Dismantling the silos: extracting new value from IT ... - CiteSeerX

Info Systems J (2001) 11, 65–84

65

Dismantling the silos: extracting new value from IT investments in public administration Frank Bannister Department of Statistics, Trinity College, Dublin 2, Ireland, email: [email protected]

Abstract. The drive for information technology-led organizational and operational change in public administration has lagged behind that in the private sector. For good reasons, central public administrations are conservative by nature. Most public administrations are bureaucracies and bureaucracies tend to resist change. Nevertheless the pressure to obtain better value from public administration information technology investments is growing and the debate as to how to achieve this is increasingly important. Part of this debate is concerned with how best to confront the formidable and specific challenges faced by the sector. These include cultural, structural, resource and technical problems as well as a legacy of isolated developments which do not interrelate. The difficulties are compounded by the problems of evaluation generally in public administration – problems which are reflected in the evaluation of public sector information systems. This drive for change is reviewed and discussed in the context of the Irish civil service, in which there is a growing awareness of the strategic importance of breaking down specialized vertical systems and providing an integrated service to the citizen. A new approach to the problem, based on adapting the concept of business objects, is suggested. Keywords: Bureaucracy, business objects, electronic data interchange (EDI), enterprise resource planning, IT value, object technology, process change, public administration, value for money

INTRODUCTION

–

BU R E AU C R AC Y A N D P U B L I C A D M I N I S T R AT I O N

Writing at the beginning of this century, the sociologist Max Weber set out the classic description of bureaucracy. A bureaucracy is characterized, inter alia by: • • • •

a dictation of labour based on functional specialization; a hierarchy of authority; a system of rules which limit discretion; impersonality;

© 2001 Blackwell Science Ltd

66

F Bannister

• a career structure based on technical competence; • written records of activities (Weber 1919). Bureaucracies exists in all walks of life, including many large commercial organizations. Much of the drive behind business process re-engineering during the 1990s has been aimed at changing this culture as well as breaking up bureaucratic procedures (Champy, 1995). But the real bastion of bureaucracy at the start of the new millennium remains public administration. It is in government – whether it is at the national, international or supranational level – that the characteristics of bureaucracy are most evident. This should not be surprising. Two key objectives of any civil service are to provide impartial execution of policy and continuity of administration (Murray, 1990). Bureaucratic structures are well suited to this purpose. Weber regarded bureaucracy as a highly effective organizational structure. Such a structure, i.e. one with hierarchical lines that executes policy impartially, also suits politicians as it enables power to be concentrated at the top. But the very structures that ensure continuity and stability are major inhibitors of change. The culture within a typical civil service is one of risk aversion (Joyce, 1985). Information technology, as a major catalyst of change, inevitably clashes with this culture. This clash can be seen in Ireland and elsewhere. In Ireland, information technology, while the subject of many pious pronouncements by successive governments, has consistently failed to receive sufficient attention (Public Service Advisory Council, 1983; Department of the Public Service, 1985; Pye, 1992). Public Administration, the leading journal of record of public administration in Ireland has published just one article on information technology in the civil service since 1952. An example of a similar lack of real interest and involvement by politicians in the UK can be found in Collingridge & Margetts (1994). There are, however, signs that this is changing and the recognition that change is necessary is percolating thinking in public administration (OECD, 1992; Muid, 1994). Civil administrations differ. These differences arise not only from their inherent structures, but also from the varying nature of political systems, history, culture and scale. Research has shown that there are considerable differences between civil administrations even within Europe, never mind further abroad (Frissen et al., 1992). Indeed it is an interesting question if there are greater difference between the IS systems of public administrations than there are between public and private organizsations as examined, for example, by Bretschneider (1990). Nonetheless, there are large areas of common ground, in particular what Bozeman (1987) describes as the dominance of political over economic authority. There is also a common thread in the struggle by public administration/management to integrate systems which are technically, structurally and geographically diverse. This combination of large, diverse, unintegrated and frequently ageing systems is at the heart of a major challenge facing public management today. It has been variously referred to as a ‘silo’ or ‘stovepipe’ problem, the legacy of decades of introspective development. The obstacles to be overcome are wide ranging and to address them new thinking and new technologies will be needed. Before we consider what form these might take and how public managers are trying to address these issues, it is necessary to discuss some value concepts. One

© 2001 Blackwell Science Ltd, Information Systems Journal 11, 65–84

IT investments in public administration

67

would intuitively expect the measurement of Information Systems value in public sector bodies, and particularly in civil services, to differ from those in private organizations. Traditionally, civil service interest in IT value has been focused on cost savings and productivity. This is now being augmented by a wider concept of quality of public service and the latter has, in turn, led to the current interest in the integration of specialist vertical systems.

VA LU E A N D VA LU E F O R M O N E Y

There is a vast literature on IT evaluation (Powell, 1992; Banker et al., 1993) and a somewhat smaller one on IT value per se (Cronk & Fitzgerald, 1997: M. Cronk and E. Fitzgerald, 1998, unpublished). When one approaches the problem from the perspective of evaluation in the public sector, it is clear that much of the current IT evaluation literature is not relevant to public administration. For a start, although the concept of value added and productivity is relatively easy to define in a manufacturing environment, doing so in a service environment is much more complex (Quinn & Baily, 1994). Furthermore, standard business measures of productivity, financial payback, return on capital employed, etc. have little and sometimes no meaning in public management. When it comes to evaluation, the primary concept used in public administration for well over a decade has been value for money (VFM). The enormous complexity of determining performance in an organization such as the civil service has been widely discussed, but there is a consensus, at least at a conceptual level, on the nature of VFM (National Audit Office (UK) (1996) Value for Money Handbook, unpublished internal handbook). In the past 20 years, three concepts, efficiency, effectiveness and economy have emerged as key criteria in public sector performance evaluation: • efficiency measures the ratio of inputs to outputs (for example the cost per hospital bed). • effectiveness is a more problematic concept. Boyle (1989), in a detailed survey of this issue, suggests that effectiveness occurs when objectives are achieved. • economy which the UK Audit Commission defines as ‘ensuring that the assets . . . and services purchased are procured and maintained at the lowest possible cost consistent with a specified quality and quantity.’ (Audit Commission, 1986). • The three Es, as they are often called, should, when combined optimally, deliver ‘value for money’. VFM has certain attractions, but it is holistic and concerned with the totality of the service provided rather than just one aspect of it. This can have the effect of blurring details. Specifically, information technology may be seen as part of an overall investment and never differentiated out. Value from IT per se is not seen as central. In public administration, IT expenditure, though large in absolute terms, is a relatively small part of overall outlay. In Ireland, civil service IT expenditure (including telecommunications) in 1997 was IR£85 million out of a total national budget of approximately IR£10 billion. The Value for Money division of the Comptroller and Auditor General’s office in Ireland aims to produce 12 research reports


68

F Bannister

Figure 1. Consumer criteria for service evaluation.

each year. Of these, one, at most, may be directly related to information technology expenditure. Since the foundation of the Value for Money division in 1994, only one study into IT in the civil service has been undertaken (Comptroller and Auditor General, 1997). There has been much debate about the VFM concept including the criticism that it is too introspective – i.e. it looks at services from the perspective of the provider (in this case the taxpayer) rather than the consumer/citizen. In more recent years, it has been argued that a better concept on which to focus would be quality of service to the citizen. The UK National Consumer Council (1986) has proposed a model (slightly adapted here) encompassing this idea as illustrated Figure 1. In this model, effectiveness is seen as an overarching concept with economy and efficiency being but two of several subsidiary issues. This model carries with it many implications about the way that public administrations should serve the public. From the above model, it is easy to see why the definition of value in public sector investments, particularly IT investments, can be difficult once one progresses beyond the ‘easy wins’ of self-financing projects. Four specific problems are: 1 Even where clear measurement of deliverables (for example cases processed) is possible, defining appropriate measures where the principal objectives of the organization are, for



69

Figure 2. Civil service IT drivers.

example, serving the public or formulating policy, is far more difficult. For example, how does one evaluate the value of information systems used to support policy making? 2 How does one define the ‘customer’ in a public service context? Often different terms (e.g. citizen, customer, consumer, service user and client) are used interchangeably whereas they imply different things (Gaster & Taylor, 1993). 3 Public organizations are constituency oriented, i.e. they focus on that segment of the population that use their services; be that farmers, health professionals, social welfare recipients, teachers or whoever. This can result in departments becoming too close to lobby groups, losing objectivity and perspective. 4 As a direct consequence of the preceding point, systems have evolved to serve highly vertical interest groups. In order to serve the citizen better and realize the pent-up value in current systems, traditional boundaries will have to be transgressed and barriers broken down. This requires a different vision of value and an understanding of how value can change with time.

E VO LU T I O N O F T H E S I LO S

Silos can evolve for a variety of reasons: technological, organizational, political and temporal to name but four. All of these can be found in any organization; however, civil service silo systems are the result of particular factors. Civil service IT systems have three primary drivers as illustrated in Figure 2. These drivers have broad parallels in any commercial organization where, for example, constituents might be considered the equivalent of customer and administrative the equivalent of internal processing needs such as the payroll or the asset register. However, the analogy is not exact, in part because of the nature of the relationship a department, and specifically its IS function has, with the political (as represented by the Minister and through the Minister the Government) and its constituents. A simple model of this is shown in Figure 3 in which the size of the arrows shows the degree of influence of one party on the other.


70

F Bannister

Figure 3. Degrees of influence/power.

Figure 4. Position of IS with a department.

For a civil service department, as the model suggests, the nature of these relationships is highly asymmetric. Furthermore, the relationship between Minister, constituents and department is not the same as that between Chief Executive Officer (CEO), customers and the rest of the company for several reasons. For example, customers cannot normally remove the CEO from office every 4 or 5 years. Few CEOs would have the degree of power and autonomy that a Minister has and a commercial company’s customers are not, generally, vying with each other for the largest slice of a finite resource. Public administration information systems are frequently driven by the current political need or most effective constituent lobby. Thus, an immediate political necessity – for example to get sheep headage payments out to farmers – can override the most thorough of long-term strategic IT plans. Further compounding the problems of the IS function can be the nature of its own position within a department. Figure 4 shows a development of the model which illustrates the difficulties in which many civil service IS departments find themselves.



71

Figure 5. Intra/inter departmental silos.

In such circumstances, an IS manager can find him/herself with very little ability to influence events. In some large departments, the most senior IT manager is only at Assistant Principal or Principal Officer level, equivalent to, say, the top layer of middle management in a commercial organization. As such, the IT manager’s ability to resist political pressure for shortterm solutions in the interests of longer term thinking is limited. Against this background, it is not surprising that so many systems have evolved in isolation. Add to this interdepartmental rivalries (at both political and administrative levels) and the unintegrated state of many public sector systems today is easier to understand. The result is, as already noted, an IS structure where separate systems are independently developed for specific purposes. This can occur within departments and, on an even greater scale, between departments (Figure 5). For example, until relatively recently, within the Irish Revenue’s systems, a citizen who owed overdue income tax, but was due a refund on VAT, could quite easily be refunded the VAT by the same government department that was taking him to court over non payment of income tax. Similarly, income tax returns could not be correlated with (say) claims for medical benefits on the basis of low income. In summary, information systems in the Irish civil service have developed as a series of isolated islands of processing without much commonality. The reasons for this are: • the nature of bureaucracy itself which tends towards specialization. This specialization is reflected in the information systems which serve it; • political control of these organizations which can lead to systems being developed or prioritized for short-term reasons; • internal politics where control of information systems is a key form of power and influence within departments; • pressure for specific solutions from lobby groups and for systems to meet short-term crises; • the relatively uninfluential position of IS managers in some departments. To this we can add:


72

F Bannister

• the fragmentation of authority which diffuses systemic decision making. In Ireland, for example, there are six state bodies and three government Ministers with responsibility for some aspect of the roads system; • the independence, particularly of larger departments of state, from central control and standards and the lack of lines of authority; • the size and breath of the range of services being offered. Today, a citizen undertaking the most routine of transactions in Ireland may have to deal with several different Government agencies. Seeking free medical aid might involve separate visits to, or dealings with, the Health, Revenue and Social Welfare departments and/or some of their operating agencies. Even obtaining a passport can, in certain circumstances, involve contact with three or four different government agencies. The silo structure is not only highly inefficient from the perspective of both citizen and civil servant, it also creates opportunities for unscrupulous citizens and civil servants to abuse it. It leads to frustration, waste of resources and a poor image of the state (Osborne & Gaebler, 1992). On the other hand, there is huge latent value locked up in these systems. The problem is how to overcome technical and structural isolation, maintained and reinforced by the prevailing bureaucratic and political culture? In order to overcome these barriers, development of new ideas about the nature of value to the citizen and a broader view of how value itself can evolve over time are required.

E X T R AC T I N G L AT E N T VA LU E

–

A PAT H D E P E N D E N T M O D E L

The concept of IT value creation having a time dimension (sometimes called the process view of IT value creation) has been the subject of a considerable research (Mooney et al., 1995; Soh & Markus, 1995). Soh and Markus synthesized five models developed by a number of other researchers and summarized these in following model illustrated in Figure 6.

Figure 6. The IT value creation process (Soh and Markus, 1995).



73

This model shows how value is created over time. We can develop this concept in a number of ways. Two that are of importance in this context are to incorporate the concepts of System lifespan and Path dependent value. Most IT systems have a life cycle, i.e. value dissipates over time (Figure 7). Systems may become obsolete or redundant for a variety of reasons including technological and/or organizational change. As Keen (1991) and others have shown, the cost of system maintenance also rises rapidly with time (Figure 7). At some point, the cost of maintaining the system may exceed the value being delivered at which point the system will (or should) be abandoned or replaced. There are several assumptions in this model including that a clear measure of value is available at all times and that the nature of the value or contribution is relatively stable over time. In practice, the value placed on any system will change over time. It may increase or decrease and the composition of the value delivered by the system may also change. The latter concept is shown in Figure 8. In this model, the mix of components that constitute value changes over time. The initial rationale for or benefits of the investment may decrease in importance (or even entirely disappear). However, it is also possible for new and possibly unanticipated benefits to emerge (Figure 9). The question is therefore whether either new benefits can be created or realized which capitalize on the existing investments and which outweigh the costs of maintaining and/or upgrading the system. If new value can be identified, the cost–value balance will change and better and more informed decisions about continuing IT investment can be made. It has already been noted that business process re-engineering seeks to unlock the value in IT systems by, inter alia, changing processes, organizational structure and culture. New value can also be realized by actually changing the perception of what constitutes value. Many bureaucracies are still wedded to mechanistic measurements of IT effectiveness. In essence, a new view of value needs to be internalized by public administration and adopted by politicians. The concept of the single source of service and information and customer centred service is now emerging as a powerful idea in public administration in Ireland and elsewhere (OECD, 1992; B. Tuohy, unpublished).

Figure 7. System net value over time.


74

F Bannister

Figure 8. The changing composition of IT value perception.

Figure 9. Identifying new value in IT investments.

D E M O L I S H I N G T H E S I LO S

–

THE CHALLENGE IN IRELAND

The central civil service in Ireland employs approximately 30 000 people. Despite its relatively small size, the Irish civil service is immensely powerful and important (Dooney & O’Toole, 1992). The Irish civil service closely adheres to the Weberian bureaucratic model. It organi-



75

zation is strictly hierarchical and its prevailing ethos is one of risk avoidance (Joyce, 1985; Murray, 1990). Given this, it is perhaps not surprising that the history of computing in the Irish public is chequered. Pye (1992) documents the development of IT in the civil service over the period 1960–90 as a continuing struggle to come to terms with IT and devise organizational structures to manage it. This pattern has continued since Pye’s account was published. Although the civil service can boast many successful IT achievements, widespread realization of IT’s full potential value has been inhibited by several factors, including indifferent senior management, inadequate staff resources, a lack of central direction, lack of political will and the semi-autonomous way in which departments can govern their own affairs, sometimes without due regard to service wide considerations (Pye, 1992). The perception of the value obtained from its IT investments amongst senior civil servants responsible for IT varies, but there is a widespread feeling that value for money is not what it should be. Willcocks reports a similar phenomenon in the UK (Willcocks, 1994). The Irish civil service may have avoided any national disasters [in contrast with, say, the UK, where there have been a number of major failures (Margetts & Willcocks, 1993; National Audit Office, 2000)], but a sense of failure to capitalize fully on IT lingers. The challenge for public administration in Ireland, as elsewhere, is to release the locked up value in its legacy systems. To do this, initiatives have been taken at several levels: • at a national level, a Forum for the Information Society has been established (Forfas, 1995); • at state level, the government has launched a Strategic Management Initiative (SMI), a wide ranging programme encompassing a more strategic view of information technology (Government of Ireland, 1996); • at civil service level, the body responsible for setting standards for IT in the civil service, the Centre for Management Organization and Development (CMOD), has initiated a major change in the way individual departments procure and justify information technology investments; • at a department level, several individual initiatives have been launched. Changing the vision of civil service managers and that of their political masters, as well as their perceptions and understanding of IT value concepts, is an essential prerequisite to integration of the diverse systems found in public administration, but integration of public administrative systems also requires, inter alia: • creation and adoption of an appropriate national framework; • suitable technologies. Creating a framework is a organizational and political issue. CMOD have produced, and continue to produce, guidelines which set standards for civil service IT systems and procurement. In the period from 1990 to mid-1998, 13 major service-wide publications on aspects of IT policy were issued by CMOD as well as numerous subsidiary sets of guidelines and regulations. Of these, perhaps the most critical is the Information Management Resources Model (Department of Finance, 1997). Unfortunately, although CMOD has considerable power, particularly in its ability to withhold approval for IT spending, its ability to control, as opposed to influence, service-wide IT strategy is limited. The larger departments of state have consider-


76

F Bannister

able autonomy in IT strategic decision making and frequently exercise it. Senior managers in CMOD argue that much of the framework is there. The question is how to get the framework accepted and, more critically, implemented? This requires technological vision and innovation. Unfortunately it also requires political will, a commodity which is generally in short supply when it comes to IT. Politicians are good at making grandiose statements about the information superhighway and electronic government, but the nature of politics in Ireland is such that, as one Government Minister put it, when it comes to political survival, getting medical cards for half a dozen constituents is infinitely more important than making a well thought out policy statement on health. With public representatives (and in some departments even senior management) showing little or no interest in civil service information systems, often the drive for change has to come from below – a less than ideal situation that some civil servants find frustrating.

A R O L E F O R E N T E R P R I S E R E S O U RC E P L A N N I N G ?

Whatever about the political problems, finding the technical means to implement even a modest degree of integration in a structure as diffuse as public administration is a formidable challenge. A comparison with commercial organizations illustrates both the problem and the difficulties and may point the way towards a solution – at least at the conceptual level. The success of Enterprise Resource Planning (ERP) in the past decade has been fuelled by the desire of businesses to integrate and standardize their systems and align processes with best practice. ERP products seek to replace commonly packaged components (such as accounting and stock control) and interface with custom components (Bancroft, 1996; Skok et al., 1997). Adopting an ERP system involves compromises, but at least most commercial operations have a considerable commonality. Most have an accounting system, a payroll, stocks, purchases, sales order processing and so on and industry specific ERP models extend this concept. ERP systems, like all packaged application software, are based on the assumption of commonality, i.e. that certain processes in business are sufficiently generic to make it economically effective to develop software which can be easily parameterized in such a way as to be usable in a wide variety of different business environments. [Even still the market leading ERP system, SAP R/3 has 1000 preset business processes and 8000 tables (Bancroft, 1996; Deutsch, 1998).] Unfortunately, in public administration commonality of this order does not exist. Apart from payroll, personnel and budgeting, almost every system is unique. A defining characteristic of the public administration IT system is that it is custom built. Traditional offthe-shelf packaged solutions to public administration processing requirements applications are relatively uncommon outside of certain well-defined areas. The question naturally arises: to what extent can ERP systems solve the problems of unintegrated public administration systems? The answer may be only to a limited extent (although this is not necessarily the view of some ERP suppliers). Of the larger ERP suppliers, both



77

Figure 10. The Sap Public Sector Model (1999 edition).

SAP and PeopleSoft have well-defined public sector ‘products’ (SAP, 1999a,b; Peoplesoft, 1999). Oracle, which has a large public sector presence for its other products, also has a natural interest in this sector. PeopleSoft’s have developed variants on their ERP software aimed at both US Federal Government and the wider US public sector. SAP have developed their own public sector model (see Figure 10). To date ERP systems seem to have achieved only limited penetration in European public administration (more research is needed to establish the extent and success of this). Unsurprisingly, SAP claim the greatest success in Europe with major implementations of their R/3 software in Austria (Department of Finance) and the Netherlands (Departments of Justice and Economic Affairs). In Ireland, the impact of ERP in the civil service to date has been marginal and at least one major implementation in a large government department had to be abandoned when the software concerned proved incapable of handing government accounting requirements. At the time of writing, a further more successful attempt at implementing this system is in progress. Notwithstanding SAP’s public sector model, given the vast diversity of civil service systems, it is difficult to see how current ERP technology can have a major impact on the totality of civil service information systems requirements although they might have an impact in certain key areas including human resource management and possibly, given certain reforms in public sector accounting, in financial management.


78

F Bannister

TO WA R D S A N E W C O N C E P T UA L F R A M E W O R K

However, even if ERP itself is at best a partial solution, a more powerful concept which has been derived from cross-fertilization of ERP, EDI and object-oriented design is that of business objects or components (Sutherland, 1995; Taylor, 1995; 1998; Skok et al., 1998). This envisions large-scale or coarse-grained ‘objects’ which are common in business (such as customer or product) with similar properties to the smaller objects more usually discussed in object technology. ERP components can be regarded as quasi objects. An ERP module is built of subprocessing units, which can be switched on or off, and the process can be routed through the internals of the object in a variety of ways that are programmable by the user. In this way, the object can be tailored to the requirements of individual organizations. Even though current ERP systems are not truly object oriented, the concept of a toolbox of such components usable across a civil service is a powerful one. The same principle, applied to data, is found in the EDIFACT standards (Colberg et al., 1995). EDI is primarily concerned with data. It is not generally concerned with the internal processes of the organizations which generate and receive data via EDI. While ERP may standardize data and processes within an organization, EDIFACT standardizes information content and format between organizations. If effective messaging is to be achieved, an EDIlike standard for government messaging will be required. Drawing together all of the threads discussed in this paper, the challenge comes down to this: is it possible, given the structural and organizational constraints of public administration, to define a Business Object Architecture (BOA) (see Casanave, 1995) which can be used throughout the civil service and which: • is based on highly reusable code and business objects? • has a standard messaging interface between as well as within departments? and • can leverage the value buried in legacy systems? Each of these is a major question in its own right. In this paper, we will focus on the first and comment briefly on the other two. Developing the object concept requires both adopting those business objects that can be usefully imported from the commercial world and defining those business objects that are unique to (or at least materially different in) the civil service. We might term the latter ‘administrative objects’ to distinguish them from business objects. Clearly many business objects used in commercial companies are usable in a civil service – for example in areas such as purchasing or personnel. Nevertheless, the civil service will have its own set of administrative objects, for example: • • • • •

citizen; licence; payment; application/claim; public information;



79

• life event; • certificate; and so on. To take a couple of these at random, the administrative object Licence could encompass: • • • • • •

television licences; drivers licences; fishing licences; dog licences; passports; export licences;

and many other forms of state permit. LifeEvent could cover: • • • • • •

birth; marriage; separation; divorce; death; change of domicile, etc.

The normal concepts of object-oriented systems can be applied to such activities including encapsulation, class, inheritance and polymorphism and so on. Figure 11 illustrates the object Licence in an object-oriented diagrammatic form.

Figure 11. The administrative object Licence.


80

F Bannister

Figure 12. A sample class hierarchy for the administrative object payment.

This is very similar to, say, a Policy object in an insurance company. The difference is that such an object would be defined for the entire administrative structure, not just for one business entity. Class hierarchies can be defined for such administrative objects. For example, one of the most common actions of civil service departments is to make payments to citizens (which, for the purposes of argument, we will here differentiate from payments to suppliers). The range and variety of payments to citizens is enormous. Nevertheless, it is possible to conceptualize a ‘Payment’ object with a class hierarchy of increasingly specialized subclasses which inherit the characteristics of their parent objects. An example of how this might work is shown in Figure 12. Apart from dismantling silo structures, there are further advantages to the business object approach. Taylor (1995) asserts that, when it comes to creating value, the ability of information systems to enable users to adapt quickly is today considerably more important than their ability to increase productivity. Taylor differentiates three stages of adaptive organization: 1 organizations that develop systems designed to react change quickly; 2 organizations whose systems try to predict change in order to position themselves to take advantage of it when it occurs;



81

3 organizations whose systems enable them create change because their systems enable them to do this. Although many of the civil service departments would be happy to be at stage 1, at least one Irish department of state is examining the possibility of using business object technology to leap to stage 3, seeing not only a potential solution to its current silo problems, but also the possibility of implementing so-called expressive systems (Pawson et al., 1995). The vision is that not only would the compartmentalization of current systems be masked from the user, but the system would enable the user to combine objects in new and creative ways in a fashion similar to that already used by pioneering users of this technology (see, for example, Haeckel & Nolan, 1993; Harmon & Morrissey, 1996). The third leg of the triangle is also important. If such an initiative is to be successful, it will have to leverage legacy systems. Object technology provides a variety of ways of integrating legacy systems in object-oriented architectures from encapsulation of stored procedures and use of procedure calls to screen scraping although some of these are more effective than others (Taylor, 1995; Harmon & Morrissey, 1996). The success of any project using object technology will hinge on releasing the value tied up in the existing legacy systems and the vast pool of data and accumulated process knowledge embedded in them, what Haeckel & Nolan (1993) refer to as knowledge of ‘the way we do things around here’. The civil service has neither the time nor the resources for lengthy experiments with technology. Short pilots are acceptable, but the ability to leverage existing systems is vital. (Aside: this point was vividly illustrated at one meeting attended by the author, when a consultant outlining a idealistic vision of future systems was brought back to earth by a senior civil servant who quietly asked: ‘this is all very admirable, but at what point does reality intrude?’.) Scale is also an issue. The sheer size of public administration systems make implementation of new technology of this type a daunting prospect. Collingridge & Margetts (1994) discuss the problems in the UK where an attempt was made over an 15-year period to integrate the systems in the Department of Health and Social Security. Writing of the wider UK public sector experience, Margetts & Willcocks (1993) are not encouraging, identifying problems also found in Ireland such as high staff turnover, the separation of policy making and IT and the isolation of IT generally. In Ireland, some departments of state have tried partial integration, integrating systems which have a high degree of commonality or which serve similar functional areas. For example, the Department of the Environment is currently in the process of integrating the licensing of vehicles and drivers. The Department of Social Welfare has integrated the payment of different short-term benefits. In arguably the most interesting of these developments, the Revenue Commissioners’ Integrated Tax Processing project, launched in 1993 and scheduled for completion in 2001, is using state of the art technology to re-write diverse legacy systems into an integrated whole. Notwithstanding these projects, it will take more than a series of isolated integrations to achieve the seamlessness envisaged by the proponents of electronic government. Much research will be needed to prove whether this concept has validity and value. As of today, the foundations of such a framework are in place. The vision exists. What is needed


82

F Bannister

now is the political will and possibly a radical re-conceptualization of public administration processes if seamlessness of service to citizens is to be achieved within the next 10 years.

C O N C LU S I O N

Like many public administrations, the Irish civil service has evolved a large number of unintegrated information systems. Discrete systems are to be found not only between departments, but within departments. These systems have evolved to serve closely defined user and/or client groups. Dismantling these silo systems and converting them into integrated systems which can serve the public efficiently and effectively is a major challenge facing public administration in Ireland and elsewhere. The vision is the easy part. To convert it to reality requires changing perceptions of value and often ingrained ideas about the nature and function of public administration. It means confronting formidable technical challenges with limited resources. Although senior IT management in the civil service is aware of the problems, the opportunities and the issues, achieving the political drive in an organization which, above all, must be responsive to innumerable political imperatives, has proved difficult in the past. If it is to be achieved, a new vision of value must be sold to both politician and mandarin. This demands both a change in mind-set about what constitutes value in a public administration information system and radical thinking about such information systems. A broader vision of what constitutes value to the citizen/consumer will be essential to continue to generate the energies and effort required to drive these changes forward during the early years of the new century.

REFERENCES Audit Commission (UK) (1986) Performance Review in Local Government. HMSO, London. Bancroft, N. (1996) Implementing SAP R/3: How to Introduce a Large System into a Large Organisation. Manning, Greenwich, CT. Banker, R., Kauffman, R. & Mahmood, M.A. (1993) A Comprehensive bibliography of the strategic and economic value of IT. In: Strategic Information Technology Management Banker, R., Kauffman, R. & Mahmood, M.A. (eds), pp. 607–657. Idea Group Publishing, Harrisburg. Boyle, R. (1989) Managing Public Sector Performance: A Comparative Study of Performance Monitoring Systems in the Public And Private Sectors. The Institute of Public Administration, Dublin. Bozeman, B. (1987) All Organisations are Public: Bridging Public and Private Organisational Theories. JosseyBass, San Francisco.

Bretschneider, S. (1990) Management systems in public and private organizations: an empirical test. Public Administration Review, 50, 536–545. Casanave, C. (1995) Business object architectures and standards. In: Business Object Design and Implementation, Proceedings of the OOPSLA ’95 Workshop, October 1995, Austin Texas, Sutherland, J., Patel, D. & Casanave, C. (eds), pp. 7–28. Springer, London. Champy, J. (1995) Reengineering Management: the Mandate for New Leadership. Harper Collins, London. Colberg, T.P., Gardner, N.W., Horan, K., Mcginnis, D., Mclauchlin, P. & Yuk-Ho, S. (1995) The Price Waterhouse EDI Handbook. John Wiley, New York. Collingridge, D. & Margetts, H. (1994) can government information systems be inflexible technology? The operational strategy revisited. Public Administration, 72, 55– 72.



83

Comptroller and Auditor General (Ireland) (1997) The Department of Social Welfare. The Development of the Integrated Short-term Schemes Computer System, Report on Value for Money Examination – Report 17. The Stationary Office, Dublin. Cronk, M. & Fitzgerald, E. (1997) A Contextual moderator of information systems business value. In: Evaluation of Information Technology, Proceedings of the Fourth European Conference on the Evaluation of Information Technology, Berghout, E. & Remenyi, D. (eds), pp.

Margetts, H. & Willcocks, L. (1993) Information technology in public services; disaster faster? Public Money and Management, April–June, 49–57. Mooney, J., Gurbaxani, V. & Kraemer, K. (1995) A process oriented framework for assessing the business value of information technology. In: Proceedings of the Sixteenth Annual International Conference on Information Systems, Degross, J., Ariav, G., Beath, C., Hoyer R. & Kemerer, C. (eds), pp. 17–27. The Association of Information Systems, Amsterdam.

131–138. Delft University Press, Delft. Department of Finance (Ireland) (1997) Delegation Supporting Material: Alignment of SMI and IT Strategy Plans, Version 2.1. Department of Finance, Ireland. Department of the Public Service (Ireland) (1985) Serving the Country Better. The Stationery Office, Dublin. Deutsch, C. (1998) Software that can make a grown company cry. New York Times, 8 Nov., http://www. Nytimes.com/librar y/tech/98/biztech/ar tic les/08 soft.html Dooney, S. & O’Toole, J. (1992) Irish Government Today. Gill and Macmillan, Dublin. Forfas (1995) Information Society Ireland: Strategy for Action, Report of Ireland’s Information Society Steering Group. Department of Enterprise and Employment, Dublin. Frissen, P.H.A., Brussaard, B., Snellen, I.Th.M. & Wolters, M. (1992) Introduction. In: European Public Administration and Informatization: a Comparative Research Project Into Policies, Systems, Infrastructures and Projects, Frissen, P.H.A., Brussaard, B.K., Snellen, I.T.M. & Wolters, M. (eds), pp. 1–10. IOS Press, Oxford. Gaster, L. & Taylor, M. (1993) Learning from Customers and Consumers. Local Government Management Board, Luton. Government of Ireland (1996) Delivering Better Government – Second Report of the Co-ordinating Group of Secretaries. The Stationery Office, Dublin. Haeckel, S. & Nolan, R.L. (1993) Managing by wire. Harvard Business Review, Sept.–Oct., 122–132. Harmon, P. & Morrissey, W. (1996) The Object Technology Casebook: Lessons from Award-Winning Business Applications. John Wiley, New York. Joyce, L. (1985) Administrators or Managers?: An Exploratory Study of Public and Private Sector Decision Making. Institute of Public Administration, Dublin. Keen, P. (1991). Shaping the Future: Business Design Through Information Technology. Harvard Business School Press, Boston, USA.

Muid, C. (1994) Information systems and new public management – a view from the centre. Public Administration, 72, 113–125. Murray, C.H. (1990) The Civil Service Observed. Institute of Public Administration, Dublin. National Audit Office (UK) (2000) The Cancellation of the Benefits System. The Stationery Office, London. National Consumer Council (UK) (1986) Measuring Up – Consumer Assessment of Local Authority Services: a Guideline Study. National Consumer Council, London. OECD (1992) Information Technology in Government – Management Challenges. OECD, Paris. Osborne, D. & Gaebler, T. (1992) Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector. Addison-Wesley, New York. Pawson, R., Bravard, J. & Cameron, L. (1995) The case for expressive systems. Sloan Management Review, Winter, 41–48. Peoplesoft (1999) Enterprise Solutions for the Public Sector. PeopleSoft, Pleasanton, California. Powell, P. (1992) Information technology evaluation: is it different?. Journal of the Operations Research Society, 43, 29–42. Public Service Advisory Council (1983). Report of the Public Service Advisory Council – 1983. The Stationary Office, Dublin. Pye, R. (1992). An Overview of Civil Service Computerisation, 1960–90. Economic and Social Research Institute, Dublin. Quinn, J. & Baily, M. (1994) Information technology: increasing productivity in services. Academy of Management Executive, 8, 28–31. SAP (1999a) Public Sector – Executive Summary. SAP, Walldorf. SAP (1999b) Applications and processes: the public sector solutions map. SAP INFOFocus, May, 1–9. Skok, W., Sims, A. & Hyam, S. (1997) An evaluation of a packaged applications strategy for information systems development. In: BIT 1997: Proceedings of the 1997


84

F Bannister

BIT Conference, Manchester Metropolitan University, Manchester. Skok, W. & Sims, A.T. & Doeringer, H. (1998) design and implementation of enterprise resource planning (ERP) systems using component based standard-software. In: BIT 1998: Proceedings of the 1998 BIT Conference, Manchester Metropolitan University, Manchester. Soh, C. & Markus, M.L. (1995) How IT creates business process value: A process theory synthesis. In: Proceedings of the Sixteenth Annual International Conference on Information Systems, Degross, J., Ariav, G., Beath, C., Hoyer R. & Kemerer, C. (eds), pp. 29–41. The Association for Information Systems, Amsterdam. Sutherland, J. (1995) The object technology architecture: business objects for corporate information systems. In: Business Object Design and Implementation, Proceedings of the OOPSLA ¢95 Workshop, October 1995, Austin Texas, Sutherland, J., Patel D. & Casanave, C. (eds), pp. 29–36. Springer, London. Taylor, D.A. (1998) Object Technology: a Manager’s Guide. Addison-Wesley, Reading, MA. Taylor, D.A. (1995) Business Engineering with Object Technology. John Wiley, New York. Weber, M. (1919) Bureaucracy. In: From Max Weber: Essays in Sociology, Gerth, H., Wright, C. & Mills, Trs. (eds), pp. 196–244. Routledge, London.

Willcocks, L. (1994) Managing information systems in uk public administration: issues and prospects. Public Administration, 72, 13–32.

Biography Frank Bannister is a Senior Lecturer in Information Systems in Trinity College, Dublin, where he is Director of Studies of the Management Science and Information Systems programme. He is a Chartered Engineer. Prior to joining Trinity, he worked for 20 years first in Operations Research in the Irish Civil Service and then with Price Waterhouse, Management Consultants. He remains a Consulting Associate with PricewaterhouseCoopers. His research interests include IT value and evaluation in Public Management/Administration, IT/IS systems acquisition and investment generally and the use of computer modelling techniques to aid decision making. He is Editor of the Irish edition of IT Policies and Procedures published by Thompson/GEE. As well as lecturing in Trinity, he is an external lecturer at the Irish Management Institute and runs specialist courses for the Irish Civil Service Training Centre.
