Jusletter IT

Modernising government means reconsidering the way that administrative work is done and how interactions between elected persons, civil servants and citizens can be improved. Governments must engage in various types of tasks such as making administrative decisions, formalizing workflow processes, collaborating with many types of stakeholders (e.g., citizens and other governments¹), making high-level decisions in negotiations and policy making, communicating directly with citizens, and executing managerial coordination and control procedures. In principle, most of these types of administrative tasks can be supported by information technology, and some types of administrative tasks can be dramatically improved by information technology (e.g., communication with citizens, cooperation within and between governmental institutions, complex administrative decision making²).

The ways in which government tasks can be supported by information technology are constantly progressing. When implementing information technology in support of government tasks various factors must be considered, such as which government functions will be supported and how they are organized and managed. Some higher-level considerations concern circumstances, restrictions and sometimes peculiarities of the administrative domain as well as the mutual influences of social, political and institutional factors with information technology.

Concerning the topic of e-Government reference is given to some earlier publications of one author (RT) - so conferring to: (Traunmüller and Wimmer, 2005 and 2007; Chen et al, 2008; Scholl et al, 2009 a, b; Traunmüller, 2010).

The public sector has several distinguishing characteristics. First, the goal structure of the public sector is complex, and must consider outcomes related to education, security and defence, justice, and many other functions³. Second, the public sector must provide a vast array of administrative services, such as issuing licenses and making laws. Furthermore, public agencies, unlike their private counterparts, are not spurred by competition. Many additional distinctions concern institutions and persons. Within governments multiple stakeholders from different cultures interact: citizens, civil servants, elected officials, high-level administrative managers, unions, etc. The users of governmental services are not clients but citizens, and they typically expect good and equal treatment. Furthermore, governments may have millions of service users within some domains (e.g., applications for collecting data for tax purposes). A further difference is that administrations function via complex structures of cooperation among very different types of entities. As a result, administrative work may appear complex and rather strange, and responsibility often seems to be diluted. Administrative culture and historical structures that may sometimes impede change, especially when inertial forces are reinforced by bureaucratic attitudes, are important factors.

Considering these distinguishing characteristics, it follows that one must be cautious when adapting commercial applications to the field of government. Administrative structures have many additional requirements: protecting the rights of citizens, ensuring procedures adhere to applicable laws, safeguarding legal validity. These goals cannot be sacrificed for the sake of process optimization. Furthermore, the domain-specific knowledge of a public agency is an important asset. Even when such knowledge is not perceivable in routine tasks, it is often of paramount importance in times of emergency. When adapting a commercial application for a government entity it is important that this domain-specific knowledge is not compromised.

When performing administrative tasks government agencies are bound by legal norms, which often rigidly harness organizational design. In consequence, administrative procedures sometimes become uneven and ineffective. However, the role of norms is to establish a framework, and they typically provide leeway for interpretation, giving rise to consensus-building as a supplementary mode of work. Furthermore, legal norms give particular meaning to administrative structures, for example by protecting citizens’ rights and safeguarding legal validity.

An important aspect of government action is a division of labour that is often congruent with a division of jurisdictions. The latter tends to delimit separate information domains within government, thereby deliberately protecting against unwanted intrusion. Various forms of coordination exist within this division of labour. They include more or less formalized workflow processes as well as various types of teamwork, negotiations, and managerial instruments such as supervision or inspection rights.

Another distinguishing characteristic of government is the diversity of its functions. There are the three principal realms - legislation, public administration and justice – which have quite different features. Additionally, there are various layers that may be considered. These include governance as a strategic-political layer, diverse administrative bodies as tactical layers and agencies as executive layers. Adding to this complexity, government must operate within diverse domains ranging from social welfare to human resources management.

In addition to serving a diverse set of functions, government must cope with a large variety of knowledge types. Basic knowledge types are stored within repositories such as registers, management information systems and legal norms. This basic knowledge is complemented by service knowledge that must be made explicit. Service knowledge may concern the policy field to be influenced, the specific means and modalities of action, the stakeholders, and distinguishing characteristics of a particular administrative system, etc. In some situations, the type of knowledge that is required may be very specific, as is often the case in legal matters. In other situations, the type of knowledge that is required may be very broad, as is often the case for planning purposes.

We shall now provide two examples of the impact of technology on administration and on administrative decision-making. The first example describes how a large-scale information system is used to collect economic and personal data in support of the income tax in France. The second example advises citizens about the services provided by government as well as their rights and obligations.

As discussed above, a national government acts within many domains: legislation, public action and administration (teaching and research, defence, security, etc…) and justice… These actions must be financed. Within Europe national governments finance their activities through three main sources: value added tax⁴, personal income tax and company tax. An interesting question is - how can governments use technology to improve the process of collecting theses taxes? The traditional process of collecting income taxes from citizens as well as income and value added taxes from companies involved mailing forms to the interested parties, which they would then complete and return⁵. When the tax authorities received these forms, they first needed to extract the information that they contained before processing and printing another form⁶ that informed the citizen or company as to the amount of his taxes.

As access to the Internet became widely disseminated, many governments decided to give their citizens and companies the option of providing their income and value added data online. In France citizens were first given the option of providing income tax data online in 2003.

Today it is still possible for a citizen to file his income declaration using a paper form, and of course it is essential to maintain this possibility because it would be unfair to require certain citizens (e.g., elderly persons, persons who are not interested in the Internet, etc) to use this means of communication.

The form that taxable citizens (i.e., those with a fiscal residence in France) initially used to declare their revenues was complemented by a very well designed four-page calculation form that explained the various steps in the computation and thereby allowed the citizen to estimate the amount of tax that he would owe before receiving the official amount computed by the fiscal administration.

Fig 1 provides a simplified view of the main steps in the computation.

As depicted in the figure, the computation takes three main sources of income into account: salary and/or other income⁷ (e.g., pensions), income from capital (e.g., interest from loans, dividends from shares), rents generated from land and/or buildings⁸. Each of these categories is subjected to various automatic⁹ or conditional parameters. In addition, it is possible to compensate within and between categories¹⁰. The sum of these three categories is the income submitted to taxation before deductions¹¹. Applicable deductions¹² are then subtracted, and finally this amount is divided by R (a variable which represents the composition and legal status of the household)¹³. This yields the taxable income to which the tax schedule is then applied. Finally, a long series of possible deductions¹⁴ are subtracted from the tax indicated in the schedule, yielding the final income tax due.

It should be noted that two main declaration forms exist: a simplified version is sufficient for citizens who have only salary or pension income and a detailed version for those who have additional forms of income. Furthermore, when income from renting of land and/or buildings exceeds a certain amount, a special form is needed¹⁵.

Fig. 1: main steps in the computation of income tax in France

Before providing the overall structure of the application, we recall that when the tax authorities mail tax forms to citizens, the data capture and computation of tax are performed electronically and using documents controlled by tax personnel. Data capture was initially performed by hand and later using optical readers. Historically, taxpayers had to provide income information, proof of this income, and proof of expenses within certain categories (e.g., building investment and maintenance expenses that can be deducted from rent, certificates for gifts that are tax deductible, etc).

Furthermore, the Ministry of Finance’s tax authorities receive information about salary amounts, pension payments, and capital gains from various institutions that are required to declare this information¹⁶.

A very simplified depiction of the application is provided in figure 2. The application is comprised of a set of web pages that exactly replicate the paper form (and provide additional interesting possibilities), an initial page that requests the identification of the user¹⁷ (this information is found on the paper form that is sent to him), his fiscal address as well as his family situation¹⁸ and fiscal status¹⁹

Fig. 2: Example of main steps in the income tax web application in a simple case

After this first step, the taxpayer is prompted to select the types of income that he had in the previous year (salary, pension, revenue from capital, rent from land and buildings, etc). This information will determine the electronic forms that he will be prompted to complete. For example, if the taxpayer is retired but still works in addition to receiving a pension and renting out an apartment, he will be prompted to complete the equivalent of two paper forms:²⁰ the main form and the special form for income from land and buildings.

The first interesting point for the taxpayer is that information that the tax authorities have received from third parties is already completed on the electronic form. For example, if he receives a salary and a pension, the amounts received will be indicated in the corresponding input zones (the tax authorities would already have received this information from the company and pension organizations). The same would be true for capital income (interest, dividends, capital gains, etc). The taxpayer can modify these figures if some information is not provided by third parties.

The second key characteristic of this application is that it automatically computes some intermediary results and transfers this information to the main form. For example, if the taxpayer declares income from rent of buildings, he must complete a four-page electronic form, because this information is unknown to the administration. In this case, the gain or loss from this rent is transferred to the main form and all fiscal rules²¹ are automatically applied such as deduction from other income to a certain limit in case of loss, carry over to the next year in the same class of income if needed , etc. If several electronic forms must be completed (for other types of income), the results are transferred to the main form in the same manner.

A third interesting characteristic is that when the input and control process is finished the system provides the amount of tax to be paid for the fiscal year before the taxpayer electronically signs the declaration. This implies that the system combines several simulation models coupled with a knowledge-based system using most of the legal and fiscal rules derived from the fiscal law of the year to compute the final tax amount.

This application has been highly successful in France. In 2011, 12.8 million citizens out of 23 million submitted their tax declaration online. The benefits can be observed from the point of view of the taxpayer and the administration.

From the administration’s point of view there is an advantage in terms of cost saving²². Tax personnel, who were responsible for data capture before 2003 are no longer needed and can be used for other tasks, in particular more control of the tax declaration. In fact, the system takes care of many controls, which not only checks computations but also applies a multitude of tax rules, thereby increasing the tax declaration’s reliability. The reduction in the number of paper forms is huge and greatly increases the efficiency of electronic storage and access to the information by all personnel of the fiscal service and by the taxpayer himself.

From the taxpayer’s point of view, the first benefits are that he immediately knows the amount of his taxes and does not need to mail his tax declaration. A small additional benefit was created during the first years when a tax deduction of 20€ was provided for filling the tax declaration electronically.

However, the greatest benefit seems to be elsewhere. It is widely observed that the system is heavily used by taxpayers to study the best fiscal strategy for them while abiding by fiscal rules. For example, the best choice between various options can be simulated in a very short time. The tax payer can, for example, simulate different strategies, mixing tax-deductible energy-saving investments (insulation, various heating systems, heat pumps, etc), investments in supplementary pension plans, donations to various organisation that he wishes to support, etc. In other words, the income tax application inadvertently provides the taxpayer with a decision support function in addition to satisfying the initial goals of providing the tax authorities with a more efficient and less costly method of data collection and income tax calculation.

It is likely that another advantage of this online approach is that it exempts the taxpayer from providing most of the justifying documents that were demanded with the paper tax forms (records of gifts, invoices corresponding to deductible expenses, etc). However, the taxpayer must be prepared to provide all of these documents if the tax authorities request them. Having more time, the personnel of the fiscal service can perform more audits to identify fraud.

Citizen Advice Systems lend a helpful hand in various situations of life, for example by providing orientation and referral services as well as information about welfare and civic rights and duties such as facts and figures about the community, information about public services, entitlements to allowances, information about counselling agencies, etc. A typical example is the Austrian Portal help.gv.at. The Austrian portal is the quite successful forerunner of many similar developments²³. It is a virtual guide to Austrian authorities and institutions. This very elaborate system won the 2003 eEurope Award for e-Government. Based on life-events, the citizen receives information about offices and institutions as well as up-to-date information on procedures, fees, deadlines, etc. In a similar manner, enterprises receive information on business situations (to be made more explicit?). The scope is broad with the portal covering 12 federal ministries, 9 federal provinces and more than 2000 municipalities.

Designing a user-friendly system is not easy. Citizens’ true needs can be quite different from those that the administration perceives. Several aspects of the system could still be improved. For example, contact with a citizen might occur rarely and his/her questions may be rather unique. Some answers may be complex, requiring a lengthy explanation. Many questions require a very urgent answer (e.g., request for a new identity card or formalities necessary when changing address). Another problem is raised by the favoured mode of access to the information especially when questions are punctual and rather personal. This occurs whenever the citizen demands not the whole service, but only some specific part of it. Examples are numerous from the purchase of a railway ticket to registration for a course at a community college.

Citizen information is envisaged in Online One-stop Government as a core part of an integrated service management. Many particular functions are included such as: citizen information about various life events, choice of the favoured access channel, aid for filling in forms, matching of the citizen’s demand with the administrative structure, invoking human mediators for help, etc. Several organizational forms are possible. Besides home and mobile access, kiosks or multifunctional service offices may be provided.

An important point is the automatic routing of citizen demands. The goal is to route the citizen either to relevant knowledge repositories or to the agency with the right legal competencies to answer the question. Automatic routing can be achieved using a diverse set of methods: a database, a sophisticated piece of software, a staffed service centre (e.g. call centre) or an official in a particular agency. Today, automatic routing is most commonly achieved by comparing key words; however, a better method involves mapping requests by means of domain ontologies.

Merely connecting citizens with relevant information is not enough however. The information must be presented in an understandable form. It follows that providing information on the web exclusively in an administrative-legal form will not fully satisfy the demands of most citizens. Rather, the designer of a web application must resolve conflicting demands, often by balancing the system’s limited explanatory capabilities with citizens’ needs for explanations.

Improving static and dynamic help is another objective. Although support can be both static and dynamic, an underlying ontology is needed. Dynamic help can be provided using software or human mediators. Providing useful static support requires thoroughly editing, commenting and illustrating the taxonomy in question. Continuous improvement of user-system interactions is desirable. This approach often progresses with small steps, for example by improving comments, then drawing clearer scenarios, later adding better help-functions.

It is highly desirable to incorporate knowledge into software. Improving the ability of citizen advice systems to comprehend requests and provide explanations is an important goal. Including knowledge can help bridge the divide between user language and the legal-administrative jargon that is often used by public agencies. The ultimate goal is to develop intelligent multi-lingual and multi-cultural personal assistants that are integrated in electronic public services

A broad range of methods and tools, such as analytical systems, information retrieval, simulation and knowledge-based systems are available to support decision-making. For details of the technology available we refer to the literature. Implementing such knowledge and tools within government can be especially difficult due to the problems posed by the intricacy and complexity of law.

Some of these problems originate within the subject domain itself. For example the large number of rules, regulations and case law coming from various sources at global, European, national and local levels make it difficult to model law and legal reasoning. This difficulty is compounded by the fact that legal terms are often not adequately defined for a variety of reasons. Sometimes legal terms are purposefully left vague to allow the possibility of future interpretation. There are also genuine inconsistencies and fuzziness in law. Legal terms can also be left vague to serve the interests of particular bureaucrats or political figures. Legal information retrieval, which has changed little in the past several decades, also poses problems. The key problem is that such systems are still keyword-oriented even thought finding a correct answer typically requires understanding the question, not just knowing the keywords. Case- based reasoning tools offer a possible solution. Prototypes using logical reasoning, including deontic logic, probabilistic methods and neuronal nets exist. However, such systems are experimental. Thus, they belong to the scientific realm and have not achieved large-scale use. An additional complication is that in some cases legal-administrative information must be used by administrations from different states. In these cases, difficulties of interpretation often occur. For example, a classical case is encountered when a European citizen works in multiple European states throughout his career, and ultimately retires in and asks for his pension from yet another state. In this situation, which is faced by an increasing number of European citizens, the pension organisations from the various states must cooperate. Problems also arise when it is difficult to find equivalent legal terms across multiple languages²⁴. This is a common problem in domains such as intellectual property, licensing, certificates and academic degrees. Furthermore, some terms may have different connotations across languages. For example, they may imply different professional boundaries as when considering the activities a of lawyer and a barrister. Sometimes, equivalent terms simply do not exist across languages. For example, this problem is often encountered when considering public honours, awards, and titles.

Collaborative decision-making adds a new dimension to conventional decision support systems and might have to include procedural support as well. Classical examples of situations in which collaborative decision support systems can be useful are negotiation and mediation procedures as well as policy formulation. Key characteristics of such domains include complex procedural regulations, activities of different types intertwined and conflicting interests of stakeholders. Generally, any cooperative work relationship is characterised by the fact that multiple actors transform or control an ensemble of mutually interacting objects and processes. To achieve their goals, the actors are mutually dependent on the quality and timeliness of the other actors. The two main types of group work methods are coordinated work and collaborative work, which are supported by workflow systems or Groupware systems. The latter category includes many different types of systems such as brainstorming, argumentation, video-conferencing, and meeting support.

Meeting support systems attempt to support the meeting process itself and perform various sub tasks, including synchronization of communications, agenda setting, problem structuring, evaluation of solutions, and discussion facilitation. It follows that such working environments must blend collaborative methods with phases of strictly structured cooperation such as distributing information by workflow and searching for information. Transitions between these methods and auxiliary functions are needed as well. An example is filtering email to remove junk messages and flag messages that may be of particular interest.

Administrative applications can often be improved through the addition of knowledge-based components, which increase the capacity to understand the meaning of queries. In particular, semantic-based technologies show great promise. A well-established example is data exchange, which is an unconditional precondition for Online One Stop Government. In EDI legal and administrative semantics must be formalised to allow global use of locally created and stored data. In praxis most modelling is done using XML and RDF formalism. This solution is sufficient for many tasks, although advanced applications require the use of semantic ontologies. Semantic technologies are also applied to improve help-systems providing advice for life events and business situations. Furthermore, semantic technologies can be used to build repositories of administrative process knowledge. Such repositories gather experience from different cases and their usage can assist new-hires in their work.

Platforms and portals offer a broad choice of methods and tools. Some tools are easy to use and low-cost, as in the case of discussion forums and mailing lists. Other more advanced solutions include brainstorming software and spatial technology for visualization. Such platforms and portals also support social media. Important applications of social media include e-Democracy and e-Participation. Governments are supposed to support the development of democratic culture. Social media provide new ways for citizens to participate in decision-making and policy formation, thereby better tapping the expertise of citizens and providing political planning processes with more input. In such a way, decisions can be improved and are more likely to be supported. Social media can also be used to support learning.

The widespread use of mobile devices makes access to the Internet quasi-ubiquitous. Social media and mobile business reinforce each other in a co-evolutionary manner. Furthermore, the fact that such a large proportion of citizens use mobile devices such as smartphones and tablets to access the Internet has serious consequences. For example, the widespread usage of such devices has necessitated changes in modes of access and communication. Furthermore, the level and the intensity of service usage has increased. It follows that businesses and governments must adapt their business procedures and communication methods accordingly.

We will restrict our discussion to a few key directions for progress. One of them concerns a more proactive government providing additional services to citizens through the Internet. The income tax application described above is a good example, since it makes it easier for citizens not only to file their taxes but also to compare the various tax options that are available to them. In a next step, data collected from citizens and organisations for several purposes could be merged and utilized to build additional systems. For example, public administration is already helping citizen access data from banks, employers, and pension funds to make their tax declarations easier. In the same way, local government may inform citizens on entitlements or possible fee reductions on recreational capabilities offered by the city.

Acknowledgment: We would like to thank professor Joseph Lajos for his stylistic comments.

Chen, H. et al. Digital Government: E-Government Research, Case Studies, and Implementation. Integrated Series in Information Systems. Springer, New York, 2008.

Michel R. Klein «Supporting Financial Strategy Development in French Towns Using a Financial DSS», Proceedings Information Systems in Public Administration and Law, Österreichische Computer Gesellschaft, G. Kirchmayr, R. R. Wagner, M. Wimmer (Eds.), 2000, Austria.

Michel R. Klein: Government Support for Energy Saving Projects. EGOV 2004:

Michel R. Klein , «Internet, Citizen Access to Town Accounting Data and Democracy», Proceedings of the 13th Bled International Conference on Electronic Commerce, June 2000, Bled, Slovenia.

Michel R. Klein , «Decision support for municipality financial planning», Proceedings 2nd IFIP TC 8 Conference on Governmental and Municipal Information Systems, North Holland, R. Traunmüller (Ed.), 1991.

Michel Klein, Leif Methlie, Expert systems, A decision support approach , Wiley 1995

Traunmüller, R., Wimmer, Maria. 2005. Online one-stop Government. Wirtschaftsinformatik 47, Heft 5, S. 383-386

Scholl, J., Janssen, M., Traunmüller, R., Wimmer, M. (eds.). Electronic Government, Proceedings EGOV 2009, LNCS Springer Verlag, Berlin et al. 2009.

Scholl, H. Jochen, Janssen, Marijn, Traunmüller, Roland; Wimmer, Maria A. (Eds.) (2009): Electronic Government: Proceedings of ongoing research and projects of EGOV 09. 8th International Conference. Trauner Druck. Bd. Schriftenreihe Informatik. Nr. 30.

Traunmüller, R. Web 2.0 Creates a New Government (Invited Paper). In: Andersen, K. et al. Electronic Government and the Information Systems Perspective. Proceedings of EGOVIS 2010, 31 August – 2 September 2010, Bilbao, LNCS 6267, Springer Verlag, Berlin et al. 2010.

Traunmüller, R. and Wimmer, M. Knowledge Management and Electronic Governance, In: Janowski, T. and Pardo, T., (Eds.). Proceedings of the 1st International Conference on Theory and Practice of Electronic Governance, ICEGOV 2007, 10 - 13 December 2007, Macao, China. 2007.

 

---

 

Roland Traunmüller, Johannes Kepler University Linz.

  

Michel R. Klein, HEC, School of Management.

 

---