Cite

Introduction

Large infrastructure projects last and deliver value to society for decades or more (Lehtinen et al. 2019). Such projects include, for instance, hospitals, tunnels, sports arenas, nuclear power plants, wind power farms, roads, railways, and bridges. A properly planned and executed project has potentially multiple positive societal effects, which sustain over a long period of time. On the other hand, poorly designed and built projects tend to have higher operational costs and result in numerous negative sustainable consequences (Liu et al. 2013). In addition, poorly planned investments are more likely to cause significant cost and time overruns, often because of conflicting stakeholder goals, interests, and time perspectives (Qiu et al. 2019). The latter is often highlighted in the public media, but less studied in research. A recent overview of empirical studies on larger infrastructure projects states that many planning issues arise from conflicting goals (Themsen 2019; Flyvbjerg et al. 2002). Thus, stakeholders have conflicting views on how to address and integrate investment goals with operational goals (Qiu et al. 2019).

The issues mentioned above receive frequent attention in the public media in Sweden. For example, a hospital project in Stockholm, “Nya Karolinska,” has resulted in at least 110 major change orders and high cost and time overruns. Each change order accounts for a fee of an additional €600,000, causing a total cost-overrun of €66 million. While the hospital was expected to be ready in 2018, the procurement of operations and maintenance will continue until 2040 (Gustafsson 2016). Another known Swedish example of cost and time overruns is the construction of the national sports arena (Friends Arena), in which the original budgeted cost of the arena accounted for €200 million (Swedbank 2013), but the final cost was €300 million (idrottensaffären.se) – a cost overrun of 47% in local currency.

Cost overruns occur worldwide and to varying degrees in different projects (Cantarelli et al. 2012; Shehu et al. 2014; Park and Papadopoulou 2012). In a study of 258 infrastructure projects, over 70 years, and across different geographical regions, Flyvbjerg et al. (2002) found that almost 90% of the projects exceed the budgeted costs, on average by 28%. In another study, 50% of the construction projects in Asia were found to result in cost overruns (Shehu et al. 2014). There is no indication that this would be less in the rest of the world. On average, the Netherlands has the lowest percentage (11%) of cost overruns, while the north-western European countries show a significantly higher figure (27%), and the remaining European countries exceed the budgeted costs by an even higher percentage equivalent (44%) (Cantarelli et al. 2010; Cantarelli et al. 2012). Cost overruns for infrastructure projects in Asia range from 2% to 98% (Park and Papadopoulou 2012). In Norway, the average cost overrun is 8% for infrastructure projects and it varies between −59% and +183% (Odeck 2004). In absolute terms, this corresponds to around €50 million (Odeck 2004). Cost overruns are directly linked to the project size (Makovšek et al. 2012). A known project failure from a time and cost perspective is the 50 km long tunnel under the English Channel. The planned budget was massively exceeded, from £2,600 million to £4,650 million, based on the monetary value of 1985. This corresponds to a cost overrun of 80% of the projected budget (Flyvbjerg et al. 2003). Yet, cost overruns are more common in smaller projects than in larger ones (Odeck 2004). The 8.7 km long Swedish Hallandsås tunnel was finished in 2015, thereby taking 23 years instead of the planned 3 years. The cost of the project was ten times more than the estimated budget, as costs soared from €100 million to around €1.1 billion (Littorin 2015). However, there are no clear explanations about why small projects have more cost overruns than larger ones. In Sweden, cost overruns are expected to correspond to 60–70% of the initial price in construction projects. Time and place are known factors for cost overruns (Odeck 2004). In addition, procurement that tends to be inconsistent with the actual outcome is more likely to exceed the cost budget. Thus, time and cost are associated.

While the extant literature has a clear view of time and cost overruns being common and significant in large projects, less is known about the drivers of cost and time overruns (Zhang et al. 2019). Difficulties in time and cost overruns remain unsolved (Themsen 2019), and the effects of poorly planned projects are still present. Therefore, we aim to contribute to this understanding and suggest a testable model of where and why (i.e., drivers) cost and time overruns take place. This aim assumes that when uncertainty in the agreements exists, it also leaves room for interpretation of change orders. A perfectly designed and planned agreement gives little room for change orders. A contract’s performance is thus a function of the room a contract leaves for additional work, i.e., change orders, in addition to the critical components of time, price, and quality.

To address the gap in contracts, but also cooperation in broader terms, we examine the client and contractor perspectives in infrastructure projects in Sweden. We examine contractual and leadership issues related to change orders. More specifically, we look for attitudes and behavior that can influence the drivers of change orders, and we compare the client and contractor perspectives on this.

Theory

A construction project is a complex process that follows several loops (Senescu et al. 2013). Theoretically, this paper is based on project life-cycle loops (Tsai et al. 2011). A project is assumed to follow several loops starting with planning and construction, and moving on to a post-process involving maintenance and learning. During these various loops, project managers meet to solve complex issues within and between the client and the contractor.

Earlier research shows that many different factors can explain project cost overruns (Segelod 2017). Various interests and perspectives may, to some extent, explain why projects cause overruns (Ruoslahti 2018). Besides the unavoidable risk and uncertainty, cost and time overruns can be vastly influenced by contractual difficulties, lack of communication, and even the opportunistic behavior of one or both parts. These factors are investigated in this paper, in relation to drivers of change orders, i.e., alterations causing time and cost addition.

Change orders

The negative sides of change orders in construction projects are often recognized in the current literature (Khanzadi et al. 2018). Change orders conceptually refer to a bargaining process in which clients (i.e., owners) and contractors (i.e., builders) meet to adjust the original scope of construction of the project after signing the contract (Ahmed et al. 2016). Such a process can be lengthy and involve substantial additions to the original plan that the client expects (Keane et al. 2010) or on-site conditional (Wu et al. 2005). As some of the change orders impact delays in time, and subsequently, cost overruns, they may also cause vivid discussions about whether these were in the contract implicitly and to what extent the client or the contractor caused the additional changes due to shortcomings in fulfilling their part of the contract. These processes of bargaining are frequently cited as a relevant project management issue in the construction industry (Johnson and Babu 2020). Following Pesämaa et al. (2018), we define interactions that occur due to bargaining as tensions. According to Pesämaa et al. (2018), these tensions imply a lack of agreed terms and that the negotiating part emphasizes their own interest. Furthermore, we assume that clients and contractors may have different expectations, which further affect drivers to change orders differently (Pesämaa et al. 2018).

Leadership drivers to change orders

Ghoddousi and Hosseini (2012) argue that insufficient quality of leadership is the key cause of non-productivity in construction projects. Adequate leadership should recognize quality issues and identify overall production issues (Durdyev and Mbachu 2018). According to Pesämaa et al. (2018), leadership issues also cause tensions that directly depend on the counterpart’s lack in leadership skills. Such lack of leadership caused by non-productive and meaningful interaction causes productivity issues and, subsequently, drivers to change orders. We argue that many of the leadership issues directly refer to issues related to weak interaction, which is defined as the effects of opportunism (Wathne and Heide 2000), competence cost, and bargaining costs (Dahlstrom and Nygaard (1999). We, therefore, hypothesize as follows:

H1a: Opportunism is related to leadership drivers to change orders.

H2a: Competence cost is related to leadership drivers to change orders.

H3a: Bargaining cost is related to leadership drivers to change orders.

Contractual drivers to change orders

Kamminga (2015) frames contractual issues as the main issue for project management. Lack of specified contracts in agreement among negotiating partners hamper the overall performance (Heravi and Charkhakan 2014). Many of the drivers to change orders are directly related to the perception that the counterpart lacks contractual skills (Ahmed et al. 2016). Such an implicit assumption not only causes tensions, but in the long run also issues of agreed contracts (Pesämaa et al. 2018). Contractual drivers to change orders thus stem from poor interaction between negotiating partners. We argue that many of the contracting issues can be deduced to the effects of opportunism (Wathne and Heide 2000), competence cost, and bargaining costs (Dahlstrom and Nygaard (1999). We, therefore, hypothesize as follows:

H1b: Opportunism is related to contractual drivers to change orders.

H2b: Competence cost is related to contractual drivers to change orders.

H3b: Bargaining cost is related to contractual drivers to change orders.

Research design

The method for this research paper is designed to explain and describe the roles of change orders. We conducted a survey in 2017 in which both clients and contractors offered their perception of change orders. The survey was sent to client project managers that prepare proposals for infrastructure projects in Sweden and contractor project managers with previous experience from infrastructure projects. The respondents, on average, have 21.9 years (SD 11.4) of work experience.

The measurements are included in full in the Appendix. Two dependent variables are used: leadership drivers to change orders, i.e., factors related to project organization and management, and contractual drivers to change orders, i.e., factors related to the specification, planning, and agreement of the project. Explanatory factors are Opportunism, i.e., action for own short-term gains, Bargaining cost, i.e., the effort of reaching agreements, and Communication cost, i.e., the effort of understanding each other. The study highlights the potential differences between the client and contractor perspectives. Each respondent is identified as being either a client or a contractor, and the analysis can thus distinguish between the two perspectives.

Results

A total of 234 complete responses were received (see Table 1). The responses were distributed among 95 clients and 139 contractors.

Mean differences between clients and contractors

Question Role N Mean SD F-value p-value
OPP1: The counterpart presents misleading information to protect their interests. Client 95 3.58 1.05
Contractor 139 3.66 0.93 0.41 0.525
Total 234 3.63 0.98

OPP2: The counterpart promises to do things without actually doing them later. Client 95 3.55 1.01
Contractor 139 3.69 0.87 1.35 0.247
Total 234 3.63 0.93

OPP3: The counterpart does not always act in accordance with our contract. Client 95 3.69 0.84
Contractor 139 3.85 0.76 2.13 0.146
Total 234 3.79 0.80

OPP4: The contractual partner sometimes breaks informal agreements to maximize their own benefit. Client 95 3.18 1.01
Contractor 139 3.24 1.00 0.19 0.663
Total 234 3.21 1.00

BC1: Negotiations of financial aspects of the contract are typically difficult and lengthy. Client 95 3.40 0.97
Contractor 139 3.45 0.93 0.13 0.715
Total 234 3.43 0.94

BC2: When unexpected changes arise, at least one party was dissatisfied with negotiated outcomes. Client 95 3.24 0.91
Contractor 139 3.24 0.91 0.00 0.969
Total 234 3.24 0.91

BC3: Our negotiations with this contracting partner are usually difficult. Client 95 3.34 0.96
Contractor 139 3.14 0.96 2.43 0.120
Total 234 3.22 0.97

BC4: Neither party is willing to lower their demands at their own cost. Client 95 3.44 0.73
Contractor 139 3.49 0.72 0.24 0.623
Total 234 3.47 0.72

CC1: Information from the contractual partner is poorly formulated and difficult to understand. Client 95 3.27 0.93
Contractor 139 3.15 0.92 1.00 0.318
Total 234 3.20 0.92

CC2: Important information from the contractual partner seldom comes at the right time. Client 95 3.38 0.99
Contractor 139 3.46 0.92 0.42 0.520
Total 234 3.43 0.95

CC3: Information from the contractual partner is either incomplete or too voluminous to understand. Client 95 3.27 0.99
Contractor 139 3.29 0.96 0.01 0.913
Total 234 3.28 0.97

LDCO1: The counterpart’s (lack of) interpersonal skills causes change orders. Client 95 3.13 1.08
Contractor 139 2.71 0.98 9.24 0.003
Total 234 2.88 1.04

LDCO2: The counterpart’s (lack of) leadership skills causes change orders. Client 95 3.18 1.03
Contractor 139 2.87 1.03 5.03 0.026
Total 234 3.00 1.04

LDCO3: The counterpart’s lack of competence causes change orders. Client 95 3.34 1.01
Contractor 139 3.49 1.14 1.11 0.293
Total 234 3.43 1.09

CDCO1: Imprecise projecting of on-site conditions causes change orders. Client 95 3.23 1.26
Contractor 139 4.20 0.91 46.78 0.000
Total 234 3.81 1.17

CDCO2: Incomplete early-stage planning causes change orders. Client 95 3.12 1.14
Contractor 139 3.63 0.97 13.50 0.000
Total 234 3.42 1.07

CDCO3: Incomplete contracts cause change orders. Client 95 2.97 1.13
Contractor 139 3.99 0.87 60.82 0.000
Total 234 3.58 1.11

CDCO4: Poor follow-up of similar projects causes change orders. Client 95 2.77 1.02
Contractor 139 3.51 0.99 0.000
Total 234 3.21 1.06

SD, standard deviation.

The decision criterion was that we look for differences between clients and contractors. If the p-value is <0.05, it is an indication that the difference in mean values is significant. For most items, there are only minor differences in the comparison of client and contractor perspectives. However, we did find differences in the items related to the drivers of change orders. The differences were prevalent for drivers related to leadership as well as contractual issues. For attitudes and behaviors, there were no significant differences. This indicates that clients and contractors, in general, do not have different views of the investment projects and the interaction therein. Yet, they do have different views on the causes of change orders. This could indicate a tendency to point fingers at the other part, and it inevitably means that the difference in perspective is fundamental for the understanding of what drives change orders and causes project time and cost overruns.

The second test was that we calculated five composite scores (i.e., summated means). These are opportunism (four variables), bargaining cost (four variables), communication cost (three variables), leadership drivers to change orders (three variables), and contractual drivers to change orders (four variables). We did a Cronbach’s alpha test to check whether the proposed measures measured the same thing. Our Cronbach’s alpha test showed that all the proposed composite scores exceeded the recommended 0.70 level, and we could thus assume sufficient reliability of these measures. The subsequent tests are thus performed on the composite scores.

The correlation table indicates that several of the proposed measures are correlated (Table 2). We can thus assume that the two types of change orders that we wanted to explain can be modeled with linear regression.

Correlation between composite scores

Composite score Mean SD N (1) (2) (3) (4) (5)
1. Opportunism 3.57 0.73 234 0.79
2. Bargaining cost 3.34 0.67 234 0.39** 0.75
3. Communication cost 3.30 0.81 234 0.41** 0.48** 0.82
4. Leadership drivers to change orders 3.10 0.86 234 0.26** 0.02 0.22** 0.82
5. Contractual drivers to change orders 3.50 0.90 234 0.10 0.17** 0.16* −0.19** 0.83

p < 0.05,

p < 0.01. Diagonal values are Cronbach’s alpha.

SD denotes standard deviation.

We further assumed that leadership drivers to change orders depend on opportunism (H1a), bargaining cost (H2a), and communication cost (H3a). We notice that opportunism and communication costs are significant and robust for the whole sample, but only communication costs remain significant for clients (Table 3). We further see that both opportunism and communication costs are positive and significant for contractors, whereas bargaining cost is negatively related to leadership drivers to change orders. There are thus two different explanations for clients and contractors.

Regression on leadership drivers to change orders

Dependent variable: Leadership drivers to change orders

All Client Contractor
H1a: Opportunism 0.24*** 0.18 0.24***
H2a: Bargaining cost −0.17* −0.07 −0.17*
H3a: Communication cost 0.21** 0.31** 0.21**
R-square 10.40% 18.10% 11.10%
Adj. R-square 9.20% 15.40% 9.10%

p < 0.05;

p < 0.01;

p < 0.001.

Beta-values reported.

Subsequently, we argued that contractual drivers to change orders depend on opportunism (H1b), bargaining cost (H2b), and communication cost (H3b). Here we notice that the overall sample reports insignificant results for all the three hypotheses (Table 4). However, turning to the two separate samples, we see that clients’ explanations of contractual change orders mainly emphasize bargaining costs. Conversely, contractors’ explanations of contractual drivers mainly rely on communication costs. The explanations are thus different for the two groups and may also explain why the R-squares for both models tend to be relatively low. It further shows how different the two perspectives are, and the risk of conflicts emerging around change orders.

Regression on contractual drivers to change orders

Dependent variable: Contractual drivers to change orders

All Client Contractor
H1b: Opportunism 0.01 −0.09 0.04
H2b: Bargaining cost 0.12 0.37*** −0.07
H3b: Communication cost 0.09 −0.10 0.27**
R-square 4.00% 11.60% 6.40%
Adj. R-square 2.40% 8.70% 4.30%

p < 0.05;

p < 0.01;

p < 0.001.

Beta-values reported.

Discussion and implications

This paper is based on an issue of contractual change orders. Such change orders are defined as additional work and are intended to show who to blame when contracts result in time and cost overruns. The results show that there are significant differences between client and contractor perceptions of drivers to change orders. While earlier research has shown the impact of change orders (Zhang et al. 2019), this research is attempted to explain such change orders.

Change orders are ultimately a critical situation (Johnson et al. 2020) in which the negotiating partners meet to discuss specific and tangible unperformed agreed tasks. To this end, the literature posits that these issues stem from poor leadership (Ghoddousi and Hosseini 2012) or contractual issues (Kamminga 2015). Based on a framework to analyze tensions (Pesämaa et al. 2018), we deduced three major determinants that may affect drivers to change orders. We argue that opportunism (Rokkan, Heide, and Wathne 2003), competence cost, and bargaining costs (Dahlstrom and Nygaard 1999) affect both drivers to leadership and contractual change orders. Furthermore, and based on earlier research (Pesämaa et al. 2018), we assume that client and contractor expectations differ in construction projects. This study confirms client and contractor differences by testing these perspectives separately in a regression model.

Through analysis of the mean differences, we found that clients and contractors do not differ in the general assessment of opportunism, bargaining costs, and communication costs. There were, however, differences in both leadership and contractual drivers to change orders. When it comes to leadership drivers to change orders, we found that clients tend to perceive these to come from competence costs, while contractors rather see this as a result of opportunistic behavior in combination with competence costs. Yet, when analyzing contractual drivers to change orders, we found that clients tend to see this as an effect of bargaining costs, while contractors rather perceive this as a result of competence costs.

The significant differences that were found indicate that although clients and contractors may have a similar understanding of the attitudes and behavior in the project interaction, the drivers of change orders are seen very differently. This difference could, potentially, itself lead to miscommunication and misconceptions, leading to further time and cost overruns.

Conclusions

While the earlier literature has identified how much time and cost overruns various projects cause (Zhang et al. 2019), this paper has tested some tentative theoretical and empirical explanations to drivers of leadership – and contractual change orders. We postulated that these types of change orders are the effects of tensions that evolve from opportunism, bargaining costs, and competence costs. Furthermore, we found that depending on the type of explanation (i.e., leadership-driven or contractual-driven change order), clients and contractors have different explanations. We believe that these tentative results could inspire further research to explore the nuances of these negotiating processes in more detail. As clients and contractors have different roles, their expectations on leadership and contracts also differ. We suggest that such a mismatch of expectations may also result in extensive blaming, which is harmful to the project.

eISSN:
1847-6228
Idioma:
Inglés
Calendario de la edición:
Volume Open
Temas de la revista:
Engineering, Introductions and Overviews, other