Evaluation of the indirect costs associated with contractual change
the influence curve approach
A key area of conflict, historically, has been the evaluation of the costs associated with contractual change and the potential for conflict transcends procurement route, this paper proposes a mechanism for the evaluation of compensation due in the event of a variation order under contracts for construction work, examines the theoretical basis for the approach, and presents a case study in which it has been successfully applied. It demonstrates how potential conflict, relating to the indirect costs of a variation, can be avoided by the use of Influence Curves. It has commonly been accepted that such costs are impossible to evaluate systematically and hence the parties to the contract have been left to argue over the cost and time effects of a variation and the compensation due, sometimes negating the trust that has been built up as a consequence of relational working. The Influence curves could be a basis for pre-contract negotiations and setting up partnering structures. The paper concludes by exploring the potential for application of influence curves in other areas of project management.
Background and Context
Worldwide, efforts are being directed towards steering organizations in construction to recognize the need to work together collaboratively rather than waste resources competing as adversaries. For example, in the UK, benchmark reports such as Rethinking Construction, Modernising Construction, Constructing the Team and Accelerating Change proposed major reforms for the construction industry to change the way members of the supply chain do business (Holti, Nicolini, Smalley & Raynsford, 2000). Encouragement to adopt the principles of integrated teams, partnering, incentivisation, and supply chains that were traditionally applied in the manufacturing industry were advocated. With regard to dispute resolution, Latham (1994) stated that the best solution is to avoid disputes in the first place by improving procurement and tendering procedures to reduce conflict, and adopting contract documents that promote teamwork and partnership to solve problems. Nonetheless, some disputes still arise over fair compensation and the need for a mechanism in place to deal with any such issues is necessary. On the same theme, Rethinking Construction (1998) led by Sir John Egan emphasised the scope for improving quality and efficiency in UK construction. The report also underscored the need for the industry to replace competitive tendering with long term relationships based on clear measurement of performance and sustained improvements in quality and efficiency. Modernising Construction (2001) and Accelerating Change (2002) highlighted good practices that have been adopted by various government departments, and demonstrated that significant efficiency gains can be achieved by changing the traditional approach to the procurement and management of construction. During the period since those major reports more benchmark reports and best practice initiatives have emerged leading to more collaborative forms of procurement with a focus on accelerating change.
The emphasis on moving away from purely price based contracts to more relational issues has certainly reduced the level of adversity in industry and should in principle create a culture that has relatively fewer claims but, none the less the nature of the industry means that changes will occur and compensation will have to be valued. Even though conditions of contract such as the Engineering and Construction Contract (ECC) acknowledge the need to provide for compensation that the contractor is entitled to in case of contractual change and the aim of the compensation event assessment is to agree the whole cost and time implications that the event has caused a systematic way of ascertaining indirect costs is not presented (Gerard, 2005). In a survey of the contractor-subcontractor relationships, Greenwood (2001) concludes that despite the trends towards change in industry and contractors’ professed interest in closer buyer-supplier relationships, these remain traditional, arm-length and cost driven from the outset; despite the code of practice for selection of subcontractors that resulted from recommendations by Latham (Greenwood, 2001).
Notwithstanding the aforementioned and many similar reforms, many industrial and public construction projects world-wide still adopt ‘admeasurement’ contracts in which the contract price is accumulated from competitively-tendered unit rates (Institution of Civil Engineers, 1996; World Bank, 1995; Langford & Rowland, 1995; Smith & Wearne, 1993). The price finally paid is based on remeasurement of the actual quantities of work completed (Merna & Bower, 1997, Bower 2003). In the event of a change the direct costs can be accounted for but compensation for indirect costs is difficult to quantify. The traditional method of settling claims for indirect costs is a ‘horse trade’ in which one party; normally the contractor, proposes a level of compensation for the variation, and then the other argues for adjustment of that amount. No attempt is made to quantify and cost all of the indirect effects. Such an approach is not systematic and has the potential to further arguments between project participants so it has commonly been accepted that such costs are very difficult to evaluate systematically (Love 2002).
In 1993 Thomas and Napolitan (1993) stated that “there have been no definitive studies reporting in quantitative terms the impact of changes”. Love (2002a) reiterates this view while drawing attention to the importance of being able to have a systematic approach to its evaluation (Love 2002). Much attention has been given to payment mechanisms that allow for the speedy and fair evaluation of payments due to a contractor, but the development of such systems, except in the case of cost-reimbursable contracts, has not addressed the problem of the evaluation of indirect costs associated with change. The effects of this are that the contractor has no certainty as to the outcome of the negotiations and hence has to allow high contingencies against the outcome. This causes contention between the parties as the contractor is constantly pushing the client to settle the claim for additional costs while invariably feeling that the reimbursement has been insufficient. This can be very damaging to relationships between all parties’ representatives especially in this era were the focus is on collaborative working.
For private ‘one-off’ clients who do not regularly do business with the construction industry and who mostly adopt the traditional procurement routes, a straight forward and fair method for determining indirect costs would go a long way to reduce conflict and ensure project success. For large regular procuring clients who tend to use collaborative procurement routes and PFI type projects, such a method would reinforce the drive towards the desired reforms.
The Case for the Influence Curve Approach
Variations are accepted as inevitable in much construction, as a project is usually only a means to an evolving end (Akinsola, 1997; Merna & Bower, 1997). Research undertaken by the authors, funded in conjunction with leading civil engineering clients, contractors and consultants has demonstrated that additional costs due to the direct effects of a variation, such as a change in resource requirements are relatively easy to estimate but it is often difficult to evaluate the indirect effects of delay and disruption (usually accompanied by a high degree of wasted effort). The indirect effects which are difficult to quantify may include:
- Rework and lost effort on work already done.
- Time lost in stopping and restarting current tasks in order to make the variation.
- Change in cash flow, financing costs, loss of earnings, etc.
- Loss of productivity due to reprogramming, loss of rhythm, unbalanced gangs and acceleration.
- Revisions to project reports and documents.
- Loss of float, therefore increased sensitivity to delay and increased cost.
A systematic approach to the assessment of indirect costs associated with a change needs to take account of those costs which arise due to a number of tasks being performed at the same time, whether they are logically linked or not, and the effects of ‘ripples’ through the programme due to logic links. These could be thought of as vertical and horizontal relationships within a bar chart programme, vertical links taking account of the control needed for a number of tasks to be live at any one time, and horizontal links being the more traditional logic links.
Thomas and Napolitan (1993) described a Factor Model which provides a model for understanding and quantifying the effects of change on labour productivity. The factor model does not take into account the cumulative or ripple effect that occurs when project conditions have deteriorated to the point where work on a task is adversely affected by another task or by the mere nature of the site environment. Another approach that has been taken in the assessment of these costs has been developed empirically by Fluor Daniel Ltd for use in process plant contracts (private communication). It is known as IMPACT and is applied to the estimation of additional direct costs to the client due to the variation. Some research has been undertaken in the USA that aims to quantify the effects of change on labour productivity, for example by Leonard (1987) and by Zink (1990), but loss of productivity is only one of the effects of a variation and in itself would not allow evaluation of the full compensation due to the contractor. Hanna, Latfalla and Lee propose the statistical-fuzzy approach to quantify cumulative impact of change orders (2002). However, the statistical-fuzzy approach adopted is limited by the difficulty of determining the right set of rules and membership functions, and the complexity of the variables restricts it to a very academic level and would make it difficult to adopt by industry.
The approach of establishing a standard ‘Influence Curve’ from which a factor could be sought in the event of a variation was believed by the authors to be a systematic, rapid and equitable technique for indicating the indirect costs. In the research, curves were developed recognising that different types of work may require different curves. Initially, research concentrated on producing a curve for road building. A numerical rather than an empirical approach was sought by which the curve could be derived. It was important that the technique for curve development was simple enough to be used by clients to derive their own curves for unique projects and could prove an excellent tool for precontract negotiations, would provide a reasonable level of certainty in case of disputes and could be used as a tool for project partnering.
Derivation of the Influence Curve
In terms of Influence Curves cost is the cost to the client, therefore, whenever possible rates tendered by the contractor should be used to assess the direct cost. If actual costs have to be used, an addition for profit should be made after the Influence adjustment, according to the terms of the contract. The Influence Curve is then used to assess the indirect part of the compensation when there is a variation. For civil engineering it seemed sensible to define a direct cost as the change in cost associated with tasks whose resource usage is changed due to the variation, that is, tasks whose duration is increased due to the variation. These affected tasks are apparent without recalculation of the network as long as the programme is clear. In cases where a subcontractor is already on site and it is not possible for him to perform other work, an allowance for standing time must be included as a direct cost. In effect the subcontractor will have to be paid time related charges as though he was working, therefore constituting an extension to the duration of the task.
Two approaches have been taken in the derivation of an Influence Curve. They have both developed from the philosophy that there are two major influences on the shape of the curve: The more live tasks at any given time, the greater the effect of a change; and the less the scope available for reworking the programme, the greater the effect of a change. For the first approach a change is introduced to a project at various points in time and then the direct costs associated with that change are assessed. The direct costs are then factored by the number of live tasks at that point in time (to take account of the ripple effect) and summed for the project. For the second approach the level of resource was assessed at various points in time then combined with the rigidity of the programme (rigidity being the lack of float). The derivation of these curves is detailed below. It should be noted that once a satisfactory curve shape has been obtained, it is used as a standard and an Influence Curve need only be derived from first principles for projects that are out of the ordinary as already stated. The simplicity of the technique means that if a client was unsure whether his project suited the criteria for a particular standard curve it would be relatively easy to derive a unique curve. A workshop for preparing for partnering on a project would provide the opportunity to agree the value of Influence Factors appropriate to that project if the parties felt that they needed review. Simulation of the effects of risks identified at that workshop would indicate the compensation for indirect costs before an event, so allowing client and contractor to assure themselves that the results would be equitable.
Define intervals at which analysis will take place. (Shorter time periods are required from 20 to 65 % of project completion time.)
At a given time examine the programme and note all of the tasks that are live. Determine how many logic links each of those tasks has and how many of the live tasks are in the same category. The number of categories is determined according to the phasing of the work.
Apply the disruption to the task with the most logic links. In the first instance a delay of five days was used, that is the duration of the task with the most logic links was extended by five days. (In instances where there is more than one task with a large number of logic links apply the delay to each in turn and note the largest change in project budget.)
The change in cost is then translated into a percentage change from the original budget which was then factored by the number of other tasks in the same category that were live at the time of the disruption (task multiplier). This gave a weighted percentage change taking into account the vertical links or ripple effect. The cumulative percentage change was plotted against time to achieve the effect of the Influence Curve increasing with time. An Influence Curve is shown in Exhibit 1. The shape of the curve was verified using Approach B.
Count maximum number of days float at day one then at specified time intervals count the number of days of float remaining. This is calculated by assuming that tasks prior to the assessment date have actually occurred as late as possible without affecting the project end date and counting how many days float are left.
For the derivation of an Influence Curve it is the inverse of this that is of interest, or the rigidity. This is calculated as follows: if the original number of days float was F, the rigidity is F-x, where x=F for the first period and x=0 for the final period. At the same points in time as the float is assessed the level of resource must also be noted so the number of labour days being worked on the site was calculated.
The rigidity and labour days were summed to give an Influence Curve. This curve was almost identical in shape to the curve derived using Approach A.
Although the scope of the research was limited in terms of the number of projects that could be evaluated it was encouraging to find that both approaches resulted in a curve of the shape that the steering group had anticipated. On a typical project, the actual Influence Factors used may be amended during negotiations if either the client or the contractor feels that the scale is inappropriate. Once the Influence Factors have been agreed, however, for a project then they should be fixed. This does not dilute the effectiveness of the approach as the outcome would still be the same, that is, there would be a systematic approach to the evaluation of variations.
Once an Influence Curve has been plotted it is used in the event of a variation by multiplying the direct costs associated with the variation by the Influence factor and adding this amount to the direct cost to give the total cost of the variation to the client. For example, consider the influence curve in Exhibit 1 above. Suppose a variation whose direct costs (plant, labour, equipment) are £1,000 is issued. If this is introduced when the project is 50% complete in terms of cost, the influence factor would be 0.4; the reimbursable value of this change would be £1,000 x 1.4 or £1,400.
Further investigation was carried out to support the approaches taken. Study of the cumulative cost curves showed an envelope if the tasks are scheduled to take place as soon as possible and as late as possible. This envelope is at its broadest when the project is two thirds complete and illustrates a number of points: the wider the band, the more float is available for rescheduling the tasks at any given point in time; and the narrower the envelope from project to project, the less scope there is for rescheduling tasks as the project progresses, consistent with the findings of Hajarat and Smith (1992). The implications for Influence factors are that as the project progresses there is a greater scope for the manoeuvre of individual tasks until a point approximately two thirds of the way through the project, but as many tasks are going on at any one time the effect of any change is large.
The analysis of the cumulative cost envelope also led to an investigation of the availability of float as the project progressed, both total and free float. It was found that the number of days available reduced to approximately 5% of the original amount with almost 2 of the 9 months of the project still to run, reinforcing the theory that although there are fewer tasks that may be disrupted, the scope for reducing the effect of change through use of float decreases.
It was also noted that the amount of control required increases steadily until the project is about 60% complete and then, as the number of tasks decreases, the amount of control required falls off. In Approach A the number of live tasks at any one point in time was measured to give an indication of this effect. The alternative approach taken here was to measure the number of men working at any one time. This required summation of the number of man-hours being worked at any given time through the project; machine operators were included in order to judge the full scale of the task.
The project used to briefly illustrate the use of the Influence Curve consisted of the construction of two kilometres of a single carriageway rural road in the UK. The contract price was £2,093,124.The example illustrates the evaluation of the indirect costs associated with the change using the Influence Curve shown in Exhibit 1.
The first change to be considered is a disruption to the disposal of unsuitable excavated material. The scheduled accumulated price by the time of the change was £370,067, or 17% of the total. Using the curve it can be seen that an Influence Factor of 0.2 is appropriate and so a multiplier of 1.2 should be applied to the direct costs to give the total compensation due to the contractor. This figure may seem low as compensation for the disruptive effects but the project is still at a relatively early stage of its development and so there is time for economic re-planning. At this time the number of other live activities is relatively low so the ripple effect will be minimal.
The direct charges associated with this variation are:
|1 wagon for 55 days||= £14850|
|2 days extra work from excavation plant||= £5011|
|Total direct cost||= £19861|
|Total amount of compensation due to the contractor||= £19861 x 1.02 = £20258|
The direct costs could be determined from a number of sources depending on the conditions of contract.
The second change relates to the reworking of the surface levels of a stretch of the road. The kerbing subcontractor was already on the site when the change was introduced, therefore a charge for standing time was incurred. As the kerbing was delayed, the surfacing was disrupted. The surfacing subcontractor was already on site, so again standing time has to be paid. When the variation was notified the project was 72% complete in terms of price, giving an Influence Factor of 0.7, this reflects the late stage at which the change was introduced. In this example there were four classes of work which needed price adjustment. These are outlined below, with an opinion as to whether or not they constitute direct charges or are included in the Influence Factor.
|Sub-base - direct charge (involves actual work being done due to variation)||= £1132|
|Kerbs - direct charge (direct delay due to variation)||= £4823|
|Surfacing - direct charge (systematic evaluation is possible)||= £14207|
|Overheads - indirect charge - Influence Factor accounts for this.|
|Total direct costs||= £20162|
|Total amount of compensation due to the contractor||= £20162 x 1.7|
Potential for Further Application
Further development of this innovative and novel approach would be an excellent opportunity to engage the sector in dialogue towards overcoming a challenge that has faced the construction industry for a long time. The adoption of the Influence Curve approach could prove an excellent tool for precontract negotiations between client-contractor, contractor-subcontractor, and contractor-supplier. Such a systematic approach would benefit clients because the certainty of fair compensation in case of contractual change would motivate contractors to make reasonable contingencies based on the assurance that a fair method of determining indirect costs irrespective of the nature of variation exists. Furthermore all supply chain members would benefit from the level of certainty associated with the influence curve approach. SMEs (mainly subcontractors and specialist suppliers) who form a large proportion of the construction industry would have an opportunity to make useful input in this critical area. The influence curve approach may also be used as a tool for project partnering. Building the parameters for the influence curve fosters confidence and trust (pillars of collaborative procurement) between supply chain members. The approach could also prove useful in exploring other areas to improve cost management and pricing mechanisms in construction, thereby providing a basis for further research and a source of reliable data in this important aspect of construction projects.
Existing methods of evaluating payment due for variations cause disputes and conflict on site and between all parties’ representatives. This conflict needs to be removed. If the reputation of the construction industry is to be enhanced, encouraging investment, then the parties to the contract must have aligned goals. A goal that the parties share is to optimise their profits or improve their business performance. For the client this may mean working to a tight budget or schedule to meet constraints external to the project. To a contractor it may mean achieving an acceptable profit margin, enhanced by reduced financing costs and a secure flow of cash into his organisation.
The Influence Curve method can remove one cause of problems. It increases clients’ certainty as to the final cost of a project. It increases the contractor's certainty as to the compensation he will be paid in the event of a change to the work. This general reduction in the financial risks to both parties should lead to lower bid prices. The rapid evaluation of the payment due to the contractor would decrease the management effort required in the measurement and valuation process, reducing project overheads. This systematic technique should help the industry achieve the team working advocated for in major reports and supported by many in the construction industry.
Akinsola, A. O., et al.. (1997) Identification and evaluation of factors influencing variations on building projects. International Journal of Project Management, 15 (4), 263-267.
Bower, D. A. (2003) Management of Procurement, London: Thomas Telford.
Construction Industry Board. (1997) Partnering in the Team, report WG12, Thomas Telford, London.
Egan, Sir J. (1998) Rethinking Construction, Report, Department of the Environment.
Gerard, R. (2005). “Relational Contracts - NEC in Perspective.” Lean Construction Journal 2: 80-86.
Greenwood, D. (2001). “Sub-contract Procurement: Are Relationships Changing.” Construction Management and Economics Journal 19: 5-7.
Hanna, A. S., Lotfallah, W. B., & Lee, M. (2002). “Statistical-Fuzzy Approach to Quantify Cumulative Impact of Change Orders.” Journal of Computing in Civil Engineering, ASCE 16(4): 252-258.
Hajarat, D. A. & Smith, N. J. (1992) Exposure Envelopes: An Assessment of the Exposure to Time and Cost Overruns during Construction Projects, University of Sheffield.
Holti, R., Nicolini, D., Smalley, M., Raynsford, N. (2000) The Handbook of Supply Chain Management – The Essentials, London CIRIA
Institution of Civil Engineers. (1996) Civil Engineering Procedure, 5th edition, Thomas Telford, London.
Langford, D. A. & Rowland, V. R. (1995) Managing Overseas Construction Contracting, Thomas Telford, London.
Latham, M. (1994) Constructing the Team. HMSO, London.
Leonard, C. A. (1987) The Effect of Change Orders on Productivity, The Revay Report, 6, Montreal, Canada.
Love, P. E. D. (2002). “Auditing the indirect consequences of rework in construction: a case based approach.” Managerial Auditing Journal 17(3).
Love, P. E. D. (2002). “Influence of Project Type and Procurement Method on Rework Costs in Building Construction Projects.” Journal of Construction Engineering and Management, ASCE 128(1).
Merna, A., & Bower, D. A. (1997) Dispute Resolution in Construction and Infrastructure Projects. Asia Law & Practice.
National Audit Office (2001) Modernising Construction. Stationary Office. London.
Smith, N. J., & Wearne, S. H. (1993) Construction Contract Arrangements in EU Countries, European Construction Institute, Loughborough.
Strategic Forum for Construction. (2002) Accelerating Change Rethinking Construction Ltd. London.
Thomas H. R. & Napolitan C. (1993) Anatomy of a Construction Change, Proceedings of CIB-W65 Conference, Trinidad, West Indies, September, 1195-1203.
World Bank. (1995) Procurement of Works, The World Bank, Washington.
Zink, D. A. (1990) Impacts and construction efficiency, Cost Engineering, November, 32, 11.
© 2006 Bower & Aritua
Originally published as part of PMI Global Congress Proceedings – Madrid, Spain