Jusletter IT

The Austrian «E-Law» system has neither been based on an international standard for systems of the kind, simply because when it was created there was none available, nor does it claim to constitute a standard of its own. It represents a solution specifically designed to meet the requirements of the Austrian federal legislative process. Nonetheless, it is based on elementary concepts of workflow management, thus providing a fairly useful example of how these principles may be implemented in managing law-making processes, and it is based on the emerging standards in document processing and preservation, as they were recognizable at the very beginning of this century, in particular the use of XML and safe electronic signatures. Briefly presenting the Austrian «E-Law» system, as it will be done in this contribution, therefore is to be perceived as a case study, nonetheless as a case study from which there may be drawn not only conclusions on how to procedurally manage a case like this, but also, and even more, conclusions on what to substantially take into account when establishing an electronic environment for the regulatory business.

Thus, this case study, while focusing on the major conceptual issues and without going into technical detail, is to

• reconstruct the creation of the «E-Law» system,
  
• outline the «E-Law» workflow system and its main functionalities,
  
• describe the «E-Law» format for legislative and legal documents,
  
• reflect on the project experiences,
  
• survey the impact of electronically supported law-making on the Austrian normative and political system, and finally
  
• derive from the «E-Law» system some basic conceptual approaches that may be deemed elementary for a «reference model» for electronic law-making support.
  

Perhaps this case study may even allow some conclusions or at least reflections on the interaction between professional communities and the academia as a prerequisite for efficiently introducing technical innovation to administrative practice. If one observes this efficiency to be high in Austria, as to be seen in the case of electronic support for law-making, this may, of course, be traced back to structural advantages like the smallness of the country, which would keep communication and interaction between the communities within a manageable scale. Nonetheless, it is also the personal impact of a few single personalities that keeps the dialogue between academia and public administration alive and fruitful, and among these personalities there is, in the first line, to be mentioned Erich Schweighofer, who this book and this contribution is dedicated to.

At this point, however, it is the author’s pleasant duty to also mention and thank those friends and colleagues, who have given him that advice and support without which it would not have been possible to prepare this contribution: Wolfgang Engeljehringer (Austrian Parliamentary Administration), Harald Hoffmann (Metadat), Karl Irresberger (Austrian Federal Chancellery), Friedrich Lachmayer (Austrian Federal Chancellery, retired), and Wolfgang Roth (T-Systems Austria).

Proceeding from a view that identifies political systems as sub-systems of the societal system, we have to put the question about the function of the political system within the societal system. Proceeding from the assumption that every system’s primary function is its own stabilization, we can further assume that its sub-systems are serving the same purpose. In this perspective, we can (with Luhmann) see the contribution of the political system to stabilization of the societal system in providing legitimization for the binding decisions that have to be made in order to ensure the framework of communication and interaction of the society’s members. In modern societies, these rules usually would be established in the form of laws. In modern democratic societies, these laws usually would be enacted by Parliaments.

If it is therefore the function of the «parliamentary complex» (Habermas, 1992, p. 448), i.e. the legislative organs along with the administrative organs supporting them in law-making and usually even initiating legislative processes, to provide legitimization for the rules to be established, what are the options to achieve this aim? Traditionally, this question has been answered by referring to Parliament being legitimized to represent the people by democratic election of its members. The legitimizing instrument of representation has, of course, been designed by lawyers or legally trained people, taking over an instrument of law: the instrument of representation entitling the representative to act on behalf of the represented, but within the scope of the mandate issued by the represented. That is why Kelsen has identified the parliamentary representation function as a fiction (Kelsen, 1929, p. 30): indeed, MPs are provided with what is called a «free mandate», which means they are, unlike the members of the former estates, no longer in any way legally bound to a substantial mandate issued to them by the represented.

The fictional character of parliamentary representation, though not always recognized as clearly as Kelsen did, therefore has always been a target for democratic criticism and scepticism, and diminished the capacity of Parliament to fulfil its legitimization function. That is why, besides parliamentary representation, there has always been another source for providing legitimization to the legal rules established, namely referring to an opinion-making process anteceding and accompanying the decision-making process within the «parliamentary complex», an opinion-making process involving civil society and basing on interaction between civil society and the «parliamentary complex». This less formalized instrument for providing legitimization has been described by Habermas as what he calls «deliberative politics», which he defines as the «interplay of the public sphere based on civil society with the institutionalised structures of opinion-making and decision-making in the parliamentary complex» (Habermas, 1992, p. 448). Of course, even Habermas is aware that the capacity of civil society to deliberate on the issues to be regulated is limited, so that in the end «deliberative politics» remains a fiction, as well.

In a somewhat provocative way, we can therefore say democratic legitimization is based on two fictions, namely

• the fiction of representation and
  
• the fiction of deliberative politics.
  

Nonetheless, what counts is the success, and we know from medicine about the efficacy of placebos. If by use of these fictions it can be ensured that the regulations established within a society would enjoy acceptance with its members, then the political system has completed its mission to stabilize society by producing legitimization. It has been observed that there is a high correlation between the acceptance decisions are finding and the way, or rather the procedure in which they have been produced; this is what Luhmann calls «legitimization by procedure». If a decision is made by the competent body in accordance with the appropriate procedural rules, it has the chance to be accepted even by those who would not be in line with the notion of the decision. If therefore a law is enacted by a Parliament elected in accordance with the electoral rules, and if Parliament has issued this enactment in accordance with the procedural rules to be applied, the law is likely to be considered legitimate. If in the process of legislation civil society has been involved, if at least the organised interests have been heard, and if there is at least a procedure for processing the opinions uttered, legitimacy enjoyed by the outcome of the legislative process is considered to be even much stronger.

In order to make procedure provide legitimization, however, procedure must not be a «black box» but has to be what might be called «transparent». In the words of Oberreuter, «transparency of the parliamentary procedure is functionally serving the inter-dependency of the (relatively) autonomous spheres of political communication, which is a basic requirement of democracy» (Oberreuter, 1986, p. 73). Only if the spheres of political communication are ensured their inter-dependency, only if the processes of decision-making and opinion-making are overlapping, and, at least, only if it is evident that the formal requirements of the decision-making process are observed, legitimization can be obtained for the outcome of the decision-making process.

Parliamentary transparency, since the 19th century, has been based on three columns:

• public access to parliamentary proceedings, in practice usually to plenary proceedings only, which thereby have got more a «rostrum» function than a substantial decision-making function, whereas substantial negotiations have been concentrated in the committees responsible for preliminary deliberation of business, which usually would take place in camera; however, it goes without saying that (physical) public access in real life necessarily has been reserved to a small minority of people;
  
  
• written documentation of parliamentary proceedings giving all those who had not had the opportunity to personally attend the parliamentary proceedings a chance to find them documented, and even those having been there an aid to support them in their references to what was said and decided; however, we must be aware that in practice traditionally it was also just a minority of people having access to these documents being published in just a small number of copies and available in large libraries only, and that it took them a lot of time if they really wanted to study them;
  
  
• mass media coverage of parliamentary proceedings therefore practically has always been the most important instrument of parliamentary transparency, giving the «ordinary» people a chance to be currently informed of what was going on in Parliament, by way of some «reduction of complexity», and thereby to stay ready to get involved.

The Internet revolution has affected all these instruments substantially:

• Public access to parliamentary proceedings no longer requires presence in person: Currently, most Parliaments throughout Europe are offering live audio and/or video streaming of plenary meetings, so that anyone having access to the Internet is enabled to follow the meetings, or parts of them, according to individual interests and availability of time. Nevertheless, of course, it would need a lot of information about parliamentary procedure as well as parliamentary business to really «understand» what is going on when following a meeting this way.
  
  
• Written documentation of parliamentary proceedings no longer requires to be consulted in libraries: Today, nearly all Parliaments throughout Europe, and many more Parliaments in the world, are publishing all parliamentary business, the committee reports, and the verbatim reports or stenographic records of the plenary meetings on the Internet, so that everyone can read and download them; new ways of (electronically) producing these documents make it even possible to have them available much sooner than in former times, so that people would find huge amounts of current information enabling them to develop positions of their own and introduce them into the opinion-making process. Nevertheless, what has been stated for electronic access to parliamentary proceedings also applies to electronic access to its written documentation: in order to be able to «understand» it, a sound information basis is required, which would not only consist of procedural knowledge but also, e.g., some knowledge of the legal language that legislative business is written in.
  
  
• Mass media coverage of parliamentary proceedings therefore has remained a basic instrument for involving the public in opinion-making and decision-making processes. In the past fifteen years, additional online information services have been established both by print and electronic mass media, also enabling the users to obtain information customized to their respective scope of interests. However, it cannot be denied that the electronic media, in the course of the last three to four pre-Internet decades, have significantly and sustainably changed the style of mass media coverage, by changing the people’s reception habits: «reduction of complexity», which, of course, always has been one of the main functions of the mass media, has significantly increased in its extent, so that today even the so-called quality newspapers would cover parliamentary proceedings much less thoroughly than they did two or three generations ago.
  

Generally spoken, the new information and communication technologies today are allowing a much easier access to legislative information than ever before. Parliamentary websites have got sort of an information broker function for legislative information, providing not only easy but in many cases even customized or customizable access, e.g. by offering SDI services allowing the users to subscribe to particular kinds of information. Nevertheless, there still remains what might be called an «information gap», a gap between the vast amounts of information on parliamentary business and proceedings available on parliamentary websites on the one hand, and the more and more reductionist coverage of parliamentary proceedings in the mass media on the other hand.

New technologies have their life-cycle, too. Among the theoretical models of technology and product life-cycles, the one Stephen Coleman has presented on different occasions and in different conceptual ways is particularly focusing on the inter-relation of technology and society; in its most convincing form, it distinguishes between the following four phases in the life-cycle of new technologies:

• Uncertainty
  
• Euphoria
  
• Institutionalization
  
• Societal adaptation
  

The phase of developing new technologies, thus, would be dominated by uncertainty: uncertainty, on the one hand, about the possible advantages of the technology in question, and, on the other hand, about its possible commercial exploitability. This phase may take quite a lot of time: the Internet, ex ARPA-net, e.g., for a quarter of a century was hidden from the general public in academic and military environments, until the «killer application» of the http protocol within a very few years boosted its societal roll-out in a way that was to fundamentally change societal communication.

This was when the phase of euphoria began, which lasted a few years only, until the burst of the «technology bubble»; these years, however, were full of substantial and sustainable economic investment, on the one hand, but also of a wide variety of technological and societal utopianism, like the concept of the «virtual communities» to self-regulate society in electronic environments (Rheingold, 1993).

After euphoria was overcome, the new technology became institutionalized, in a smoother but in the end more sustainable way. More focus is now directed on functionality and interoperability, but also on design; pragmatism has its way. More and more it becomes «business as usual», more and more it becomes part of the human life-world, less and less it is perceived as something new, as something revolutionary.

Thus, the logical consequence is societal adaptation: once trust in the new technology has assumed a level on which society would be ready to have patterns of action and interaction, which so far were supported by the new technology in a «mimicry»-like way, transformed into new patterns making unrestricted use of the technological potential, without feeling the need any more to keep up traditional phenomenological styles, this last phase in the technology life-cycle has commenced.

While for good reason we may call the large macroeconomic cycle, which we are currently involved in, the «age of information technology» or briefly the «information age» (Castells, 1996-98) – in a Kondratiev perspective this would be the fifth Kondratiev cycle –, we have to be aware IT is not a uniform technology but composed of multiple hardware and software technologies, so that their life-cycles would anyway overlap: Whereas, e.g., the classical Internet technologies would already have reached the stage of transition from institutionalization towards societal adaption, the new semantic technologies might just be on the jump from uncertainty into euphoria.

The process of institutionalizing Internet-based communication technologies being well advanced, we must not be surprised to observe institutionalization efforts even in conservative environments like those of regulatory processes. Due to the particularly high demands on these processes being transparent, traceable, and well documented, they are particularly apt to systematic use of the new technologies, but at the same time dependent on an extraordinarily high level of trust in the way they are performed, and in the tools used to support them. Thus, whereas on the administrative level in the law-making bodies throughout Europe and the U.S. the new technologies meanwhile are regularly applied, very often there would remain some elevated process steps of high symbolic importance that would still be performed in a traditional, i.e. paper-based way, e.g. enactments to be signed manually, in hand writing, instead of electronically.

When conceptualizing the development of the IT support of law-making, one might distinguish six major approaches, which to some extent may occur consecutively but needless to point out may and will overlap, too:

• Information management
  
• Workflow management
  
• Knowledge management
  
• Improving the procedural quality
  
• Improving the output quality
  
• Improving the participatory quality
  

If the focus is directed on managing the legislative process and its documentation, in most cases IT use would start up from the management, mostly the ex-post management of legislative and legal information within the framework of (mostly relational) databases. Process modelling, in this case, is done in a purely descriptive (and often somewhat reductionist) way. Nonetheless, the process analysis done in the course of establishing an information management system would make available in a well structured form a lot of structural knowledge, which then can be made use of in the following phases: When establishing a workflow management system, the ex-post approach of documentation is replaced by an ex-ante concept of support, and process modelling must not be reductionist any more, if the whole process is to be modelled within an integrated system, without system breaches, and without process stages to be further performed outside the system. Moreover, there will also be a step from descriptive towards prescriptive modelling, since if the full complexity of the legislative process is to be modelled some process stages will come up that will be easier to change than to model, even though the stages within a conservative process like a legislative one will always be hard to change. Finally, knowledge management will have to be introduced, which would imply contextualizing the information already available, transcending the single organization-focused way of managing information, and inter-relating the pieces of information horizontally and vertically, i.e. over different classes of information as well as different time-layers, so to enable the respective user to obtain information in a customized and contextualized way, and turn it into knowledge (Schefbeck, 2009a).

If the focus is turned to the impact of IT use on the legislative process and its setting, there would quite necessarily be some improvement of the quality of legislative procedure once workflow management is to be introduced: As already indicated above, the comprehensive analysis of said procedure, which is the prerequisite for its modelling, will unveil suboptimal procedural features. As far as they are based on administrative practice, it will in many cases be possible to adapt or even omit them in the course of initial process modelling, or at least within a reasonable period of time. In case, however, they are based on the legal rules of procedure, it might be more difficult to change them, since then it needs to convince the political decision-makers about the change requirement.

Beyond mere procedural quality, IT, however, may also impact the output quality of the legislative process, i.e. the quality of the legal enactments. Their formal quality, in first instance, may already be improved by the requirement of adhering to the pertinent formal styles and structural rules as imposed by editors to be used, e.g. an XML editor enforcing such rules to enable the users to produce valid XML documents. The potential scope of advantages of (future) legimatic drafting systems would also include, e.g., support for other kinds of legislative drafting rules, e.g. rules pertaining to language usage; once such systems were to include simulation applications, they might also be used to regularly integrate regulatory impact assessment in the legislative process.

Finally, the participatory quality of legislative processes might be sustainably improved by introducing new IT tools, as well (Schefbeck, 2009b):

Such tools might be aiming at

• making available relevant public knowledge resources, and facilitating the sharing of knowledge, within legislative processes,
  
• empowering co-operative action in these processes aiming at producing aggregated and well-structured opinions,
  
• providing the public discourse with a more transparent setting by visualizing the distribution of arguments and opinions, and
  
• structuring the outcome of policy modelling and consultation processes in a way enabling government to process and take into account even large quantities of contributions from civil society’s side.
  

Thus, IT tools, in particular semantic and social web tools, might be applied to enhance participation of citizens and civil society in the legislative processes, by supporting interaction and dialogue between societal groups and individual citizens, on the one hand, and between them and governmental structures, on the other hand. In particular, they will facilitate social networking capacities, to support co-operation and utilizing shared knowledge in formulating comments on draft legislation as well as policy principles, and in formulating legislative petitions; at the same time such tools will enable public administrations to efficiently keep track of public discourse on policy issues and its outcome, and to integrate the latter within draft legislation. Thus, they will also facilitate transparency of policy-making by making traceable the way of arguments in the decision-making process.

It is true, the primary motivation for introducing authentic electronic promulgation of federal laws was finding a way to save money, namely the considerable printing costs for the Federal Law Gazette, i.e. the authentic media to promulgate the federal laws enacted. That was why in 2001 the «E-Law» project was initiated, in the course of which, by joint effort of the Austrian Federal Chancellery and the Austrian Parliamentary Administration, the electronic workflow system was established that was the prerequisite for having the Federal Law Gazette authentically published in electronic form.

When in February 2000 the Austrian People’s Party and Freedom Party formed a coalition government, a significant reduction of the government deficit was one of its explicit objectives, an objective that was intended to be reached primarily through reduction of expenditure. That was why within the federal administration any suggestion for a measure having the potential to effect an economy had a fair chance to be acted on. The staff responsible for legal informatics in the Federal Chancellery and the Parliamentary Administration used the opportunity: Their suggestion, which was based on considerations and talks already started the year before, to replace print by electronic publication of the Federal Law Gazette was seized, and they were able to convince the political decision-makers of the necessary implication of having the whole legislative workflow put on a new electronic basis – only if the documents to be promulgated in the Federal Law Gazette were produced and processed within a uniform technical framework, it would be possible to publish them without any additional formatting work (which would have raised personnel cost) and without any delay. Thus, as Friedrich Lachmayer has put it, an «ex-ante» approach of correctly formatting the legislative documents and adding the required metadata was chosen instead of an «ex-post» approach (Lachmayer & Hoffmann, 2005).

On 6 June 2001, the Council of Ministers adopted a resolution on the «E-Law» project, which outlined the project’s main objectives:

• Establishing a workflow system to support the whole federal legislative process
  
• Adapting the layout styles of the Federal Law Gazette to the requirements of electronic document production and processing
  
• Shifting from print to electronic promulgation of the federal laws
  

A working group of the Federal Chancellery and the Parliamentary Administration was to coordinate the project, and develop joint solutions where necessary, which was the case in particular with regard to the common document format, while, of course, there were project groups within both administrative structures that had to develop the internal solutions. Programming, for the most part, was outsourced.

Since the project was to be based on the preliminary talks that had already taken place before, within half a year only it was possible to establish the workflow system, which was operational as of 1 January 2002, and to agree on the common document format, which underwent only slight changes later.

Development and implementation of the «E-Law» system may be broken down to three dimensions:

• Technical development and implementation
  
• Organizational implementation
  
• Legal implementation
  

Having in mind the speediness of the project and its motivation in cost-cutting, one will not be surprised about the technical solution being not too revolutionary. Wherever possible, it was based on the already existing infrastructure, which also included the requirement to put up with a system breach between the workflow systems, in the technical sense, to be applied on government and Parliament levels: The Federal Government had just adopted the FabaSoft suite to support the administrative workflow within the federal ministries (with the exception of the Federal Ministry of Defence), and therefore decided to apply it on the pre- and post-parliamentary stages of the legislative process, too, which anyway was an economically reasonable choice. Parliament, however, from the early 1990s on, had built up a comprehensive Oracle environment for all its information management purposes, including a huge application called «Parliamentary Business» to support clerical and documentation services in managing parliamentary documents; thus, it would have been quite unreasonable to also make use of the FabaSoft workflow system for supporting the legislative process, whereas it was a much more economic solution to just add the required workflow functions to the already existing information management system, within the Oracle environment.

That was why the Federal Government’s and the Parliament’s legislative workflow systems had to be connected through interfaces, which however did not cause major difficulties. Whereas this system breach was easily to be accepted, much effort was devoted to avoiding a format breach, i.e. a breach in the chain of document formatting within the legislative process. Austria, in this regard, was lucky enough to learn from the Swiss experiences: Switzerland already in 1998 had established an electronic workflow to support its federal legislative process; the Swiss Federal Government and the Swiss Parliament, however, had been unable to agree on a common legislative document format, since the requirements on both sides seemed to be too different, with the consequence that all the formatting done at the pre-parliamentary stages of the legislative process was lost once a government bill was submitted to Parliament, and had to be redone on the legal enactment once parliamentary procedure was completed.

Conceiving a common format for legislative and legal documents, therefore, was one of the main tasks of the joint working group by the Austrian Federal Chancellery and Parliamentary Administration. Whereas it would, of course, also have been an option to develop several schemas or templates for different document types, it was very soon decided to only develop only one to cover all existing document types, thus raising the level of abstraction. Another objective was to reduce the overall number of admissible elements as much as possible, which also contributed to a high level of abstraction. Indeed, after some negotiation it proved possible to restrict the number of admissible «E-Law» styles to about 90, which was a reduction by some 25 % in comparison with the original concept.

The main question to decide was what editor, and thereby what kind of format to use for producing and editing the documents, since from the very beginning it was evident that the target format would be XML.

Even though XML, in those days, was a very young standard, it had already been widely recognized that it combined the potential of SGML with much simpler manageability. There had indeed been a project to develop an SGML DTD for the Austrian Federal Law Gazette, and use SGML for producing it, but finally this project was not further pursued, when XML came up.

Among the major advantages of XML, the following ones were particularly to be taken into account:

• XML, as an open standard, served the technical and legal requirements for electronically signing the documents, in accordance with the Austrian Electronic Signature Act; proprietary formats like Word would not.
  
  
• For several reasons, choosing XML as the storage format for the legal documents was the most appropriate approach to ensure, the best possible way, their long-term preservation, which in case of the documents forming the normative system was to be deemed even more important than in other cases. Among these reasons, the platform independence and presentation openness were to be mentioned, just as the simple fact that as text-based, human-readable format XML would even provide some interpretability if the schema (or the DTD) got lost, at least by statistical means.
  
  
• Its openness for presentation and its platform independence would not only have their value in a long-term perspective, but anyway provide for comfortable current access to the documents, regardless of individual platform and format preferences.
  
  
• As a meta-language like SGML, XML would provide the full scope of options to develop languages appropriate to structurally describe specific classes of documents, like, in our case, legislative and legal ones. Based on such structure-orientation, layout requirements were easily to be met, too, but not only those of one particular publication format; the re-use of documents within other publication environments (e.g., legal commentaries to follow the specific layout styles chosen by a particular publisher) was easily to be supported, as well, simply through additional stylesheets.
  
  
• Ingest of procedural metadata to XML documents (besides the structural metadata they would anyway contain) was an option to support the «ex-ante» approach in currently describing legislative documents and their respective legal status, instead of an «ex-post» approach only adding metadata after particular procedural stages were completed. Thus, the describing metadata were to be delivered along with the documents from stage to stage within the workflow, even through the system interfaces. As a matter of fact, the practice of keeping procedural metadata outside the documents, in relational databases, was retained, and is likely to be retained in a foreseeable future, but in the long run these «external» databases will be deemed superfluous, and XML documents will bear all required metadata within themselves.
  

Instead of writing a DTD to define the XML structure of the «E-Law» document type, the young XML Schema standard was used. The XML schema definition finally established was the result of the analysis and negotiation process taking place in the working group, and mirrors this process in its evidently heterogeneous concept: It consists of several separate schema definitions covering different requirements, which are kept together in a control document by inclusion statements. The metadata defined would primarily serve the purposes of promulgation in the Federal Law Gazette and consolidated publication in the Legal Information System of the Federal Chancellery, whereas the vast majority of other kinds of (procedural) metadata would not be included but have to be kept externally, in specific data bases. From the organic growth of the «E-Law» XML schema definition it is also easily to be concluded why this definition cannot serve as a reference model for other jurisdictions – it is not the result of an abstract analysis of the requirements of an XML format for legal and legislative documents, which then would have been broken down to the Austrian federal legislation, but it is the result of a very concrete analysis of the requirements of describing the Austrian federal legal and legislative documents in a way appropriate to facilitate achievement of the main project objective, namely authentic electronic promulgation of the federal laws, and further processing of the texts within the Legal Information System.

Fig. 1 Main components of the XML schema definition of an «E-Law» document

While XML was chosen for warranting correctly formatted and safely signable documents, as well as their retrievability in a long-term perspective, safe electronic signatures were to warrant the authenticity and integrity of documents, or, to be more precise, to support

• the authentication of authorship (i.e., what might be called the authenticity of documents), as well as
  
• the authentication of content (i.e., what might be called the integrity of documents),
  
• both in a long-term perspective (which would require a technical and organizational concept for regularly re-signing the documents, to cope with the cryptographic development).
  

Needless to point out that electronic signatures, of course, would only indicate any breach of the authenticity and integrity of documents, so that additional security solutions have to be implemented, in accordance with the applicable standards in the field, to protect them against such breaches.

Within the Austrian e-government infrastructure, standard electronic signature applications had been developed by the Federal Chancellery that now were also to be made use of within the framework of the «E-Law» system. Out of these so-called MOA («modules for online applications»), which are open source, the modules MOA SS to create a server signature, and MOA SP for signature verification, were to be applied.

Having defined XML as the target format, it would have been a logical consequence to use it as the production format for legislative documents, too. The prerequisite, of course, would have been to avail of an appropriate editor. Unfortunately, the editors available in those days, like XMetaL, did not seem to be comfortable enough to inflict them on the legislative drafting staff in the federal ministries and the Parliamentary Administration. As a matter of fact, at the same time there was a project conducted in the Federal Ministry of Social Affairs to adapt XMetaL to legislative drafting purposes, and also modify its user interface in a way making it look similar to the well-known Word interface, but this remained a stand-alone solution that never was rolled out on a wider scale, even within the mentioned ministry (Zach et al., 2005).

The «E-Law» working group, however, decided to have MS Word used as the editor for producing and editing the legislative documents, and only have them converted into XML afterwards, through a converter specifically built on the ground of MS COM technology, and made available within the «E-Law» workflow as web service. To ensure by such conversion process creating well-formed and valid XML documents, and avoid manual post-editing at the conversion stage, the use of the common Word functionalities had to be widely restricted, thus forcing the legislative drafters to apply the specific «E-Law» styles only. This was accomplished by modifying Word through a macro solution, which further on, of course, had to be adapted to each new version of Word, this being quite a time-consuming and costly procedure, once and again. That is why the «E-Law» macro solution so far has been adapted to Word 2007 only, with its well-known performance disadvantages, but not yet to Word 2010.

The restraints in the common Word functionalities, however, have become more and more sophisticated: whereas, e.g., in the early years of «E-Law» it was still possible to override «E-Law» paragraph styles by common Word character styles, which only would become visible once the document was to be converted into XML, meanwhile these functionalities have been disabled, and thereby the conversion process has become much smoother. Assigning the correct «E-Law» styles to particular paragraphs or characters/character strings is supported by specific toolbars created to visualize the set of styles available and admissible. Thus, not only the Word functionalities but also the appearance of the Word interface have been more and more modified.

Fig. 2 Formatting an «E-Law» document by means of a modified MS Word interface

Simultaneously with restricting the legislative drafters to the use of the predefined «E-Law» styles, the macro solution also provided for more and more elaborate tools assisting them in checking the correctness in the use of styles already within the Word environment, in advance of the conversion into XML. Thus, the modified Word editor comes with a tool to check whether only the admissible «E-Law» styles have been used within drafting a document, and to indicate all styles not being in accordance with the «E-Law» set of styles. Whereas this tool is regularly and successfully used by legislative drafting staff, another tool has proved to be less useful: it would allow the legislative drafter to have the correct «E-Law» styles automatically attributed to the respective structural elements of a draft bill, but even though the recognition rate is high there would always be some need for manual editing. Thus, the latter tool usually would only be made use of at the beginning of a drafting process, whereas using it on a document already formatted before, e.g. to assign the correct styles to amendments of the document, would be quite risky, because there might occur some incorrect style recognition again, in elements that had already been manually reformatted before.

Even though legislative drafters, by more and more being bound to the «E-Law» styles, had to accept restrictions in their freedom of editing their documents, they were yet given the opportunity, by the «E-Law» editor solution chosen, to keep to Word as the editor nearly all of them had been working with already before, as of the mid-1990s, when MS Word was rolled out throughout the Austrian federal administration, and were not forced to switch to an XML editor to be used specifically for producing and editing the «E-Law» documents, whereas in all their other everyday business they would anyway have continued in using Word.

This was simply done to seek for acceptance of the new «E-Law» system with its primary group of users, i.e. the legislative drafting staff in the federal ministries as well as the Parliamentary Administration, and thus facilitate the system’s organizational implementation. Nonetheless, in the very first year of the system being operational, i.e. in 2002, resistance within the government agencies was manifest, and thus, in this first year, only 8 % of the government bills were submitted to Parliament as safe «E-Law» documents, within the «E-Law» workflow. This resistance was to be overcome only by a massive information campaign launched be the Federal Chancellery, in the course of which the project officers paid personal visits to nearly all legislative drafters in the federal ministries, to train them in how to use the system, but even more to convince them about its advantages. This campaign was quite successful, so that in the second year, i.e. in 2003, already 70 % of the government bills were submitted as safe «E-Law» documents, and since 2004 a rate of 100 % has been achieved.

Resistance within the Parliamentary Administration was anyway smaller, which may be due to three circumstances:

• First of all, it is a much smaller organization, where the actors within the legislative process, in particular the committee secretaries and the plenary staff, were easier to approach personally, from the very beginning of the project, and a reasonable number of them were even members of the project group and thereby personally involved in the system development.
  
  
• Secondly, a new organizational unit called «Competence Center E-Law» was created that was to support the legislative drafters, in particular in applying the «E-Law» styles, so that in case of difficulties they also could lean back and forward the document in question, within the workflow, to the mentioned unit, to have it reformatted there. This option took some pressure from the legislative drafters, and indeed very soon most of them managed to do all the regular formatting work on their own. One important function of this new organizational unit was also to ingest other kinds of draft bills than the government bills, in particular private members’ bills, into the «E-Law» workflow, and prepare the documents for further processing, through formatting them, if necessary, in accordance with the «E-Law» styles.
  
  
• Finally, training of the legislative drafting staff was supported by an e-learning application developed within the Parliamentary Administration, which was technically based on the ECDL concept. Developing the «E-Law ECDL» brought about two major advantages: On the one hand, conceiving it offered the opportunity to reconsider the parliamentary part of the workflow system, and modify it to the extent deemed necessary; on the other hand, once the e-learning system being available, it could be made use of in a very flexible way – to obtain basic knowledge on the system, to obtain training on specific functions of regular practical need, and to be retrained once and again on functions more rarely required, all this to be done within the framework of the respective individual schedule, but not bound to a classroom training which is difficult to coordinate in an organizational environment like a parliamentary one, but was also offered additionally.
  

The «E-Law» e-learning application was made available to the MPs, too, but among them found less interest than two other e-learning applications developed at the same time, one of them offering training in the parliamentary information management system, the other one in the technical functionalities of the notebooks handed out to the MPs, in timely coincidence with implementing the «E-Law» system. Indeed, rolling out the notebooks to the MPs was not only a case of coincidence, but of micro-political correlation with the «E-Law» project. Obtaining notebooks (which, of course, were only lent to the MPs for the duration of their respective mandate) was the condition under which the MPs were ready to further on do without the paper copies of parliamentary business, which so far had been regularly distributed to all of them (but kept by a few of them only, so that not rarely a single MP obtained a particular document more than once). Abandoning the production and distribution of paper copies, in turn, was required to amortize the investment in developing the «E-Law» system.

Indeed, taking into account the savings in paper consumption within the Parliamentary Administration, it may be stated that the return of investment was accomplished after about one year already: The reduction in paper consumption being some 60 tons of paper per year, and the typesetting and printing costs as to those kinds of documents so far printed by the State Printing Office being considerable, the overall savings per year may be estimated at about 1 million €, which is tantamount to the additional expenses for developing the parliamentary part of the «E-Law» system; the expenses for procuring the notebooks are not taken into account in this calculation, but in order to provide the MPs with contemporary equipment they would have to be bought sooner or later anyway.

The saving of labour, within the administrative environment, cannot be exactly determined but is to be deemed considerable, too. Having in mind that so far, e.g., four paper copies of each legal enactment by the National Council had to be produced manually, literally by «cut and paste», i.e. by applying scissors and adhesives, which in case of major amendments adopted might take some hours, henceforward just one electronic document was to be produced, in a much quicker, easier and, last but not least, cleaner way.

Even though the MPs were sustainably affected by the implementation of the «E-Law» system, through being the primary addressees of the documents produced, and already before through being invited to provide non-authentic versions of the parliamentary business they were producing, like private members’ bills, motions for amendments, motions for resolutions, etc., they, and their assistants, were not made actors in the system, in the formal sense. Legally, they still had, and have, to submit all business on paper, and authenticate it by hand-written signature. As a matter of fact, the Electronic Signature Act, Federal Law Gazette I no. 190/1999, in principle provides for electronic signatures being recognized as authenticating documents, but the rules of procedure of the National Council, as well as those of the Federal Council, so far have not been amended in a way allowing parliamentary business to be authentically submitted in an electronic way.

This, first of all, is due to the political circumstance that each and every amendment of the rules of procedure is subject to a compromise between the parliamentary groups, in particular because such an amendment would require a two-thirds majority, so that opposition parliamentary groups in such matters have their say, too. In the past years, all discussion on amending the rules of procedure of the National Council was centred around the issue of entitling parliamentary minorities to have established a parliamentary committee of enquiry, and thus no comprehensive amendment was achieved at all. Besides, it must also be stated that the parliamentary groups at the federal level, unlike those in the provincial parliament of Styria, where the outcome of a recent procedural reform allows for parliamentary business being authentically submitted in an electronic way, have not shown an active interest in obtaining such an option; there may also be some political reason for this lack of interest, since the Styrian example has shown that the freedom to submit parliamentary business through electronic channels has made MPs more independent from their parliamentary groups and their infrastructure, which parliamentary group leaders and staff at the federal level might want to beware of.

It there is not felt an urgent need to include the MPs in the scope of active «E-Law» users, this however might also be due to the experience that everything works well anyway; with maintainable effort, Parliamentary Administration staff would integrate all legislatively relevant parliamentary business in the «E-Law» workflow, so that the project’s primary objective, to have all legal enactments ready for authentic electronic promulgation is easily achieved. The same even goes for the post-parliamentary procedure of authentication by the Federal President and countersignature by the federal Chancellor, which is also still today not performed by electronic but hand-written signature. Thus, there still is a logical, and in fact a legal divergence between the whole legislative process being administratively and technically performed in an electronic way, whereas legal authentication on all stages in advance of the final one, i,e. the stage of promulgation, is done in a traditional, paper-based way – quite a significant instance of the simultaneity of the non-simultaneous, to paraphrase Ernst Bloch.

Authentic electronic promulgation of the federal laws was finally achieved as of the beginning of the year 2004, thus making the Republic of Austria a pioneer state in electronic legislation, within a European and even global scope.

The Federal Law on the Reform of Promulgation, Federal Law Gazette I no. 100/2003, provided for authentic electronic promulgation of the federal laws with effect from 1 January 2004, by amending the Federal Constitution accordingly and issuing the required implementing legislation. In accordance with section 6 of the new Federal Law on the Federal Law Gazette 2004 (art. 4 of the above-mentioned act), this is to be done within the framework of the Legal Information System of the Federal Chancellery. Section 7 precisely defines the URL where the Federal Law Gazette is to be published (www.ris.bka.gv.at ), and section 8 defines the criteria to be met in order to safeguard the authenticity and integrity of the documents published in the electronic Federal Law Gazette: In accordance with these criteria, and the state of the art, the file format chosen is XML, and the files are electronically signed. It may be mentioned that in the more than six and a half years past no violation of the integrity of a federal law electronically published has been noticed.

Section 9 of the new Federal Law on the Federal Law Gazette, at last, entitles anyone to freely and without prove of identity access the Federal Law Gazette, and make printouts free of charge. Besides, non-authentic printouts may still be purchased, as well, which, of course, is a service meanwhile rarely demanded. Besides the authentic XHTML representation bearing the electronic signature, three additional non-authentic electronic versions of the Federal Law Gazette are offered on the same site, to make possible different ways of re-use and further processing, namely a PDF, an HTML and an RTF version. According to practical experience, the non-authentic versions, in particular the PDF version, nowadays are even more often used than the authentic one, which would indicate electronic publication of federal laws in Austria has become a trust-based system.

Legislative processes would regularly follow a cyclic procedural pattern:

• Awareness (agenda setting)
  
• Specification (policy modelling)
  
• Drafting (normative modelling)
  
• Consultation
  
• Deliberation
  
• Decision
  
• Promulgation
  
• Consolidation
  
• Evaluation
  

The legislative workflow in the narrower sense, however, i.e. the legislative business process embedded in a comprehensive set of both administrative and legal procedural rules, would usually only commence at the normative modelling stage: Legislative drafting and the way the draft bills are handled, very often is already regulated in quite a specific way. The closer the legislative process approaches the stages of formally enacting and promulgating a law, the more detailed and strict the procedural rules to be observed would be. Consolidation of legislation, even though of great practical importance, in many normative systems already would backslide into a less regulated sphere of convention, and the same goes for evaluation of legislation. Thus, the formal procedural requirements are most restrictive at those stages of the legislative process that would represent the decision-making process that is to add legal effect to a set of provisions; not surprisingly, it is this formal decision-making process which is to provide legitimization to the normative provisions by means of «legitimization by procedure».

From what has been said above it will become evident why the «E-Law» workflow process was conceived to support the legislative process stages from drafting through promulgation, along with providing an interface to the consolidation stage. It was this more or less highly regulated core part of the legislative process that proved particularly appropriate for being formally modelled, even though this model, on the other hand, was to grow quite complex; solely within the framework of parliamentary procedure, there were no less than some 1000 different process steps to be identified and integrated in the model, most of them, of course, belonging to the facultative ones. Nonetheless, we have to be aware of the fact that legislative processes are multi-layered; besides, beyond, or behind the highly formalized legislative procedure as being laid down in the procedural rules, there are informal layers, too, which would interact with the formal layer. Out of these informal layers, there would be some to be supported by the workflow system, as well, which in particular include the administrative procedures behind the scenes; others, however, would elude any IT-supported tracing, which goes for those informal meetings and negotiations of political decision-makers (along with lobbyists and representatives of societal interests, too) that would necessarily take place in advance of and in parallel to the meetings of the decision-making bodies.

To very briefly recapitulate the constitutional rules governing the federal legislative procedure in Austria (as laid down in art. 41-49 of the Federal Constitution), it is to be pointed out that even though there are two organs of the federal legislation, i.e. the National Council whose members are elected by the people, and the Federal Council whose members are delegated by the provincial parliaments, the task of substantive law-making is exclusively incumbent on the National Council, whereas the Federal Council in most cases just has the option to raise a suspensive veto to the legal enactments made by the National Council, while in a few cases its consent is required; anyway, in none of these cases the Federal Council has the power to modify the legal enactments. Bills may be submitted to the National Council by the Federal Government (but not by the individual ministers), by the members of the National Council, by its committees, by the Federal Council or one third of its members, and by means of a so-called «people’s initiative» if supported by at least 100.000 voters or one sixth each of the voters of three provinces. To become a federal law, a legal enactment by the National Council, once the Federal Council procedure is completed or an eight weeks’ deadline expired, is to be authenticated by the Federal President, countersigned by the Federal Chancellor, and, under the latter’s responsibility, promulgated in the Federal Law Gazette.

Fig. 3 Way of the Austrian federal legislation (very simplified view)

In legislative practice, a significant majority of the legal enactments, usually about 70 %, would trace back to government bills. This is one more reason why the «E-Law» project from the very beginning had to focus on the pre-parliamentary phase of the legislative process, i.e. the preparation phase of a government bill, no less than on the parliamentary and post-parliamentary phases; it is just to be mentioned there are a few differences in the procedural rules applying to government bills vs. other types of bills at the parliamentary stage, whereas in the post-parliamentary phase of the process it does not make a difference any more in what way the bill was submitted that a particular legal enactment traces back to.

In the majority of cases, therefore, a particular law-making process would start within the federal ministry having the executive competence in the respective subject matter. Whereas in some federal ministries it is incumbent on the respective department responsible for executing the subject matter to also do the legislative drafting in the same policy field, other federal ministries have established centralized legislative drafting departments doing the drafting on the ground of the substantive information received from the respective executive department (which, of course, might make a difference in the quality of drafting, and anyway made a difference as to how quickly and effectively the respective ministerial staff became used to the «E-Law» workflow system and document format).

Once a draft bill has been accorded within the federal ministry, and very often with major stakeholders, too, the draft bill would be sent out for a formal consultation procedure, within the framework of which the other federal ministries, the provincial administrations, and the relevant interest groups are invited to provide their comments. For sake of information of the parliamentary groups, the draft bills, as well as the comments, are also sent to Parliament.

Once the deadline for consultation has expired, the ministerial administration produces a synopsis of the comments given, which is the basis for the political decision by the minister on the final version of the draft bill to be submitted to the Council of Ministers. If the latter adopts the draft bill, it is submitted, through the Federal Chancellery, to the National Council as a government bill.

Within both the National Council and the Federal Council, legislative procedure, in principle is bipartite, consisting of a committee and a plenary stage. On the ground of the results of the competent committee’s preliminary deliberation on a bill or a legal enactment, resp., as laid down in the committee report, the plenary would debate the issue and take the decision. The legal enactment would be forwarded to the Federal Chancellery, where promulgation as a law in the Federal Law Gazette is prepared and executed once authentication and countersignature were completed.

Fig. 4 Main steps of the parliamentary legislative process (simplified view)

Thus, the legislative workflow would connect several organizational actors, i.e. the administrative and legislative organs involved, and in practice a considerable number of individual actors, who traditionally communicated with each other through letter mail processed by administrative delivery services, and in the last years before introducing the «E-Law» system already to some extent through e-mail. When modelling the business process, these interaction patterns and the roles interacting were to be analyzed and implemented in the workflow system, and then, in accordance with the actor/role paradigm quintessential for systems of the kind, actors and roles were to be mapped. By doing so, it became possible to assign the process steps to be performed to the respective performers.

The «E-Law» workflow system may be accessed within the intranet applications of the federal ministries and the Parliamentary Administration only, via user ID and password; where extranet applications have been implemented, like in case of the virtual private network solution of the Parliamentary Administration, users physically may be located outside the respective office premises. Secure communication within the workflow is ensured through applying the SSL, or meanwhile TSL protocol.

The two major advantages of workflow support for the legislative process became evident very soon:

• First, communication became faster. Documents, from now on, were to be exchanged not only within but also between the different organizational bodies without any delay of time, as it was necessitated before by physical delivery. Even more time than in administrative communication was saved when it came to publishing documents, because from now on the documents produced within the system were to be published immediately, whereas before they had to be typeset, proofread, and printed.
  
  
• Secondly, task assignment became predefinable, precise, and clearly arranged. Once implemented within the actor/role setting, the responsible actors simply had to open their customized or customizable desktop interface, or more specifically their respective work basket, to find there the documents to be processed, along with metadata information and accompanying material. The check-out/check-in paradigm applied made sure that one particular document within the workflow system was processed by one actor only at one point in time, thus avoiding any ambiguities in document processing.
  

Having in mind the multi-stage and multifaceted legislative process as only outlined above, one can imagine that not only the number of roles to be distinguished, and the number of process steps to be foreseen in conceiving the workflow system, but also the actual quantity of documents to be produced, processed, and stored was quite considerable. Unlike former times, when indeed in most cases only the final versions of documents as adopted on the different stages of the legislative process went to the files, whereas all the various drafts produced at the administrative level were sorted out, from now on each and every version of a document, once checked in, was stored, and was to be stored to be able to completely trace the document life-cycle, and take precaution for eventual mistakes now traceable to their originator (or even intentional but unjustified document changes, which would, from all experience, not occur in practice but cannot be totally excluded). A major asset of the versioning facilities is the functionality of text comparison, which can be used to compare two discretionary versions, even though there are also some routines for comparing the official document versions, which in practice are much more used, simply because these are the versions of political relevance.

Needless to point out there is to make a distinction between documents in the legal or logical sense, and documents in the physical sense, i.e. physical files. Very often, several files would form one document in the legal sense: A government bill, e.g., regularly consists of

• the legislative proposal (i.e., the text proposed to be turned into law),
  
  
• explanatory remarks (usually tripartite: a cover sheet, a general part outlining the objectives of the bill, and a specific part containing comments on the different structural elements of the legislative proposal), and finally,
  
  
• in case of a bill for an amending act, a text comparison (i.e., a table contrasting the text of the law in force with a consolidated version anticipating the proposed amendments).
  

Nonetheless, it may also contain, e.g.,

• normative annexes (i.e., annexes to be comprised by the legal enactment, e.g. tables, model forms, maps, or depictions of signs),
  
  
• non-normative annexes (i.e., annexes not to be comprised by the legal enactments, e.g. further explanatory material, like the detailed key to the budgetary posts regularly enclosed with the draft budget), or,
  
  
• in case of a state treaty contained in a government bill, different language versions, which in turn may be normative or non-normative.
  

To keep together, within the workflow, the different files belonging to one and the same logical document, it obviously needs a container. In the «E-Law» XML container, not only Word files being in accordance with the «E-Law» format, but also other file types may be embedded, like image files (e.g., TIFF, JPEG, GIF, or PDF images); in practice, not only maps and illustrations would have to be introduced to the workflow as images, but also foreign language versions of state treaties, even though this was not originally intended, often would be submitted as PDF images, in particular when typed in a non-Latin set of characters.

At the end of the workflow, the actor performing the pertaining role in the Federal Chancellery will publish the final document, i.e. the legal enactment to become a law, in the Federal Law Gazette, after having converted it to XML and put his electronic signature on it, to ensure technical authenticity and integrity. At the same final workflow stage, however, there is also an interface to the Federal Chancellery’s Legal Information System, through which the document is made available for producing the consolidated version of the law in question (if it is a question of amendment), which is still done manually; along with the document, of course, the metadata required for the Legal Information System are provided, after having been produced at the different preceding process steps already.

Even though, as indicated above, about 70 % of legal enactments, on an average, are based on government bills, the remaining 30 % have to be prepared within the «E-Law» system, too. As long as authentic electronic submission of private members’ bills would not be admissible under the rules of procedure of the National Council, they have to be integrated in the «E-Law» workflow by parliamentary staff, in advance of the committee deliberations commencing. In nearly all cases, nowadays, non-authentic electronic files would be made available by the parliamentary groups, but not all of them would be correctly formatted (with the important exception of private members’ bills submitted by MPs belonging to the governing parties, which often are prepared by or with the assistance of ministerial drafting staff); thus, it is incumbent on parliamentary staff, in particular on the «Competence Center E-Law» mentioned above, to take care of these bills being made processable for the committee and plenary workflow stages. As to the committee bills, they would anyway be prepared, along with the committee reports, by the committee secretaries. Thus, all legal enactments by the National Council are produced within the «E-Law» workflow, and ready for being promulgated as laws in the Federal Law Gazette without any further editing, except a new header being added.

Whereas from the very beginning of the «E-Law» system being implemented, the whole legislative process as sketched above was supported, it should be mentioned that at the pre-parliamentary stage of the process, in preparing government bills, only from the (relatively advanced) process step of submitting the draft bill to the Council of Ministers utilization of the system was centrally enforceable and enforced. That was why in the first years of the system being operational, while already 100 % of the government bills were submitted to the National Council through the system as of 2004, only about 50 % of them also were produced and processed within the system from the very first step of conceiving the first draft in the responsible ministerial departments; the others were integrated in the system somewhere in between this first drafting step and submission to the Council of Ministers, mostly because the responsible ministerial staff were unwilling to use the prescribed document format at these early stages already. This practice, of course, meant a lot of reformatting work for the staff of the Federal Chancellery, which had to bring the documents in full accordance with the «E-Law» format before forwarding them to the Parliamentary Administration.

Meanwhile, this has changed, because, on the one hand, the legislative drafters became used to the document format more and more, and, on the other hand, in 2007 an additional interface was created which allowed the federal ministries to directly submit the draft bills, when being sent out for consultation, to Parliament, whereas before they had to do so outside the system anyway.

The next step to be taken will be to extend the workflow system to the non-governmental bodies involved in the consultation process, so to enable them to directly integrate their respective comments in the workflow system. This, on the one hand, will spare the administrative staff the time now required for relating the comments to the draft bills through adding metadata, and, on the other hand, if an approach is chosen that will allow the commentators to structurally relate their comments to the respective draft bill’s structure, it will also offer the opportunity to automatically generate the synopsis of the consultation results. Currently, however, this project is postponed due to budgetary restrictions.

Since 1849, authentic publication of laws in Austria has been in a print Law Gazette, though, of course, on the ground of legal provisions undergoing some changes. The last major change, in advance of introducing the «E-Law» system, was issued in the Federal Law on the Federal Law Gazette of 1996 (FLG no. 660/1996), by which the Federal Law Gazette was given a new structure consisting of three series for (roughly spoken) laws, ordinances, and state treaties, resp.

The Federal Law Gazette was obtainable from the State Printing Office, on a subscription basis as well as by purchase of single pieces. Altogether, about 6000 copies of the Federal Law Gazette were printed, a number that had been decreasing over the years. It was available in the larger public libraries, as well as at the municipal offices of the about 2350 Austrian municipalities, where people had the opportunity to consult it for free, during opening hours. There were, of course, some individuals, mostly lawyers, subscribing for the Federal Law Gazette on their own (and at their own expense), but most people were restricted in their access to the Federal Law Gazette by being bound to public libraries and municipal offices.

This situation has fundamentally changed.

Having in mind that, according to recent polls, about 76 % of the Austrian population older than 14 years are using the Internet (Integral, Austrian Internet Monitor-Consumer, Q1/10), and about 70 % of the Austrian households have Internet access (Eurostat, Data in Focus, 46/2009), the shift not only in quantity but even in quality of (potential) access to legal information achieved through authentic electronic promulgation becomes evident: In the former paper environment, there was, on an average, one copy of the Federal Law Gazette for more than 1300 people. Access free of charge was obtainable but coupled with some trouble and investment of time, as well as restrictions in availability. Nowadays, an Internet connection will do.

As a matter of fact, we have to be aware that already a few years before this large-scale free access to the authentic federal laws was facilitated by means of the «E-Law» system, people were enjoying free access to non-authentic kinds of electronic legal information through the Legal Information System of the Federal Chancellery, and they used these kinds of information much more intensively than the authentic Federal Law Gazette, simply based on trust instead of a warranty of authenticity. Above that, it is due to the very nature of the normative system and its dynamics – and this observation, in principle, applies to every modern normative system – that the authentic publication of laws in the respective law gazette would attract less attention than non-authentic (or, in a very few normative systems, authentic) consolidated versions of the law in force, at a given time. Nonetheless, introduction of authentic electronic promulgation of federal laws in Austria, based on the «E-Law» system, has brought a new quality to the accessibility and transparency of the Austrian legal system, and it has done so in a sustainable way, since there will be no way back behind this quality, simply because the Austrian citizens would not be willing to do without free online access to the legal texts any more.

Beyond this primary impact, introduction of the «E-Law» system has significantly affected the normative system in general and the legislative process in particular in several ways, as well:

• First of all, the electronic support of the law-making procedure was an appropriate opportunity for some preliminary «document re-engineering» that made government bills and committee reports more clearly arranged and easier to survey; the outward structure of the documents was made more distinct by dividing them into different files (e.g., legal text, explanatory remarks, comparison of texts). A more efficient use of XML within the production process of the legislative documents will later on facilitate texts to become better structured even internally; this will be accompanied by a systematic replacement of layout-oriented text formats by structure-oriented ones, which will also enhance the machine-readability of legal texts, and thereby support legal knowledge management.
  
  
• Secondly, quick availability of parliamentary business is strongly supported by the «E-Law» system. Whereas in former times often some weeks passed by until more comprehensive items of business were available in print, now they would be available within a few days or even hours. This effect met the need felt by political decision-makers to speed up the legislative process, which on the one hand is due to the ever increasing societal dynamics requiring an increase in the adaptability of the legal system to developing societal demands, whereas on the other hand political decision-makers have also recognized that by accelerating a particular law-making process they can easier escape some unwelcome public discourse. The increase in velocity of the legislative process, thus, may be assessed in a somewhat ambivalent way: While increasing societal dynamics would make it to some extent inevitable, a tactical speed-up may also lead to a lack in democratic discourse and a decrease in legislative quality. It should, anyway, be noted that the tendency of accelerating the legislative process was already observable long before IT tools supporting it were made available; these tools, like the «E-Law» system, have just reinforced, if not boosted, this tendency.
  
  
• Thirdly, new types of documents are made available that make the results of the parliamentary process more easily intelligible: Whereas the legal enactments of the National Council before introducing the «E-Law» system just existed in a few internal paper copies which were not available to the public (that therefore either had to wait for promulgation of a law in the Federal Law Gazette or to reconstruct the amendments adopted both on the committee and the plenary stages, in order to obtain the text of the enactment as finally adopted), now the legal enactments are also published electronically. Remembering the considerable number of citizens that in former times regularly (but in vain) expressed their demand for obtaining the text of a legal enactment already in advance of the promulgation of the respective law, to be able to prepare for the provisions to enter into force, this new class of publicly available documents has also significantly improved the transparency – in this case, the «ex-ante transparency», so to say – of the legal system in Austria.
  
  
• Finally, the usual interval of time between completing the parliamentary procedure and promulgating a federal law in the Federal Law Gazette has been reduced, as well. Unlike the acceleration of the legislative process up to adopting the legal enactment in Parliament, shortening the post-parliamentary phase of the legislative process may not affect the quality of legislation any more but also contributes to the ability of the normative system to respond to societal dynamics by quickly modifying legal provisions if required.
  

Theoretically, there would be two options for a Parliament or rather the organizational structures cooperating within the «parliamentary complex», when aiming at introducing an electronic system to support the legislative process: either to develop a new system from the scratch, which then, if such a project was completed successfully, might perfectly meet the respective specific needs, or to adopt and adapt a system, or a standard, already available; in this case, it might even more be necessary to not only adapt the system to the processes but also the processes to the system. Such requirement for process re-engineering, however, would also offer a major opportunity to detect and remedy process defects.

In a realistic perspective, the latter approach in principle seems to be more appropriate to meet the usual budgetary margins, but also enable further development in line with transnational developments. However, it requires appropriate systems or standards to be available, which was not the case when the «E-Law» system was created. Needless to point out Parliaments worldwide are forming a very small market only, in comparison to the markets IT vendors usually are serving, so that the latter would not deem it economically justifiable to develop an off-the-shelf management system for legislative information. Thus, until a few years ago a Parliament intending to establish a system of the kind either had to develop it on its own, on the ground of tools available, e.g. a relational database like Oracle, or had to have it developed by a system integrator. Even in the case of outsourcing the software adaptation and integration, the Parliament's administration would anyway have to provide the procedural knowledge required to be able to map the parliamentary procedures as well as drafting rules or practices to the respective system components, in particular the workflow system and the legislative document format.

Meanwhile, the standardization process, which is quintessential for the institutionalization phase in the technology life-cycle, has already entered the area of legislation, too: Within the framework of the «Africa i-Parliaments» initiative, which was launched by the Pan-African Parliament along with national African Parliaments in 2005 and is being implemented by UNDESA, an XML-based standard parliamentary document format, which is named AKOMA NTOSO, has been developed and is already in use for publishing the national laws in Kenya; a comprehensive management system for parliamentary information, which is named Bungeni, is still under construction, but the alpha release for testing by the partner Parliaments was presented in February 2010, so that it may well be assumed that it will be operational in 2011. The AKOMA NTOSO standard was also taken into account when conceiving a standard interchange format for legislative and legal documents named CEN MetaLex, which was even standardized within a formal procedure under the auspices of CEN and issued as CWA 15710:2007, and which is under further development.

Thus, when nowadays conceiving a management system for legislative information, one will not escape at least considering the emerging AKOMA NTOSO standard and the Bungeni prototype. Nonetheless, in particular since there are not yet any practical experiences with the latter, still a lot of effort will have to be dedicated to designing an application specific to the respective requirements. In doing so, the concept of a system already implemented and operated successfully may, however, offer some useful reference. That is why in the following the main features of a solution based on the concept and experiences of the Austrian «E-Law» system, or rather the principles of a generic management system for legislative information (as elaborated by Harald Hoffmann together with the author of these lines) are outlined; they may be useful as sort of a reference model, which any concept of a system to be specifically developed, as well as a future off-the-shelf system, may be checked against.

Fig. 5 Abstract management system for legislative information (core elements)

There are to be described three core elements of a state-of-the-art management system for legislative information:

• the workflow system,
  
• the legislative editor, and
  
• the open archive.
  

The object of such a system is what might be generally called the legislative document, which comprises data (in particular, legislative text) and metadata (like layout information, information on the legal status of the document as well as document history, and semantic annotations). Nowadays, the syntax of a legislative document will have to be defined in XML. Style sheets will define the respective sets of metadata required for specific types of legislative documents, e.g. a draft bill or a legal enactment to be published in the law gazette.

A workflow system, metaphorically spoken, acts like a conveyor belt for transporting documents between different stages of production, for archiving the results of each stage, and for guaranteeing technical authenticity. The parliamentary workflow system may be conceived as the middle part of the overall workflow for producing laws, which implies that it is anyway to be recommended to cooperate with the other actors in the legislative process, when establishing a management system for legislative information. The parliamentary workflow receives an input from the bodies responsible for drafting laws, in particular from administrative agencies, or it creates its own input. It outputs authenticated versions of legal enactments, ready for publication on paper (e.g., in the law gazette) but also in electronic form, which is to supersede paper publication more and more.

Within the parliamentary system processing mostly consists of producing and/or registering, editing, and storing draft versions, and authenticating, storing, and forwarding final versions of legislative documents, in accordance with the parliamentary rules of procedure. Specific steps within the workflow might branch into sub-workflows, within the Parliament or outside. Certain interim outputs usually are made publicly available, in the course of the parliamentary process, as is the final output, once it is completed.

Having in mind the potential variety of document types and formats, even in the legislative process, the workflow system has to provide for kind of a container transporting one or more logically/procedurally connected files (e.g., image files, which not rarely would be attached to draft bills or state treaties, but in some cases even audio files and other multimedia data), requiring additional metadata. It furthermore requires a management system for the workflow itself.

A workflow typically just allows for sequential processing of a document. This means that at any given time the document has one and only one owner, and that only the owner is allowed to process the document. Exceptions from this general rule, if required by the rules of procedure of a parliament (e.g., in case a bill is referred to several different committees, which then would deliberate and work on it in parallel), have to be specifically defined. However, even if several sub-workflows are taking place simultaneously, the current status of the document in question can be tracked precisely, on the parallel layers, as well, like it would be the case if the classical sequential workflow paradigm was followed.

Assigning ownership of a particular document to a particular actor would imply application of the check-out/check-in model, or the input/output paradigm, from the sub-workflows' perspective: Working on a document (e.g., editing it) would then require to check the document out of the workflow system. Once editing is finished, the document has to be checked in again, to become available for further steps in the workflow. If there is a precisely defined process chain, the next actor, in such a case, would by a system message be informed about the document being ready to be checked out.

An important consequence of the simple input/output paradigm, in the sequential legislative process, is that as to the legislative workflow Parliament would remain organizationally independent from those (administrative) agencies responsible for legislative drafting and from those (administrative) agencies responsible for publication. All these organizational structures may use separate (even different) workflow systems, just interfaces would be required.

However, the format of the legislative document is, in accordance with all practical experience, required to be a uniform one from document creation through publication. Otherwise, a lot of formatting work would be lost once and again in the course of the legislative process, and efficiency of the process would be impaired. The document format, therefore, needs agreement from all parties involved. Workflows and workflow systems may differ.

Proceeding from the requirement of a common format to be used for producing, editing, and storing the legislative documents, as a prerequisite for an efficient organization of the legislative process, it would also be recommendable (even though not compelling) to provide the different (parliamentary as well as extra-parliamentary) actors in this process with the same type of legislative editor, i.e. the same tool in the technical sense, or a plug-in adapting different tools (e.g., different versions of an editor) to the same standard required to process the legislative documents in a uniform way.

Drafting and editing legislative texts is to follow legislative drafting rules, mapped to templates (style sheets) within the environment of the legislative editor. The legislative editor would therefore confine drafting and editing of legislative texts to the legislative drafting rules applicable on layout and structure of legislative documents. An XML editor would deliver native XML files, and therefore avoid the problem of a general purpose editor whose input/output has to be converted from/to XML. Nonetheless, once again we have to be aware that the legislative XML editors available would not yet fulfill the requirements of standard solutions, and generic XML editors would anyway need some adaptation; thus, general purpose editors like Word continue being a reasonable alternative, and in particular the wide-spread Word editor might find easier acceptance with drafting staff than a less well-known XML editor.

A quintessential issue in producing and processing legislative documents is the modification of existing laws. Usually, Parliament would only process and adopt the modifications, resulting in a legal enactment being an amendment of a law already existing. For efficient (public) access to the laws in force, in most normative systems consolidated versions of the laws would be produced, even if not being rendered legal authenticity. Whereas even in many technically advanced countries it is incumbent on legal editors to intellectually produce the consolidated versions, which is a time-consuming task, the production of a consolidated version of a law may also be conceived a specific issue for automation, which would require, within legislation, strict adherence to legislative drafting rules, as well as authentic metadata being added to the legislative documents at each stage of production. Only such an «ex-ante approach» would allow for the automated production of a consolidated version of a law, which then could claim technical authenticity.

The same consolidation function also applies to editing bills in the course of the legislative process: Regularly, amendments adopted at two stages at least, i.e. at the committee and plenary stages, are to be inserted in the text of a bill. This kind of text incorporation, within the framework of a management system for legislative information using an appropriate editor, may be processed automatically, too, if the mentioned prerequisites are fulfilled. Nonetheless, even if processing of amendments of bills is still to be done intellectually, the legislative editor would provide a much more comfortable environment for doing so than available in the traditional paper-based «cut and paste» culture.

Among the prerequisites of (future) automated processing of amendments and producing consolidated versions, XML conformance is to be named in first place. However, XML conformance is also required to enable publication, in accordance with the style guide to be applied on the respective publication media, without re-formatting. The syntax defined in an XML schema (or formerly a DTD) would describe the structural elements specific to a particular type of document. Through a well-defined style sheet, the respective appropriate publication format may be selected. A conformance check is required for each document submitted to the workflow, in particular from external sources. Documents detected not to be conformant to the schema will be rejected, and have to be corrected, which however partly may be done in an automated way.

XML, as an open standard, also would enable the legislative documents to be electronically signed and thereby authenticated, in accordance with standard legislation in the field. If this option is chosen, one or more authorized persons may electronically sign an XML document at the legally proper point in time, in the course of the production process, thus making it the legally authentic version. Needless to point out the concept of legal authentication is to be separated from that of technical authentication.

As an open standard, once again, but due to its capability of storing metadata inside the documents, too, and for some other reason, XML is a format particularly appropriate not only for storage but even for long-term storage of documents. Within the system concept, a specific core element is to serve as permanent storage, i.e. the open archive.

Being «open» in the technical sense means that information in the archive is not hidden in a possibly proprietary part of the workflow system. Information has to be openly accessible from outside the workflow as well, e.g., using a web front-end for public access. More generally speaking, the open archive has to be able to act as a docking point for all kinds of access including those from, e.g., sub-workflows implemented by different vendors. Whereas it may seem acceptable or even unavoidable to make use of proprietary software in any other part of the system, the archive, for simple reason of ensuring long-term preservation the best possible way, should anyway be open-source; hidden source code, in particular after a vendor has disappeared from the market, may seriously impede future accessibility of document content, and even concluding a source code escrow agreement in practice would not guarantee the source code to be available if needed in the future.

The open archive being the permanent storage facility for all documents produced in the legislative process, i.e. the internal draft versions as well as the final versions made accessible to the public at once (while the draft versions would have to be made accessible only later, in accordance with archival legislation in force), it is to be separated from another facility that usually would make accessible the final output of that process only, i.e. the legal information system. The latter might offer (usually non-authentic) consolidated versions of the laws in force but nowadays also be utilized for (authentic) electronic publication of the laws adopted. Even in case a legal information system is not used for authentic publication of laws but just, e.g., for providing non-authentic electronic versions of laws authentically published on paper, or non-authentic consolidated versions of laws, the legal and legislative information systems will have to be cross-connected: In particular, the legal enactments by Parliament will have to be linked with the respective promulgated laws and, if available, the consolidated versions of the laws in question; this linking will have to be processed automatically.

Whereas the focus so far has been on a management system for legislative information, this being the very core of a parliamentary business application infrastructure, it is to be pointed out that the system easily can be extended to a comprehensive management system for parliamentary information by introducing additional document types and specific workflows related to them, in particular with regard to the other main areas of parliamentary business, like oversight of the executive. The legislative process being of particular complexity, usually these additional workflows would be much less complex, and the respective documents less deeply structured, so that this extension will not mean to be a qualitative but merely a quantitative challenge. This is the Bungeni approach, too, so that one may proceed from the assumption that electronic law-making support in the future will be embedded in integrated management systems for parliamentary business.

The Austrian «E-Law» system was a pioneer, and, given the dynamics of IT in general and legal informatics in particular, is already a dinosaur among the electronic systems to support law-making. As indicated above, the new information and communication technologies meanwhile have reached the institutionalization phase in the technology life-cycle; that is why in the field of IT support for law-making not only research but already implementation efforts, too, in the near future will centre around standardization as a main objective to be typically followed in this phase (Schefbeck, 2010).

Standardization in this field will have to cover different, even though interlinked areas. In principle, one might distinguish:

• Technical standards: e.g., encoding standards, cryptographic standards, layout standards, or naming standards, which are implementable by standard formats
  
  
• Drafting standards: structural standards, content standards, and procedural standards, which are supportable by standard formats, editors, and workflow systems
  

Thus, the everyday work of the future legislative drafter will be performed within a smart drafting environment supporting the drafter in implementing the drafting standards:

• Structural standards would refer to document structure in the formal sense, based on structural rules laid down in legislative drafting guidelines, which nowadays usually would not be binding, but may be declared binding and put under legal sanction in the future, once tools are available and in use that will allow such structural drafting rules to be enforced in the drafting process. A binding standardization of a predefined formal hierarchy of classes of provisions, within a legislative/legal document, and an obligation to follow a block-wise amendment technique, i.e. an amendment technique only replacing indexed elements by others, would easily allow for automatic consolidation of legal texts, once the legacy text corpus has been converted to XML.
  
  
• Currently, the formal structure of a legislative/legal document usually is just a formal hierarchy roughly following the concept of clustering specific provision types, e.g., starting from definitions, and ending up with transitional provisions. Ideally, however, the formal structure would map the logical structure of the legal text; such a relation was to be enforced by a prescriptive standard to be supported by future semantic drafting tools assisting the drafter in identifying and visualizing the logical structures, and transforming them into formal structures. These structures may even be made better conceivable, e.g., by means of visualization or hypertextualizing of references.
  
  
• Content standards would link up with the future way of more clearly expressing the logical structure of legal text in its formal structure, but even go beyond the structure into the language of legal texts. Indeed, normative language is often perceived as an impediment to «understanding», i.e. interpretation of legal texts. Future technical support may, e.g., assist the legislative drafter to implement syntactical rules, or terminological standards, or even force the drafter to do so. Inconsistencies in terminology may be revealed, and the terms and concepts already available within the legal system may be put at the drafter’s disposal, once a legal ontology is implemented within the drafting environment. Thus, legislative drafting will be systematically contextualized in a comprehensive way, and legal knowledge management will improve the content quality of legislation.
  
  
• Needless to point out that the content of legal provisions is reflecting societal interests articulated in the legislative process, and would usually represent a compromise between such interests. Political argumentation, and thereby the content quality of the output of the legislative process, is dependent on the knowledge of the actors involved in said process, on their knowledge about what solution would best serve their own interests, and on their knowledge about the parallelogram of societal interests forming the framework for the negotiation process on a specific legal solution. The new techniques of knowledge management will therefore substantially affect legislative processes, too, and simulation technologies will enable the actors in these processes to better assess the potential impact of particular normative models.
  
  
• Procedural standards would take into account these future knowledge management and simulation techniques, and, e.g., provide for regulatory impact assessment as a necessary step to be taken within legislative procedure. In the past decade, the need for regulatory impact assessment has already been recognized, and various sets of guidelines, like those of the European Commission, SEC (2009) 92, have been compiled; nonetheless, the actual take-off will require quantitative methods and reliable and transparent algorithms to be implemented in technical tools available to legislative drafters as well as bearers of societal interests. Only in this case regulatory impact assessment will sustainably affect and improve «deliberative politics», or the intercourse between civil society and the law-making agencies. This intercourse will be put on a more regular and transparent basis by being embedded in new electronic platforms supporting consultation and petitioning procedures, such procedures being assigned a more prominent and transparent role within the legislative process.
  

The normative system is facing two challenges, that might be called, in the figurative sense, the «syntactical» and the «semantic» one:

• The «syntactical» challenge is to address and access the norms (and draft norms) and may be met by legal and legislative information management, whose IT support already has a long tradition, but may be sustainably advanced by the new technologies, too, e.g. by introducing (global) naming standards to be based on a URN concept.
  
  
• The «semantic» challenge is to interpret norms (and draft norms) and may be met by legal and legislative knowledge management, whose IT support has only been growing in the past years, and will open up new opportunities in the years to come, e.g. by systematically using semantic technologies to contextualize legal information.
  
  
• Legal and legislative information management as well as knowledge management will improve the access by the members of civil society to the normative system, and thereby also their ability and readiness to be actively involved in the process of law-making. At the same time, the new technologies will enable the «professional» actors in said process to improve the quality of normative modelling, and to better take into account the outcome of their intercourse with civil society. Electronic support for law-making will thereby facilitate and promote the ability of the political system to provide legitimization for the normative decisions made.
  

The Austrian «E-Law» system represents an early stage of development in electronic support for law-making. Whereas the document format created to produce, process, publish, and store legislative and legal documents was defined in a form very specific to the tradition of Austrian federal legislation, the concept of the overall solution may well be used to derive from it kind of a reference model of a management system for legislative information, consisting of a workflow system to support all the stages of the legislative process, a legislative editor to draft and edit the documents conveyed within this process, and an open archive to sustainably store these documents. Such a reference model may help in developing contemporary solutions, and even inspire efforts to conceive standard solutions for electronic law-making support.

Bachmann, S. (2003), e-Legislation?, Journal für Rechtspolitik 11, pp. 69-73
Barotanyi, B. (2007), E-Recht: Law Making in a Contemporary Way, Masaryk University Journal of Law and Technology 1, pp. 355-360,http://storage02.video.muni.cz/prf/mujlt/storage/1205310341_sb_g08-barotanyi.pdf
Castells, M. (1996-98), The Information Age: Economy, Society and Culture, Vol. I-III, Cambridge, MA
Eberhard, H. (2003), Die Kundmachungsreform 2004, Juristische Ausbildung und Praxisvorbereitung 2003/04, pp. 187-192
Engeljehringer, W. (2004), Das Projekt e-Recht – eine Erfolgsstory, Forum Parlament 2, pp. 52-57
Habermas, J. (1992), Faktizität und Geltung, Frankfurt/Main
Kelsen, H. (1929), Vom Wesen und Wert der Demokratie, 2nd ed., Tübingen
Lachmayer, F., & Stöger, H. (1996), Austrian Legal Information System, in:Ciampi, C., et al. (ed.) , Verso un sistema esperto giuridico integrale, Milano, pp. 577-581
Lachmayer, F. & Hoffmann, H. (2005), From Legal Categories towards Legal Ontologies, inLehmann, J., et al. (ed.) , LOAIT – Legal Ontologies and Artificial Intelligence Techniques, IAAIL Workshop Series, Vol. 4, Nijmegen, pp. 63-69
Laurer, R. (2004), Neues vom Bundesgesetzblatt, Österreichische Juristen-Zeitung, pp. 521-533
Oberreuter, H. (1986), Parlament und Öffentlichkeit, inLangenbucher, W.R. (ed.) , Politische Kommunikation, Wien, pp. 70-79
Sander, P. (2003), eRecht – Das Gesetz im Internet, Journal für Rechtspolitik 11, pp. 73-79
Saurugger, E. (2004), Die Genese des elektronischen Gesetzgebungsverfahrens im Parlament, Forum Parlament 2, pp. 48-52
Schefbeck, G. (2003), E-Recht: Auf dem Weg von der Workfloworganisation zum elektronischen Gesetz, inSchweighofer, E., et al. (ed.) , Zwischen Rechtstheorie und e-Government, Wien, pp. 231-242
Schefbeck, G. (2009a), Per Anhalter durch das Legiversum: Rechts- und Legislativinformatik 2.0, inSchweighofer, E. (ed.) , Semantisches Web und soziales Recht im Web, Wien, pp. 53-61
Schefbeck, G. (2009b), Electronic Support for the Legislative Consultation Process: Theoretical Concepts and Practical Requirements, inProsser, A., & Parycek, P. (ed.) , Proceedings of EDEM 2009, Wien, pp. 79-90
Schefbeck, G. (2010), Standards und «Best Practice» in der elektronischen Rechtsetzung, inSchweighofer, E., et al. (ed.) , Globale Sicherheit und proaktiver Staat – Die Rolle der Rechtsinformatik, Wien, pp. 427-434
Schwarz, M. (2004), Elektronische Gesetze in Österreich – Zur authentischen Kundmachung des Bundesrechts im Internet nach dem Kundmachungsreformgesetz 2004,www.rechtsprobleme.at/doks/elektronische_gesetze-schwarz.pdf
Szep, C. (2003), Rechtsfragen von e-Parlament (thesis), Wien
Weichsel, H. (2002), Der elektronische Rechtserzeugungsprozeß (e-Recht), inSchweighofer, E., et al. (ed.) , IT in Recht und Staat, Wien, pp. 191-196
Zach, M., et al. (2005) , LEXWAS-Begutachtung – Konzept und einsatzfähiger Prototyp eines online Begutachtungsverfahrens in der Sozialversicherungslegistik, inSchweighofer, E., et al. (ed.) , Effizienz von e-Lösungen in Staat und Gesellschaft, pp. 132

---

Günther Schefbeck, Head of department «Parliamentary Documentation, Archives, and Statistics», Austrian Parliamentary Administration, Dr. Karl Renner-Ring 3, 1017 Wien, AT,guenther.schefbeck@parlament.gv.at