This post is written by Patryk Ciurak, an assistant in the Department of Legal Informatics, University of Gdańsk, and a member of The Legal Engineering Scientific Coalition (legalengineering.eu).

  • The problem: Digitalisation of public services is excluding people.

  • Why it matters: Without full representation, we overlook service users.

  • The solution: User-centred law is a vital requirement.

The ongoing digitalisation of the judiciary and public administration is a fact, even though its pace and final results may be unsatisfying in some aspects. One of the issues is that in order to create a system that facilitates the realisation of rights or fulfilling obligations, one has to translate the existing law into a code.

This gives rise to a number of problems such as 'gaps in the translation', the incompatibility and 'siloed' nature of the solutions, and, finally, the high maintenance costs resulting from the previous two.

Above all, however, it brings to light an often overlooked and negative feature of many of the regulations: they are intended for the use of public authorities, not for common citizens.

How to avoid a dead-end

Therefore, if digitalisation should avoid running into a dead-end, the priorities of law-making should change. This means focusing on the end-user (user-centred design) which is a concept that originated in the late 1970s and early 1980s and primarily was limited to computer software development.

While drafting the legislation, it is necessary to unequivocally determine who will be the subject to the regulation, whether a right will be granted or an obligation imposed, what is the goal that addressee wants to achieve, what situations the law will be applied in, what entities the addressee will interact with throughout the process etc. These questions may seem obvious and reading the explanations to the drafts of normative acts allows one to assume that the answers are plain to see.

However, the smallest attempt to translate a fragment of the normative act into a computer code usually leads to different conclusions. Despite the fact that the addressees of the norms are explicitly declared to be, for example, entrepreneurs, the existing law is primarily intended to regulate the actions taken by public administration. The legislator specifies which body, when and towards whom they may take action to enforce an obligation or to enable the exercise of a right. The regulations are more focused on the manner of achieving a goal rather than on the goal itself. There is a tendency to indicate what technical means should be used and in what form the action should be taken. Often the addressee should provide statements or facts that the administration can easily obtain from public registers.

Legally speaking

As a result, to create an IT system that is to be used by citizens, the law must be interpreted and it is the interpretation, not the actual text of a normative act, that lays at the base of creating the digital solution. In other words, incorporating the law into computer systems used by citizens is so difficult because the law doesn’t describe actions taken by citizens. Needless to say, interpretations of the same legal act may vary which can lead to differences in the design of the systems and their possible incompatibility (for more information see this paper).

Diverting from the existing approach to drafting legislation and focusing on the addressee of the norms and on the goal that is to be achieved would allow designing an automated (to a certain degree) process of applying the law. The role of public administration bodies (and courts) would be to supervise the process instead of initiating and performing most of the stages. However, it must not be understood as the complete exclusion of the interpretation of the legal text, which would be more harmful than beneficial.

It may also bring to light situations in which the legislator's intentions are not expressed clearly enough. This often leads to confusion of both the addressee and public administration and may result in different, often contradictory, interpretations of the same provisions.

Creating end-user-oriented legislation seems to be necessary for the widespread use of law in the form of computer code. Machine-consumable legislation or Rules-as-Code are supposed to guarantee the same meaning of code and natural language text (which is sometimes referred to as isomorphism) and allow both versions to be treated as authentic texts. Rules-as-Code could serve then as a “single source of truth” for IT systems used by public administration and provide the necessary logic behind the actions taken (see Wong, 2020). For this to be possible, there must be a common starting point as any action related to the translation of the text into code (or vice versa) will lead to "translation gaps" (for more information see this report).

Finding solutions

A possible solution is to write down the assumptions of the proposed regulation in the form of pseudo-code which is a simplified way to model stages of a certain process. The sentences written in this way retain the syntax typical of computer code but are devoid of implementation details needed to execute the code. Some of the commands are also written in natural language to increase readability.

The result could (or even should) be illustrated using diagrams created following the Business Process Model and Notation (BPMN) or the Unified Modeling Language (UML) - depending on the level of detail needed. This approach has a lot in common with creating requirements in the process of software development and facilitates a common understanding of the objectives set for the team working on the legislation. Moreover, it’s relatively easy to read and can be verified without much effort by subject-matter experts who are non-lawyers. An additional benefit is the possibility to identify errors or loopholes in the already existing law itself of which one may not be aware at all.

Societies and economies are becoming more and more data-driven and so should be the law. The priorities of creating normative acts should change which means, among others, focusing on the end-user and introducing a law in the form of computer code. This will be an extremely complicated task, mainly because of necessary changes in legislative technique and the legislative process as well as adapting rules of interpretation of legal texts. But it must be undertaken if the law is to stay effective in securing citizens’ rights and enforcing obligations.

👋 You can create a post like this one! Share your thoughts with a community of public servants. Learn more

(Image Credit: Unsplash)