The real price of everything, what everything really costs to the man who wants to acquire it, is the toil and trouble of acquiring it.
Adam Smith, The Wealth of Nations, vol. 1, bk. 1, Ch. 5 (1776).

There are some things which cannot be learned quickly, and time, which is all we have, must be paid heavily for their acquiring. They are the very simplest things and because it takes a man’s life to know them the little new that each man gets from life is very costly and the only heritage he has to leave.
Ernest Hemingway, Death in the Afternoon, ch. 16 (1932).

1.     Introduction

The establishment of the costs of IT is an extremely complex matter which is not well addressed by practitioners, consultants or academics through the literature. In fact there is almost an implied assumption in the IT evaluation literature that benefits may be difficult to estimates but that cost are really rather straight forward. This is quite untrue. The determination of IT costs can be just as difficult as IT benefits and in general the cost issue is not handled well by IT practitioners, or for that matter by accountants, academics or even consultants. The cost issue is considered to be the relatively easy, unglamorous end of the evaluation equation which may be left up to the bookkeepers or the bean-counting accountants. But even fully qualified cost accountants are often challenged by the problems, which they face when they come to understand and record the costs of IT. On reflection on this is hardly surprising as cost accounting even more than general financial accounting is a subject, which is open to a very high degree of interpretation and is well know to be very much more an art than a science.

To add to this complication there is the fact that cost may be seen through the lens of accountants or through the lens of managers or economists – and there is a significant differece between these two mid-sets. In general, accountants accumulate data in order to be able to  answer the question of what was actually paid to achieve a specific objective while a manager or an  economist considers the question of investment or action alternative including what would have happen if some other objective was pursue i.e. considers, inter ali, the case of opportunity costs. Sometimes these two different lens or mind-sets are hard to reconcile. To make this situation even more challenging there is the issue of dis-benefits or soft costs which sometimes accompany IT investment. These types of costs, just like intangible benefit are quite real but are very difficult to quantify and as a result they have largely been ignored By those developing business cases or producing cost benefit analysis statements.

This is a theoretical paper in which the evidence to support the arguments presented is drawn both from the general literature on evaluation (Price Waterhouse 1990; Strassmann 1990; Keen 1991, Hochstrasser 1995, Hobgin and Thomas 1994; Willcocks and Lester 1994; Bacon 1994, Sherwood-Smith 1989, Bannister 1995, Remenyi et al 2000) and from conceptual and empirical work done by the authors in the information systems field. It reviews the issue of IT costs from a number of different perspectives and examines some of the key issues pointing out areas of controversy. The paper argues that like other areas of corporate investment and expenditure there are no simple right and wrong ways of applying the principles of cost accounting. Considerable care needs to be taken when addressing the issues of IT cost.

The paper proposes the development of an IT Cost Reference Model as an aid to understanding the nature of IT costs and the challenges in assessing them and it offers this  framework for discussion.

It is hoped that this paper will trigger a debate in the IT community about the crucial importance of understanding IT cost issues and that it will provide a stimulus for further work in this area.

1          Managers need to know

It is something of a truism that managers and perhaps especially senior managers in many organisations are concerned about the very substantial amounts their organizations are spending on IT.  Furthermore there is ample survey evidence to support the fact that this concern has been persistent (Computer Weekly 1991, Datamation  1992, Price Waterhouse 1994, 1996; Wentworth 1997, Computing 2000). However, this problem does not stop with a concern that the IT budget may be too big or even that the budget, whatever its size, may be exceeded. The roots of this problem are much deeper and many managers have an even more fundamental concern about IT and its costs.  In simple terms many mangers just have no idea what IT is actually costing them. Despite this lack of accurate knowledge many mangers and executives have intuitively felt for quite some time that the cost is too high and that they are not getting value for money from their IT investment. This was clearly reflected by Computer Weekly 1991 when it wrote:-

Time has run out for IT managers who act like a protected species. Their three-fold failure to understand the business they are supposed to be part of, to communicate with their business colleagues, and to deliver cost effective systems has led to a collapse of faith in the IT department itself.

And this situation is really no different today.

The worry that the costs are excessive is further supported by Sing (1993) when he said:-

Problems are more abundant than solutions, organisations experience rising costs instead of cost reductions, IS misuse and rejection are more frequent than acceptance.

Although it would be wrong to say that  this concern for cost is true for all organisations it is actually very common, a fact which is supported by the research of Lacity and Hirschheim (1995):-

 ……only two of the thirteen companies that participated in the study agree that their IS departments are critical to corporate success. The remaining eleven companies all see their IS departments as necessary, but burdensome, cost pits.

Clearly the need to understand IT costs  is indeed becoming a most important issue. There is a new wave of technology being offered to organization in the form of e-Commerce and e-Business opportunities and it is clear that the adaption  of this technology will result in the total expenditure on IT to continue to grow substantially. Thus the understanding of IT has now become quite critical and needs urgent attention.

2          An underpinning assumption

A considerable part of IT evaluation, be it pre-investment or retrospective, is predicated on knowing what the IT costs[1] actually are.  Accurate knowledge of costs is critical given the pervasive use of capital budgeting, cost benefit analysis and variance accounting as IT pre and post evaluation methods (Peat Marwick Mitchell 1989; Willcocks and Lester 1994; Bacon 1994)[2].   Much of the work on costs found in the literature is concerned with directly visible costs, i.e. those costs which appear in IT financial budgets. This is clearly demonstrated in the work  of Lockwood and Sobol (1989).  In an investigation of 159 organisations, Lockwood and Sobol analysed IT expenditure under a variety of headings such as personnel, hardware, software, etc.   Their analysis was, however, based on budgets and it is the contention of this paper that IT budgets give at best a partial view of IT costs.  Research has shown that up to 40% of IT costs may be incurred outside of the traditional IT budget (Price Waterhouse 1990; Strassmann 1990; Keen 1991, Remenyi et al 2000). In fact Hochstrasser estimates that as much 30-50% of IT costs occur outside the official IT budget, commenting that:

 “… a need has been identified to be better aware of the true costs of IT projects…”

(Hochstrasser 1995, p156).  Very little of the IS literature is concerned with how cost is determined, possibly because this is considered an accounting problem and thus the province of another academic discipline or field or study.  Based on the survey evidence, it can be seen that only a minority of organisations have any real understanding of their IT costs.  Surveys of IT evaluation practice discuss in depth the various types of evaluation method used and how to choose the best evaluation method, but very few concern themselves with the basic mechanics of arriving at the actual cost figures against which any benefits should be measured.

 3          Three Important Challenges

In order to evaluate anything, it is first necessary to define clearly what is to be evaluated and in the field of IT evaluation,  this is not always easy. Consider three generic types of  IT evaluation problems:

1.      Total Corporate IS expenditure;

2.      A defined subset of its IT operations;

3.      A defined project.

While these three generic types of  IT have much in common, each of the above presents different challenges to the evaluator in defining the expenditure whose outcome or value is to be assessed .

3.1         Total Corporate IS  Expenditure

At first glance to those who are not experienced in IT costing, total IT expenditure may appear be the easiest to define.  Total IS Cost bill should simply require identification of all IT related external expenditure on capital goods, consumable goods, and services. Then add to this the internal wage and corporate accommodation costs.  To help with this there are many lists and categorisations of IT costs available in the literature (for example LAMSAC 1978; Parker and Benson 1988 Appendix C; Hobgin and Thomas 1994; Bannister 1995, Remenyi 1999, 2000).  Unfortunately, arriving at the total cost of IT is not quite as straight forward as this.  The full or true cost of any item needs to include the overhead[3] it attracts.   Gartner Group (1997) estimated the typical three year cost of running a networked PC to be in the range $27,000-$36,000 (although these figures have been disputed[4]). This so-called total cost of ownership concept has been extended beyond PCs to any identifiable unit (such as software packages, remote access workstations, mainframe, etc.).  It is at its most complex and debated in the area of networks (see, for example, Leach & Smallen 1998; PC Week 1997; Dudman 1999). 

Even if one narrows the problem to the direct cash cost per period, many costs may not be captured by the accounting system.  To take an oft cited example, introduction of a new software package might require training many users.  Training costs are notorious for not being allocated to the cost of the use of the IT solution. Furthermore there are those which claim that not only the cost of the course should be attributed to the IT but also the time of the individuals being trained[5]. In such cases unless the company has a time recording system the cost of the users’ time spent in training will almost certainly not show up as an IT cost.  Even if it does, informal on the job training, time spent reading manuals and impromptu mutual self help activities are simply untrackable in most organisations.  In theory, professional organisations such as accountancy, consulting and legal practices which record staff activity by time can track and therefore highlight such activities.  In practice to do so with any degree of accuracy requires both an appropriate coding system and a high degree of user discipline which is sometime absent[6]. However in reality  even good data capture systems are prone to manipulation and bias for a number of reasons (Lukka 1998). Furthermore agency theory also suggests that people massage reported data to meet the criteria by which they perceive that their performance will be judged (Puxty 1985; Morgan 1991; Koford and Penno, 1992; Anthony and Govindarajan 1995) and this make the identification of the real cost even more difficult to establish..

To add to this problem there are other categories of soft costs, which are sometimes referred to as dis-benefits (Khalifa 1999, Remenyi et al 1995), that do not generally find their way  into the total IT cost equation.  These soft costs or dis-benefits normally involve losses in productivity for a variety of reasons including:

·        badly designed systems

·        mistakes made due to inadequate training,

·        slow system response,

·        loss of business during down time,

·        IT enabled time wasting,

as well as factors such as costs of system failure/unavailability, the impact of inaccurate information etc..  It is sometimes argued that it is never possible to know the full cost of an activity as ubiquitous as IT.

As noted above, conventional accounting computation of the total cost of system ownership may yield a considerable underestimate.  Some IT professional would argue that it is because of these underestimates of the cost of IT in the organisation which has allowed much IT investment to take place. The idea is that if senior management had know in advance what the total cost would actually have been then it is unlikely that they would have authorized that expenditure[7]. This ironically goes against the advice of  Cervantes (1605) who suggests that :-

That which costs little is less valued

Clearly, in a sense the IT treads a delicate line between being two extreme positions, which are being too expensive, and having adequate importance to the organisation.

In summary it is not trivial to estimate the organisations total IS cost and this has been a contributing factor to the suspicions of business executives and mangers that too much is actually being expended on this function. It could actually be argued that it might be impossible to establish an accurate figure for total corporate IS cost and certainly the question of whether it is actually worth while to be able to do this needs to be carefully considered.

3.2         Defined Subsets of IT Operations

To answer the question: ‘How much is system “A” costing per annum?’ we need first to carefully define system “A”.  In an age of integrated and distributed computing this is rarely easy and, in certain circumstances, may be close to impossible.  A subset of operations might be defined in a number of ways, including, for example:

·        A business unit (e.g. a bank branch);

·        A department (e.g. production);

·        A process (e.g. order processing);

·        An application (e.g. office automation);

·        A piece of software (e.g. the Executive Information System);

·        An infrastructural component (e.g. the network)

and so on.  There are many such possibilities and the costing problems vary with the nature of the system.  There is frequently a considerable amount of overlapping and very few IT sub systems can be conveniently  isolated for costing purposes. Even if the front end cost[8] of a system is clear, as soon as it is put into production, a series of other costs immediately arises. As mentioned above users have to be trained[9]; there may be disruption to existing operations; existing systems may need to be interfaced. Such cost consequence may lead to a ripple effect the boundaries of which are far from clear.  For example, Snyder and Davenport (1997) in their guide to costing library services, incorporate many of the factors noted above, but even they omit some of the more subtle cost impacts. Hochstrasser (1994) provides a worksheet for such costs, but like the other authors who provide lists of costs, does not discuss the problems in computing them.

Thus the task of defining a cost of a system which is a subset of the IT operation is a challenging one and this has resulted in there being problems is making decisions as to how to manage investment in specific applications which are a subset of the whole IT operation. Also for most organisation it may be quite important to know what different parts of their IT operation is costing them so that decentralised manger can make appropriate decisions about their budgets.

3.3         IS Project Costs

IS Project costs are no less difficult to identify and estimate than either of the other two categories. In theory an IT  project should be easily definable and this should be a major factor in ensuring its professional management.  Project management theory states that a well-managed project should have clear boundaries (Turner 1993, 1995; Kerzner 1995; Field and Keller 1998), as one pair of authors puts it:

A project is a unique venture with a beginning and an end, conducted by people to meet established goals within parameters of cost, schedule and quality

Buchanan and Boddy (1992, p8).  Thus if the project boundaries are clear, so should be the costs. However, not all IT projects are clear cut (Lock 1996; Remenyi et al 1997; Remenyi 1999).  In particular, few IT projects proceed in isolation. They are often part of a much larger IT programme. Furthermore almost all IT projects use and share existing resources and this immediately gives rise to the question or cost apportionment. , To add to this IT projects frequently disrupt other activities and these disruptions push up the organisation’s total cost bill by generating   a variety of indirect costs.  

Of the three challenges, IS Project evaluation is most frequently the object of an investment decision. This causes the question of  what cost data are relevant to this decision to be asked?  This is often not a simple question to answer.

For a start, costing of IT projects is frequently distorted by accident or design. Key costs are frequently overlooked.  One example is pre-acquisition costs[10] which are commonly regarded as ‘sunk’ costs and not relevant at the point at which the decision is made[11].  This is a traditional marginal cost approach[12] (see Drury 1997; Horngren et al 1997); while it employs theoretically sound DCF logic by concentrating on the future incremental costs, it may fail to capture the aggregate long term costs of a project[13]. An unscrupulous manager can easily manipulate numbers to make the cost of an investment in the short term appear quite low, hiding from view longer term costs.

The weaknesses of traditional marginal costing were amongst the factors that led to the development over the past decade of Activity Based Costing (ABC) (Cooper and Kaplan 1988; Drury 1994, 1997; Horngren et al 1997; Atkinson et al 1997). ABC is based on the concept of cost drivers, i.e. that any cost which is a consequence, even if indirect, of an activity is a cost of that activity.   It attempts to capture intertemporal cost consequences through recognition of the fact there is often a significant lag between decisions or initialising activities and the associated escalation of organisational costs.  The growth in interest in ABC is a recognition of the fact that it is to some business executives a more meaningful measure.

In general project costs are frequently fudged in order to be accepted by management and this approach seems to have serves IT management rather well over a long period of time. However this approach is not conducive to improving corporate decision making or corporate learning.

4          Other important IT cost issues

Besides the three generic types of IT evaluation problems there are several other important which need to be addressed in order to understand the complexity of IT cost estimation.

4.1         Opportunity cost approach

As mentioned above managers and economists tend to view costs differently to accountants. This is especially obvious by the way that investments are sometime discussed. Whereas the accountant will frequently ask what is the cost of a project, a manger will often ask what is the cost of not investing in the project which is actually asking what will happen to the organisation if the proposed action is not taken. This approach is described as considering the issue of opportunity cost.  Opportunity costs arise in a number of different ways, one of which is when there is no spare capacity and resources needs to be diverted from other established activities to the project under evaluation.  Any benefits lost from the activity impoverished or deferred are regarded as costs of the driving project.   Opportunity costs have become more important with the success of business process engineering.  Organisations with plenty of slack can often commit considerable resources without impinging on other core activities.  As BPR makes organisations leaner and meaner, such flexibility decreases.

4.2         The time horizon

Another problem with project evaluation is the time horizon which is used in the cost calculations.  A project needs to have a termination point when it is deemed to be complete.  In many IT projects this point is not clear. Even where termination is achieved, additional post acquisition or post commissioning costs almost invariably emerge – for example where additional resources are necessary to ensure the successful management of the new system or unanticipated knock on effects are encountered which create dis-benefits.

Some authors and organisations have developed sophisticated IT project cost estimating methodologies. Wellman (1992), for example, presents a detailed methodology for estimating the cost of software projects encompassing both Total Installation Cost (TIC) and Life Cycle Cost (LCC).  Wellman takes into account pre- and post- acquisition costs and incorporates into his model such refinements as inflation, exchange rates and the cost of staff replacement.   However indirect costs are only partially considered and how such costs might be captured is not discussed except by reference to Block’s Accomplishment Cost Procedure (Block 1971).  The difficult problems of allocation are not addressed.

5          Three persistent  concerns

It is clear that each of the above problems presents some common difficulties and some which are unique to its class.   Three main common issues emerge:

·        Data capture,

·        Costing and accounting principles and

·        Dis-benefits.

Figure 2 gives an overview of these problems which will now be considered in more detail.

From Figure 2 it may be seen that there are several different opportunities for cost to escape from the accountants net and thus not be fully taken into account.

Figure 2: Some Aspects of a Costing System Weaknesses

One of the key features of Figure 2 is the fact that cost leakage can occur in all three major strands of the model. This is potentially a major concern for management and thus requires careful implementation of control systems.

5.1         Weak Data Capture

We have already mentioned that one of the major problems in tracking IT costs is failure to capture the information in the first instance.  Data capture weaknesses fall into two categories: areas where the data are not captured at all and areas where data capture is inadequate or inaccurate.

5.1.1        Missed costs?

We term the first of these two categories missed costs.  Missed costs may be defined as those which should have been Identified and allocated to a particular cost point but which for one reason or another have not been. However at which cost point an item should be  identified will depend upon a number of issues including whether the organisation is following a direct costing approach or a full costing approach. For example in a full costing approach the payroll costs of non IT staff engaged in IT related activities could be regarded as an IT cost whereas in a direct costing environment they would not.

Some obvious missed costs which are often not captured by the typical accounting system include:

·        Disruption costs:  Research has shown that new systems implementation is often accompanied by a short term loss in productivity (Kaplan 1986) as systems are disrupted and users adapt;

·        Displacement costs:  These can occur when people and operations have to be moved to accommodate a new system – for example, data has to be re-distributed over servers or staff have to re-locate;

·        Dis-benefits (these are discussed below).

Missed costs occur because there is no mechanism whereby the accounting system can capture the data.  This is particularly true of people costs and dis-benefits.

5.2         Miss-assigned costs

The second problem is mis-assignment of costs. The Database Administrator goes on a three day course in Paris.  The cost of the airfare is recorded by the accounting system as a travel expense, the accommodation as miscellaneous and other expenses as entertainment when in reality all three are IT training costs. 

Accurate data capture requires two things:

1.      First, the chart of accounts needs to be properly designed, i.e. it needs to have the right headings.  If it does not, costs will be charged into the nearest convenient heading. If an organisation has not changed its chart of accounts in many years, it may be quite unsuitable for the way the business currently operates. We refer to this as structural missassignment.

2.      Secondly it requires staff to use the accounting system correctly.  Even where the chart of accounts is well designed staff sometimes miscode costs deliberately, a process sometimes called ‘burying’ costs.  A common phenomenon is to code costs to a heading where there is still some unused budget rather than to a correct heading which is over budget.  Another is to deliberately mis-categorise goods to circumvent company purchasing rules.  We refer to this as behavioral missassignment.

Where data capture is poor, subsequent management information can be potentially misleading.  Poor management information can have various undesirable consequences.  In this context, it may lead to poor IT investment decisions and/or incorrect evaluations. 

6            Overhead Allocation

In general there are few more troublesome problems in business that that of overhead allocation.

Overhead allocation is the subject of both theoretical debate (Ahmed and Scapens 1991) and sometimes intense practical contention.  As mentioned earlier it is certainly regarded as an art rather than a science and it is often a question of corporate politics as much as a question of good accounting policy. Inappropriate overhead allocation not only results in poor management information, it can also distort user and manager behavior leading to counterproductive or sub-optimal actions.  Overhead allocation can be the cause of very bitter disputes in organisations. And what is even more problematical about overheads is that there is seldom any right or wrong way to allocate these costs and thus the judgements made are always open to critical comment.

6.1         Overhead and Allocation Bases

Within the IT function itself there is a large range of potential overheads cost for a  system including:

·        Shared devices (printers, storage, tape drives, etc.);

·        Shared infrastructure (network, cabling);

·        Shared services (helpdesk, system administration, backup);

·        Central costs (insurance, legal, etc.).

To apportion these costs in a reasonable and fair way the accountant or manager is  faced with the problem of which criterion or which criteria to use as the basis of overhead allocation. This ius seldom an easy choice and one criterion will frequently advantage someone and disadvantage others. The list of possibilities is formidable and includes:

·        Headcount;

·        Number of screens;

·        Number of devices;

·        Total capital investment;

·        CPU usage;

·        Disk usage;

·        Network usage

All of these can give rise to different distortions in information and behaviour.  The following case history illustrates the type of problem which can arise. 

In a large professional services organisation, a small technical department which made up under 5% of the staff had approximately 8.5% of the organisation’s PCs.  The cost of central IT services (including the mainframe, the user support group, the help desk and the network) was allocated on the basis of PC count thus assigning the department almost twice as much IT overhead cost per head as other departments.  In fact, due to the high level of technical expertise in the department, the staff tended to solve most of their own problems and, pro rata, made less use per capita of help and support services than other departments.  Because they were PC oriented, they also made less use of the mainframe so the cost allocation system was doubly unfair.  As the overhead cost was a charge on the department’s profitability and the bonuses of the department’s managers were linked to their departmental profit performance, the allocation of IT costs became a personal and highly political issue.  As a result, a large amount of time was squandered in meetings and tempers sometime rose, further damaging productivity and morale.

6.2         Early Adopters and the Hotel Night Problem

Another problem which can distort IT costs is the so-called ‘hotel night’ problem, This occurs when a new system is implemented, but initial take up is slow (or is restricted to a small group).  As the use of the technology spreads, the cost per user drops.  Failure to account properly for this phenomenon leads to both poor decision making and corporate game playing. The former can arise when the long term cost per user of a system is misunderstood leading to short-sighted decision.  The latter arises when ‘clever’ managers let other departments absorb the expensive up front costs of a new system and then ‘piggy back’ on the system when it settles in.

6.3         Chargeback

Cost allocation may be effected in a number of ways.  One common method involves the assignment of a periodic amount of cost to user departments. More complex systems of allocation may involve a formal system of charge-back to user departments. In essence, an internal market is created whereby the supplying department becomes a profit centre aiming to maximise revenues while user departments aim to use the minimum of such service that is consistent with the achievement of their own objectives.  These systems are usually more controversial than their more basic counterparts noted above, not least because they aim to affect managerial behaviour and decision-making.  Few such systems manage to promote perfect goal congruence. 

Many charge-back systems use a standard cost approach (i.e. they do not attempt to deal with the hotel night problem noted above).  The standard cost is based on a budgeted or anticipated level of usage.  Like any standard costing system, charge-back may under or over absorb costs.  Thus a department which makes heavy use of a network or a mainframe may end up paying more than the cost of the resource used.  This is largely due to the fact that standard charge-backs normally involve a single charge (per ‘unit’ of consumption) that combines the cost of fixed and variable expenses incurred in the supplying department.

7          Dis-benefits or soft costs

Computer systems can have an adverse impact on company performance leading to what we term dis-benefits which are in some instances referred to as soft costs.  Dis-benefits are undesirable effects of an IT-investment and as such are a cost to the organisation.  Dis-benefits are often difficult to identify let alone quantify or are therefore disregarded because they are regarded as just too difficult to handle.  Dis-benefits are discussed by a number of commentators (e.g. Hochstrasser 1987, Remenyi 1991, 1995, 2000).  Strassmann (1988) comments that many IT investments are driven by considerations of efficiency rather than effectiveness and he regards this rather narrow view of the impact of IT to be in itself to be a dis-benefit. Chalcraft (1997) discusses the concept of ‘uncertainty’ and Gurbaxani and Whang (1991) takes the ‘opportunity cost’ of poor information as a variation on the dis-benefit theme.

In general dis-benefits are normally thought of as being of  two sorts:-

1.      Those which are caused by a badly executed implementation.

2.      Those which are inherent in the process change, which no amount of planning or forethought could completely eliminate.

In the first case the problem might be a failure to achieve the level of benefits forecast, or it might involve some unforeseen negative factor of the implementation, poor communication leading to industrial action for example.

In the second case the task of the implementation team is to recognise the existence of dis-benefits, and to manage them. These inherent dis-benefits are not the same as a lack of benefits. They might more usefully be thought of as soft or intangible costs.

Inherent dis-benefits occur for a number of reasons. They can loosely be grouped into three categories; the people, the job and the information system itself. Only the third group are peculiar to computer systems, the first two are a feature of any change programme.

7.1         Dis-benefits related to people

People dis-benefits are largely caused by the way individuals react to change, they are not specifically anything to do with the nature of the IT investment itself. However, because of a concern about IT, these reactions will probably be more marked when a computer system is being introduced than with most other forms of change.

These problems occur mostly because staff frequently mistrust management intentions or motives. People fear loss of their jobs, as the new system will presumably be more efficient than the old. They may very well fear potential discomfort or health problems, although these are less likely now, as more implementations involve changing an existing system, not the initial introduction of terminals.

Staff may very well see increasingly sophisticated computerisation as giving management more information about their working patterns, and thus more control, about the day to day work. Data entry systems which give key stroke counts are an obvious example. A similar example is an electronic mail systems which time stamp all messages. This could encourage staff to send messages either very early in the morning or very late at night.

These concerns are not directly to do with IT itself. The level of distrust will be determined as much by the prevailing industrial relations climate as by the amount and quality of the communication undertaken for this change. But, however good a firm, at least some staff will always suspect anything that the firm does especially with IT investment.

A new, different, information system will mean re-training. Apart from the cost and time away from the job, this re-training will upset the working of the group in other ways. Staff will not be as efficient until they have been working the new system for some time. Mistakes will occur no matter how much training is undertaken. Whilst trying to come to grips with the new system, staff will be under more pressure and may snap at each other or at customers. Dismissing these manifestations as trivial is all too easy for the implementation team who have probably been chosen for a somewhat broader outlook than the majority; of staff who enjoy the atmosphere in their department. These types of situations can be very de-motivating and a significant dis-benefit.

7.2         Dis-benefits related to the job

Many jobs only work because of informal procedures that mangers know little about and the systems analyst did not discover. This is often a difficult concept for managers to accept but can most commonly be seen in the way staff obtain materials to complete jobs. A change to a new system will almost certainly destroy these practices. It may take months or years for the informal procedures or networks to be re-established. During that time the work group's performance will be degraded for no very obvious reason.

IT based systems are often less flexible than traditional ones. They do not normally offer the facility of scribbling special instructions or explanations in the margin of a form. Especially in the early days, this will be seized upon by some staff as further proof that the change will result in a worsening of customer service.

Even if the system can cope with unusual procedures, the initial training will probably have concentrated on the run of the mill, ensuring that staff are as efficient as possible for the majority of cases so the staff may not know how to persuade the system to accept unusual orders or amendments. There is little more calculated to enrage staff than standing in front of a customer staring at a meaningless error message on a computer screen.

7.3         Dis-benefits related specifically to IT

Dis-benefits in this category tend to be risks. They will occur if things go wrong, either the system suffers some major catastrophe, or performs well below expectation.

Computer systems are by their nature more vulnerable to catastrophe than manual systems. The more you rely on the system the more helpless you are when it is not there. Recent research suggests that 80% of firms which have suffered a major IT catastrophe did not survive. Many of these firms did have some disaster recovery plan and backups. But in the event they either proved inadequate or too expensive to implement. The cost of backups and preparing for disaster recovery is of course a well known dis-benefit in its own right.

When the IT function is not performing well, it can absorb an inordinate amount of senior management time; time they should be spending on their core business. If some part of the business is not working well you would expect managers to be devoting a fair bit of their time to the problem. This can be a significant distraction for core business issues.

Of course there will be situations where the solution is simple; more terminals, better lighting, what‑ever. But when the real problem is that the system is not performing as expected, or the hardware is overloaded, the solution is more likely to lie with the IT professionals than with line managers. The line managers may have to spend a considerable amount of time learning about the issues.

It could be argued that the root cause of this weakness is the very low appreciation of IT amongst too many managers, and that this is an issue more correctly categorised as a people problem; i.e. that the level of ignorance displayed by many senior mangers is just as much a cause for concern as the fears demonstrated by the shop floor workers. No matter how much truth there is in this view, it has to be admitted that an information system will always be more complex, and less susceptible to immediate management alteration, than a traditional manual system.

Dis-benefits or soft costs can be external or internal, direct or indirect. .External dis-benefits arise when investment in technology leads to a loss in benefits/profit from the operations of the products or services that the technology is supporting.  This leads to some difficult

 Figure 3: A 2 X 2 Matrix of dis-benefits showing external or internal, direct or indirect dis-benefits or soft costs

questions.  Where a new technology alters the structure and the economics of a market the result may be a much less profitable business, but a company may have no option other than to invest in this technology if it wishes to survive. How to account for both sides of this is not clear.

Dis-benefits are not restricted to external competitive effects.  Introduction of technology can lead to more time wasting by employees, greater stress, new health problems, greater exposure to increased security problem and so on. 

Indirect dis-benefits can occur in a number of ways including the adverse effect of factors such poor system design, system disruption and system failures.  These can lead to problems such as incorrect invoices, failure to collect/delays in collection of receivables and loss of customers.  All of these lead to cash losses and possible loss of future business.

While dis-benefits are often difficult to quantify this is no reason for disregarding them.  Kaplan’s observation about “conservative accountants who assign zero values to many intangible benefits prefer being precisely wrong to being vaguely right” (Kaplan, 1986) can be applied with equal validity to difficult to measure dis-benefits.

8          Capitalisation  and  Amortisation

Two other issues which affect the quantification of IT costs are capitalisation of expenditure and its subsequent amortisation.  Capitalisation refers to the treatment of certain expenditures as assets which then appear on the corporate balance sheet  Amortisation relates to the manner in which this expenditure is subsequently written off in the income statement over a number of years[14].  Different firms are likely to approach capitalisation in different ways.  Most will prefer a conservative approach, capitalising the minimum that is consistent with the dictats of accounting practices and standards in their industry. To this end the advice of auditors is usually sought. This accounting frame of reference, which is governed by various formal statements of accounting practice, can give a significantly erroneous view of the aggregate cost of an IT investment[15].  Even the balance sheets of those firms which adopt a more expansive capitalisation approach - something which is more likely to happen if current profits are under pressure - will fail to capture a rounded view of the cost of such investments[16].  Further complications arise from the fact that internal management accounting reports need not follow external reporting standards and the amounts allowed for amortisation by tax authorities[17] are also likely to differ from both these.  It is therefore essential that caution is exercised in the use of both capitalised and amortised calculations or measures; as a general rule these need to be challenged and augmented along the lines already suggested by the ABC cost paradigm[18].

9          Cost of Risk

Another cost that is associated with  many IT projects is derived from the risk of the project. The risk of a project may be defined as the propensity of the actual outcomes of the project to vary from the planned or budgeted outcomes. In this IT investment environment this is normally related to the project being late and over budget and also not delivering the benefits which were suggested by the project stakeholders. High-risk projects encounter additional costs in at least two different senses. In the first place high-risk projects can be quite stressful on the staff and this stress can simply reduce staff performance. Any reduction is staff performance is de facto a cost to the organisation. Secondly high-risk projects require to be managed differently. If the risks which these projects face are not directly addressed then it is quite probable that risks will develop into problems which will either delay the project or escalate the costs or both. Risk management is not as difficult as is often suggested but it does require funding and thus escalates the cost of the project.

Risk is not normally converted to a monetary cost, but, as already noted, computerisation can open up an organisation to a range of risks to which it would not otherwise be exposed. If the company has to insure against such risks it will be an IT cost (on the assumption that the risk would not otherwise be incurred).  If the company chooses not to insure against increased risk, it is, in effect, self insuring.  This risk premium is clearly an IT cost.

10     Towards a IT Cost Reference Model

Over the life of any IT investment, the range of cost spreads out over time to incur more and more remote effects while the ability of current systems to capture those costs decreases.  This ripple pattern phenomenon is illustrated in figure 4.

Figure 4:  The ripple pattern of cost accumulation

We have shown how capture and accounting for these costs is complex and requires clear and focused accounting policies.  If there is to be clarity and consensus on IT evaluation, a consensus on IT costs is required.  We propose the development an IT cost reference model.  From the preceding discussion, we can outline some of the characteristics of such a model:

1.         there needs to be agreement on what costs are to be included in it;

2.         it needs to have a properly designed data capture system;

3.         there need to be agreed standards for allocation of costs;

4.         it needs to be able to filter out irrelevant data or noise;

5.         it should take account of external rules and standards.

An outline of the proposed model is shown in figure 5.

Figure 5: An IT Cost Reference Model

We argue that without an agreed basis, comparative evaluations are suspect and individual evaluations may be invalid. 

11     Conclusion

Thus the question “How Much Did We Pay for That?” is not a trivial one and there is frequently no one simple answer.

Certainly there is much concern about the cost effectiveness of IT. However there is not an adequate understanding of IT cost issues. This can sometimes result in costs being significantly underestimated in the evaluation of IT investments. This is due to both misunderstanding of what is involved in costing as well as to the fact that relevant costs throughout the wider organisation over an appropriate multi-period timeframe are inadequately captured either within the firm’s historic cost accounting system or its IT budgets.   Costs which tend to be overlooked range from pre-acquisition costs through training/learning costs and overheads within and beyond the immediate IT area of the organisation, to a variety of internal and external dis-benefits that arise as consequences of such investments. 

Weak data capture (whether for structural or behavioural reasons), misassigned costs, inappropriate cost allocation, conservative capitalisation conventions and the inability of conventional accounting to capture disbenefits and opportunity costs all contribute to this problem.  The use of traditional marginal cost logic, which is still a strongly favoured theme of economic and business education, may exacerbate the issue.  It is also suggested that the problem of underestimation has become more acute over the past ten years or so as organisational slack has been progressively eliminated through BPR and delayering initiatives, thus giving rise to higher opportunity costs (due to greater levels of disruption, displacement and diversion of limited resources).  A further, compounding, problem is seen to be the relative rise in the indirect costs of IT initiatives, since these often occur well outside the organisational locus of such investments and are consequently subject to more significant capture and allocation difficulties.

This paper has proposed the development of an IT Cost Reference Model and put forward an initial framework for discussion.  We argue for further research into an area which is both relevant and challenging.

References

• Ahmed, M. N. and Scapens, R. W. (1991): ‘Cost allocation theory and practice: the continuing debate’, in Ashton, D., T. Hopper, and R.W. Scapens (Editors), Issues in Management Accounting, Prentice Hall Europe, Hemel Hempstead.

• Anthony, R. N. and Govindarajan, V. (1995): Management Control Systems, Irwin, Chicago.

• Atkinson, A. A., Banker, R. D., Kaplan, R. S. & Young, R. S. (1997): Management Accounting, Prentice Hall, New Jersey.

• Bacon, J. (1994): “Why companies invest in information technology” in: Leslie Willcocks (editor), Information Management, The evaluation of information systems investments, Chapman Hall, London, 31-47.

• Bannister, F. (1995): Purchasing and Financing IT, GEE Publishing, London.

• Block E.B. (1971): Accomplishment/cost – better project control, Harvard Business Review, May-June, 1971.

• Buchanan D. and D. Boddy (1992): The Expertise of the Change Agent: Public Performance and Backstage Activity, Prentice Hall, London.

• de Cervantes M,  Don Quixote, pt. 1, bk. 4, ch. 7 (1605; tr. by P. Motteux) cited In The Columbia Dictionary of Quotations 1995, Columbia University Press, NY.

• Computer Weekly, Does the IT Department have a future? July 25, 1991.

• Cooper, R. &  & Kaplan, R. S.  (1988): ‘Measure Costs Right: Make the Right Decisions’, Harvard Business Review, September-October, 96-103.

• Drury, C.(1994): Costing: An Introduction – 3^rd Edition, Chapman & Hall, London.

• Drury, C. (1997): Accounting for Management Decisions, International Thomson Business Press, London.

• Dudman, J. (1999): Counting the Cost of Networks, Computing, 1 July 1999.

• Chalcraft, C. (1997): The justification and costing of information systems, Pitman, London.

• Gartner Group (1997): Gartner Group's 1997 PC/LAN TCO Model -- The Basics, Gartner Analytics, December 19, 1997.

• Gurbaxani, V. and S. Whang (1991): The Impact of Information Systems on organisations and markets, Communications of the ACM, 43,1, 59-73.

• Hochstrasser, B. (1987):  Does Information Technology Slow You Down?, Kobler Unit, Imperial College, London.

• Hochstrasser, B. (1994): ‘Justifying IT Investments’ in: Leslie Willcocks (editor),

• Information Management, The evaluation of information systems investments, Chapman Hall, London, 151-170.

• Hogbin, G. and D.V. Thomas (1994): Investing in Information Technology: Managing the Decision-Making Process, McGraw-Hill, Maidenhead, England.

• Horngren, C. T., Foster, G. & Datar, S. M. (1997): Cost Accounting – A Managerial Emphasis, Prentice Hall International, New Jersey.

• Kaplan, R. (1986), Must CIM be Justified on Faith Alone?, Harvard Business Review, March-April 1986, pp 87-95.

• Field L. and M. Keller (1998), Project Management, The Open University Press/ International Thompson Business Press, London.

• Khalifa G Irani Z and  Baldwin L, Factors Impacting IT/IS Evaluation:
  Linking Investment Barriers To Life Cycle, ECITS 99, Brunel University
• Keen, Peter G.W. (1991): Shaping the future : business design through information technology, Harvard Business School Press,  Boston, Mass.
• Kerzner, H. (1995): Project management : a systems approach to planning, scheduling, and controlling (5th edition), Van Nostrand Reinhold, New York. – London.

• Koford, K and M. Penno (1992): ‘Accounting, Principal-Agent Theory and Self Interested Behaviour’, in Bowie, N, and R. Freeman (Editors), Ethics and agency theory : an introduction, Oxford University Press, New York.

• Lacity, M.C. & Hirschheim, R. Information System Outsourcing. p 190. John Wiley & Sons.Chichester.1995

• Lukka, K. (1988): Budgetary Biasing in Organisations: Theoretical Frameworks and Empirical Evidence; Accounting, Organisations and Society, 13 (3).

• PC Week (1997): TCO Fact or Illusion?, PC Week, 14,27 June 1997.

• Leach, K. and D. Smallen (1998):, What Do Information Technology Support Services Really Cost?, Cause/Effect, 21, 2, 1998, pp38-45.

• LAMSAC (Local Authorities Management Services and Computer Committee Computer Panel) (1978): Costing the use of computers in local government : report of  the Computer Panel of LAMSAC, Local Authorities Management Services and Computer Committee, London.

• Lock, D. (1996): Project Management – 6^th Edition, Gower, Aldershot, UK.

• Lockwood, D and M.G. Sobol (1989): IS spending survey: Communications technologies dominate growth areas, Journal of Systems Management, December , pp 31-37.

• Morgan, G. (1991): The Bias Factor, Certified Accountant, November 1991

• Parker, M. and R. Benson (1988): Information Economics: Linking Business Performance to Information Technology, Prentice-Hall, Englewood Cliffs, New Jersey.

• Price Waterhouse (1990): Information Technology Review (1990/91), Price Waterhouse , London.

• Price Waterhouse (1994): International Technology Review (1993/94), Price Waterhouse ,

• Menlo Park, California.

• Price Waterhouse (1996): Technology Review (1995/96), Price Waterhouse , London.

• Puxty, A. (1985): ‘A Critical View of Agency Theory’ in:  Issues in Accountability XII: Critiques of Agency Theory in Accountancy,  Puxty A. (Editor), Strathclyde Convergencies, Glasgow.

• Sing, S.K. "Using IT Effectively", Information and Management, Vol 24, 1993.

• Strassmann, P. (1988):  Productivity, Computerworld Extra, June 20^th, pp10-13.

• Strassmann, P. (1990):  The Business Value of Computers, The Information Economics Press, New Canaan, Connecticut.

• Strassmann, P. (1997):  The Squandered Computer, Information Economics Press, New Canaan, Connecticut.

• Snyder, H. and E. Davenport (1997): Costing and pricing in the digital age : a practical guide for information services, Library Association Publishing, London.

• Wellman, F. (1992): Software costing : an objective approach to estimating and
  controlling the cost of computer software, Prentice Hall, New York.

• Wentworth Research (1996), Assessing IS Performance, Research Report, Wentworth Management Programme, November 1996.

• Willcocks, Leslie and Stephanie Lester (1994): “Evaluating the feasibility of Information Systems Investments: Recent UK Evidence and New Approaches”, in: Leslie Willcocks (editor), Information Management, The evaluation of information systems investments, Chapman Hall, London, 49-80.

EJISE is published by Academic Conferences Limited
Curtis Farm, Kidmore End, Nr Reading RG4 9AY, England
Tel: +44 (0)1189 724148, Fax: +44 (0)1189 724691, Email: info@ejise.com

Send mail to info@academic-conferences.com with questions or comments about this web site.
Copyright © 2002-2006 Electronic Journal of Information Systems Evaluation
Last modified: November 25, 2004
ISSN 1566-6379