Computerised Legislative Drafting and Access to Statutes

Ian Brightwell and Richard Dixon Hughes
DH4 Pty Ltd
ianbri@dh4.com.au , and richard@dh4.com.au

Delivered at the 2nd AustLII Conference on Computerisation of Law via the Internet , Australasian Legal Information Institute ( AustLII ), University of Technology, Sydney, Australia, 21-23 July 1999.

Contents Abstract 1. Introduction   1.1 Context of the Project 1.2 Objectives of the LDDS Project   1.2.1 Objectives of Stage 1 1.2.2 Objectives of Stage 2 1.3 Scope of the Project 2. Key Planning Issues   2.1 Why SGML? 2.2 Project Risks 3. Project Status 3.1 The Project Plan 3.2 Tasks Completed or Underway 3.3 Data Conversion Contract 4. The Case for a National Standard   4.1 Current Situation 4.2 Problems with Current Approach 4.3 Move towards a National DTD References   Download

Abstract

This paper discusses issues related to the development of a legislative database for NSW based on use of Standard Generalised Markup Language ( SGML ). In particular it addresses the issues related to selecting SGML as a markup environment and discusses identified project risks. The paper also provides an update on the progress of the project to date and the next stages.

The paper also looks at the benefits of adopting a standard approach among all Australian jurisdictions to reduce the long term development costs of legislation and advocates the use of uniform SGML/XML structures for this purpose.

1. Introduction

This paper largely stems from work undertaken by DH4 Pty Ltd on behalf of the Parliamentary Counsel's Office (PCO) of NSW. DH4 is currently assisting the PCO prepare a tender for the supply and installation of a proposed new Legislative Drafting and Database System (LDDS).

1.1 Context of the Project

The PCO's computer system is critical for the efficient operation of its core functions of drafting and publishing NSW legislation and in enabling it to meet the key objectives in its Corporate Plan.

The current system was designed in 1989, following the closure of the Government Printing Office. Under this system, legislation is drafted on network PC's running WordPerfect 8 and is stored in a repository as WordPerfect 5.1 documents. Until recently, all documents were prepared for publication using CAPS, a specialised Unix application, which runs on SUN/Solaris workstations and generates hardcopy output of typeset quality. This approach had the disadvantages that additional proof-reading was required to ensure that any drafting changes to a published document are reflected in both the CAPS system and the WordPerfect document. Also, the earlier versions of Wordperfect had problems in the representation of some tables, formulae and graphics. More recently, hardcopy approaching typeset quality has been generated directly from documents in the repository by using Wordperfect 8 as a publishing tool.

The current system has enabled the PCO to produce all NSW legislation and ancillary publications in-house on a cost-effective basis. More importantly, it has enabled the development of an electronic database of all NSW Legislation, which is both comprehensive and up-to-date and which was the first such collection of State legislation in Australia and was one of the first collections made available to Australasian Legal Information Institute for publishing on their Internet site.

The PCO's strategy as described in its recent Annual Reports to Parliament, is to seek a development path for its IT systems and Legislative Database that will not be restricted by the limitations of proprietary software or document formats. In particular, the Office is seeking a long-term path for at least the next five to ten years that will:

• allow the use of cost-effective IT tools for its core legislative drafting and publishing functions
• support more streamlined and automated publishing processes
• provide greater accessibility and portability of the Legislation Database

The development path is intended to ensure that the PCO retains its position in terms of best practice in the use of IT for legislative drafting and publishing.

Towards this end, the PCO engaged DH4 in 1997 to undertake a strategic assessment and analysis of available options. A business case for a new Legislative Drafting and Database System (LDDS) was subsequently prepared and endorsed by the NSW Government.

1.2 Objectives of the LDDS Project

The overall objectives of the LDDS Project are to:

• provide a cost-effective solution for the replacement of obsolete drafting and publishing tools
   
• streamline and automate the legislative publishing and updating process
   
• reduce the life cycle cost of converting data as new drafting and publishing tools are implemented
   
• improve the quality and usefulness of the electronic product
   
• improve public access to legislation

The project is being addressed in two stages, with a target completion date of mid-2001.

1.2.1 Objectives of Stage 1

The objectives of Stage 1 of the LDDS Project are to:

• select and install a new drafting tool to support SGML based drafting,
   
• develop DTD (Document Type Definition) and Layout files,
   
• convert most of the current legislative database to SGML ensuring that the converted data is suitable for incorporation into an automated consolidation environment,
   
• reduce costs associated with electronic publishing for web applications through the provision of legislative data in an SGML format (the web publishing language HTML is derived from SGML),
   
• reduce production cost of printing legislation by providing final documents in an electronic format thus allowing the use of print-on-demand technology.

1.2.2 Objectives of Stage 2

The objectives of Stage 2 of the LDDS Project are to:

• reduce cost of managing the consolidation of legislation through the installation of an automated consolidation system,
   
• improve management of electronic information through the installation of electronic records, document and workflow management systems,
   
• improve work allocation and management reporting through the installation of a practice management system,
   
• improve the effectiveness of legislative data for web applications by providing SGML data which intrinsically supports point-in-time searching.

1.3 Scope of the Project

The LDDS project is designed principally to implement the new legislative drafting and database system described in PCO's business case ( DH4 Pty Ltd ). The business case described the 'Initial Project' as having the following major tasks:

• installation and configuration of the new integrated drafting and publishing tool to replace WordPerfect 8.
   
• conversion of the NSW Legislation Database to an SGML format suitable for electronic publishing and distribution and the support of any future automatic consolidation of amendments,
   
• selection, installation and configuration of records and electronic document management system and practice management system,
   
• selection, installation and configuration of a document repository to provide a secure environment for the storage of documents with integrated search and document publishing capabilities,
   
• design, installation and configuration of an automatic consolidation of amendments module.

The scope of Stage 1 of the LDDS project is limited to implementing a subset of the current legislative database, which represents most of the NSW Statute Book.

The scope of Stage 2 includes the construction, testing, training and implementation of the automated consolidation system, document, records and workflow management for and Practice Management System.

2. Key Planning Issues

2.1 Why SGML?

Standard Generalised Markup Language ( SGML ) is a standards-based document format available for storage of textual data. SGML offers the following advantages to legislative drafters:

• system independent - document content is stored along with a user-defined document structure which is independent of the editing tool or computing platform being used.
   
• non-proprietary - the documents can be directly used by a wide range of drafting tools with less risk of incompatibilities due to vendor-driven changes in internal data representation.
   
• allows 'smart' drafting tools to be developed through the use of explicitly represented document structure combined with macro programs
   
• allows 'intelligent' searches of legislative data
   
• easily converted to HTML (or potentially XML) for use with Internet-access allowing easy distribution by electronic means.

Currently, the Tasmanian PCO is the only jurisdiction in Australia using SGML in the legislative process. However, there are several projects in progress overseas and locally which are either using or investigating the use of SGML to manage their legislative database. These include: HMSO, Formex (EC documents), Kluwer (Dutch legislation) and the Singapore Government (which uses FrameMaker +SGML). Also legal publishers such as Butterworths, CCH and Desktop Law have selected SGML as their preferred data storage format.

2.2 Project Risks

The principal risk associated with the use of SGML is possible obsolescence, which may occur if it is not supported actively enough in the market place. The potential contender to replace SGML is XML. Nevertheless, SGML appears to be no more disadvantaged than any proprietary data structure with respect to its life expectancy.

Other factors that need to be carefully managed during implementation of an SGML-based legislative drafting and database system include:

• Conversion - Converting documents from proprietary formats to SGML will be difficult - but not significantly more difficult than converting between proprietary formats or between the formats used with subsequent versions of a proprietary product.
   
• Balancing support and stability of WordPerfect 8 against its relative low cost.
   
• Low cost SGML editors include WordPerfect 8/9, XMetal from SoftQuad (a recent release) and more traditional Framemaker +SGML, ArborText and 3B2, most of which are locally supported. However, as yet, most low cost SGML editing tools lack sufficient flexibility to cope with the requirements of legislative drafting, particularly macro language, forms, and graphics.
   
• In situations where legislative drafters have been using word processing in a relatively unstructured way, they may initially find it difficult to adapt to a more structured drafting regime.

3. Project Status

The current project was conceived over twelve months ago with $1.5M funding given to the PCO for a three year period commencing with the 1998/99 financial year.

3.1 The Project Plan

The major dates for each major component of the LDDS project are as follows:

Stage 1 (Sep 1998 - Dec 2000)

• Arrange external technical advice and assistance
• SGML/XML Training
• Assess and select core drafting tool (including evaluation of WordPerfect 8 and 9 and Framemaker +SGML)
• Pilot document management and workflow
• Legislative review & changes
• RFT Development and tendering process
• DTD Development and Data Conversion Contract
• Drafting Tool Contract
• Quality Assurance Contract
• Document repository and web server

Stage 2 (Apr 2000 to Jun 2001)

• Prepare project plan and manage RFT
• Implement automated amendment and consolidation system
• Implement records, document management and workflow systems
• Implement practice management
• Post implementation review

3.2 Tasks Completed or Underway

A number of project initiation tasks have been undertaken. These include:

• the engagement of consultants to assist the PCO supervise the conversion contract
   
• amendment of older legislation to remove inconsistent structures and formats
   
• review of the current legislative database
   
• identifying the location of tables, graphics, forms and formulae in current legislation
   
• developing policies to increase the standardisation of current data and reduce the effort required for data conversion
   
• Office-wide SGML familiarisation presentations
   
• participating in community awareness program through establishment of an LDDS Advisory Panel and through conference presentations
   
• prepare RFt for DTD Development and Data Conversion

The most significant task commenced to date is the evaluation of a drafting tool, which has (so far) focussed on assessing WordPerfect Versions 8 and 9 (because WordPerfect is the only fully-functioned word processor that supports SGML as a native document format and is the PCO's incumbent environment) and Framemaker +SGML. The selection of the drafting tool is critical, as this is a core component of the LDDS and its capabilities must influence the overall system architecture, PCO training/support needs and on the tasks required of the data conversion contractor. At the time of publishing the final selection had not been made.

In addition to the above tasks the PCO has been pro-active in providing legislative documents on the web in PDF. The PCO's own web site contains several publications about the PCO and the preparation of legislation and the status of legislation eg. 'Legislation in Force' which is updated weekly.

In addition, the PCO has collaborated with the NSW Parliament and is currently publishing Bills on the web. These can be found at the NSW Parliament web site .

3.3 Data Conversion Contract

During Stage 1 a contract will be let for conversion of the existing legislative database into SGML this will occur in two phases.

The first phase will require the contractor to provide a fixed price for the creation of a DTD to satisfy the legislative drafting and publishing requirements of the PCO. The resulting data types will be subjected to evaluation and acceptance testing before proceeding to next phase. The PCO has reserved the right to discontinue the contract at this point.

The second phase will entail the creation of marked-up SGML documents from existing WordPerfect 5.1 documents and some paper based information for legislation selected from the NSW Legislative Database.

An additional contract will be let to implement the converted data using the selected drafting tool for the PCO and to configure style/layout files to generate final art work or electronic file. The specific responsibilities of the contractor for this contract will be:

• Configuration of the drafting tool to satisfy the drafting and printing requirements of the PCO.
   
• Configuration of drafting system to produce printed documents which satisfies the requirements of the PCO in terms of 'final artwork'.
   
• Preparation of documentation sufficient to maintain and operate the supplied system.
   
• The training of all staff to level sufficient to allow the PCO to produce required new legislation with the new system using SGML data at all stages.

The PCO will undertake the conversion of miscellaneous graphics, formulae, tables, forms from a paper format to electronic. Also the supply and installation of a search engine suitable for retrieving and displaying converted legislation both over and Intranet and the Internet will be created.

4. The Case for a National Standard

4.1 Current Situation

All jurisdictions in Australia are now preparing the bulk of their legislative information as electronic documents using common desktop word processing tools such as MS Word or WordPerfect. These tools have led to significant cost savings and productivity gains in drafting offices and have also provided legislative information in electronic forms which have facilitated electronic delivery via services such as AustLII and the Commonwealth's SCALE+ system.

The internal savings from using these tools stem from the reduction in data handling compared to older paper based methods. Due to the availability of a low cost drafting tools, legislative drafters can create a document which is almost ready for mass printing. We say 'almost' because, at present, there are some editorial and publishing technicalities which are often more effectively handled by clerical support staff.

4.2 Problems with Current Approach

Unfortunately, there is little commonality between each jurisdiction's method of using drafting tools. Each jurisdiction has its own style sheets which to a greater or lesser extent reflect the structure of the local legislation and assist the offices format the final printed product. When legislation is 'applied' from one jurisdiction to another the process of incorporating the legislation into the receiving jurisdiction's statute book can be onerous, particularly where electronic documents have to be converted from one form to another and then manually verified by proof-reading.

Similarly, incompatibilities between different versions of proprietary word-processing products or file conversions between different products now incur significant costs in editing, correcting and validating legislative data.

Additionally the current ad hoc approach adds significant cost to the delivery of the law to the community. This is reflected in the fact that an industry has emerged which does little more than convert legislative documents from the various jurisdictions' formats to a common format for electronic publishing and in some cases value adding.

Moves toward electronic publishing and the automation of Parliamentary support processes have the potential to expose any weaknesses in the technology chain used to produce legislation. This point is developed in the recent Victorian Law Reform Committee Report ( Parliament of Victoria ) which made many far reaching recommendations regarding the use of information technology and the law. It is interesting to note that in section 11.42 of the report the following statement was made 'The Victorian (web) site is the bi-product [sic] of a larger integrated legislative management system. Called the 'Legislative Document Management System' (LDMS), it provides support for drafting and automates the process of the progress of a bill through Parliament.'. This statement supports the concept that the efficient operation of 'front office' legislative publishing systems (particularly web based systems) are dependent on 'back office' support systems and can potentially fail due to any lack of control or effectiveness in those systems.

4.3 Move towards a National DTD

Various Australian Governments and legal commentators have postulated that access to the law is a fundamental right for all Australians. The NSW Law Society ( Access to Justice Task Force ) found that if a community is to exercise their right to participate in the justice system, it is important that the justice system is seen to be and is:

• accessible and affordable;
• readily easy to understand; and
• fair, efficient and effective.

The ability for citizens and their legal advisers to have accurate, up-to-date and cost effective access the statutes is becoming an increasingly important aspect of their ability to identify the law and to obtain access to justice.

The development of a single national Document Type Definition (DTD) to satisfy many (if not all) of the features of Australian legislation is an achievable goal. The adoption of a common national DTD would be a significant achievement for Australia and have the following benefits.

• improve access and affordability by:
• • reducing the cost of downstream electronic publishing
  • encouraging value adding to legislation
  • reducing duplicated system development effort
     
• additionally it will allow 'applied' legislation to be more easily transported across jurisdictions again reducing the cost of producing and maintaining legislation to government.

A major benefit for all jurisdictions would be the ability to conjointly develop or share sophisticated technology for handling amendments to legislation and to provide historical views of legislation as it stood in previous times. Australian industry and research organisations have already developed some expertise in providing such capability and a national focus may provide opportunities for our IT industries.

It is interesting to note that in the USA a Multistate Legislative Document Management Project has recently been established to consider the possibility of a multistate replacement of TextDBMS and the development of a comprehensive legislative document management system. The following goals for this project are:

• Develop a document and information model for legislative documents
• Determine whether a multi-state text processing standard is feasible and desirable.

Additionally the Legal Information Standards Council sponsored by the Law Foundation of NSW is developing some best practice standards for the document naming and formatting of legislation in electronic form. Although this is a tentative first step, it is important that a strategy be developed to encourage the national development of a suitable standard for all jurisdictions.

It is suggested that the first step in achieving this goal is to prepare a business case quantifying the benefits of standardising legislative mark-up. Given that sufficient benefits are demonstrable, a national forum should be established to prepare and manage a strategy for the implementation of suitable standards. The ultimate objective of this strategy should be to define a nationally acceptable DTD.

It is not expected that a DTD of this type is going to provide the detail required by each jurisdiction to fully implement their drafting environments. It should however, provide a common subset of information that publishers can rely upon for down-stream processing and that jurisdictions can use for data interchange and as a basis for systems design.

References

DH4 Pty Ltd: Business Case - Legislative Drafting & Database System, Unpublished report for NSW PCO, September 1997.

SGML: Standard Generalised Markup Language – ISO 8879-1986.

Parliament of Victoria: LAW REFORM COMMITTEE REPORT, Melbourne, Government Printer, May 1999, No. 52 Session 1998–99, ISBN 0-7311-5272-7, < http://www.parliament.vic.gov.au/lawreform/tech/default.htm >.

Access to Justice Task Force: New South Wales Law Society, Access to Justice Task Force, Access to Justice Final, Report December 1998, < http://www.lawsocnsw.asn.au/about/papers/access_to_justice/access_to_justice-3 .html >.

---

This is a Conference Paper published on 29 February 2000.

Citation: Brightwell I et al, 'Computerised Legislative Drafting and Access to Statutes', Conference Paper, 2000 (1) The Journal of Information, Law and Technology (JILT). <http://elj.warwick.ac.uk/jilt/00-1/brightwell.html>. New citation as at 1/1/04: <http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/2000_1/austlii/brightwell/>

---