Through Liquid Democracy to Sustainable Non-Bureaucratic Government

Alois Paulin, 2014

CeDEM 2014, Krems,


We explore how the principles of liquid democracy could be applied in Self-Service Government (ss-Gov) for collaborative decision making as an alternative to modern parliaments. We summarize the principles of Sustainable Non-Bureaucratic Government (SNBG) and explore how liquid democracy could be utilized to enable democratic end-to-end self-management of jural relations.


Liquid Democracy, Sustainable Non-Bureaucratic Government, Self-Service Government, e-Government, e-Democracy


A recent study by Paulin (2013) explored the computability of jural eligibilities by means of modern ICTs as a method for enabling what is named there Self-Service Government (ss-Gov). SS-Gov is a model for governing a society in which a dedicated public administration system (a bureaucracy) is not required for asserting a subject’s (e.g. citizen’s) jural eligibilities in a particular context, but rather the eligibilities can be calculated by means of relational algebra based on raw data about the subject’s jural facts.

This raw data, which serves as a basis of a subject’s jural eligibilities, is read and written by active and passive jural subjects, who again do have the calculated eligibility to consume, respectively provide, this data. Thus, in theory, a closed circuit is established in which subjects of various jural statuses interact with a relational system of jural data, which through self-service manipulation of the raw data stored within enables transformation of jural eligibilities of subjects in juropolitical societies. Paulin (2013) names this approach to the technical determination / calculation of jural eligibilities in a juropolitical society Constellation-Based Reasoning (CBR), whereby he compares this methodology to “a key opening a pin-tumbler lock, where the key due to its specific shape moves the pins into the right constellation, which allows the lock to be opened” (ibid., p.1775). The lock, then, defines the constellation and definition of the required data (defined as a relational set), which must be satisfied by the key, i.e. the data of a stakeholder and/or context in a situation, to unlock a particular eligibility in a given context. Thus, ss-Gov enables a new model of government in which eligibilities (e.g. rights) are not obtained in form of credentials from state authorities through administrative proceedings, but are rather determined by means of CBR. The mathematical basis for the determination of eligibilities enables homogeneous, standardizable technical storage, rule-based generation and –access to the raw jural data, and hence its sustainable storage as structured data in digital systems. This approach makes the existence of an administrative middle-layer (the bureaucracy) hypothetically obsolete, without however systemically rejecting or disabling such system.

In this paper we are to explore if CBR can be utilized to enable non-bureaucratic collaborative decision-making through liquid democracy, such as it might be used in republican juropolitical systems for empowering political leaders and -representatives, or for creating common policies and jural regulations. In chapter 2 we shall first elaborate the theoretical framework by describing the concepts of Sustainable Non-Bureaucratic Government (SNBG) as a vision encompassing non-bureaucratic collaborative decision making, and describe the principles and history of Liquid Democracy (LD); in chapter 3 we shall explore how modern parliamentary decision-making could be handled either through the principles of SNBG without changing the existing structure, as well as how the same would be transposed to the realm of LD.

Theory Frame - Concepts for Self-Management

The modern system of public administration and its dependent stakeholders relies on an ever-increasing influx of capital (through e.g. taxes or public resources) to sustain itself, which it does by constantly increasing its legitimacy that bases on an increasing self-imposed handling of new regulations, responsibilities, and taxes (cf. Shleifer and Vishny 1993, 616). Within such bureaucratic ecosystem, informal networks take control, which Banfield (1975) terms machines. These machines are communities, which exist based on a system of exchanges of favours (such as jobs, opportunities to make money by legal or other means, perks, etc.) amongst officials and external interest groups. Such hierarchies, which “arise from extra-legal, if not illegal, arrangements, are ad hoc, and must be continually renewed by ‘deals’ in order to prevent them from collapsing” (ibid.).

Increasingly demanding state machines are an everlasting issue in any civilization – limits of bearable growth of government requirements (manifested through taxation) however are easily reached and, as Adams (2001) argues, have caused the dusk of many once strong civilizations, including Ancient Egypt, Greece, Rome, the Aztec Empire, and the European Empires.

A promising attempt to curb the impact and burden of state machines on the society was the introduction of office automation technologies and bespoke ICT systems introduced in the last couple of decades to automate routine tasks of government agencies and to provide self-service access to government information and services. Computerization of the public sector has in large parts of everyday bureaucratic chores shifted bureaucratic discretionary power from a predominantly street-level bureaucracy with “large numbers of faceless officials whose freies Ermessen (discretionary power) could cause an open society to be smothered in the bud” (Bovens and Zouridis 2002, 174) to a system-level bureaucracy in which information is automatically collected from various sources and applications from citizens can be instantly handled – e.g. approved, rejected, or set aside for manual inspection.

The digital era without doubt influenced the public sector. The available technology makes it possible for bureaucratic networks to easier manage their duties, and to better supervise their subjects and its own kin. The digital era however seems to be merely another noteworthy change in environment which the bureaucracy aims to survive – the modern bureaucratic culture, which began in the mid-17th century (Walter 2011, chap. 2), after all, is too big to fail, or is it?

The continuity of the bureaucracy, “which survived the changes from monarchy to republic, from republic to dictatorship, from dictatorship to democracy” (König in: Walter 2011, 27, own translation), has in the recent past been challenged by two novel concepts – new public management (Osborne and Gaebler 1992) as an organizational, and e-government as a technological challenger. Latter, driven by powerful myths (Bekkers and Homburg 2007), has constructed a hype portraying technology as the enabler to new and better government featuring transparency, rich participation, and self-service one-stop-shops. Alas, a deep change in structure, away from the well-established bureaucratic approach, has not been part of its vision.

The e-government approach, however, has many flaws: As Bekkers & Homburg (ibid.) emphasize, e-government artefacts frequently require the coordination of a multitude of heterogeneous back-offices within the public administration. Aiming for progress in this regard often results in what they call battle of the back-offices, which prevents the development of a sustainable, goal-oriented e-government system-of-systems.

Aside from this, Paulin (2013) elaborates three hazards of e-government, which make this approach unsustainable: hazard I (expiry date) targets the dependency of monolithic e-government artefacts on law – such artefacts are developed according to law which is valid at design-time, but which will inevitably change sooner or later, requiring either a costly reengineering of the artefact, or making free changes to law unlikely due to systems that are simply too big to be changed; hazard II (monopolization, corruption and exclusion) targets the gap between the legislator defining the functional characteristics of e-government artefacts and their possible many heterogeneous, non-interoperable, technical instantiations – Paulin (ibid.) gives examples of the European e-ID and the Slovenian system for electronic registered mail delivery, which both led to nationally favoured technical instances, which discriminated other providers; the 3rd hazard (legal certainty) finally targets the challenge how to provide e-government artefacts whose internal processes and interfaces would follow the core jural principle of legal certainty, whereby it is emphasized that users of such systems should be able to rely on jurally clearly defined and stable interfaces and system behaviour.

At the end of the day, e-government remains a bureaucracy-driven approach that supports the continuation and influence of the latter in the digital era. In search for a sustainable self-management of juropolitical societies, we shall further below explore concepts for self-managing jural relations, as well as self-managed collaborative decision making.

Sustainable Non-Bureaucratic Government (SNBG)

Self-Service Government (Paulin 2013) through its Constellation-Based Reasoning (CBR) concept represents a scaffolding for creating, storing, retrieving and changing jural facts based on which eligibilities of jural subjects can be determined.

However, while this model provides a feasible approach towards a sustainable base infrastructure for storing and communicating jural data, it represents only a part of the complexity required to bring the vision towards a form of government that does not require a bureaucratic machine for administering jural relations in a juropolitical society, into reality. Thus, if constellations of jural data enable eligibilities, then naturally one must ask how to recognize such constellation – thus, domain-specific semantics, data structures, etc., must be defined, which make it possible to recognize for example a specific constellation of data representing a university degree, a driving permission, a land parcel, or a political representative’s mandate. All these however must remain independent from the infrastructure responsible for creating, reading, updating and deleting (CRUD) the jural facts, in order to ensure sustainability of the infrastructure.

Let us therefore continue thinking how self-managed, sustainable non-bureaucratic government (SNBG) could be established by means of ICTs. Expressed through a technology stack, we may see SNBG as an architecture comprising five layers:

The first, bottom-most layer is a technical communication network, such as e.g. (but not mandatory!) the modern Internet. This layer is about exchanging arbitrary messages required for telecommunication.

Figure 1: Five layers of SNBG: Layer #2, where the jural data is stored, is by definition sustainable. The individual eligibilities are determined by means of layer #3 semantics, whose range shifts through time. Layers #4 and #5 are unsustainable and adopt to current fashion, without influencing the concepts from layers #2 and #3. The right-facing arrows illustrate the evolution of the respective artefacts through time.

The second layer is about a content-agnostic technical infrastructure that enables arbitrary communication and manipulation of jural facts. An instantiation of the ss-Gov model would be a suiting approach to deliver infrastructure for this layer.

On the third level, a contextualization-layer would provide artefacts that would define domain-specific data structures, semantic conventions, identity, etc. This layer would enable interoperability between nodes that would constitute the network defined on the 2nd level, and provide the corresponding semantics. This layer then would answer questions such as the one posed above, defining for example the structure and semantics of a constellation that would denote a land parcel, a university degree, a diplomat’s jural status, etc.

A clear separation of this layer from layer#2 is crucial, as the semantics and structures of layer #3 will change through time – for having a university degree for example, the requirements of tomorrow might be slightly different compared to the requirements of today or yesterday; nonetheless, the complex concept of a university degree, which entitles individuals to certain eligibilities (e.g. only individuals with a university degree are permitted to compete for civil service jobs) may survive many changes in its intrinsic composition, until perhaps in some point in the future this concept might lose its original value. (An example of such complex concept which through time became obsolete are for example the aristocratic titles of the Austro-Hungarian monarchy – once they enabled the access to government jobs, but became after the break-up of the monarchy suddenly of no value.) These complex concepts of the contextualization layer act as locks in CBR reasoning, which are unlocked through the fulfillment of the required data constellations.

The contextualization layer might be established and governed by professional guilds, who would find proper definitions and micro-architectures for complex jural concepts. Thus for example, a guild- or de-facto-standard could emerge, which would define on a European, or global level, how a bachelor degree is to be represented by layer#2 jural facts. This would enable a subject, which graduated from a British university to enjoy eligibilities associated with having tertiary education in Slovenia without the need for additional homologation – the British university in this case would be the technical host of the layer#2 jural facts, which could be referred to in order to utilize them as a key (or part of it) to unlock eligibilities in other countries.

On the 4th layer, a unified approach to describing processes needs to be found, which would engage the contextualized constellations from layer#3 into business processes (level#5), that would constitute the business logic of complex information systems, which could be used by lay (i.e. not adequately literate in terms of data-level command of ICTs) subjects to interact with the network of jural relations stored on layer#2. A process here is to be understood as a system consisting of multiple stages of hierarchically interdependent CBR locks, where unlocked earlier locks present part of the key for later ones. (E.g.: to be selected for a civil service job, one must have first applied for such job, whereby in order to apply for such job, one must have prior fulfilled all requirements for having an appropriate university degree.) A modeling technique that might be feasible for describing layer#4 processes is the diagraming tool as proposed by Paulin (2013, 1780).

Layer five, finally, is about technical artefacts (such as information systems, in whichever form) that would provide means for lay interaction with the network of jural relations from layer#2. Graphical user interfaces, m2m APIs, technologies for planning, visualizing, analyzing, etc. of layer#2 data would enable a rich environment for subjects/citizens to interact with the state and service-providers, whereby latter might be either subsidized by the state or a local community, or be purely commercial providers of solutions for accessing layer#2 data.

The mistake of modern e-government was that it immediately went to providing monolithic layer#5 artefacts, which turned out to have at least issues with sustainability and interoperability, if the complex jural implications of system-level bureaucracy are left aside. Also approaching the design and development by e.g. starting at layer#4, as e.g. by developing a methodology for describing business processes on a high level and automatically translating them into the business logic of e-government artefacts, would, without considering layers#2-3 inevitably result in an unsustainable approach that might well satisfy acute needs (such as e-government does), but would not be prepared for future.

Liquid Democratic Collaborative Decision-Making

LD is a weighted way of making collaborative decisions, which does not depend on elected representatives, but rather on the transient delegation of votes. We can describe this processes mathematically as follows (cf. Jabbusch 2011, 35–7): each member A of a society can delegate its power to another member B (and withdraw it again at any time), whereat A – assuming each member’s power is v and the sum of all v is V, has thus (vA-vA)/V = 0 influence on voting on a decision, while member B thus has (vA+vB)/V influence on a decision made by all who are eligible to influence the given decision.

It is not clear who came up with liquid democracy first, and it appears that this idea arose from many minds independently. In terms of functioning information systems, to the best of our knowledge, three instantiations are known: in 2010, Paulin (2010) applied liquid democracy for executive empowerment to a student union, which was established mid-April 2010; a few weeks later, in Mai 2010, the German Pirate Party approved Liquid-Feedback, a system based on liquid-democracy that served as a backbone for their inter-party decision making process; in Russia, Leonid Volkov & Fyodor Krasheninnikov launched in 2011 the portal, which instantiates their specific vision on liquid democracy - cloud democracy (cf. Velikanov 2013).

The roots of the concept as such have been researched by Jabbusch (2011, 30–33), who aims to summarize the history of various concepts associated with liquid democracy based on an article by James Green-Armytage (2010). We shall partly repeat Jabbusch’ summary for sake of completeness – translated to English – here:

In 1912 the New York Times reports of William S. O’Ren, who demands for interactive representation whereby each elected politician’s – the so-called proxy’s, influence would be weighted with regard to the amount of votes received. His idea was picked-up more than half a century later in 1967 by the mathematician Gordon Tullock, who in passing suggests that voters could “by wire” chose their representative or vote themselves (in the parliament) while the debate would be broadcast by TV. In 1969 James C. Miller argued that everybody should have the possibility to vote on any question themselves, or appoint a representative. This idea was welcomed by Martin Shubik in 1970, who calls it an “instant referendum”, but is concerned that the speed of decision-making might influence the time available for a public debate.

Further roots of the LD-concept, Jabbusch argues, are to be found in the ideas developed by “sayke”, an anonymous user of the web, according to whom “liquid democracy can be thought of as a function that takes a question as an argument, and returns a list of answers sorted by group preference [… or] as a voting system that migrates along the line between direct and representative democracy.” (“sayke” in : Jabbusch 2011, p.31)

Sayke’s idea was developed further through a wiki until 2003, resulting in the concept that a decentralized information system (software) should enable citizens to participate in political decision-making, thus making parliaments obsolete. Each citizen shall have one vote, whereby the system would provide the citizen with all proposals on the question at stake. If the citizen would not want to do his own research on the particular question, he could subscribe to the opinions of friends instead. The system would further provide that decisions could be made automatically by the system (based on the friend’s suggestions?). The idea was further developed by the anonymous user “Kragg”, who dropped the subscription to friends’ decisions and instead proposed that voters could be delegated transitively, which is vital for LD.

Liquid democracy is thus a method of collaborative decision making that allows equal members of a community to either express their decisions on a matter directly, or empower a proxy to act on their behalf, whereby thus given power is transient. Thus for example, if Ann represents Bob and Carl, and Carl represents Dave and Eve, then Ann holds the power to act on behalf of all, including Dave and Eve, whose power transitively shifts from Carl to Ann. If this community would further consist of Franck and Gaby, thus in total seven people, then Ann’s decision would represent 5/7 = 71.4% of the community’s will, provided that nobody in Ann’s network votes for herself. Would however Carl decide to vote on an issue himself, then Ann would suddenly be representing only (5-3=2)/7 = 28.6% of the community.

Liquid Democracy in the German Pirate Party: a path down the rabbit hole

Jabbusch, himself a former high-ranking party member of the German Pirates, sees the Internet as an opportunity to get rid of “inefficient and costly organizations like parties and parliaments” and sees liquid democracy as an enabler to “exchange or completely remove representative democracy” (Jabbusch 2011, 8–9). Through liquid democracy, he argues, “the people would assume all legislative roles of parliament – committee debates, amendments, formation of opinions and resolutions regarding even complex wordings of laws” whereby “citizens would have the right to propose new laws at any time” (ibid., p.35). Jabbusch further envisions that individuals could delegate their voting power “temporarily to organizations (such as political parties, NGOs, associations, etc.) or to individuals (politicians, experts, friends)”, or, optionally, the citizen’s vote could be delegated only for particular topics (idib., p.35).

As Jabbusch notes (ibid., p.41-2), liquid democracy was a priority of the German pirates ever since their formation and rose to a wide-debated issue in the years 2007-2009. In 2009, when the pirates experienced nation-wide publicity and a strong rise in membership, several working groups were founded to investigate how to introduce advanced collaborative decision-making approaches for inter-party decision-making processes.

One such group founded the association Liquid Democracy e.V., which developed the software Adhocracy, which was later used by the German parliament (Bundestag) – albeit with no transitive delegation, for an e-participation pilot (Bundestag 2013). Adhocracy however, the Pirates found, did not suit their requirements, whereupon within two weeks the system LiquidFeedback (LQFB) was developed (Jabbusch 2011, 42), which continued playing a vital role in the party’s policy-making processes.

LQFB has been designed as a virtual place for forming opinions by the party base, which are to serve the party organs as recommendations and feedback (ibid., p.53). Opinion forming in LQFB takes place through initiatives, which can be proposed by any registered member, however, no discussions like they take place on other web forums are permitted in order to prevent trolling – i.e. counterproductive contributions. Once an initiative has been proposed, it must first receive support of at least 10% of the registered users within a certain time span; if it succeeds in doing so, time for discussion and eventual modifications of the initiative is allocated, whereby modifications are allowed only up to a certain time span before the end of the discussion period. After this, members can vote on the final proposal. The interesting point here is that discussion is deliberately excluded from the system and thus has to take place in wikis, other forums, or in real-world discussions (ibid., p.58-60).

The Meinungsbilder (“opinion-pictures/frames”), which are formed by the party base through LQFB however are not binding for the party leadership. Jabbusch (ibid., p.75ff) demonstrates how a Meinungsbild is formed on the example of universal basic income (UBI), a political idea arguing that the state should give every citizen a living-costs covering income unconditionally. A Meinungsbild-forming is initiated by a proposer posting an initiative. Each initiative can receive counter-initiatives, which are then competing within the Meinungsbild for dominance. It may happen that many initiatives within a Meinungsbild are accepted, which results in a fuzzy representation of the party base’s opinion. In addition, it is not impossible to initiate many similar Meinungsbild-forming processes, which further diversifies the results. At the time of Jabbusch’s report, seven UBI-related Meinungsbilder were formed (not including the failed initiatives); as of summer 2013, another more than ten Meinungsbilder on this very topic existed.

Despite that is up to the party leadership how to consider the opinions and petitions from the base during their mandate, the Meinungsbilder are further processed in party meetings and committees and are eventually included in official policy documents. In the case of the UBI, Jabbusch reports, the leadership accepted a petition for supporting a demonstration, which the party did by publishing an appeal for support on YouTube.

This kind of collaborative opinion-gathering as exercised by the Pirates however cannot be regarded as more than an admittedly elaborate form of deliberation. The non-binding nature of the Meinungsbilder does not prevent the party leadership from acting against the will of the party basis, as indeed has happened in 2013 (Herwatz 2013). However, even if the party leadership would be de-jure bound to the collaboratively expressed will, the Pirates might run danger to drawn in a torrent of different interpretations, contradictions, and juridical tricks that would allow the leadership to have its way in the end anyhow.

The approach chosen by the German Pirates does present a significantly novelty in the intra-party policy creating process, which could be eventually applied at a national level for a more democratic way of making political decisions. This way of collaborative opinion-gathering however does in no way affect the continuity of the bureaucracy but rather strengthens its legitimacy by making it appear more accountable and more participative.

Liquid Democratic Decision-Making in SNBG

Modern parliamentary systems institutionalize collaborative decision-making and regulate it through strict procedures conducted by elected representatives of the political community. National assemblies, legislative councils, Russian Dumas, Muslim Majlis’, or Western Parliaments, etc., are then but different names for the same concept of elected (or hereditary, or appointed) representatives deciding on rules, investments, and other matters from the public domain (of the republic so to speak).

The process of making a decision by any such legislative assembly can be broadly divided into four distinct phases:

  • first, the proposal is elaborated and presented to the assembly,
  • next, the proposal is deliberated (often involving many instances),
  • then the assembly votes on the last version of the proposal, and
  • lastly, the proposal (if elected) is enacted and steps into action.

In Slovenia for example, the Constitution of the Republic defines a bicameral legislative system with a state assembly (Državni zbor, hereinafter: parliament) as the legislative body and a state council (Državni svet) as a second chamber with a right to request a second round of deliberation on an already accepted proposal from the former (§91). The parliament consists of a fixed number of 90 members (§80), who decide in most cases with majority of the present assembly, whereby more than half of all members must be present for a decision to be valid (§86). A bill can be proposed either by the government, any member of the parliament, or by at least five thousand voters (§88). The process of deliberating and deciding on a proposal is defined as a multi-phase procedure regulated by the Rules of Procedure of the State Assembly (poslovnik Državnega zbora) of the parliament (§89, §94).

The Rules of Procedure of the State Assembly define how the legislative procedure is conducted, as well as the structure of the proposal. Latter must be sent to the president of the parliament (§114) and must contain (§115) an explanation of the causes of the proposed statute, its aims and goals, an estimate of the financial implications for the state budget if the proposal was to be enacted, a review of similar regulations in other legal systems and the conformance of the proposal with European Union law, and a discussion of other consequences the enacted law would imply. The president of the parliament initiates the legislative procedure by immediately distributing the received proposal amongst the members of the parliament.

At least ten members of the parliament can request within 15 days a deliberation on the reasons for the proposal (§122), as a result of which the proposal can be rejected preliminarily if the parliament finds that the proposal is not fit for further consideration. If the proposal is to be considered further, it is delegated to a taskforce, where it undergoes deliberation and where it can be brought through amendments into a further stage of ripeness, before the task force presents the thus updated proposal to the parliament. The parliament then deliberates a second time on the updated proposal, where further changes can be made. After that, a third round of deliberation takes place after which the voting on the final version of the proposed statute is done.

Could this procedure be translated into the domain of SNBG? We shall think how a multiphase legislative procedure could be handled by SNBG in two distinct ways – first, how it would be translated to remain virtually the same, and later, how it would be conducted by means of liquid democracy.

If we were to keep the same bodies, their characteristics and powers, then we had four distinct jural subjects that would contribute to the enacted statute – the proposer, the state assembly (parliament), the president of the former, and the assistive taskforce.

In such scenario, the proposer would generate a proposal by writing it into the respective ss-Gov registry. The eligibility to write proposals into that registry would be given based on the qualification of the proposing subject – either the subject would be a member of the parliament, a representative of the government, or it would be a proposal signed by at least 5.000 subjects whose membership in the voting registry would be valid. Once registered in the respective ss-Gov registry, the proposal would be available to the members of parliament (MPs) for deliberation. Within 15 days, 10 members of parliament might flag the proposal as blocked, whereupon the majority of the MPs together would either unblock or reject it by declaring their vote.

The proposal would then remain in a status where the parliament could not decide on it, until the responsive taskforce would flag it (or an updated version of it) as ripe. The taskforce would be set-up and empowered by the president of the parliament, and based on the membership in this taskforces the majority of its members would be able to set the flag.

Once flagged by the taskforce, the president of the parliament would flag the ripe proposal so to denote that it passed the second deliberation (the first deliberation happened if the 10 MPs flagged it at the beginning). Provided this flag being set, the president would have to once further flag it in order to denote that it passed the third deliberation, making it ripe for voting. After that, the majority of the MPs would be able to flag the statute as either enacted or rejected.

Through the flow of changes in the status of the proposal, the proposal (as updated by the taskforce) would be finally enacted. However, unlike in the present-day procedure, the role of the subjects taking part in this flow, would be focused on setting flags – i.e. changing information that describes the readiness of the proposal. The transactions thus change form push to pull, i.e. the subjects do not push documents to each other, but rather pull the status of the proposal and change it once the status permits it. This implies the need for periodic pull-requests into the corresponding registries which would enable the subjects to be notified on changes, such as e.g. the appearance of a new proposal, the changes in its status, and so on.

Also the final enactment of the proposal would be a pull-action, which then could be done by a subject in whose interest it is to enact the accepted statute – this for example could be any MP, or the proposer itself. The enactment of a statute today is limited to its publication in the official journal or similar media; in SNBG however an enactment of a new regulation would imply the immediate change of law. Aside from this, complex constraints such as for example a certain required minimum time between the approval of a proposal and its enactment might be possible as well.

How then would the MPs deliberate on the proposal in SNBG? As SNBG does not deal with deliberation, but merely with the government of eligibilities, it does not impose any constraints on how human beings exchange their opinions. Thus, if so desired, the Parliament / Duma / Majlis may remain as a place or institution where members and the public present their opinions in formal or informal ways, behind the lectern or in the lobbies, according to strict rules or traditional customs, in any way, so to speak, that pleases the expectations of the society. The same liberty applies to other aspects of the procedure which leads to the enactment or rejection of a proposal, such as the flow of information about a new proposal, its status, change, etc. It is reasonable to assume that humans remain to interact with each other in an SNBG-enabled system and thus information is conveyed with at least the same efficiency between subjects, which then may act upon the received news.

How about secret voting? Also here, SNBG does not set any constraints – thus, although CBR would require the identity of the voter to verify that the subject is a member of the eligible body, the identity does not need to be stored with the vote preference. This would be analogous to checking the identity of a voter entering the polling place.

How about then liquid democracy? – So far we described the translation of a classical legislative process into its SNBG-enabled clone. This way we argue that SNBG is capable to support the existing context without much ado. Next, we might think how to further evolve collaborative decision-making into the novel domain of liquid democracy (LD).

LD is by no means restricted to a certain predefined community of eligible members – thus, in the here discussed case of parliamentary decision-making, three scenarios might be feasible:

  • One would be that only the (here: elected) members of the parliament take part in LD-decision making, whereby after they are elected they can delegate their voting power to their colleagues (e.g. to the presidents of their fractions), which might result in a significantly reduced amount of powerful MPs, who would transparently represent the power structures within the parliament. Such approach might make more sense in larger parliaments (such as e.g. the European Parliament), where the sheer amount of members make decision-making less clear.
  • Another option would be to treat the entire community of voters in the nation as members of the LD system, which might make political parties obsolete, as the delegated eligibilities would shape into a network with a couple of very powerful nodes to which large numbers of individual subjects would link transitively.
  • The third option would be to take a hybrid approach where voters might have an opt-out option from the LD system by voting for MPs. Latter then would in total represent the community of voters who cast their vote in the elections, either in equal shares (i.e. each MP would have one vote in 90), or relatively in accordance with the amount of votes received. The voters that would remain in the LD modus might still be able to delegate their votes to MPs, or contribute their own decisions.

In either case, the multi-phase mode could be maintained, or at will transformed into a full LD-style decision making single-phase activity. While in the former every changed state in the process of making a decision would be collaboratively decided through the logic of LD, in latter a proposal would be accepted as soon as the required majority through LD would be found. In order to prevent hasty decisions, constraints regarding the time required for forming a decision might be set on an elevated legislative level (e.g. on the constitutional level) – a statute thus could be passed for example after at least one hour once the consensus of all members of the voting body is reached, or after at least one week once a two-third majority is given.

For storing the LD-relations between subjects, a dedicated ss-Gov registry would be required, in which each subject would be able to address the attribute denoting to whom the voting power was delegated (if at all). The proposal being voted on might then contain a snapshot of the network of relations at the time of its enactment / rejection for sake of accountability / documentation. The snapshot of the decision on a particular proposal would remain dynamic and modify-able until it would be frozen at the point where a decision is made. Thus, each subject would be able to either actively change its preference at any time until the final decision, or remain passive and such support another member or do not participate in the decision-making at all.

As this scenario shows, SNBG is compatible with both conventional and progressive collaborative decision-making techniques, which implies a fluent transition from one mode to another being feasible.


We summarized Self-Service Government (ss-Gov), a model for determining jural eligibilities based on jural facts stored in digital form in a dedicated ICT network. We described our vision of Sustainable Non-Bureaucratic Government (SNBG) which we structured into a five-level technology stack featuring ss-Gov as an essential basis, though not a complete enabler of SNBG. Further, we described the principles and beginnings of Liquid Democracy (LD) and presented an overview of the liquid-democratic approach taken by the German Pirate Party for intra-party agenda setting, which we criticized as unfeasible to generate systemic change towards self-managed government.

In chapter 3 we applied the principles of SNBG to the existing parliamentary decision-making procedure in Slovenia and examined how latter would perform under the joint principles of SNBG and LD. We found that the key/lock approach (i.e. Constellation-Based Reasoning – CBR) of ss-Gov can be feasibly applied to the existing parliamentary procedure, and further that an introduction of LD as an approach to increase inclusion and accuracy in shaping / determining the common will in democratic systems is conceptually doable by means of CBR.

We conclude that a combination of SNBG and LD would yield a powerful approach that would enable the emergence of truly self-managed juropolitical societies, where law, as well as the empowerment of active subjectivity in public functions would become a matter of collaborative decisions made based on ultra-democratic principles.

The here presented work however is purely theoretical. Further research is required, which would evaluate the principles of CBR and/or LD to detailed real-world scenarios, as well as such research, which would contribute to the searching-for, the development and improvement of particular technologies in the SNBG technology stack.