inscription of behaviour and flexible interpretation in information infrastructures: the case of...

18
Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs Stefan Henningsson , Helle Zinner Henriksen Center for Applied ICT, Copenhagen Business School, Howitzvej 60, 2000 Frederiksberg, Denmark article info Article history: Received 11 November 2009 Received in revised form 12 April 2011 Accepted 19 May 2011 Available online 1 July 2011 Keywords: Information Infrastructure e-Customs e-Government Inscription of behaviour Interpretative flexibility Digital infrastructure abstract Despite the increasing importance of large IT-based solutions binding actors and processes together across institutional borders, still little is known about how these Information Infrastructures (IIs) assume their shapes and potentially may be reshaped towards specific ends. We focus on the duality of organizations using the IT artefacts of the European e-Cus- toms IIs to inscribe harmonised behaviour into the operation of the infrastructure and how the IT artefacts are divergently interpreted by the users of the II. We find a tension between the need for artefacts with strong inscription to regulate the II and a need for flexibility for II users. The consequence is diluted inscriptions and, in contradiction to what has previ- ously been concluded about interpretative flexibility, the European e-Customs II does not show signs of settling down in a stable phase with consensus. This lack of stability forces the users of the European e-Customs II to continously modify and investment in IT participate in the infrastructure, explaining an expressed longing for the ‘good old paper days’. Based on theoretical integration and empirical findings, we develop a model of the duality of inscription and interpretation in IIs. Ó 2011 Elsevier B.V. All rights reserved. 1. Introduction The information processing capacity, supporting control of European trade, is comprehensive and complex (Rukanova et al., 2010). It consists of both computerised and non-computerised parts that are located at both commercial organizations, such as traders and shipping companies, as well as at governmental bodies, such as customs and health authorities. Two mil- lion traders and governmental organizations from 27 countries take part in the information exchange that spans organiza- tional and national borders. Altogether, the traders in 2007 completed more than 183 million customs declarations, representing more than 2000 million tons of cargo to be processed, assessed and verified by the more than 3000 customs offices spread over Europe (TAXAUD, 2010). The information processing capacity is the backbone on which control mecha- nisms operate, enabling Europe’s customs organizations to monitor trade and to collect import and export duties (Tan et al., 2010). That the information processing capacity is effective is essential for customs to perform their duties and for traders to carry out time critical business transactions (Rukanova et al., 2010). The IT based solution, supporting control of European trade, is an example of what has been denoted as a new type of IT artefact – Information Infrastructure (II) (Tilson et al., 2010; Hanseth and Lyytinen, 2010; Gal et al., 2008). Also referred to as digital infrastructure, II can be defined as ‘‘the basic information technologies and organizational structures, along with the related services and facilities necessary for an enterprise or industry to function’’ (Tilson et al., 2010, p. 748). IIs are charac- terised by the number and heterogeneity of included components and relations, and their dynamic and unexpected interac- tions (Hanseth and Lyytinen, 2010; Gal et al., 2008; Monteiro and Hanseth, 1996; Star and Ruhleder, 1996; Turner et al., 0963-8687/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jsis.2011.05.003 Corresponding author. E-mail addresses: [email protected] (S. Henningsson), [email protected] (H.Z. Henriksen). Journal of Strategic Information Systems 20 (2011) 355–372 Contents lists available at ScienceDirect Journal of Strategic Information Systems journal homepage: www.elsevier.com/locate/jsis

Upload: stefan-henningsson

Post on 09-Sep-2016

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

Journal of Strategic Information Systems 20 (2011) 355–372

Contents lists available at ScienceDirect

Journal of Strategic Information Systems

journal homepage: www.elsevier .com/ locate/ js is

Inscription of behaviour and flexible interpretation in InformationInfrastructures: The case of European e-Customs

Stefan Henningsson ⇑, Helle Zinner HenriksenCenter for Applied ICT, Copenhagen Business School, Howitzvej 60, 2000 Frederiksberg, Denmark

a r t i c l e i n f o a b s t r a c t

Article history:Received 11 November 2009Received in revised form 12 April 2011Accepted 19 May 2011Available online 1 July 2011

Keywords:Information Infrastructuree-Customse-GovernmentInscription of behaviourInterpretative flexibilityDigital infrastructure

0963-8687/$ - see front matter � 2011 Elsevier B.Vdoi:10.1016/j.jsis.2011.05.003

⇑ Corresponding author.E-mail addresses: [email protected] (S. Henningsso

Despite the increasing importance of large IT-based solutions binding actors and processestogether across institutional borders, still little is known about how these InformationInfrastructures (IIs) assume their shapes and potentially may be reshaped towards specificends. We focus on the duality of organizations using the IT artefacts of the European e-Cus-toms IIs to inscribe harmonised behaviour into the operation of the infrastructure and howthe IT artefacts are divergently interpreted by the users of the II. We find a tension betweenthe need for artefacts with strong inscription to regulate the II and a need for flexibility forII users. The consequence is diluted inscriptions and, in contradiction to what has previ-ously been concluded about interpretative flexibility, the European e-Customs II doesnot show signs of settling down in a stable phase with consensus. This lack of stabilityforces the users of the European e-Customs II to continously modify and investment inIT participate in the infrastructure, explaining an expressed longing for the ‘good old paperdays’. Based on theoretical integration and empirical findings, we develop a model of theduality of inscription and interpretation in IIs.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

The information processing capacity, supporting control of European trade, is comprehensive and complex (Rukanovaet al., 2010). It consists of both computerised and non-computerised parts that are located at both commercial organizations,such as traders and shipping companies, as well as at governmental bodies, such as customs and health authorities. Two mil-lion traders and governmental organizations from 27 countries take part in the information exchange that spans organiza-tional and national borders. Altogether, the traders in 2007 completed more than 183 million customs declarations,representing more than 2000 million tons of cargo to be processed, assessed and verified by the more than 3000 customsoffices spread over Europe (TAXAUD, 2010). The information processing capacity is the backbone on which control mecha-nisms operate, enabling Europe’s customs organizations to monitor trade and to collect import and export duties (Tan et al.,2010). That the information processing capacity is effective is essential for customs to perform their duties and for traders tocarry out time critical business transactions (Rukanova et al., 2010).

The IT based solution, supporting control of European trade, is an example of what has been denoted as a new type of ITartefact – Information Infrastructure (II) (Tilson et al., 2010; Hanseth and Lyytinen, 2010; Gal et al., 2008). Also referred to asdigital infrastructure, II can be defined as ‘‘the basic information technologies and organizational structures, along with therelated services and facilities necessary for an enterprise or industry to function’’ (Tilson et al., 2010, p. 748). IIs are charac-terised by the number and heterogeneity of included components and relations, and their dynamic and unexpected interac-tions (Hanseth and Lyytinen, 2010; Gal et al., 2008; Monteiro and Hanseth, 1996; Star and Ruhleder, 1996; Turner et al.,

. All rights reserved.

n), [email protected] (H.Z. Henriksen).

Page 2: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

356 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

2006) in IT solutions that are the shared responsibility of several organizational entities (Northrop et al., 2006; Janssen et al.,2009). IIs are not conceptualised as static entities, but dynamic and evolving, as technological innovations are introduced orsocial practice is changed (Tilson et al., 2010; Ciborra, 2000).

Deregulation of markets and a general process of internationalisation are fuelling the development of IIs that extend be-yond organizational and national borders. Effective IIs hold considerable benefits for business and society at large, in thatcross-organizational information sharing is enabled, thereby reducing costs of already existing interactions and enablingnew ones (Hanseth and Lyytinen, 2010). An effective II can here be defined as an II that is robust (limitations in downtime,errors, security flaws, etc.) and that enables efficient information exchange between parties using the II. However, design anddevelopment of effective IIs has proven difficult, incurring huge losses in foregone investments, opportunity costs, and polit-ical and social problems (Hanseth and Lyytinen, 2010), exemplified by the difficulties with developing a nationwide e-Healthsystem in the UK (Sauer and Willcocks, 2007; Greenhalgh et al., 2008) and the struggle to introduce international standardsfor EDI (Damsgaard and Truex, 2000). The evolution of IIs has been compared to ‘‘drifting’’, in many cases completely out ofmanagerial control (Ciborra, 2000). The problems have been attributed to existing software engineering principles and de-sign methodologies being ill-suited for development of IIs, with its dynamic characteristics and often shared responsibility(Northrop et al., 2006). There is thus a need for research focusing on the processes that shape IIs, and the steps that can betaken to design and develop effective IIs.

Previous research on IIs has shown that one of the problems with designing and developing IIs is that IT artefacts intro-duced into IIs, frequently do not have the anticipated effects when adopted by the IIs users (Gal et al., 2008; Monteiro andHanseth, 1996; Star and Ruhleder, 1996; Turner et al., 2006): A standards specification, itself triggers further standardisingneeds, in a seemingly endless cycle (Hanseth et al., 2006); an inter-organizational system requires both organizational andtechnical changes in the internals of adopting organizations (Gal et al., 2008); and a collaborative IT application gets rejectedsince it does not fit with the practices of intended adopters (Star and Ruhleder, 1996).

When IT artefacts are introduced into an II, this can be understood as organizations using the IT artefacts of the II, to in-scribe harmonised behaviour into the operation of the infrastructure (Hanseth and Monteiro, 1997; Kallinikos, 2004b). Whenthese introduced IT artefacts are not used in the intended way this can be understood as a consequence of the need to alignthe introduced IT artefacts to other entities in the II, and thus interpret them into different contexts by adopting organiza-tions (Monteiro and Hanseth, 1996). The latter is also known as flexible interpretation of IT artefacts (Orlikowski, 1992;Pinch and Bijker, 1984).

With this paper we approach these two processes as a duality. The duality view is a way to conceptualize that the twoprocesses interact with and influence each other. Our interest in the interaction of inscription and interpretation processes,is the seeming tension that exists between the two processes (Hanseth and Monteiro, 1997). Tilson et al. (2010) describesthis as the ‘‘paradox of control’’. On the one hand organizations seeking to inscribe behaviour into IT artefacts, are interestedin coherent adoption of the artefact to control the use of the II (Tilson et al., 2010; Hanseth and Monteiro, 1997); they aretherefore interested in minimising the interpretative flexibility. On the other hand, for the adopting organizations to be ableto align the artefact with their different installed IT bases and processes, there is also a paradoxical need for maximising theinterpretative flexibility (Gal et al., 2008; Monteiro and Hanseth, 1996). For combining and effectively balancing the twoneeds we require better insight into the interplay between the two processes (Tilson et al., 2010). In addition, the balancingact is dependent on our knowledge of how to inscribe behaviour into IT artefacts, and how it is interpreted. In this area muchis still unknown. Our argument is that the inscription into IT artefacts can only fully be understood if the inscription act isstudied in combination with how the inscriptions are interpreted by the intended users.

Given the limited knowledge that exist of how and why II assume their shapes (Tilson et al., 2011; Hanseth and Lyytinen,2010), this article aspires to contribute additional knowledge on the role of IT artefacts in the design and development of IIs.Specifically we aim to bring further understanding to the following research question: What is the role of IT artefacts ininscription of behaviour into IIs?

With this study we make an inquiry into the socio-material duality of II, following the view on the socio-material dualityof technology as articulated by Orlikowski (1992). The inscription of behaviour into IT artefacts can be seen as a socio-con-structivist view of II where the II is shaped by the context (Monteiro and Hanseth, 1996). The interpretation of IT artefacts isin turn a partly technologically deterministic view, since characteristics of the artefact set limitations for how it can be inter-preted (Kallinikos, 2004b; Rose and Jones, 2004). Through the combined view of the two processes, we acknowledge theduality – that IIs are both shaped by, and shape the context. Although the socio-material view of technology was emphasisedby Orlikowski almost two decades ago (Orlikowski, 1992), research that takes this dual view into account is generally under-represented in the IS literature, yet considered as particularly relevant to increase our understanding of the link between ISand our society (Orlikowski and Scott, 2008).

Empirically, this paper is based on a study of the II supporting control of European trade (henceforth referred to as theEuropean e-Customs II), through the perspective of Danish Customs and a Danish company with pan-European operations.The II is currently going through a transformation that represents the general trend towards creation of IIs, with dramaticallyincreased number and heterogeneity of included components. The European e-Customs II provides an ideal setting for thestudy of inscription of behaviour and flexible interpretation in II, since it is an area where a number of actors attempt toinfluence how the II develops with mixed results. For national governments a transition from paper-based customs processesto e-Customs, is a possible strategic move to increase control in international trade and at the same time increase the com-petitive power of national traders (Tan et al., 2006; Rukanova et al., 2010).

Page 3: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 357

As the European e-Customs II spans many different hierarchies and jurisdictions, the development can as consequencenot be influenced by direct command (Henriksen et al., 2008). Instead, one of the approaches used to promote the e-Cus-toms II development in a harmonised and interoperable direction, is to inscribe behaviour into the IT artefacts of the infra-structure. However, Henningsson and Henriksen (2009) note an inconsistency between the influencers’ ambitions, and howthe users of the infrastructure perceived the work that triggered our interest. Most notably, the exporters and customs offi-cers expressed longing for the ‘good old paper days’, describing the move from a paper-based II to a digital one as increas-ing differences between national customs processes and eventually leading to decreased efficiency. This paper extends thework of Henningsson and Henriksen (2009) with additional empirical data and a focused theoretical frame crafted tomatch this papers’ aim to explain the role of IT artefacts in the problematic development of IIs, such as the European e-Customs II.

This paper seeks to make two principal contributions. First, it contributes to the general understanding of how IIs evolveby integrating theoretical and empirical findings explaining the duality of inscription of behaviour into IT artefacts and flex-ible interpretation in IIs. Second, it draws conclusions on how the introduction of IT innovations may contribute to the erod-ing of working II based inscription and interpretation processes. Our knowledge contributions can be used by businessorganizations, governmental agencies and supra-national bodies to understand how they will be affected by future develop-ments of an II and how they may and may not influence the development of IIs. Understanding the processes of inscriptionand interpretation in II should be particularly relevant to governmental and supra-national organizations that are showing akeen interest in IT’s role in strategies for national and regional development. In the field of e-Government, IT has been putforward as ‘‘a catalyst for social, economic and political change’’ (Fountain, 2002, p. 45), but claims have also been made thatin reality little transformation of the public sector can be observed (Kraemer and King, 2005; West, 2004; Coursey and Nor-ris, 2008).

The remainder of this paper is structured as follows. Section 2 examines previous research on II and positions the e-Cus-toms II in this landscape. It also presents our study design, which we apply to get an understanding of the development of theEuropean c-Customs II. We introduce the theoretical concepts of inscription of behaviour into IT artefacts and flexible inter-pretation, and we outline what the different types of IT artefacts are in IIs and their capacity for flexible interpretation. Sec-tion 3 presents the structured case study approach that was employed in the study and motivates methodological choices.Section 4 describes the European e-Customs II, and attempts to inscribe behaviour. The analysis in Section 5 integrates fielddata and the theoretical concepts of the duality of inscription and interpretation in II. Finally, implications of the study arediscussed, followed by conclusions and implications for future research and practice.

2. Related literature and study design

The notion of II first gained attention in the 1990s as a political notion in the rhetoric of an information-centric society. AsII faded from the political vocabulary, the notion was picked up by the community of IS researchers. During the last two dec-ades II has proven to be valuable to frame a number of extensive case studies (e.g. Star and Ruhleder, 1996; Gal, 2008; Hans-eth and Ciborra, 2007). The concept changed the perspective from systems to infrastructures, and from organizations toorganizational networks (Bygstad, 2008). One of the lessons learned in these studies is that large II, compared to traditionalIS, pose a set of new challenges regarding design and use, which can be attributed to the infrastructural and organizationalrelationships (Star and Ruhleder, 1996). This section gives the theoretical background for approaching inscription of behav-iour and interpretative flexibility in II. The section ends by presenting our integrated study design.

2.1. The concept of Information Infrastructure

II is also referred to as digital infrastructure (Tilson et al., 2010). Although II in principle can refer to a non-digital infra-structure for information exchange, in reality all works the authors have been able to identify using the term within theinformation systems field, refer to digital varieties (e.g. Hanseth and Lyytinen, 2010; Gal et al., 2008; Monteiro and Hanseth,1996; Star and Ruhleder, 1996; Turner et al., 2006, Broadbent and Weill, 1997). In addition to define digital infrastructure,Tilson et al. (2010) draws on Hanseth and Lyytinen (2010), who use the term II. Therefore we make no distinction betweenthe two terms in this paper.

The increasing importance of II comes from its fundamental role in the phenomenon generally referred to as digital con-vergence – the merging of devices, technologies, industries and activities following digitalization (Tilson et al., 2010). Digi-talization and technological innovations have made possible increasingly integrated and versatile digital solutions thatunderpin contemporary organizational life (Hanseth and Lyytinen, 2010; Yoo, 2010). Like Tilson et al. (2010), we conceptu-alize II as the relational constructs binding together the heterogeneous sets of IT artefacts. Consequently, ‘‘these infrastruc-tures cannot be defined through a distinct set of functions (unlike specific systems), or strict boundaries (unlikeapplications). In contrast, they are characterised by dynamism and longevity and are relational in nature.’’ (Tilson et al.,2010, p. 749).

The literature presents two distinct views on II and how they may be reshaped towards specific ends (Gal et al., 2008).One stream of research depicts II as large amalgamations of technological components and human skills. This conceptu-alisation assumes that IIs can be differentiated and distinguished from what is not an infrastructure, and that since

Page 4: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

358 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

infrastructures can be identified, they can also be controlled and managed in a fairly straightforward fashion (Gal et al.,2008). The presumption that IIs can be defined and assessed is demonstrated in works by Broadbent et al. (1999), Broad-bent and Weill (1997) and Henderson and Venkatraman (1992). Further, it becomes visible in the conclusion made byByrd and Turner: ‘‘The IT infrastructure concept can be divided into two related, but distinct, components: a technicalIT infrastructure and a human IT infrastructure’’ (Byrd and Turner, 2000, p. 172).

The separation of the human elements from the technical elements of IIs may facilitate creation of an analytical constructand managerial principles, but oversimplifies the processes that are involved in the shaping and functioning of IIs (Gal et al.,2008). That view misses the relational character of II (c.f. Tilson et al., 2010). Several researchers argue in favour of an‘‘ensemble view’’ of II (Orlikowski and Iacono, 2000), representing the argument that the technological components and hu-man skills of IIs cannot easily be distinguished from the social and organizational context in which they exist (Gal et al.,2008; Monteiro and Hanseth, 1996; Star and Ruhleder, 1996; Turner et al., 2006). Star and Ruhleder (1996) describe thisas IIs being sunk into structures, social arrangements and technologies. Edwards notes that ‘‘. . .although ‘infrastructure’ isoften used as if it were synonymous with ‘hardware’. . . all infrastructures. . . are in fact socio-technical in nature. Not onlyhardware but organisations, socially-communicated background knowledge, general acceptance and reliance, and near-ubiquitous accessibility are required for a system to be an infrastructure’’ (Edwards, 2003, pp. 187–188). Drawing on Ed-wards, 2003, Gal, 2008, p. 18 states: ‘‘Technically, the construction of an infrastructural system requires the establishmentof protocols and standards that enable the system to be used and seamlessly connect with other systems. Socially, its con-struction necessitates the elaboration of a system of classifications that symbolically represent and organise things in soci-ety: people, classes, geographical areas, religions, civil status, and so on.’’

As noted in the introduction, our view of II attends to the second stream of writings. We conceptualize II as a specific typeof IT artefact that is characterised by the number and heterogeneity of included socio-technical components, relations, andtheir dynamic and unexpected interactions (Hanseth and Lyytinen, 2010; Gal et al., 2008; Monteiro and Hanseth, 1996; Starand Ruhleder, 1996; Turner et al., 2006). They are IT solutions that are the shared responsibility of several organizationalentities (Northrop et al., 2006; Janssen et al., 2009). Furthermore, IIs are not static entities, but evolving as technologicalinnovations are introduced or the social practice is changed (Ciborra, 2000). To the extent that planned development takesplace, it is frequently the result of collective action and politics rather than clear-cut strategic rationales (Rukanova et al.,2010). Although a few researchers have already inquired into the processes that shape the II (e.g. Hanseth and Braa,2001; Monteiro and Hanseth, 1996; Kallinikos, 2007; Gal et al., 2008; Star and Ruhleder, 1996), the collective understandingis in its formative stages. Partly due to the concept of the II’s origin in the practical, rather than academic domain, theorydevelopment in the area of II is still limited (Bygstad, 2008).

IIs, itself being conceptualised as a specific type of relational IT artefact, should not be confused with the IT artefacts thatconstitute the building blocks of IIs. Given that we define IIs as relational artefacts, it is not the individual IT-based systemscovered by the II themselves that are in focus, but the IT artefacts binding the systems together in an II. Conversely, the prop-erties of the constituting IT artefacts that are of interest are their ways of binding the II together. For example, Gal (2008)identifies the establishment of protocols and standards that enable interoperability as the technical side of II development.Similarly, Hanseth and Monteiro (1997) identify that the artefacts regulating the communicative patterns constitute thetechnical basis of II. In coherence with our ‘‘ensemble view’’ of II as relational IT artefacts, we define the constituting IT arte-facts of II as the socio-technical components that bind the infrastructure together, making the heterogeneous technologicalbase interoperable, and in relation to the social practice, form the II. It is through these regulating elements the II may bemanipulated. In the literature (Gal, 2008; Hanseth et al., 2006) it is mostly standards, protocols, and other ‘template’-arte-facts that have been ascribed this regulating role, but we will, based on our empirical work, suggest that there are other typesof IT artefacts in which II behaviour can be inscribed.

2.2. e-Government Information Infrastructures

A large variety of artefacts are being called II: from corporate infrastructures to international ones (Tilson et al., 2010).Examples of settings where IIs are prominent range from cross-border police co-operation (INTERPOL, 2010) to financial col-laboration in the SWIFT-network (SWIFT, 2010); from co-creating architects (Gal et al., 2008) to a geographically dispersedcommunity of geneticists (Star and Ruhleder, 1996). The kind of II we approach in this paper, the European e-Customs II, canbe labelled as an e-Government II.

Over the past decades, government organizations have embraced IT-solutions to realise electronic service delivery to cit-izens and businesses and to increase their operational effectiveness. These initiatives are believed to enable ‘‘a longer-termtransformation of government that goes far beyond online service delivery’’ (OECD, 2005, p. 164). Instead of merely digital-izing current government processes and activities, this transformation aims at leveraging the value of e-Government initia-tives (Irani et al., 2007). It is a general belief that IS holds the capacity for strengthening efficiency, providing tools forsecurity, and furthermore being an instrument for streamlining of e-Government procedures (Irani et al., 2007; Lenk,2002). Reduction of the governments’ administrative tasks is welcomed by businesses that often consider their interactionwith the public sector to be a burdensome task (Fountain and Osorio-Urzua, 2001).

Along with the benefits that e-Government service can offer, lie a number of overarching challenges. These include thecreation of inter-domain infrastructures that enable private and public parties to share information electronically. E-Govern-ment Infrastructures are typically used by large numbers of different users (Fountain, 2002) who assume that the

Page 5: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 359

infrastructures exist and operate as expected. A fundamental characteristic of infrastructures is that they are used by manydifferent users, with the usage evolving over time, as may the type of users. (Janssen et al., 2009).

The European e-Customs II has this characteristic inter-organizational span that is not the case for the corporate II (c.f.Hanseth and Braa, 2001; Kayworth and Sambamurthy, 2000; Weill and Broadbent, 1998). Furthermore, using organizationsinclude both private and public organizations, which can be expected to have diverging objectives for infrastructure devel-opment (Janssen et al., 2009; Henningsson et al., 2010). Governmental agencies have a holistic interest in the II, to ensurethat it in all its heterogeneity, operates in a standardised or at least interoperable way that allows for efficiency in interactionwith business organizations (Janssen et al., 2009; Rukanova et al., 2010). It is in this context we approach the processes ofinscription of behaviour and flexible interpretation in II.

2.3. Study design

Following the advice of Webster and Watson (2002) on how to explain the way an article builds on previous research, westart by presenting the general logic of the study design and then explain and motivate the concepts used to understand thestudy. Our study design presented in Fig. 1 conceptualizes the inscription and interpretation of behaviour in II as a dualitythat involves developing and using organizations (Cadili and Whitley, 2005; Doherty et al., 2006; Orlikowski, 1992). The pro-cess starts with one or a group of actors’ attempt to inscribe behaviour into the IT artefacts that will influence the operationof the II. The nature of the inscription and the room for interpretation are dependent on the nature of the IT artefact (Dohertyet al., 2006; Kallinikos, 2004b; Rose and Jones, 2004). Gradually the inscription is enforced and the flexible interpretation islimited as a common understanding of the artefact is established among the members of relevant social groups (Pinch andBijker, 1984).

As a starting point, to understand the inscription and interpretation of behaviour in II, we gain inspiration from Orlikow-ski, 1992 who in her seminal article provides a robust understanding of how organizational design influences technology andvice versa. Orlikowski provides an insight into the time–space discontinuity and hence the shift in interpretative flexibilityover time. Our conceptualisation of inscription draws on actor network theory (ANT) (Latour, 1991) and the notion of ‘pro-grams of action’. Our interest is in how the IT artefacts of IIs carry programs of actions, thus containing stronger or weakerinscriptions. As we do not carry out a full ANT-analysis, we refer to this as inscription strength in our theoretical discussion,case presentation, and case analysis, rather than use the vocabulary of ANT. However, we retain the pertaintent logic of pro-grams of actions to use in the analysis.

We use the term interpretative flexibility with reference to the work of Pinch and Bijker (1984). In their view IT artefactsare socially constituted and are open to more than one interpretation (Pinch and Bijker, 1984). Pinch and Bijker refer to inter-pretative flexibility as a first stage in a process where the view of artefacts evolves from multiple interpretations to sharedunderstanding.

As will be motivated in the following sections, there are two aspects of the study that are of particular interest. We returnto these two aspects in the analysis. First, the capacity of IT artefacts in the II as inscription carriers of visions and objectives,is fundamental to any attempt to use IT artefacts to inscribe behaviour. Previous literature highlights differences in this po-tential to carry inscription (Cadili and Whitley, 2005: Kallinikos, 2004a), but little is known about this potential for IT arte-facts in IIs. Second, we will focus on the interplay between inscription and interpretation. Traditionally there have been twopolarised views on how to conceptualise the relationship between technology and society. On the one hand, technologicaldeterminism holds IT as the casual agent of change where the properties of the technology determine its use (DeSanctis andPoole, 1994). On the other hand, social reductionism or constructivism holds a view that technology is a consequence of soci-ety and its actors (DeSanctis and Poole, 1994). Today most researchers adhere to a position somewhere between the twoextremes (Hanseth and Monteiro, 1997; Doherty et al., 2006; Rose and Jones, 2004), recognising that the competing viewsshould be conceived as a duality: technology is both shaped by the social, and will influence the behaviour of users (Orlikow-ski, 1992).

The next two sections describe the entities of the study design in more detail, how they motivate the construction, anddiscuss how integration of inscription and interpretation in IIs tentatively can be understood.

Phase 2 Phase 3

Common understanding of IT

artefact, stable II

Developing organizations

Using organizations

IT artefact

InterpretationInscription

Visions and objectives

Installed social and technical base

Use modeDesign mode

Phase 1

IT artefact with divergent

interpretations destabilize II

Fig. 1. Integrative research model for inscription and interpretation in Information Infrastructures.

Page 6: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

360 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

2.3.1. Inscribing behaviour into IT artefactsThe conclusion that IT artefacts have both enabling and restricting implications has been reached by the use of several

different theoretical frameworks, including for example structuration theory (Orlikowski and Robey, 1991), phenomenology(Boland and Greenberg, 1992), and actor-network theory (Hanseth and Monteiro, 1997). The enabling and restricting impli-cations can be seen as inscription of behaviour into the artefact (Akrich and Latour, 1992; Latour, 1992), keeping in mind thatthe inscriptions always are interpreted by the artefact’s users (Orlikowski and Robey, 1991). Inscriptions are the scenario inwhich the designed artefact is intended to be used (Hanseth and Monteiro, 1997), or in the terminology of actor-networktheory, an assigned program of action (Latour, 1992). In structuration theory (DeSanctis and Poole, 1994; Orlikowski,1992) structure is embodied in technology through human agency, while in turn, human agency is shaped by the embodiedstructure (Giddens, 1984).

In the terminology of ANT, technologies inscribe strong/inflexible or weak/flexible programs of action (Latour, 1991). Inhis often cited example of the design of hotel keys, Latour (1991) describes how the artefact defines the roles between theuser and the artefact. The example describes how the same artefact shifts from having a weak and flexible program of actionto a strong and inflexible program of action, by adding a heavy piece of metal to the key, instead of having a sign asking theguest to leave the key at the reception, the program of action changes. Because of the added weight the hotel guest leaves thekey in the reception instead of carrying it with her, thereby risking losing it. The hotel key is an example of an artefact with astrong and inflexible program of action because it imposes a particular use.

When users of an artefact do not follow the inscribed program of action Latour calls this an anti-program. How strong aninscription is can be stated as the likelihood that a user will follow the program or an anti-program respectively. In the exam-ple with the door key, the attractiveness of the anti-program (to carry the heavy key with you) is decreased, making theinscription of the desired program stronger. A modern version of the hotel key, the key card, does not have this undesirableanti-program. Supposedly, fewer guests leave the key card at the reception. On the other hand, since replacement of a keycard is low cost, there is little desire for hotels to try to inscribe such behaviour. Rather, the inscription of the desired pro-gram in the design of the key card appears to be that the hotel guest should carry the card in the wallet, as the key card nor-mally is given the shape of a credit card and fits well in the wallet.

In Fig. 1 we outline the IIs developing organizations attempts to inscribe behaviour into the II. This mirrors the inscriptionof programs of action in ANT. The program of action delegates roles and competencies to the users as well as to the compo-nents of the system (Hanseth and Monteiro, 1997). The program of action thus serves as the syntax of the intentions behindthe artefact. The more detailed the syntax is, the easier it is to follow or directly disobey.

There are several examples of how IT artefacts, deliberately or unintentionally, have been used by developingorganizations to inscribe behaviour into IIs. When a standard for communication is established in an II, a certain patternof use is inscribed into the II (Hanseth and Monteiro, 1997). The inscription (attempts to) pre-defines what informationwill be communicated and/or how this communication should be carried out. In a study of II in the constructionindustry, Gal et al. (2008) identified that a shift in the technical elements that mediated inter-organizational collabora-tion had direct implications for the collaboration practice. Another example is how the SWIFT standards define how aninternational bank transfer should be carried out and which information is considered relevant for a bank transfer(SWIFT, 2010). The standards actually go so far as to presuppose that banks have a certain organizational structure, withnational identity, branches, regional departments and front offices. However, the standard allows for flexible use, sincebanks that are lacking this structure just assign the same identifier where the supposed structure is not matching. Athird example of how behaviour is inscribed in II, is how the technical architecture can trigger or inhibit innovation.Zittrain (2006) argues that the success of the Internet is closely related to its ‘generativity’ – the fact that the intelligenceof the Internet lays in the endpoint – the computers, and not the network itself, is what makes it exceptionallygenerative.

ANT-based approaches have been criticised for assuming a use-centric position in which the role of the IT artefact and itsfeatures are downplayed, if not completely ignored (Doherty et al., 2006; Kallinikos, 2004b). In ANT, where nothing but het-erogeneous networks exists, how the IT artefact is used is a matter of its relations to other actors in the network (Latour,1992). This is a standpoint that may detract from the features of the IT artefact itself (Doherty et al., 2006; Kallinikos,2004a). Similar to other technologies, IT artefacts present differences in flexibility of programs of actions that can be attrib-uted to how the artefact is designed and constructed. That is for example the case with word processing software which hasmultiple uses (Kallinikos, 2004a). ERP systems on the other hand, normally have a strongly inscribed pattern of use (Kallin-ikos, 2004a). For IT artefacts, little is known about their potential to sustain inscribed behaviour (Doherty et al., 2006; Kal-linikos, 2004b; Rose and Jones, 2004). How ‘strong’ the inscription of behaviour is, may vary from artefact to artefact. Someartefacts may have a very specific use or assigned program of action, as ANT would phrase it, while others leave much roomfor interpretation (Kallinikos, 2004a). It can be assumed that some IT artefacts, for example a word processor, may be usedfor many different purposes, while an ERP system may sustain a stronger behavioural inscription (Kallinikos, 2004a). It isbased on this foundation that we see a need to further investigate the capacity of IT artefacts as inscription carriers of devel-oping organisation’s visions and objectives.

2.3.2. Interpretative flexibility in Information InfrastructuresThe other side of inscription into IT artefacts is the interpretation of those inscriptions, depicted in Fig. 1 as process, where

the II’s using organisation interprets an IT artefact to fit into its technical and social practice. From a social constructivist

Page 7: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 361

perspective where phenomena such as IT artefacts are socially constituted, it has been broadly accepted for decades thatartefacts are open to more than one interpretation (Pinch and Bijker, 1984). Pinch and Bijker refer to this as interpretativeflexibility.

With reference to the work of Pinch and Bijker (1984), the term interpretative flexibility has received attention in the ISliterature among researchers focusing on socio-technical aspects of IS. The IS literature has referred to interpretative flexi-bility as ‘‘the capacity of a specific technology to sustain divergent opinions’’ (Sahay and Robey, 1996, p. 260). Other exam-ples include Orlikowski (1992), Doherty et al. (2006) and Cadili and Whitley (2005). Doherty and colleagues (2006, p. 571)state that ‘‘Interpretive flexibility expresses the idea that technological artefacts are both culturally constructed and inter-preted’’. Adopting this view, we recognise that interpretative flexibility thus caters to the idea that technological artefactsrepresent different things to different people. Along similar lines, Orlikowski (1992) emphasises that interpretative flexibilityis an attribute of the relation between humans and technology which is influenced by the characteristics of the artefact, thehuman agents, and the context.

We are here interested in the interplay between inscription and interpretation in IIs. In the context of construction ofscientific knowledge, Pinch and Bijker refer to interpretative flexibility as a first stage in a process where the view ofartefacts evolves from multiple interpretations to shared understanding. The second stage in this evolution shows thatsome degree of scientific consensus is reached. In that stage the interpretative flexibility is limited because of anagreement on a shared ‘truth’ among relevant social groups. The third and final stage is the closure and stabilization,where a common understanding of the artefact is established among the members of relevant social groups. In their1984 article Pinch and Bijker emphasise that the first stage is well described in literature whereas the second and, inparticular, the third stage is less studied. The number of relevant groups involved in the closure and stabilization, makesit complicated to actually reach that stage; in particular in relation to technology that constantly undergoes development(Pinch and Bijker, 1984). For an II, a similar pattern would be that in a stable II, when a new IT artefact is introduced toinscribe a behaviour leading to certain objective, it initially destabilizes the II as the users of the II are divertingly inter-preting the new artefact. Eventually a common interpretation is reached and the II assumes a new stable phase. How-ever, this is not entirely matching the view of IIs as constantly evolving (c.f. Ciborra, 2000; Bygstad, 2008; Hanseth andLyytinen, 2010). We therefore recognise these diverging views, and will use the empirical investigation to reinforce thisunderstanding.

Previous research on II indicates, that to understand inscription of behaviour into IT artefacts of an II, it is necessary toalso understand the interpretation process of those IT artefacts. Organizational interpretation of II artefacts means aligningthe II with the internal installed base, both the social and technical (Hanseth and Braa, 2001). In Gal et al.’s study (2008) theinterpretation of a new IT artefact mediating the II triggered a shift in organizational identity for the parts of the organisationexposed to the new IT artefact. Technically, the IT elements of the II are highly interconnected. This is explicitly stated byHanseth: ‘‘The individual elements are very interdependent, and their size and complexity make them extremely difficultto control and manage’’ (Hanseth, 2000, p. 56). The consequence is that future development becomes significantly pathdependent, meaning that the possibilities for future development are, to a high degree, dependent on what currently exists.Drawing on Hanseth (2000), a view to further investigate is that there are limitations to what and/or how strong inscriptionsrealistically can be introduced into an II.

Further, on the relationship between inscription and interpretation, it has been noted that in which part of its life cyclethe IT artefact exists influences its perceived flexibility (Cadili and Whitley, 2005). At earlier stages in the development pro-cess there is a perceived flexibility, whereas later, the degree of flexibility decreases (Cadili and Whitley, 2005). Here there isa connection to be made to Orlikowski’s (1992) work on interpretation that distinguishes between development and utilisa-tion, which she refers to as time–space discontinuity. Orlikowski suggests that interaction with technology should have twoiterative modes: the design mode and the use mode. She stresses that the distinction is an analytical convenience rather thana practical distinction because the two modes of interaction are tightly coupled.

The strength of working with the two modes is that the two temporal stages invite different levels of interaction. In thedesign mode users are naturally engaged in the development of technology; at this stage actions constitute the technologybased on the inherent qualities of the technology. In the use mode on the other hand, users interpret, appropriate and manip-ulate it in various ways under the influence of other factors such as contextual, social and human factors. At this stage actionsare constituted by the technology, and humans assign shared meaning to the technology, ultimately leading to stabilization,which is the situation where relevant social groups see their problems as being solved by the technology in question (Pinchand Bijker, 1984). In an empirical study of evaluation of a nurses management system using elements of interpretative flex-ibility, Wilson and Howcroft (2005) find that especially the notion of ‘‘relevant social groups’’ represent a challenge in rela-tion to stabilization because the success of the system depends on whether the relevant social groups are persuaded inrelation to the claimed benefits of the system and if they can be harvested by adoption of the system. This conclusion con-firms the findings of Barley (1986), that similar technologies adopted in different organizations are subject to different inter-pretations leading to different outcomes.

This section has presented our study design, motivated its components and outlined two main areas for further investi-gation: IT artefacts as potential carriers of inscription, and the interplay between inscription and interpretation process.Fig. 1 depicts our tentative understanding of the interplay. We return to the two areas in the analysis. In the next sectionwe present our research method.

Page 8: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

362 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

3. Methodology

The empirical part of this paper is the case of how the II supporting European trade evolved from a paper-based infra-structure to a digital II – the European e-Customs II. To research such an object as the European e-Customs II is an interestingmethodological challenge. The infrastructures have ties to two million users, both business and authorities, in more than 27countries. Previous research has shown that much of the essence of these large infrastructures lies in the scale and complex-ity (Hanseth and Ciborra, 2007; Hanseth et al., 2006; Kallinikos, 2007; Ciborra, 2002; Star and Ruhleder, 1996). The meth-odological challenge is to preserve the large scale perspective, but at the same time be able to draw conclusions fromdetails and specific findings. We approach the object of study with a structured case study approach (Carroll and Swatman,2000).

3.1. Research context, approach and empirical limitations

The essence of the structured case study approach (Carroll and Swatman, 2000) is an iterative research cycle upon a for-mal theoretical framework. The structured case approach is an interpretative approach that has its major strength in sense-making of complex and rich empirical settings, as multiple perspectives can be accumulated in a process towards theoreticalsaturation. In initial stages a broad theoretical frame can be used, whereas in later stages the object of study and the theo-retical concepts are narrowed down and precisely defined. Following the guidelines by Carroll and Swatman (2000), westarted the study presented in this paper wide, and attempted to narrow down in several study designs, including theone presented in Fig. 1 that addressed the general themes and processes encountered in earlier stages.

Study of the inscription and interpretation processes in II that we are interested in, required in-depth information fromdeveloping organizations on what visions and objectives the IT artefacts supposedly should carry, and then similar in-depthinformation from the II’s users to assess how intentions were matched by interpretations. These processes, we discovered,are not always explicit or the organizations are not even aware of their actions. The reasoning and motivation behind actionswere essential to understand why IT artefacts were designed the way they were and what drove the interpretations. Amongexisting alternatives, we considered the need for rich and contextually embedded information best being met by an inter-pretative approach based on qualitative data. Thus the choice of the structured case study approach, which is especiallydeveloped for this research approach and in addition has a primary strength in dealing with complexity in research settings(Carroll and Swatman, 2000) – a key characteristic of II (Hanseth and Lyytinen, 2010).

The reported study was part of a project on e-Customs implementation throughout Europe, involving multiple stakehold-ers from government, businesses, a standardization body (UN), and academia. The project, initiated in January 2006 and ter-minated in December 2010, was built around the concept of Living-Labs (LLs). LL is a concept that has been used in variousoccurrences to depict a research setting that emphasises the use of real world environments instead of isolated labs as thesetting for research (Klein and Higgins, 2010). It is frequently used in a common sense meaning, but has also been used morespecifically as a research approach in various research fields (e.g. McNeese, 2004).

LLs build on the tradition of action research and have been put forward as platforms for user-driven innovation and co-creation of knowledge in IT (Klein and Higgins, 2010). In this specific project, LL was defined as a test setting with definedborders where the different actors could approach specific problems with suggested solutions that ended in proof-of-conceptimplementations. User involvement and co-creation of knowledge meant that stakeholders with interest in the developmentof European e-Customs, both developing and using organizations, were brought together to try to solve specific e-Customsrelated problems. The roles of researchers, including the authors of this paper, were to moderate the processes, fuel the workwith knowledge from academia, and to ensure cumulative knowledge creation in the problem area through critical studiesand the writing of papers like this one. The active user involvement directed the selection of research problem and reflectsthe inability of existing theory in the area to provide practitioners with sufficient guidance to manage II development.

The present study reports the specific experiences gained in the project’s third LL situated in Denmark. The Danish LL,named ‘Food LL,’ focused on export of dairy products and involved the dairy producer Arla, Danish Customs, the Danish soft-ware provider Resultmaker, as well as researchers from national and international institutions. Denmark is a suitable basefor studying e-Customs since Danish authorities are relatively advanced in their IT use (Bjørn-Andersen et al., 2007) and arefrequently showcased as a role model for other countries. Studying customs and trade in Denmark is also particularly suit-able due to Denmark’s position among the world leaders in logistics and economy based on foreign trade.

The II of interest in this study is the European e-Customs II. Empirically, to limit our approach in a way that enabled fur-ther understanding of the theoretical concepts in our study design, the study in this paper took its starting point in Arla’sexport of dairy products from Denmark to Russia. We then limited empirical investigations to the parts of the II that affectedArla’s export processes and the actors that were identified as major influencers of these parts of the II (see below for a com-plete listing).

3.2. Data collection and analysis

In a first research cycle, general workshops and project meetings in the Food LL of the e-Customs project were used. In thefirst data collection phase, 10 meetings were attended, recorded and analysed for information on the II and its evolution. The

Page 9: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 363

project workshops and meetings brought together traders, governmental agencies, standardization bodies and IT providersthat discussed how the current II could be reshaped to improve and facilitate trade. The combined view of these stakeholdersand the explicit objective to discuss how to work with and build on the existing infrastructure was ideal for creating a broadview of the current infrastructure and the major stakeholders in its transformation. In a 2-day workshop, 29 participantsfrom the stakeholder organizations collectively synthesised a report on barriers on drivers for future e-Customs initiativesin Europe. Another outcome was a technical description (data models, security measures, connection form, etc.) of the exportsystems of all 27 EU member countries. These documents, recordings, transcriptions, notes, and reflections were structuredin a research folder. Data was coded in general themes, representing for example major issues, involved stakeholders, pro-cesses, events and objectives. The extensive and broad, rather than narrow and focused, empirical material made the struc-tured case-study approach suitable (Carroll and Swatman, 2000).

The first research phase was broad and focused on creating a relatively neutral story with events, processes, actors andoutcomes. We used the concept of II to focus data collection, but were not explicit in which theoretical approach to use to tryto explain the transition. Reflecting on the findings we searched for theories that could help us explain how the infrastruc-ture had developed the way it had. In the empirical material the processes of inscription of behaviour and interpretation of ITartefacts emerged as highly relevant. Already at an early stage, national and supra-national authorities expressed frustrationover difficulties to achieve harmonisation, and on the other side, interpretative flexibility seemed to be promising in thatregard since it put emphasis on the interplay between different components in a complex socio-material setting, where pathdependency, emergence and unintended consequences seemed to be the characteristics of the case at hand. Thus, based onthe initial case story and the theoretical concepts of behaviour inscription and interpretative flexibility, 12 semi-structuredinterviews were conducted in a second data collection phase with the case company Arla (Business relationship manager, ITmanager, Responsible for Danish export), Danish Customs (Head of Customs unit, Manager e-Customs project, Developer e-Export project), the EC’s unit for e-Customs (Head of unit, Head of infrastructure, Developer), WCO (Head of IT unit), the ITprovider of Arla (IBM: Developer of Arla’s IT infrastructure), and Arla’s logistics provider (Maersk: Supply chain systems de-signer). These interviews specifically addressed the study design depicted in Fig. 1 with the ambition to support theoreticalconstructs with empirical data. The interviews were recorded and key sections transcribed. Our interpretations, findings anduse of data were continuously shared with the contributing organisation to ensure correct interpretations and avoid disclo-sure of potentially sensitive information. It should be recognised that the development of the European e-Customs II is apolitical and potentially sensitive process. Our intentions were to enforce the process, not to harm it with disclosure of sen-sitive information.

In line with the structured case method approach several theoretical perspectives were used to code and analyse theempirical data; one of them the study design of this paper. Although one single study design is presented in this paper,in accordance with the structured case approach, variations and differences in included sub-concepts were part of the iter-ation. Case analysis implied investigation of congruence between the e-Customs II case and the study design’s sub-conceptand interactions.

4. From paper-based to electronic customs processes in Europe

As explained in the following section, the European e-Customs infrastructure is currently going through radical changesin the transition from paper-based processes to e-Customs. The transformation is an attempt to improve security in tradethrough better control of traded goods due to more comprehensive and reliable trade data in combination with the potentialto automatically process data in the search for abnormalities and risks. At the same time e-Customs is assumed to lower theadministrative burden for traders that currently are spending a lot of time and resources to comply with customs require-ments (TAXAUD, 2007). This section describes the European e-Customs II and developments related to the processes ofinscription and interpretation.

4.1. The paper-based customs infrastructure

In the 1980s, a company that was an active exporter or importer on the European market had to deal with more than 200different forms to cover the various customs procedures in different countries and for different modes of transport. In re-sponse to the call for a reduction of the administrative burden for economic operators in the increasingly globalised businessarena, the single administrative document (SAD) and related legislation was introduced in the EU.

The original SAD was introduced 1 January 1988. Initially 18 countries used the document as a standard for their customsdeclaration. In 1993, the SAD became obsolete for trade within the EU as a result of the creation of a common internal mar-ket, but by then the standard had gained acceptance outside the EU and was used for extra-EU trade. In 2007, the format wasadopted by more than 30 countries, including, for example, Russia.

In terms of inscription, the SAD document, which originally was a set of eight copies of the same document printed oncarbon paper, regulated which data should be reported, to which authority and when during the journey of the goods.The first copy was for the country where export formalities were carried out, the second was used for statistics in the exportcountry, the third copy was returned to the exporter, and so on. Within the same operation a company could accomplish

Page 10: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

364 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

export clearance, issue the transport document, and prepare the clearance in the country of destination. The SAD inscribedan efficient, paper-based, process for declaring export goods.

4.2. The European e-Customs initiative

The first tentative steps towards e-Customs in Europe were taken in the 1980s when pioneering customs organizations,inspired by the industries’ efficiency improvements with EDI, decided to enable electronic submission of export and importdata. Initially all e-Customs initiatives were undertaken nationally without respect to other e-Customs initiatives in Europe.

e-Customs became a European concern at the end of the 1990s. The EU-driven e-Customs 2002, 2007, and 2013 projectswere launched to implement the European transition to e-Customs in a harmonised process. The European e-Customs ini-tiative is outlined in the Multi-Annual Strategic Plan (MASP) (TAXUD, 2008), which describes visions and objectives for cus-toms in line with the desire for improved security and decreased administrative burden through harmonised e-Customs.‘Harmonisation’ is a key word here that frequently recurs in the EU-driven discourse. As the EU has a common inner marketwithout internal customs control, any deviation in customs processes within the EU affects also the rest of the union’s mem-ber states. Secondly, for pan-European businesses operating in several European countries, harmonisation would also lead toimprovements. Having to deal with customs organizations through different processes and through different IT solutionsdepending on where the production unit was placed, hampered scale and consolidation benefits in pan-European companies.

The key point of the strategic ambitions of the EC is that the foreseen improvements in customs controls with regard tosecurity and efficiency cannot be achieved if an effective e-Customs II is not established. Hence the EC’s strong interest ininfrastructure development. The resources set aside for the Customs 2013 project manifest the strength of the interest. Dur-ing a 5-year period the budget amounts to € 300 million. Note that this is only for the central EU initiatives and does notinclude resources spent on national level to incorporate the EU initiatives.

4.3. Inscribing behaviour into the European e-Customs Information Infrastructure

Introduction of IT artefacts into the European e-Customs II triggers a range of dynamic reactions with consequence forrelated IT artefacts and their social context. The European e-Customs II is deeply embedded into multiple layers of legislationthat affect how IT artefacts can be shaped and assimilated into the II. Customs laws that affect European traders exist asinternational treaties, EU law, and national law in a hierarchical structure where the latter are supposed to incorporatethe former. International organizations such as the WCO and the UN work through Conventions, which are voluntarily ac-cepted by the member countries.

At a regional level, EU acts through the EC using Communications, Decisions, and Regulations that conveys to what stan-dards e-Customs in Europe should conform. Relevant EU law regulates through the ‘‘Modernised Customs Code’’ in generalterms how EU member countries’ legislation should enable e-Customs. National legislation of relevance is law directly re-lated to export and import of goods, but also law related to accounting, tax, agriculture, health, pesticides, and many otherareas.

When actors such as UN, WCO and EU introduce standards and other specifications, the artefacts have to be accommo-dated on three different levels: EU, national, and trader. In principle, these are interrelated hierarchically with the develop-ments at EU level as decisive for national and trader developments. Something decided at EU level sets off an adaptationprocess of national projects, which in turn trigger changes at trader level to comply with new requirements. It should,however, be noted that the inscription process is a two-way process – any decision at EU level has to be an agreement be-tween all member states, and what the member states argue for in the negotiation is largely dependent on the characteristicsof the business in their country and what would improve conditions for these companies. Fig. 2 schematically outlines the

EU DomainStrategy/Programme Projects

NCTSMulti-Annual Strategic Plan

Customs 2002

Customs 2007

Customs 2013

ECS

ICSAEO

EORI...

Danish DomainProjects/Subprojects

e-Customs Project

NCTSECS

ICSAEO

EORI...

Arla Domain

Adaption projects

ECSICS

AEO: Authorized Economic OperatorEORI: European Operators Registration and identification

system

NCTS: New Computerized Transit SystemECS: Export Control SystemICS: Import Control System

Fig. 2. Organisation of e-Customs projects at EU, national, and trader level.

Page 11: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 365

organisation of e-Customs related projects at the three levels, given the example of Denmark as national domain and Arla astrader domain.

Standards are common tools in the attempts to inscribe harmonised behaviour into the European e-Customs II. However,other artefacts have been used in this role. The tight connection between legislation and e-Customs was manifested in thework with the New Computerised Transit System (NCTS), which was first launched in 2001 and then re-launched three timesduring the first decade of its existence. Transit goods are goods that only transit a country on their way to a final destination.How the goods are defined and should be treated by national customs organizations is defined by the 200-page long TIR(Transports Internationaux Routiers) Convention stemming back to 1975 as UN-driven prolongation of the Kyoto Conventionfrom 1920. All goods that are transited through one or more countries on their way to the final destination are since 1975followed by a TIR Carnet, a physical paper document, that allows transiting without payment of local tax or being affected bylocal legislation. Since changing the TIR Convention is a monumental task, when implementing the NCTS, the new systemhad to comply with the process inscribed into the existing TIR.

With the NCTS the EU member states jointly developed a common system to ensure that the transit process was carriedout similarly in each country. The NCTS was comparable to a standard package of business software that is locally custom-ised and implemented. In the end, the NCTS was a pale compromise that did not really fit the needs of any customs orga-nisation. The national customs authorities had to spend more resources than expected customising and refining thestandard package to fit into the national context. First of all, it had to fit into the national legislative context. Althoughthe TIR convention specified transit goods regulation, customs systems are interrelated to many other legislative bodies.The second aspect that influenced the customisation need of the NCTS were the, in many countries existing, legacy systemsthat were integrated with the national systems. These ‘‘other systems’’ included for example systems that compile statisticsbased on the export declarations, deal with agricultural control, VAT control, and control of hazardous goods. Rather thanadapting all these systems, Customs organizations generally customized their NCTS implementations to work with interfac-ing systems, interpreting the artefact differently. It was however, not a straightforward task to do so. Only when implement-ing the new NCTS did it became apparent exactly which other systems and organizations were affected by the changes. Thisresulted in several adaptation iterations in both NCTS did and surrounding systems.

‘‘Especially the countries with advanced e-Government solutions in place had an extensive job in understanding how thenew e-Customs component would have to interface to other systems, and how other systems would be affected by thenew component.’’ (DG/TAXUD).

Another example of an attempt to inscribe behaviour in the European e-Customs II is the work with risk management.With the formation of an inner European market the protection of the border between the inside and outside of the EU be-came more important. Once inside the EU any malicious goods could be distributed to any location in the EU. As a response,risk management became the core focus of the first joint customs initiative in the EU. Rapidly, guidelines for how to handlerisk were expressed in a Risk Framework (RF). However, as this framework could be differently interpreted at national oreven customs office level, it was not considered enough to ensure harmonised risk levels. The electronic Risk InformationForm (RIF) system was launched in April 2005. The system was implemented for disseminating and exchanging risk infor-mation to all customs control points of the EU. The purpose of the RIF is to exchange risk information dealing with routinecontrol concerns. A RIF should raise the awareness of the offices concerned with regard to a potential irregularity. A RIF canbe prepared following a finding of an irregularity (for example a faulty declaration or finding of counterfeit or undeclaredgoods). The RIF could give information on the technique used to find the irregularity, for example the result of a physicalexamination or a classification decision. The RIF system is implemented centrally in the EU and customs officers interactwith the system using a web interface.

The next step to improve risk management in the EU was the construction of a Customs Risk Management System(CRMS). Although the official documents from the EU speak about the CRMS as one system, it is actually 27 individual sys-tems developed by the member states. These systems contain two parts, one common and one individual. The common partis functionally specified by the EC. The description includes which data elements should be submitted to the customs for riskmanagement purposes in relation to an import or export declaration. The common instructions do, however, not say any-thing on how the risk data elements should be analysed and what the premises are for something being categorised as ahigh-risk item. The individual part is the national preference of what is needed for risk management in each respective coun-try. This can be because a country has had problems with specific types of goods or specific exporters.

Several other attempts to inscribe behaviour in the European e-Customs II have been made, including the development ofcommon databases (EORI) and system functionality specifications (ECS). However, it is beyond the scope of this case-presen-tation to include all initiatives. The next section describes the effects of the attempts to inscribe behaviour into the Europeane-Customs II for exporting businesses and what these companies face when they have to handle e-Customs in their dailypractices.

4.4. Consequences of inscription and interpretation processes

Arla is one of Europe’s largest dairy companies with export from several European countries. For several years, Arla hassubmitted data needed for customs procedures, when exporting from Denmark, electronically through the Danish ‘‘e-Exportsystem’’, which has been massively diffused among Danish exporters:

Page 12: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

366 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

‘‘The previous e-Export system was not well received by the traders so we decided to redesign our system based on theexporters’ demands. Our goal was that 95% of all transactions should be electronic. Today we have reached a level of 98–99%’’ (Danish Customs).

The data model in the e-Export system is a field-by-field translation of the SAD-fields that the Danish customs used priorto computerisation, as what data to include in export declarations are described in national legislation. A different data mod-el, with only data of relevance today, would have to be accommodated on the national legislative level, and to do so it wouldalso require changes at EU level.

Arla has interfaced its internal enterprise system (SAP) to Danish Customs via an EDI link. Export declarations are pre-pared in the SAP environment as one integrated part of a sales process:

‘‘We normally don’t have to add any data. What is needed for the declaration we already have in our system and can feedautomatically’’ (Arla).

In the process of digitizing SAD it was noted that the meaning of submitted data could be an issue, since it was not auto-matically inscribed into SAD. Much effort was devoted to agreeing on what was meant by fields such as ‘‘weight’’ (with orwithout packaging?) and ‘‘exporter’’ (the selling company or the truck driver? The company or the individual?). The e-Exportsystem benefits from being a direct derivation from the SAD. A substantial amount of work has already been undertaken byUN/CEFACT, ISO, and the EC/TAXAUD to specify the meaning of data. Also, for such fairly interpretable fields such as ‘‘productdescription,’’ there are appropriate guidelines on how to provide and interpret data. Adding new data would require a newharmonisation process on data meaning in the EU.

Another heritage from the SAD is the division between mandatory and optional data fields. In the SAD, this was not aproblem. The form looked exactly the same wherever used. Only some fields were not used. In the electronic versions ofthe SAD, the data models for export declaration, it is a problem. The national data models are omitting data not used, pre-senting Arla with different models depending on the country of export. In reality, this has led to diverging interpretation anddifferent XML schemas being used for data transfer related to export declarations. In addition to the differences in the datamodel, e-Customs also differ in the legislative aspects that define the use of IT artefacts.

For each country where from Arla exports, it has to develop a specific e-Customs interface for the national e-Customssolution. Arla investigated the cost of developing such a module for Sweden and had an offer of about € 100,000.

‘‘Based on that offer, which we by the way think is a too low estimation – the real cost will be much higher. [. . .] In addi-tion, we will have extra maintenance and upgrading costs in the future’’ (Arla).

The effects of the European e-Customs II are this far, that almost every country that earlier used the very same standard-ised SAD for customs clearance, now have developed individual e-Customs solutions that enforce companies to set up indi-vidual export processes for each country where the company is exporting from.

‘‘If we would like to do e-Customs in Poland that would also require investments above 100,000 Euro. If we would likehave e-Customs in France that means yet another investment of the same size, and so on. . . I don’t think there are any twocountries where we could use the same interface’’ (Arla).

The contemporary history of e-Customs in Europe is a story of how the introduction of IT has constrained a well function-ing, paper-based II. The non-coordinated European initiatives are leading to higher costs for the pan-European exporterswhich might even contest the transition, due to costs which are not directly serving as a means for improving productionprocess, but rather are contributing only to the improvements of the infrastructural technologies.

5. The duality of inscription and interpretation

Based on the situation expressed by Arla, it is clear that not all attempts to inscribe behaviour into the European e-Cus-toms II have been completely successful. The previous section described the European e-Customs II and attempts to inscribebehaviour into the IT artefacts of the II. This section analyses the inscription attempts and explains the consequences of theattempts. We structure the analysis according to the two main areas of interest identified when developing the study designin Section 2. We then integrate our findings into a model of inscription and interpretation in II, and discuss how the modelexplains the traders’ expressed longing for the good old paper days.

5.1. IT artefacts as inscription carriers

Over the years a number of technical artefacts have been introduced in attempts to increase coordination in the Europeane-Customs II. Standards and specifications, developed through the UN, WCO and EU have played key roles. Examples includethe WCO data model, UN eDocs, and the various Regulations by the EC. However, attempts have also been made with stan-dard software (NCTS), software modules (RIF), and databases (EORI). These artefacts show differences in the strength ofinscription, which partly can be attributed to the temporal position i.e. in the design or use mode (Orlikowski, 1992). Theartefacts used in design mode are in our case more abstract template artefacts with limited patterns of inscription relativeto the implemented artefacts already in use.

Page 13: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 367

The difference between template and implemented artefacts shed some light on why Arla perceives the digital II as lessstandardised than its paper-based precursor, based on the SAD. The SAD is an implemented artefact with a strong inscription.A digital SAD template equivalent would be a common instruction on how to construct a paper-based form. In the develop-ment of national electronic export systems, the SAD as was used as specification for the systems – the implemented artefactwas used as a template artefact that was interpreted differently in EU’s member states. The overall e-Customs agenda is toalign export procedures in the 27 EU member-states and to establish a robust e-Customs II. However, the programs of actionin the SAD template are weak resulting in an anti-program with respect to the implementation of a universal e-Customs II.Returning to Latours’ (1991) key example where strong inscription in the form of a large cumbersome weight to the key,meant that the key was left in the reception. Our case demonstrates that it is too easy to deviate from the template artefact.One reason can be that the common instructions are too vague and not intuitively understood even by the organisation thatwould like to follow the inscribed program of action. Another, and perhaps more likely explanation, could be that IT stan-dards are difficult to impose, in particular at an international level – it is not possible to agree on a way to attach the weightto the artefact.

In the implementation process, the interpretation of the SAD differed leading to different export systems in Denmark (ECSDK), Sweden (ECS SE) and any EU country (ECS X). These alternations are depicted in Fig. 3, which also describes the alter-nations of the risk management artefacts. Similar to the SAD, a general risk management framework (RMF) was translated toan implemented artefact (RIF) that was common for the whole e-Customs II. The RIF artefact still in use, was extended with atemplate artefact for further risk management functionality – the template artefact CRMS. The CRMS in turn was partlyimplemented as a common artefact, and partly in national implementations (Denmark: CRMS DK, Sweden: CRMS SE, etc.)

Template and implemented artefacts are rough categories in Fig. 3. Abstraction is more correctly seen as a continuum inour case. Template artefacts include both functional (‘‘do risk analysis’’) and technical specifications (‘‘an export declarationshould contain the following data fields. . .’’), where the latter leaves less room for interpretation. The level of specificationdetail also stipulates degrees of freedom in interpretation. Further, the NCTS was something like a semi-template artefact inthat it was a standard system that allowed for customisation.

When addressing inscription strength as a matter of possible programs of actions associated with the IT artefact, it shouldbe recognised that abstraction level is one aspect of the inscription strength. As noted by Kallinikos (2004a), a word proces-sor can be said to have weaker inscription than an ERP system, although both are implemented artefacts. However, in ourcase the implemented artefacts, such as the RIF, doubtlessly leave less room for interpretation than does its specificationcounterpart, the CRMS specification. When the CRMS was introduced, divergence in terms of objectives, functionality, pro-cesses, and data increased compared to the RIF.

Combining our observations regarding inscription strength and interpretative flexibility reveals a dilemma related to theparadox of control (Tilson et al., 2010) for the one who wishes to influence the European e-Customs II. On the one hand an ITartefact with strong behavioural inscription is required to enforce harmonisation and interoperability in the II. Such an arte-fact would require an enormous amount of work in each country to accommodate the artefact in the existing technical andsocial base. On the other hand, an artefact with weaker behavioural inscription may be too edulcorate to render the benefitsthat come with harmonisation. This creates a tension between the processes of inscription and interpretation that leads totrade offs and pale compromises.

The paradoxical dilemma is well illustrated in the work with the export declaration systems. The approach with a jointlydeveloped standard system, such as the NCTS, was consciously rejected because of the undoable of incorporating such anartefact in the nationally installed social and technical base. The technical integration with other government systems,and the incorporation into national legislation and customs practice would simply be too much for the EU member states.Therefore a functional specification, specified in Regulation 648/2005, was used. The choice and expressed motivation re-veals that it was a conscious choice to be vaguer in the inscription, allowing flexible interpretation. As a consequence het-erogeneity in the II increased, leading to increased costs for the traders that had to interact with Customs organizations indifferent ways.

5.2. Duality of inscription and interpretation processes

Drawing on previous literature, our study design as depicted in Fig. 1 assumed that new artefacts go through a stabiliza-tion process that ends in a common understanding. Closure and stabilization of a technical artefact represents the stage

Fig. 3. Alternations between template and implemented artefacts for export systems and risk management.

Page 14: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

368 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

where there is a common understanding among members of relevant social groups (Pinch and Bijker, 1984). In the e-Cus-toms II case the relevant social groups consist of multiple players. Broadly speaking the relevant social groups can be dividedinto two: the regulatory bodies which consist of both standardization bodies such as the EU, UN, WTO and national govern-ments of the 27 European EU member states. The second group consists of the organizations in the supply chains of theexporting businesses (producers, intermediaries, and transportation agencies) as well as the national customs offices. Withinthese two groups there are vested interests that go beyond the individual IT artefact to an effective e-Customs II based onshared interpretations of constituting components.

Closure and stabilization seems to be hard to achieve in the European e-Customs II. At the level of the users the Arlaexample of how they have to implement different systems in each EU country demonstrates that no common understandinghas yet been reached. Nor have they reached the characteristic of the Pinch and Bijker’s third stage where there is a consen-sus of the components or standards of the e-Customs II. In retrospective the SAD qualified as an artefact that had reachedconsensus and closure among the relevant social group. The SAD had clear defined fields that were understood by exporters,transporters, and customs authorities within and beyond Europe. However, the introduction of e-Customs II has disruptedthat stable stage.

In terms of Orlikowski’s (1992) analytical distinction between design and use mode, the European e-Customs II has forseveral decades been in a continuous design mode. We see two primary reasons for this. First, one of the obstacles whichgoes beyond the very complex issue of a common agreed standard is the continuous shift in focus which makes the e-Cus-toms II stay in the design mode rather than move onto the use mode (c.f. Fig. 1). As presented in the case, the first steps in thee-Customs development were related to creating the possibility of submitting export data electronically. Over time, the tech-nological development raised the expectations towards of e-Customs solutions, and external events gave rise to new require-ments, e.g. risk due to security threats leading to the CRMS. These ongoing developments combined with the constant needto carry out export activities regardless of the introduction of new IT artefacts create a situation where it is challenging toalign the II with both the social and technical installed base, because it is a moving target. That is in particular clearly illus-trated in the Arla part of our case-presentation. Instead of making e-Customs the standard mode of operation Arla only ap-plied it if it was requested to do so by business partners. Arla developed e-Customs as a consequence on a modular basiswhere they add one more country when necessary. Visions and objectives (c.f. Fig. 1) are to accommodate needs of businesspartners not to work towards an installed social and technical base.

The second reason has to do with the dynamic character of II. In the European e-Customs II there are several levels ofimplementation: international, regional (the EU), national, and corporate. What is designed at one level is used in a designprocess at the next level. To further complicate matters this is not a one-way, top-down assimilation process. That is, designat regional and even international level is using artefacts designed at national and corporate level in the design process.When an specification introduced by the EU is used by the national customs organizations, it is used in a design processto design another IT artefact, that then in turn influence how the EU design the next IT artefact, and so on. Thus the inter-pretation processes is not evolving towards stabilization, rather the European e-Customs II displays a continuous interplaybetween design and use.

Returning to Fig. 1, our study reinforces the idea that flexible interpretation is dependent on the utilisation in a design oruse mode, but suggests that this, in the development of II, should be seen as an ongoing dynamic interaction rather than alinear evolution and growth towards stabilisation. We thus concur in the recommendation of Cadili and Whitley (2005)when they suggest that interpretative flexibility should be understood as originally proposed by Pinch and Bijker (1984)where interpretative flexibility does not include any consideration of the material features of the technology in the initialphase of the social construction of the technology.

During the time we followed the European e-Customs II, only marginal steps were taken towards what Pinch and Bijker(1984) refer to as common understanding of how to interpret artefacts – the third phase. It might be that our timeframe wastoo narrow, but some of the harmonisation efforts had been ongoing for more than a decade when we finalised data collec-tion. Ten years is a long time with regard to innovations in IT. During this time we have seen hardware innovations such asdigital seals, and GPS enabled tracking devices (Tan et al., 2006) as well as software innovation such as mobile services andservice-oriented architecture. It is a legitimate question to ask if such an object as the European e-Customs II ever will reacha stable phase. What we find in our case, which of course should be regarded with consideration to the limitations of ourstudy, is that stability and complete harmonisation is an infrastructural Utopia that might serve a symbolic value in definingdevelopment direction, but never will become the reality. What instead may be the reality is a somewhat harmonised II withpartially aligned IT artefacts and processes that operate together supported by interoperability tools.

5.3. A model of inscription and interpretation in Information Infrastructures

The empirically triggered adjustments of our understanding of the duality of inscription and interpretation in II are pre-sented in Fig. 4. Following the analysis of IT artefacts as inscription carriers we introduce the tension between inscriptionstrength and accepted adjustment as attributes of inscription and interpretation processes. Further, based on the analysisof the duality of inscription and interpretation processes we depict the model’s second phase as a recursive process, andthe model as a whole as open and without a final stable phase.

The model explains through the lack of stabilization phase why Arla expresses a longing for the ‘‘good old paper-days’’. Inshort, the introduction of e-Customs has led to a perceived de-standardization of European export processes, ultimately

Page 15: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

Phase 2

Developing organization

Using organization

IT artefact Template - Implementation

Interpretation- Accepted adjustement

Inscription- Inscription strength

Visions and objectives

Installed social and technical base

Use modeDesign mode

Phase 1

IT artefact with divergent

interpretations destabilize II

Phase 3

Co-existing interpretations,

bridge by interoperability

tools

IT innovation and contextual changes

Fig. 4. Duality of inscription and interpretation in Information Infrastructures.

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 369

leading to a breakdown of an infrastructure which had demonstrated its robustness during the paper-based days, but whichwas lost in the process of digitalization. In other words: digitalization and the introduction of IT artefacts have eroded aworking infrastructure that had strong inscriptions and clear defined programs of action to a situation with weak inscrip-tions and limited programs of action.

The European e-Customs II case highlights that even though extensive international standardization efforts have takenplace there is reason to believe that continuous interpretative flexibility rather than consensus, and stabilization will bethe norm of the European e-Customs II. The development of the IT artefacts we followed takes place in an asynchronous pacerather than parallel with the work on developing standards. The search for an inherent truth which leads to consensusaccording to Pinch and Bijker (1984) is disrupted by the ongoing development of more advanced IT artefacts which leadsto a continuous interpretative flexibility in the domain of e-Customs II.

The introduction of IT has led to a perceived de-standardization of the II supporting customs processes. In the transition toe-Customs regulators were faced with a dilemma: to achieve harmonisation of the European e-Customs II an artefact withstrong behavioural inscription was needed, but such an artefact was rejected by the adaptors since a completely inflexibleartefact would not fit with the installed social and technical base. An artefact with weaker inscription, permitting higher de-grees of interpretation, was developed. This artefact could be accommodated with the installed social and technical bases inthe EU, but did not render the benefits for companies like Arla that a full harmonisation would have led to.

6. Conclusions

This paper uses theories related to inscription of behaviour into IT artefacts and flexible interpretation to make two prin-cipal knowledge contributions to the domain of II. First, it contributes to the general understanding of how IIs evolve, by inte-grating theoretical and empirical findings using an interpretive frame of the duality of inscription of behaviour into ITartefacts and flexible interpretation in IIs. Second, it draws conclusions on how the introduction of IT innovations may con-tribute to the eroding of a working II.

Our main conclusions are summarised in Fig. 4 showing the duality of inscription of behaviour and interpretative flexi-bility in II based on our theoretical and empirical investigations. We note that in the case of the European e-Customs II thereis limited success in establishing an effective II by inscribing behaviour into IT artefacts. At the individual member-state le-vel, some are successful in inscribing patterns of use among exporters and customs authorities whereas the cross-nationalinscription of behaviour similar to the paper-based SAD-practice is still absent at large.

The situation for Arla illustrates the problems in an exemplar manner. The national differences between EU membercountries involve different e-Customs practices for identical products simply because of differences in the IT base supportinga process. This is true also for areas that explicitly have been addressed with standard and harmonisation efforts. These ef-forts faced a dilemma that reveals the tension between processes of inscription and interpretation in IIs: on the one hand anIT artefact with strong behavioural inscription was required to enforce harmonisation and interoperability in the II. However,such an artefact would require an enormous amount of work in each country to accommodate the artefact in the existingtechnical and social base. On the other hand, an artefact with weaker behavioural inscription may be too edulcorate to ren-der the benefits that come with harmonisation.

The contributions we make to the domain of II can be used by organizations to better understand how they will be af-fected by future developments of the IIs they are part of, but also by organizations that have an interest in the developmentof an II to understand how they may and may not influence the development of the II. The duality-view gives a frame foraligning IT investments and information system developments to external processes influencing the organisation’s use ofIT. In the specific case of e-Customs, organizations with an interest in the e-Customs II include traders, governmental agen-cies and supra-national bodies. E-Customs was introduced in Asia already in the late 1980s. The TradeNet in Singapore waspresented as a successful case (King and Konsynski, 1990a), whereas the TradeLink in Hong Kong had a less successful out-come (King and Konsynski, 1990b). Our work can contribute to understanding why these and other e-Customs initiatives are

Page 16: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

370 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

more or less successful in their outsets. Considering that more than two million European traders are directly dependent onthis II and that trade-related administration costs are estimated to as much as 2.5% of the total GDP in some European coun-tries (Bjørn-Andersen et al., 2007), this is an significant implication in its own.

With growing importance of II that extends national and organizational borders, our contributions have impact in a con-text wider than e-Customs. Understanding the processes of inscription and interpretation in II should be particularly rele-vant to governmental and supra-national organizations that are showing a keen interest in IT’s role in strategies fornational and regional development. In the field of e-Government, public–private II has been put forward as ‘‘a catalyst forsocial, economic and political change’’ (Fountain, 2002, p. 45), but claims have also been made that in reality little transfor-mation of the public sector can be observed (Kraemer and King, 2005; West, 2004; Coursey and Norris, 2008). The presentstudy demonstrates that the obstacles to achieve major transformation are embedded in the IT innovations that constitutethe II. This represents a vicious circle because more flexible artefacts invite to larger variation in anti-programs and thus nostandardisation, which is prudent to implement e-Customs in a satisfactory manner.

We started this article by pointing out the increasing relevance of II for individuals, organizations and society at large.This article has reconfirmed the importance of regarding II as a specific type of IT artefact characteristics by scale, complex-ity, heterogeneity of components and dynamics between components (Hanseth and Lyytinen, 2010; Gal et al., 2008; Monte-iro and Hanseth, 1996; Star and Ruhleder, 1996; Turner et al., 2006). We further the understanding of II developmentexplaining how these characteristics through the processes of inscription in and interpretation of IT artefacts leads to con-stant re-alignment needs, reinforced by IT innovations and changes in the social practice into which the constantly changingII is embedded. Concluding that II have specific characteristics and behave differently in development projects reemphasizethe need for research and practice to start developing methodologies to approach the specific challenges of the large, com-plex, heterogeneous, and dynamic IIs that steadily gain more prominent positions in our lives.

Acknowledgments

An earlier version of this paper (Henningsson and Henriksen, 2009) was presented at the ECIS 2009 conference in Verona,Italy. Comments from reviewers and delegates were appreciatively used in revising the paper before submission to the Jour-nal of Strategic Information Systems. We also thank the five anonymous reviewers and the senior editor for their suggestionsand constructive criticism. Based on the feedback given, this paper is repositioned, applies a different theoretical perspective,draws on an extended empirical foundation, and consequently makes distinct and better grounded contributions.

This research is part of the integrated project ITAIDE (Nr.027829), funded by the 6th Framework IST Programme of theEuropean Commission (see www.itaide.org). The ideas and opinions expressed by the authors do not necessarily reflectthe views/insights/interests of all ITAIDE partners.

References

Akrich, M., Latour, B., 1992. A summary of a convenient vocabulary for the semiotics of human and nonhuman assemblies. In: Bijker, W., Law, J. (Eds.),Shaping Technology/Building Society. MIT Press, Cambridge, MA.

Barley, S.R., 1986. Technology as an occasion for structuring: evidence from observations of CT scanners and the social order of radiology departments.Administrative Science Quarterly 31, 78–108.

Bjørn-Andersen, N., Razmerita, L., Henriksen, H.Z., 2007. The streamlining of cross-border taxation using IT: the Danish eExport solution. In: J. Macolm, G.O.(Ed.), E-Taxation: State & Perspectives - E-Government in the field of Taxation: Scientific Basis, Implementation Strategies, Good Practice Examples.Trauner Verlag, Linz.

Boland, R.J., Greenberg, R.H., 1992. Method and metaphor in organizational analysis. Accounting, Management and Information Technologies 2, 117–141.Broadbent, M., Weill, P., 1997. Management by maxim: how business and IT managers can create IT infrastructures. Sloan Management Review 38, 77–92.Broadbent, M., Weill, P., ST. Clair, D., 1999. The implications of information technology infrastructure for business process redesign. MIS Quarterly 23, 159–

182.Bygstad, B., 2008. Information Infrastructure as Organization: A critical realist perspective. ICIS 2008. Paris, France.Byrd, T.A., Turner, D., 2000. Measuring the flexibility of information technology infrastructure: exploratory analysis of a construct. Journal of Management

Information Systems 17, 167–208.Cadili, S., Whitley, E.A., 2005. On the interpretative flexibility of hosted ERP systems. The Journal of Strategic Information Systems 14, 167–195.Carroll, J.M., Swatman, P.A., 2000. Structured-case: a methodological framework for building theory in information systems research. European Journal of

Information Systems 9, 235–242.Ciborra, C., 2000. From Control to Drift. Oxford University Press, Oxford UK.Ciborra, C., 2002. The Labyrinths of Information: Challenging the Wisdom of Systems. Oxford University Press, Oxford, UK.Coursey, D., Norris, D.F., 2008. Models of e-government: are they correct? An empirical assessment. Public Administration Review 68, 523–536.Damsgaard, J., Truex, D., 2000. Binary trading relations and the limits of EDI standards: the procrustean bed of standards. European Journal of Information

Systems 9 (3), 173–188.Desanctis, G., Poole, M.S., 1994. Capturing the complexity in advanced technology use: adaptive structuration theory. Organization Science 5, 121–147.Doherty, N.F., Coombs, C.R., Loan-Clarke, J., 2006. A re-conceptualization of the interpretive flexibility of information technologies: redressing the balance

between the social and the technical. European Journal of Information Systems 15, 569–582.Edwards, P.M., 2003. Infrastructure and modernity: force, time, and social organization in the history of sociotechnical systems. In: Misa, T.J., Brey, P.,

Feenberg, A. (Eds.), Modernity and Technology. MIT Press, Cambridge, MA.Fountain, J.E., 2002. Information, Institutions and Governance. Cambridge, MA, John F. Kennedy School of Government, Harvard University.Fountain, J.E., Osorio-urzua, C.A., 2001. Public sector: early stage of deep transformation. In: Rivlin, A., Litan, R. (Eds.), The economic payoff of the internet

revolution. Brookings Institution Press, Washington, DC.Gal, U., 2008. Boundary matters: The dynamics of boundary objects, information infrastructures, and organizational identities. Doctor of Philosophy, Case

Western Reserve University.Gal, U., Lyytinen, K., Youngjin, Y., 2008. The dynamics of IT boundary objects, information infrastructures, and organisational identities: the introduction of

3D modelling technologies into the architecture, engineering, and construction industry. European Journal of Information Systems 17, 290–304.

Page 17: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372 371

Giddens, A., 1984. The Constitution of Society: Outline of the Theory of Structuration. University of California Press.Greenhalgh, T., Stramer, K., Bratan, T., Byrne, E., Mohammad, Y., Russell, J., 2008. Introduction of shared electronic records: multi-site case study using

diffusion of innovation theory. British Medical Journal 337, a1786 (Clinical Research Ed.).Hanseth, O., 2000. The Economics of Standards. In: Ciborra, C. (Ed.), From Control to Drift. The Dynamics of Corporate Information Infrastructures. Oxford

University Press, Oxford.Hanseth, O., Braa, K., 2001. Hunting for the treasure at the end of the rainbow: standardizing corporate IT infrastructure. Computer Supported Cooperative

Work 10, 261–292.Hanseth, O., Ciborra, C., 2007. Risk, Complexity and ICT. Edward Elgar Publishing.Hanseth, O., Jacucci, E., Grisot, M., Aanestad, M., 2006. Reflexive standardization: side effects and complexity in standard making. MIS Quarterly 30, 563–

581.Hanseth, O., Lyytinen, K., 2010. Design theory for adaptive complexity in information infrastructures. Journal of Information Technology 25, 1–19.Hanseth, O., Monteiro, E., 1997. Inscribing behaviour in information infrastructure standards. Accounting, Management and Information Technologies 7,

183–211.Henderson, J., Venkatraman, N., 1992. Strategic alignment: a model for organizational transformation through information technology. In: Kochan, T.A.,

Useem, M. (Eds.), Transforming Organizations. Oxford University Press, New York, N.Y.Henningsson, S., Budel, R., Gal, U., Bjorn-andersen, N., 2010. Itaide information infrastructure (I3) framework. In: Tan, Y.H., Bjørn-Andersen, N., Klein, S.,

Rukanova, B. (Eds.), Accelerating Global Supply Chains with IT-Innovation. Springer, Berlin.Henningsson, S., Henriksen, H.Z., 2009. A SAD Story. European Conference of Information Systems, Verona, Italy, June, 2009.Henriksen, H., Rukanova, B., Tan, Y.H., 2008. Pacta Sunt Servanda but where is the agreement? The complicated case of eCustoms. In: Wimmer, M.A., Scholl,

H.A., Ferro, E. (Eds.), Electronic Government: 7th International Conference, EGOV 2008. Springer-Verlag, Turin, Italy.INTERPOL, 2010. Operational Data Services and Databases for Police. <http://www.interpol.int/Public/ICPO/corefunctions/databases.asp> (accessed

07.05.10).Irani, Z., Elliman, T., Jackson, P., 2007. Electronic transformation of government in the UK: a research agenda. European Journal of Information Systems 16,

327–335.Janssen, M., Chun, S., Gil-garcia, J.R., 2009. Building the next generation of digital government infrastructures. Government Information Quarterly 26, 233–

237.Kallinikos, J., 2004a. Deconstructing information packages: Organizational and behavioural implications of ERP systems. Information Technology & People

17, 8–30.Kallinikos, J., 2004b. Farewell to constructivism: technology and context-embedded action. In: Avgerou, C., Ciborra, C., Land, F. (Eds.), The Social Study of

Information and Communication Technology: Innovation, Actors and Contexts. Oxford University Press, Oxford.Kallinikos, J., 2007. Technology, contingency and risk: the vagaries of large-scale information systems. In: Hanseth, O., Ciborra, C. (Eds.), Risk, Complexity

and ICT. Edward Elgar Publishing.Kayworth, T., Sambamurthy, V., 2000. Managing the information technology infrastructure. Baylor Business Review 18, 13–15.King, J.L., Konsynski, B., 1990a. Singapore TradeNet: A tale of one city. Harvard Business School, Case 9-191-009.King, J.L., Konsynski, B., 1990b. Hong Kong TradeLink: News from the second city. Harvard Business School, Case 9-191-26.Klein, S., Higgins, A., 2010. The concept of Living Labs. In: Tan, Y.H., Bjorn-andersen, N., Klein, S., Rukanova, B. (Eds.), Accelerating Global Supply Chains with

IT-innovation: ITAIDE Methods and Tools. Springer.Kraemer, K.L., King, J.L., 2005. Information technology and administrative reform: will e-government be different? International Journal of Electronic

Government Research 1, 27–49.Latour, B., 1991. Technology is society made durable. In: Law, J. (Ed.), A Sociology of Monsters? Essays on Power, Technology and Domination. Routledge,

London.Latour, B., 1992. Where are the missing masses? The sociology of a few mundane artifacts. In: Bijker, W., Law, J. (Eds.), Shaping Technology/Building Society.

MIT Press, Cambridge, Ma.Lenk, K., 2002. Electronic service delivery – A driver of public sector modernisation. Information Polity 7 (23), 87–96.Mcneese, M.D., 2004. How video informs cognitive systems engineering: making experience count. Cognition Technology & Work 6 (3), 186–196.Monteiro, E., Hanseth, O., 1996. Social shaping of information infrastructure: on being specific about the technology. In: Orlikowski, W., Walsham, G., Jones,

M.R., Degross, J. (Eds.), Information Technology and Changes in Organisational Work. Chapman & Hall.Northrop, L., Feiler, P., Gabriel, R.P., Goodenough, J., Linger, R., Longstaff, T., Kazman, R., Klein, M., Schmidt, D., Sullivan, K., Wallnau, K. 2006. Ultra-Large-

Scale Systems: The Software Challenge of the Future, Carnegie Mellon Software Engineering Institute, Ultra-Large-Scale Systems Study Report.OECD, 2005. E-Government for Better Government. OECD E-Government Studies Paris: OECD.Orlikowski, W.J., Iacono, S., 2000. Research commentary: desperately seeking the IT in IT research – a call to theorizing the IT artifact. Information Systems

Research 12 (2), 121–134.Orlikowski, W., Robey, D., 1991. Information technology and the structuring of organizations. Information Systems Research 2, 143–170.Orlikowski, W., Scott, S.V., 2008. Sociomateriality: challenging the separation of technology, work and organization. Academy of Management Annuals 2,

433–474.Orlikowski, W.J., 1992. The duality of technology: rethinking the concept of technology in organizations. Organization Science 3, 398–427.Pinch, T.J., Bijker, W.E., 1984. The social construction of facts and artefacts: or how the sociology of science and the sociology of technology might benefit

each other. Social Studies of Science 14, 399–441.Rose, J., Jones, M., 2004. The double dance of agency: a socio-theoretic account of how machines and humans interact. In: Goldkuhl, G., Lind, M., Cronholm,

S. (Eds.), Action in Language, Organisations and Information Systems. Linköping University, Linköping, Sweden.Rukanova, B., Van stijn, E., Henriksen, H., Baida, Z., Tan, Y.H., 2010. Understanding the influence of multiple levels of governments on the development of

inter-organizational systems. European Journal of Information Systems 18, 387–408.Sahay, S., Robey, D., 1996. Organizational context, social interpretation, and the implementation and consequences of geographic information systems.

Accounting, Management and Information Technologies 6 (4), 255–282.Sauer, C., Willcocks, L., 2007. Unreasonable expectations – NHS IT, Greek choruses and the games institutions play around mega-programmes. Journal of

Information Technology 22, 195–201.Star, S.L., Ruhleder, K., 1996. Steps toward an ecology of infrastructure: design and access for large information space. Information Systems Research 7, 111–

134.SWIFT, 2010. The Global Provider of Secure Financial Messaging Services. <http://www.swift.com/about_swift/company_information/index.page?>

(accessed 07.05.10).Tan, Y.H., Klein, S., Rukanova, B., Higgins, A., Baida, Z., 2006. e-Customs innovation and transformation: a research approach. In: Bled eCommerce

Conference, 2006, Slovenia.Tan, Y.-H., Bjørn-Andersen, N., et al, 2010. Accelerating Global Supply Chains with IT-Innovation. Springer, Berlin.TAXAUD, 2007. Electronic Customs Multi-Annual Strategic Plan – TAXUD/477/2004 – Rev. 8 – EN. Brussels, Belgium: European Commission/DG TAXUD.TAXUD, 2008. Electronic Customs Multi-Annual Strategic Plan – 2008 Yearly Revision. Brussel, European Commission/DG TAXUD.TAXAUD, 2010. 2009 customs detentions of goods suspected of infringing Intellectual Property Right (IPR) – MEMO/10/353 – EN. Brussels, Belgium,

European Commission/DG TAXUD.Tilson, D., Lyytinen, K., et al, 2010. Research commentary – digital infrastructures: the missing is research agenda. Information Systems Research 21 (4),

748–759.

Page 18: Inscription of behaviour and flexible interpretation in Information Infrastructures: The case of European e-Customs

372 S. Henningsson, H.Z. Henriksen / Journal of Strategic Information Systems 20 (2011) 355–372

Turner, W., Bowker, G., Gasser, L., Zackland, M., 2006. Information infrastructures for distributed collective practices. Computer Supported Co-operativeWork 15, 93–110.

Webster, J., Watson, R., 2002. Analyzing the past to prepare for the future: writing a literature review. MIS Quarterly 26, xiii–xxiii.Weill, P., Broadbent, M., 1998. Leveraging the New Infrastructure. Harvard Business School Press, Boston.West, D.M., 2004. E-government and the transformation of service delivery and citizen attitudes. Public Administration Review 24, 15–27.Wilson, M., Howcroft, D., 2005. Power, politics and persuasion in IS evaluation: a focus on ‘relevant social groups’. The Journal of Strategic Information

Systems 14, 17–43.Yoo, Y., 2010. Computing in everyday life: a call for research on experiential computing. Management Information Systems Quarterly 34 (2), 3.Zittrain, J., 2006. The generative internet. Harvard Law Review 119, 1974–2040.