hard to tell (interview with administrator). The interviewed researcher sees the potential as well, but agrees it is difficult to assess: "I think it [is] a really well-organised, transparent process, which clearly states the different roles and duties and rights of the different stakeholders in the process. And which also gives information on where the money comes from, where [it goes], what we do with it and why. So I think that maybe it's more in the potential way. It has the potential to increase trust. [. . .] But if this actually is the case is another question" (interview with researcher). Those who participate in these types of processes often already have a certain level of interest in politics, "So it is very difficult to reach those who have lost any contact with the political atmosphere" (interview with administrator).

Introduction
In the Netherlands, the official governmental website www.internetconsultatie.nl has been built as a platform to organise online consultation in the national legislative process. The public is consulted about draft bills, general orders in council, ministerial decrees and policy notes. E-consultation is part of the preparatory stage, the stage before the Council of State assesses the legislative proposal and the parliamentary debate about it.
The first e-consultations took place in 2009. After 2 years, the experimental e-consultation procedure was evaluated. The positive evaluation led to a structural implementation of e-consultation within the legislative process. The aims of the official website for online consultation are to increase the transparency of the lawmaking process, to offer new opportunities for participation and to improve the quality and practical feasibility of laws and regulations (website). The site has a function for "monitoring citizens" (van den Hoven 2005) (and monitoring organisations) as well. There are a lot of people that visit the site without contributing to the online consultation but only look for information about legislative proposals. Before, the lawmaking process was a black box and the consultation often included only a limited group of organisations they always consulted (interview with two administrators). An official procedure to consult individuals and organisations did not exist.
Two administrators of the Ministry of Security and Justice have been the main initiators and organisers in the past. Later on, most of the ministries were involved in the development and use of the tools (interview with administrators). One of the two originators is still responsible for the management of the website. For the specific consultations administrators involved with the specific legislative proposal are in charge. The website to consult the public is used by different ministries. As of our reference date (October 2018), 1036 online consultations have been completed (Internetconsultatie.nl). The House of Representatives has used the site only ten times to consult the public.
The website shows all online consultations that are running in chronological order. The user can look for specific topics via the search function, which results in a list of running and closed consultations. Clicking on a specific consultation leads to information about: the aims of the legislative proposal, the target group(s), the expected effects, the aims of the consultation, the procedure of the online consultation, a link to the proposal and sometimes links to additional information on the subject (as policy documents, parliamentary documents or even media articles).
In another click one can participate in the online consultation. Sometimes specific questions are formulated around the legislative proposal; in other cases the general question "what do you think about this proposal" is posed. One can participate by providing a response to the question(s). In the next screen, participants are invited to upload their own document with their reaction on the legislative proposal. One can skip one of these two. In the third step, participants are asked for their name and email address and if they want their reaction to be published.

Participants
In principle, the target group is as broad as the Dutch population, private individuals as well as staff of businesses and civil society organisations. More specifically, the target groups of actors that are affected by the particular legislative proposal naturally vary amongst proposals. The e-consultation is especially directed at groups of individuals and organisations beyond the "usual suspects" that would normally be included in consultation processes of ministries (interview with two administrators). Broek et al. (2016) record more than 17.000 public contributions of participants in e-consultations from 2009. The actual number of contributions is, however, much higher, because visitors can decide whether they want their response to be published or not (Ministerie van Veiligheid en Justitie 2011). The number of responses per bill differs considerably and depends in part on the subject, the anticipated effects of the bill and the stakeholder groups that are affected by the legislative proposal. On average around 20-public-reactions per consultation are measured by Broek et al. (2016). Most-public-reactions seem to be placed in relation to legislative proposals around education (Broek et al. 2016: 41-42). A survey among users of the site (N ¼ 171) showed that the group of participants consists mainly of highly educated individuals and that white males dominated the group of participants (Broek et al. 2016).
It is not allowed to respond completely anonymously in the e-consultation. Participants are obliged to enter their name and email address (which can only be used once). The official Digital Identity system of the government is not used. The administrator explains that they (the organisers) sometimes wonder: "Shouldn't we do more to check a person's identity, with regard to security? That is, I think one of the issues that do ... At the same time, our main goal in an internet consultation is collecting ideas and it does not really matter if the ideas are offered by let's say Shelwin or the grocer at the corner." The name of the organisation or initials of the individuals, their last name and place of residence are published on the website unless participants object against the publication of their input (website internetconsultatie.nl). Publication could be an obstacle for people or organisations with relevant specialist knowledge to participate if they don't want to share the information publicly with their competition. For instance, "tax issues are always sensitive issues" where confidentiality is often appreciated. And, "in economic affairs, the know-how is often located outside the ministry. So, if you want to come up with legislation to bring to an end certain trends or developments in the industry people within that industry know more about that than people at the ministry" (interview with researcher). With regard to this input confidentiality is clearly appreciated, but the public and parliament do want to know what information has been included from whom in the formulation of the legislative proposals and request publication. The confidentiality has thus become controversial.
Communication strategies show enormous variation in the different trajectories of legislative proposals (Broek et al. 2016). In some cases, there was no publicity at all; in other cases specific organisations that are contacted via letters, emails, phone or social media as LinkedIn are used. Social media are not only used to gain attention, but also to organise and stimulate an interactive discussion with citizens and professionals around the policy issue. For example, social media have been employed to engage individuals and organisations to discuss proposed regulation around drones or proposed regulation about energy (Broek et al. 2016). However, these interactive strategies are only used once in a while. The website internetconsultatie.nl itself is not interactive at all.
In general, the communication and engagement strategies to involve particular target groups can be strengthened, conclude Broek et al. (2016: 74). The administrator argues this is largely a task that must be tailored for every specific proposal (interview with administrator). She also stated that they work on a connection of the e-consultation site to the website that provides more information on legislative processes, which is "a system of a legislation calendar where you cannot only see 'this is a bill and it's in this phase of the procedure' [but] that you can also see when an e-consultation is organised, and which consultation responses are entered to that proposal. That would enable people to follow the different phases of the legislative process" (interview with administrator). This has not been realised yet.

Participatory Process
Participants can respond to proposed bills, general orders in council, ministerial decrees and policy notes by answering questions posed by the administrators involved or by raising questions or concerns themselves (with the possibility to upload their own document). Subsequently, participants can subscribe to receive information on the further process: "Then you'll get a notification when a report is placed on the website or when the proposal is submitted to parliament or when it proceeds from the Lower House to the Senate" (interview with administrator). It is unknown how many participants make use of this service.

Lack of Interaction
As discussed before, the website for the e-consultation doesn't entail possibilities to interact with stakeholders. Some administrators that apply Internet consultation feel a need for more interaction between participants and between participants and administrators, and participants feel this need even more (Broek et al. 2016). In contrast, most administrators involved would rather see interaction in earlier stages of the legislative process than in the phase of the online consultation. Broek et al. (2016: 75) note that a general strategy with regard to interactivity would probably be undesirable; this needs to be deliberately decided upon per consultation.
The researcher in the interview suggests that a categorisation of reactions can stimulate interaction and the quality of the contributions. "I actually think that the dialogue that would arise if people see their reaction to the site, that others see the reactions and some kind of interaction could occur, then I think that would lead to a much richer result" (interview with researcher). Earlier reactions may inspire other participants, the researcher argues, and this could lead to a kind of co-creation. However, the administrator has her doubts in relation to such moderation. She questions whether such moderation could be done neutrally-without steering the discussion in a certain direction-and whether capacity within the ministries can be reserved to moderate such interactions (interview with administrator).

Interpretation of the Input
Procedural conditions are formulated about the Internet consultation process. These procedural requirements range from securing privacy to communication targets: that actors who might be affected by new regulation should receive information about the consultation and participants should receive feedback about the results. The requirements could be further improved and complemented, Broek et al. (2016) argue, on how to process the input. Some civil servants feel the need for a clearer assessment framework for taking the decision to deploy Internet consultation or not.
The input is interpreted by the administrators involved with the particular legislative proposal. How the input of participants is handled depends on contextual factors like the number of reactions, the quality of the input, the administrators involved and the time available to process the input (Broek et al. 2016). Sometimes an organisation is invited to the ministry on the basis of its input to further discuss the issue (interview with administrator). Further procedures could make the processing of input more solid.
The process of interpreting the input is not always publicly clarified by the administration. And although the goals of different consultations vary, the researchers question whether this interpretation process is as professional, systematic and methodically sound as it can be. "I spoke with someone who said 'I got 1100 responses, I went to my boss and asked 1100 comments, how many do you think I can process per day?'. I think that's an example of a totally non-systematic approach. But others say, no you just get 1100 or some number of responses and you are going to read these one by one but you will first cluster these reactions, are these from organisations and individuals; in supporting and not supporting reactions; new arguments or existing arguments. Others said we put everything in an Excel file and then we deal with them systematically before we read them in detail. I think it's important that other ministries would do that as well" (interview with researcher). Hence, one of the recommendations in the evaluation of Broek et al. (2016: 74) is a professionalisation of the interpretation method with which the input of all participants is handled. More attention is needed for a systematic procedure to process the input, for the analysis and the interpretation. Civil servants at the ministry need to have these research skills.

Results
The results of the participatory process are somewhat ambiguous. On a positive note, participants, as well as administrators, do indicate that Internet consultation contributes to the transparency of and participation within the legislative process. The transparency of the legislative process has increased, as argue 87% of the administrators and 65% of the participants surveyed by Broek et al. (2016). Moreover, in the studies of 2011 and 2016 two-thirds of the administrators evaluated the contribution to the quality improvement of legislation positively; they received useful input via the e-consultation. At the same time, the administrators indicate that often little room exists to manoeuvre: not much can be done with the input of participants, since the proposals are already seen as almost finalised.
Although the policy is to structurally e-consult legislative proposals, the researcher and the administrator question this default. "That leads to several ministries thinking 'well okay then, but then it shouldn't be too complicated [for us]. We just throw it on the site with a general question and then we'll see" (interview with researcher). The administrator of internetconultatie.nl confirms that civil servants within different ministries want to "tick the box" within the legislative process, without seeing the utility. Moreover, public consultation is not always that appropriate, as for technically complicated juridical law proposals or the implementation of laws, as the researcher suggest: "That can sometimes be really hard ... Firstly, to consult about it because there is not always a lot of room to manoeuvre." And secondly, with regard to the time schedule: "The European Commission has of course penalties for non-timely implementation which are considerably high, so the government has made special arrangements to deal with implementation proposals more quickly" (interview with administrator). Internet consultation could frustrate this tight schedule.
On the side of the participants, they feel that they do not have that much insight into what has been done with their input. Often a report about how the input of the e-consultation has been used in the legislative process is missing (Broek et al. 2016: 76). However, the output of the e-consultation process is often described in the explanatory memorandum alongside the legislative proposal for parliamentary debate, though the reflections in the memorandum are not always as elaborate and transparent as they could be. The extent to which it is explained how the participatory input is exactly used varies considerably. And participants are not always able to find this information themselves. It is thus understandable that participants have great doubts about the impact of their contribution on the quality of laws and regulations (Broek et al. 2016: 64). In an earlier evaluation this was also a point of concern. Often quite some time elapses before the website report on the most important results and the changes made to bills as a result. "Participants indicate that participation should be rewarded, for example by ensuring that responses are published on the site without delay" (Edwards and De Kool 2016: 42). "From the perspective of learning processes, the system can be considered as too slow. People respond, but [information about] what's been done with that reaction reaches them way too late. [. . .] I think you can learn so much from the dynamics of that process, around which a dialogue should be organized" (interview with researcher).
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.