Performance-based contracting in military supply chains and the willingness to bear risks

2.1 Performance-based contracting in the service supply chain

A service supply chain (SSC) consists of a network of suppliers, service providers, consumers and other supporting units. Parties in a SSC perform the functions of the transaction of resources required to produce service, the transformation of resources into services and the delivery of these services to customers (Baltacioglu et al., 2007). In the beginning of this century, it was noted that the traditional view of (manufacturing-oriented) supply chain management is not entirely applicable to SSC (Kathawala and Abdou, 2003; Sampson, 2000; Sampson and Spring, 2012b). The customer has an important role in the SSC covering different characters, amongst others: the supplier, design engineer, production manager, quality assurance manager and competitor (Sampson and Spring (2012a). This phenomenon is called: customer-supplier duality. The customer also has an important role in the specification of the expected performance (Oflaç et al., 2012). Kim et al. (2007) were among the first to study the use of PBC in an SSC. They analyzed two important issues of contracting in SSCs: the performance requirement allocation and risk sharing.

PBC is as a contractual approach of tying the service provider’s payment to specified performance, rather than merely paying for its activities and tasks (Kim et al., 2007). Characteristic of PBC is the emphasis on specification and evaluation of outcomes rather than required inputs, activities or processes of the supplier (Martin, 2007). At the core of PBC lies the contracting mechanism that the payments of buyers depend on the actual performance of suppliers. This results-oriented contracting method focuses on the outputs, quality or outcomes that are tied to (at least partially) suppliers’ payments (Selviaridis and Norrman, 2014; Liinamaa et al., 2016). Typical for PBC is the use of financial incentives and penalties which are linked to specified outputs or outcomes (Selviaridis and Wynstra, 2014).

Although PBC has received much attention in the last decade, it is not a new phenomenon. As the 1960s initiatives were taken with academic publications focusing on incentivizing the supplier to provide good performance for the customer. An often-mentioned example of PBC in the aerospace industry is the Roll Royce’s Power-by-the-Hour™ business model where the company is compensated for engine availability[1] (flight hours) rather than the cost of labor and spare parts (Neely, 2009; Ng et al., 2009; Selviaridis and Wynstra, 2014; Voss et al., 2002). Inherent inefficiencies and strained relations between the customer and service provider have driven a recent move away from the traditional sustainment contracts. PBS aims to improve this by linking the service provider’s profit to the amount of post-production support services they sell. The more parts break down, the more service or parts the service provider is able to sell. This return on sales model demonstrates a negativity associated with a transactional arrangement where the majority of the risk is absorbed by the customer (Sols et al., 2007). The customer ends up bearing the financial burden associated with the uncertainties in reliability. The aim of PBC is to align the customer’s and the service provider’s objectives. PBC transfers financial risks to suppliers (Selviaridis and Norrman, 2014) and can lead to improved service and reduced costs (Kim et al., 2007).

2.2 Principal-agency theory

To better understand the relationship between the service provider and customer, we take a look at the principal-agency theory (PAT). PAT analyzes problems where one party (the principal) delegates work to another party (the agent) to perform a certain task (Eisenhardt, 1989). In the case, which we will describe in the next section, the principle is the customer and the agent is the service provider. When the principal contracts an agent to perform a service on his behalf (and delegates some decision-making authority to the agent in the process) (Jensen and Meckling, 1976), the principal allows the agent’s actions to directly affect his own well-being (Toivonen and Toivonen, 2014). This addresses two specific problems that can occur in agency relationship (Eisenhardt, 1989). The first is the agency problem that arises when the desires or goals of the principal and agent conflict, and it is difficult or expensive for the principal to verify what the agent is actually doing. The latter is the problem of risk sharing that arises when the principal and agent have different attitudes toward risk. The problem is that the principal, and the agent may prefer different actions because of the different risk preferences. Eisenhardt (1989) studied goal misalignment in relation to outcome-based contracts. She concluded that with an outcome-based contract the agent is more likely to behave in the interests of the principal. For the customer, this would reduce the risk of opportunistic behavior by the service provider. Essentially, whenever the agent’s actions deviate from the principal’s interest/goals, there is lost agency. The degree of that deviation determines the amount of agency lost. The ultimate goal is to align interests/goals of both parties. Agencies can do that (or come close to doing that) with outcome-based contracts and performance incentives.

2.3 Risks associated with performance-based contracting

Activity-based or transactional contracts are very different from performance-based contracts. A simple cost-plus contract is activity based as it covers the costs of a supplier, providing no incentives for cost reduction or improved performance (Kim et al., 2007). The Federal Acquisition Regulation (FAR) outlines three variants of a cost-plus contract: a cost-plus-incentive-fee, cost-plus-award-fee and a cost-plus-fixed-fee contract. These are all cost-reimbursement types of contracts and provide payment to the contractor of allowable incurred costs, to the extent prescribed in the contract (FAR subpart 16.3. A fixed-price contract is defined by the FAR as a contract that provides for a price that is not subject to any adjustment on the basis of the contractor’s cost experience in performing the contract. It places upon the contractor maximum risk and full responsibility for all costs and resulting profit or loss. And it provides maximum incentive for the contractor to control costs and perform effectively and imposes a minimum administrative burden upon the contracting parties (FAR subpart 16.2 (Administration GS, 2018). Performance-based contracts link payments to performance indicators. All payments can fully relate to the performance indicator and also parts of the payment can form an incentive (Glas et al., 2013). For instance, a fixed-price incentive contract provides a fixed payment, but also offers additional incentives (Sols et al., 2007). We follow Glas et al. (2013) who identified alternatives as performance-based contracts (types A-D in Figure 1). A performance-based contract in the narrow sense fully links payments with performance (Type A, pay for performance). Performance-based contracts in the wider sense have either a cost-plus basis with additional incentives (Type B, cost-plus incentive fee) of fixed prices with incentives (Type C, fixed price incentive). Even a fixed price contract is considered performance-based, whenever the contract is not activity-based (Type D, fixed price).

PBCs alter the allocation and degree of financial risks and operational risks between the service provider and the customer in the SSC (Doerr et al., 2004; Ekström and Selviaridis, 2014; Selviaridis and Norrman, 2015). For most service providers the primary concern in PBC is the financial risk (Gardner, 2008). For the customer, there is also the operational risk of not being able to produce the desired effects (Doerr et al., 2004; Ekström and Selviaridis, 2014). Kim et al. (2007) studied which contracts (Cost-plus, Fixed Price or PBC) provided the best results depending on the risk appetite of the SSC members, using a principal-agent model.

In Figure 1, moving from the left with the cost-plus contract to the right with the fixed-price contract, the amount of risk is gradually transferred from the customer to the service provider (Gruneberg et al., 2007). In a cost-plus contract, the customer owns most of the risks while in a fixed-price contract, the service provider owns most of the risk. There are several other contract possibilities between these extremes, characterized by moderate distributions of risks between customer and supplier. The most commonly performance-based contracts are: cost-plus incentive fee (CPIF, type B), cost-plus fixed fee (CPFF, type D) and fixed price incentive (FPI, type C) (Berends, 2000; Glas et al., 2013; Roels et al., 2010; Sols et al., 2007). Cost-plus-award-fee (CPAF) contracts are frequently used incentive contracts in DoD and other agencies. Especially in cases where it is difficult to establish pre-negotiated targets to cost, schedule or technical performance (for instance service acquisitions).

2.4 Management of risk associated with performance-based contracting

Risk management in PBC is about risk identification and allocation (Gruneberg et al., 2007) and sequentially developing strategies for managing the probabilities of negative events and/or their consequences should they occur (Cohen and Kunreuther, 2007). Little is known about the real-world implications of how PBC changes risk (Gruneberg et al., 2007). Incentives are normally used in industrial service contracts to transfer risks and to measure compliance with performance measures (Datta and Roy, 2011). Furthermore, Prendergast (2000) considered the tradeoff between risk and incentives as a central principle of agency theory. Cohen and Kunreuther (2007) state that in risk management, providing incentive plays an important role to assure an appropriate level of investment.

In PBC, both parties, the service provider and the customer, assume a certain amount of risk that is different in nature for each. For most service providers, the primary concern is financial risk; providers want to ensure that the customer will give them sufficient business with an adequate return on investment. One of the best ways for service providers to ensure profitability and reduce financial risk is to secure longer contracts. They must also carefully weigh their (operational) risks in determining the level of service they are willing/able to provide (Gardner, 2008). In this study, the willingness to bear PBC-induced risks refers to the extent that the service provider will agree on and accept terms and conditions in the contract within service supply chains.

In addition to the negotiated price, performance-based contracts often include financial penalties and incentives to promote performance. Contractual incentives are mechanisms that produce compliance with performance measures (Datta and Roy, 2013). In order to be effective, the incentives must be aligned to the overall project objectives, and not just the risk involved. Moreover, incentives should be attractive to the contractor over the entire contract duration to meet the desired performance. The design of incentive mechanisms depends upon the type of inter-organizational governance approach used for managing buyer–supplier relationships (Datta and Roy, 2013). Factors influencing incentive design for government contracts are: develop services (quality and quantity), minimize costs and minimize budget uncertainty (Hooper, 2008). Wiseman and Gomez-Mejia (1998) studied the alignment of incentives. They noticed that aligning incentives generally involved four issues: the allocation of compensation between base and variable; the design of variable pay forms of compensation; the setting of performance targets for the awarding of variable pay and the selection of measures used in evaluating performance.

Economists use the term risk preferences to describe risk tolerance (Wu, 2007).This risk preference can be stated in a degree of risk aversion. Firms or individuals who are risk averse are willing to give up some possible positive outcomes to avoid risk (March and Shapira, 1987). Risk preference is studied on numerous subjects: students, animals, individuals and organizations. Principals will tend to be risk-seeking while dealing with risky options whose outcomes are generally poor. In contrast, principals will generally be risk averse when dealing with risky options whose outcomes are generally good (Kahneman and Tversky, 1979). When a service provider is risk averse, it might be an inhibitor of PBC (Selviaridis and Wynstra, 2014). Selviaridis and Norrman (2014) found four influencing factors: the ability of the service provider to:

1. measure and control the service performance (“attributability”);

2. balance risk and reward across the SSC;

3. transfer risk to subcontractors; and

4. the importance of relational governance in SSC relationships.

Attributability in this study consists of two dimensions of the service provider’s ability to measure, control, and influence the service output. First, the measurability of the service (performance) output and second the service provider’s (lack of) control over input and behavior of the customer and the subcontractors (Selviaridis and Norrman, 2014)). Low attributability means that the service provider cannot be held responsible and accountable. Performance measurement is one of the key elements in PBC (Datta and Roy, 2011). A problem could occur when the customer lacks the necessary information to specify the decision-making activities. According to Selviaridis and Norrman (2014) and Oflaç et al. (2012), the ability to attribute service performance seems to depend on the measurability of the performance. The ability to control the service performance reduces the service provider’s risk (Fearnley et al., 2004; Selviaridis and Norrman, 2014), with more control translating to less risk for the service provider.

Incentives and penalties can be used to balance and mitigate risks. Selviaridis and Norrman (2014) state that the incentives should reflect a good balance of risk and reward for buyers and suppliers. In addition, risk sharing can be considered an important risk mitigation mechanism in PBC (Cohen and Kunreuther, 2007; Heinrich and Choi, 2007; Li et al., 2015; Towse and Garrison, 2010). As a critical note however, Agrell et al. (2004) noticed that the problem with risk sharing/balancing is that companies might place their own interest above the SSC’s and consequently leading to opportunistic behavior. In this study, balancing of risk refers to the extent that both the service provider and the customer are able to balance risks and rewards in their service supply chain, including subcontractors (Selviaridis and Norrman, 2014). Critical dimensions are the customer views on performance-based incentives (for instance, a customer can be primarily price focused and not be willing to pay performance bonus), and the sharing of risks and rewards by the service provider with the subcontractors. Subcontractors play an important part in PBC in the SSC. In most cases, the service provider is not able to deliver the service performance alone and depends on subcontracters for significant share of the performance delivery. Datta and Roy (2013) consider uncertainty sharing between the service provider and the subcontractors the basis for sustainable PBC. However, Selviaridis and Norrman (2014) question the added value of the service provider if the risk transferred from the customer will be passed on to the subcontractors. In this study, the transfer of risk is operationalized as the transfer of (financial) risks from the customer to the service provider and on to the subcontractors.

Relational governance can reduce the risk associated with low performance attributability. The importance of relational governance is confirmed by Lu et al. (2015). They concluded that relational governance is important for improving project performance and mitigating opportunism. Zheng et al. (2008) found evidence that contractual governance should be seen as crucial but not sufficient “qualifier” for effective exchange and therefore needs to be complemented with pro-active relational governance. In this study, relational governance refers to the interfirm exchange relationships, including specific assets in combination with high levels of inter-organizational trust. Relational governance implies trust-based, flexible collaboration wherein the work is monitored while relying on social norms and personal relationships, rather than on formal governance mechanisms (contracts and authority). Building on the model presented by Selviaridis and Norrman (2014), we entered the study with the following conceptual model (Figure 2).

3.1 Data collection and data processing

Different sources of complimentary evidence were used, and several measures were taken to strengthen the validity of the study (Yin, 2014) (Table I). The case study involved in-depth semi-structured interviews with key informants of the service provider and the customer, as well as an analysis of documentation. These documents included the contract between the service provider and the customer and supporting documents as there were among others the performance-based logistics plan (including PWS), incentive fee plan, performance measurement process document and the metric taxonomy and calculations documents. These documents were an important source of data as the contract content could be related and compared to the data gained from the interviews. The interview questions covered several themes such as the (actual) risk transfer to the service provider, the incentives, and the factors that influence the willingness to bear PBC induced risk by the service provider. The interview process was conducted as described in Figure 4. After identifying the candidates to provide the essential case study data, an interview date was set. During each interview the same structured interview protocol was used to safeguard the reliability of the study (see Appendix Interview Guide with the structured interview protocol). After the interview, the data were processed and the draft report was adjudicated by the interviewee. Depending on the comments, the report was either adjusted and sent back to the interviewee for endorsement or made final and stored in the case study database. The interviews were conducted following the guidelines stated in Interviewing as Qualitative Research (Seidman, 2013).

Interviews were held with key managers spanning several positions in the PBC process: procurement officers, contracting officers, account managers and business development managers. Four managers from the customer side were interviewed and three managers from the service provider side, together they formed the group that worked the PBL construct from the customer and service provider side. The other sources of data were organizational documents, such as the PBL plan and the PWS of the contract. These documents describe the essence of the contract, namely, the required performance, measurement of the performance and incentive structure.

Coding was used for data reduction to make the analysis of the large amount of data more efficient. Codes have been assigned to the fully transcribed interviews and documents. The qualitative data analysis software program NVivo was used for this coding process to label and index-related data. NVivo allows you to add codes (or nodes as they are called in NVivo) before, during and after processing the data. For the first cycle of coding, a provisional and descriptive coding process was used followed by structural and sub-coding (conform Saldaña, 2009). Pattern coding was used as a second cycle. In line with the coding process presented by Miles and Huberman (1994), the initial codes were related to the factors that influence the service provider’s willingness to bear PBC induced risks, such as control of the service performance and balancing risk. As new insights emerged and part of progressive refinement of the codes, sub-codes and new codes were added to new findings and observations, like the uncertainty due to accurate forecasts and the length of the contract. NVivo was also used as the database to store all the used data, namely, recordings, transcripts, articles, etc., to maintain an evidence trail. Afterwards all the coded material was presented in one clear overview, allowing better and easier pattern matching. Following the three concurrent stages of Miles and Huberman (1994) after the data reduction and data display the next step was analysis.

During the interviews and after transcription attempts were made to recognize patterns between the different interviewees and organizational documents. According to Almutairi et al. (2014), pattern matching techniques can be used to reconcile the multiple perspectives of knowledge in case-study research. Pattern matching also helped to insure the internal validity of this research (Yin, 2003). We have applied within-case analysis. The codes helped with a within-case reliability check as the codes had a large number of agreements in relation to total number of agreements (Miles and Huberman, 1994). According to Eisenhardt (1989, p. 533): “within-case analysis gains familiarity with data and preliminary theory generation”. Within-case analysis was applied within the unit of analysis and between the unit of analysis and the documents.

4.1 The PBC case and characteristics

The study consisted of conducting a single case study that exemplifies the PBC construct. The customer and service provider in this case study worked together for a considerable amount of time on this PBC construct. The service provider is the OEM of the engine with a long-standing relationship with the customer. This is the first performance contract between the two parties and is considered a transition PBL contract. The PBL plan contains the PWS, considered as the working end of the contract describing specifics who, what and how of the required service performance. The case is part of one of the biggest US military development and sustainment projects in history (Vucetic, 2013), with a total expected program cost of US$1.5tn (through 2070 in then-year dollars).

In Firm Fixed Price contracts and in performance-based contracts, financial risks are primarily transferred to the service provider. However, the studied documents do not mention any transfer of financial risk. It appeared that the service provider was only partly paid on a performance basis for his sustainment activities. The interviewees agreed that a limited financial risk was transferred to the service provider in this contract. The operational risk however, is supposed to remain the customer’s responsibility. The discussions tended to get a more philosophical as to whether one (governmental) organization can contract another organization to be responsible for its own core business. According to a manager of the customer:

You can hold the service provider financially accountable for financial risks but if he fails, the customer is still responsible for the operational risk. The operational risk of not being able to conduct the (fighter/bomber) mission is still on the customer.

In PBC, the customer is supposed to incentivize the performance of the service provider, while balancing the risks with these incentives. The contract shows two incentivized metrics, one on performance and one on costs. The performance metric is the non-mission capable (NMC) rate. The costs metric is on the total cost of the contract. The contract does not literally state risk is transferred to the service provider and incentives are used to balance this. The interviewees provided similar answers:

There are financial incentives in place to balance the change in allocation of the risks. We have one incentivized metric on performance. The incentive is fee or profit.

As a service provider, we are willing to accept an increased amount with risk, but you need to be able to control at least a portion of the performance outcome. And it needs to be rewarded.

The service provider is supposed to behave in the interest of the customer in accordance with the number of PBC constituents in the contract. The case study provided evidence to support this assumption. A manager of the service provider:

[…] [red. on another contract] we improve the reliability of the parts that resulted in an improved Time on Wing. For the service provider this meant fewer parts to be repaired (equals less costs), for the customer this meant higher serviceability of the engines (aircraft) and significantly improved readiness.

Still, there was a critical note from the side of the customer:

Right now, we have a cost-plus incentive fee, so the contractor has to give us insight in their spending (amount of man-hours). Once you move to a fixed price, you will not have that insight. So you’re paying for certainty. But you might be paying too much, which is not in the best of our interest.

4.2 Factors influencing the service provider’s willingness

The model showed four factors that influence the service provider’s willingness to bear PBC induced risks. These factors and the extent to which the data from this study supported these factors will be discussed below.

4.3 Additional factors influencing the service provider’s willingness

During the interviews, coding and pattern matching process, three additional major factors emerged. These factors were supported by the interview data from both, the service provider and the customer and the documents. These three additional major factors are ability to make accurate forecasts in relation to the maturity of the system, the growth path toward full PBC and the length of the contract.

5.1 Discussion of main findings

In this section, we will describe the significance of the main findings in light of what was already known and discuss the impact of our new insights. This insight consists of the three additional factors that were found that strongly influence the willingness: the chosen growth path towards full PBC, the ability to make accurate forecasts and the length of the contract. The other new insight that was found is the way a lack of total performance attributability can be managed.

5.2 Managerial and theoretical implications

This study has managerial implications for entities operating in a PBC environment. Managing risks induced by PBC is a complex matter. Companies should be aware of the investigated factors that might influence a service provider’s willingness to bear PBC induced risk. Awareness of these factors could improve the ability to align the goals of the customer and service provider, one of the aims of PBC.

Particularly, the service provider’s performance attributability is a factor to take in to account in PBC relationships. This study provides two methods to manage this factor: exclusions and SLA’s. Exclusions take area’s where the service provider has (very) limited control over required performance, out of the scope of the contract. SLA’s with third party providers (or in some cases the customer himself) gives the service provider a greater level of certainty that he will able to provide the desired performance. This limits the amount of risk to the service provider, which should translate in a more affordable contract for the customer. Although more study on this subject is recommended, these methods can be taken in consideration when performance attributability is an issue in the PBC relationship.

Our study confirmed the importance of collecting and analyzing accurate historical data necessary for making reliable forecasts. Accurate historical data and corresponding accurate forecasts are critical to manage performance risks. Customers should have sufficient data to accurately forecast future costs, as these are a considerable risk factor in PBC. A stepped or phased growth path could be selected to manage the risk of not being able to make accurate forecasts.

A stepped of phased growth path can be taken in several dimensions. Perhaps the most common approach is to gradually move from a transactional contract to the full PBC contract (Figure 1): gradually moving from the left (transactional) to the right (full PBC). Another (hybrid) approach is to contract those sustainment activities where the risk is acceptable under a PBC. Sustainment activities which still pose a significant risk are better to keep on a transactional contract, as high risk will translate in a high price for the customer. The amount of incentive fee could be considered another dimension of the growth path. Particularly in the beginning of a PBC relationship there could be uncertainty on both sides whether the desired performance is achievable and to what effort. Keeping the incentive fee relatively small in the beginning of the PBC relationship allows both parties to get familiar with new construct without high financial risks. The same can be said about the objective value of the incentivized metric. When the sustainment activities are in a more mature and steady state and thus risk is reduced, the incentivized metric and the incentive fee may be adjusted.

5.3 Limitations and future research

This study has limitations which could be addressed in further research. The use of a single case study limits the external validity of the results. This case has specific characteristics unique by time (maturity of the technical system and supply chain) and place (market). The results should therefore be interpreted with caution. The results could be tested in another research setting. Another recommendation would be to perform a similar study when both the technical system and the supply chain are in a stable, full grown state. It will be interesting to see how this case changes over time, particularly the growth path in relation to the maturity of the technical system and the global service supply chain. Further research is needed to extend the knowledge in this field and to identify whether findings can be generalized to other PBC projects in the defense industry.