Collective action problems in the contracting of public services

Abstract

In this paper, we examine collective action problems in the UK government in the process of contracting public services to the private sector. In particular, we examine the Ministry of Justice (MoJ) and its evolution in contract monitoring as part of a larger effort of the government to join up departments in contract management. By analyzing MoJ’s management of the electronic tagging contract with G4S and Serco, we show that a lack of coordination within the department and with other departments was a major reason for the overbilling by the two companies. Recent efforts to join up contract management efforts throughout government show promise in rectifying these contracting issues.

Introduction

The contracting of public services from government to the private sector is an increasingly important part of public sector governance in the 21st century (Levin and Tadelis, 2010). Governments contract out public services in a wide variety of policy areas, many of which are judged to be simple and ripe for contracting in the eyes of economists, but also in many complex areas that require higher levels of attention from both the government and the contractor (Brudney et al., 2005). Contractors are likely to possess significant informational advantages over those in government who award contracts. In the absence of monitoring, private contractors can negotiate contracts with exceedingly favorable terms and yet may fail to hit performance targets or deliver projects on time or on budget (Brown and Potoski, 2003; Brown et al., 2008, 2013).

Delegating the provision of public services to the private sector can improve efficiency and service, but governments develop methods of monitoring contractor behavior in order to mitigate the cost of high information asymmetries (Brown and Potoski, 2003, 2004). Recent research has provided clues for how government and private contractors can work together to minimize losses when contracting for public services. Governments can try to stimulate markets into creating more service providers and, therefore, competition (Warner and Hefetz, 2008). They can also reduce their information disadvantages by networking and sharing information with other governments (Brown and Potoski, 2004). In the provision of complex public services, contractual rules need to be more detailed and need to ensure that the incentives of government and contractor are as harmoniously aligned as possible (Brown et al., 2016).

In many studies that examine the contracting of public services, the government is typically treated as a monolithic entity, with the underlying assumption that actors within government are well organized enough to experience minimal collective action problems and fully exert their collective bargaining power with potential contractors (i.e. Brown and Potoski, 2003, 2004; Brown et al., 2013, 2016). Under this image of a monolithic government, we also assume that levels of monitoring and commercial expertise rise and fall in tandem with efforts to mitigate information asymmetries. However, governments consist of multiple organizations and contracting happens in particular departments, often pushed down to junior civil servants with little oversight from senior departmental managers or from representatives in agencies further up the chain (NAO, 2014a). Thus, the challenge of maximizing expertise and bargaining leverage in fact represents a collective action problem within government (Box, 1999). To coordinate the sharing of information and expertise, collective action problems must be overcome both between and within agencies.

In this paper, we examine collective action problems in government in the process of contracting public services to the private sector. In particular, our research question focuses on how collective action problems within government departments influence the management of complex contracting initiatives. But we also look at how recent attempts to share information and overcome such coordination problems influence contract management as well. We argue that government departments acting on their own when contracting exacerbate information asymmetries, as such departments are only able to utilize a fraction of the government’s overall bargaining power and expertise in contract monitoring and negotiation. For government to utilize fully its bargaining power as a single customer, there must be a high level of commercial expertise for monitoring contract performance, but there must also be communication between the relevant people within and across government departments.

In order to assess such contracting collective action problems, we examine a case from the government of the UK. We choose to analyze one particular case in the UK government for three primary reasons. First, a qualitative study is appropriate here because it is difficult to capture the dynamics of inter- and intra-agency collaboration with simple quantitative indicators (Brower et al., 2000; Groeneveld et al., 2015). Second, the UK government offers a particularly appropriate venue to study the challenges of contracting out, as it has witnessed over the past 15 years major collective action problems in contracting initiatives, proposed reforms to overcome these problems and there is a wealth of government documentation available to scrutinize such data. Finally, the UK is a large nation and billions of pounds are involved in these contracts, so the stakes are high. In fact, every year, the UK Government spends over £187bn on goods and services provided by private contractors (NAO, 2013b), yet it has experienced high-profile failures of contract management, often at a major cost to the taxpayer.

Here, we examine the UK’s Ministry of Justice (MoJ) and its monitoring of the electronic tagging contract with two major outsourcing companies: G4S ¹ and Serco. ² Relying upon semi-structured interviews, government documents and media accounts, we examine the MoJ’s evolution of contract monitoring. This case represents part of a larger effort of the government to join up departments in contract monitoring in order to utilize government’s bargaining power. Previous research has tended to ignore this need for coordinated efforts within government when contracting for public services. This case contributes to existing work on contracting and public management by showing that the incentives of contractors and government agents can only be aligned if collective action problems within government are overcome.

Our paper proceeds as follows: we first present a basic discussion of markets, hierarchies and networks, to explain what these concepts are and how they have overlapped and converged in studies of public service delivery. In this regard, collaboration within government in the process of contract management represents all three concepts to some degree. Second, we discuss transaction costs and how they define the contracting literature. Third, we discuss our research methods and our case of contract management in the UK MoJ. Finally, we discuss actions taken within government to improve contract management since the MoJ case and conclude by reiterating the importance of collective action within government and how it will continue to evolve in contract management.

Managerial challenges of contracting out initiatives in the public sector

Governance of contracting

Contracting out was originally developed as an attempt to include market governance in the delivery of public services (Greve, 2007). Kettl (1993) argued that the inclusion of contracting for policy delivery would increase competition, and might be a substitute for the classic hierarchical approach of governments towards the delivery of public services, in which political principals define the policies and government agents implement them. We are then faced with the key question of what sort of governance arrangements define the relationships among different actors delivering public policies (Peters and Pierre, 1998). Academic literature has traditionally differentiated between three modes of governance: markets, hierarchies and networks (Powell, 1990). Williamson (1975) proposed the distinction between markets and hierarchies to account for the governance differences between transactions among two or more firms – hierarchies within a firm. This reflected the ideas of Coase (1937) who argued that when analyzing a specific economic transaction, one could identify two distinct forms of organization among the actors in the transaction, either through a market or within a firm. Powell (1990) argued that there are other non-hierarchical forms of coordination behind markets, and he proposed the idea of networks as a third main form of governance. Networks were thus presented as “discrete forms of governance, characterizing them as having unique structural characteristics, modes of conflict resolution, bases of legitimacy, etc.” (Provan and Kenis, 2008: 232).

In studies of these three forms of governance – markets, hierarchies and networks – each one was treated separately in academic analyses; but in recent years, the boundaries between the different models have become blurred, due not only to public managers’ growing willingness to make hierarchical structures more flexible, but also to the increasing use of management tools based on principles inherent in the three paradigms (Agranoff, 2007). Recent debate focuses not so much on defining the conceptual differences between market, hierarchy and network in terms of independent and impermeable models, as on assuming that current models of public governance are combinations of all three (Klijn and Koppenjan, 2016).

The gradual convergence of the three meta-narratives of governance into just one has given rise to the need to find a name for the new theoretical model. Despite several attempts to establish unused terms able to convey the new way of thinking adequately – Agranoff (2007), for example, refers to collaborarchy – most authors still refer to it as “governance”, or occasionally “network governance”. Klijn (2008) recently upheld that the two concepts refer to the same thing. Along the same line, Walker et al. (2007: 741) define network as

a pattern of interdependence among social actors in which at least a portion of the links are framed in terms of something other than superior-subordinate relations. Parts of this network may include hierarchical arrays, but at least some portions of the pattern are linked in another fashion.

Contracting public services to the private sector is considered a classic example of a market-based activity in public service delivery, yet the process also incorporates hierarchies and networks as well. Markets are involved, as the government looks to contractors in the marketplace who compete to offer the best bid. Once the contract is awarded, however, government actors are responsible for monitoring the performance of the contractor, resulting in a hierarchical relationship. And there are, in fact, multiple hierarchies, as the contractor has its own corporate governance structure, the department contracting out the service has junior civil servants responsible for day-to-day contract management, but senior managers in that department are responsible for overseeing the street-level civil servants. Additionally, executive departments are responsible for monitoring individual department performance, but also for sharing information and trying to improve capacity across government. Sharing information and increasing expertise resembles more the behavior of nodes in networks, yet this is a vital part of contract management, as it helps government to coordinate its resources and maximize its bargaining leverage.

It is this coordination and collaboration across government that we believe is important to investigate. Previous research on contracting public services has taught us a great deal about how government actors and private contractors can align incentives (Brown et al., 2006, 2016) and minimize transaction costs (Brown and Potoski, 2003, 2004), but the government has been treated as a unitary actor in these studies. Government is made of different branches and numerous different departments within the bureaucracy and these agencies have different priorities, administrative procedures and technological systems. Maximizing bargaining leverage when negotiating a new contract requires the coordinated expertise of multiple departments. When street-level civil servant’s individual departments are left on their own to manage contracts without proper expertise or guidance, the problems of moral hazard within hierarchies become more obvious. Thus, we contend that coordinated action within government (or the lack thereof), as exemplified by both hierarchies and networks, requires greater study as it is a crucial component of contracting for public service delivery.

Transaction costs in contracting

Transaction cost theory has been widely used to explain why governments decide to contract out public services to private or non-profit providers (see, among others, Bel and Fageda, 2007; Brown and Potoski, 2003). The basic assumption of transaction cost theory is that when decision-makers have to implement a public policy, they will consider the costs of producing a service internally (within their organizations), and they will then compare this cost to the resources that would be needed to contract the provision of the service to another organization (Globerman and Vining, 1996). Brown et al. (2015) describe three main factors affecting the overall transaction costs of a contracting initiative: service-specific characteristics, market conditions and the organizational type of service providers. They conclude that “contracting is more likely to be successful when the service is easy to describe and specify in a contract, the service is easy to produce and does not require specialized investments, and the market has plenty of alternative suppliers” (Brown et al., 2015: 240).

A major limitation of this perspective is that most of the time the organization contracting the service will inevitably face high levels of uncertainty and limited information (Coase, 1937; Williamson, 1981, 1996). A lack of knowledge in contract management makes it more difficult to forecast those costs when negotiating a new contract. This lack of information regarding the development process might impact the final costs of the public service when governments decide to contract out the activity to a private sector operator. An information asymmetry between the government and the service provider which favors the service provider can facilitate misbehavior by the vendor; as Brown and Potoski warn: “under conditions of information asymmetry, when vendors have more information about their activities and performance than the contracting organization does, the vendor can inaccurately report high performance” (Brown and Potoski, 2003: 443). Governments must, therefore, monitor contractor behavior in order to mitigate the potential damage from such information asymmetries.

When governments engage in complex service contracts, the problem of incomplete contracting becomes more acute. We speak of incomplete contracting when it is impossible to account for all possible future outcomes in a service delivery project (Segal, 1999). Governments find themselves contracting for goods or services for which the production costs are relatively unknown, or difficult to forecast (Brown et al., 2016). The contract managers in government must be ready to do more than simply ensure the most rudimentary and easily measurable tasks are carried out. The managers must be ready for adverse situations, in which outcomes are not achieved and government must hold the contractor to account. The businesses carrying out the contracted services have more information than government and are in a better position to plead their case in such situations.

The problem of incomplete contracting tends to become more acute the more complex and specialized the service to be provided by the contractor. And as incomplete contracting becomes more problematic, information asymmetries that disadvantage the government also tend to grow. Amongst complex services, Williamson (1981) focused on the measurability and the asset specificity of the service or good being contracted. Asset specificity refers to whether specialized investments are necessary to deliver the service or create the good. They can increase the transaction costs of a contracting venture “because specific and specialized investments and skills are difficult to redeploy in alternative uses in public service production” (Rodriguez et al., 2012: 618).

The literature shows varying effects of asset specificity in contracting out activities. Some authors have argued that activities with high asset specificity have a greater chance of being contracted to a public provider (Wassenaar et al., 2013). In a similar vein, Brown and Potoski (2003) argue that high asset specificity will push government to contract out activities as these will require high-capital structures to be created or produced. More recently, Shrestha and Feiock (2011) have described the relationship between asset specificity and contracting out activities as an inverted U-shape. When asset specificity is low, governments will tend to contract out the activity to a private provider; whereas, if it is high, they will tend to develop the service in-house. Finally, for mid-levels of asset specificity they will seek to contract out the service to other governmental organizations or departments. Despite this uncertainty, the total amount spent on contracting to the private sector doubled under the Coalition Government (2010–2015) to £88bn from £45bn under the previous Labour Government (Plimmer, 2014). This illustrates the government’s willingness to contract with the private sector for complex public services.

On measurability, Williamson (1981) referred to the degree of difficulty when measuring the results of the service or good being produced. Wassenaar et al. (2013) also found empirical evidence suggesting that when the service or asset was difficult to measure, public organizations preferred to deliver the service or produce the good in-house, instead of engaging in a contracting venture with another public or private organization. This finding supports previous studies showing that when the results of the service or good to be contracted are difficult to measure, governments will be more inclined to avoid contracting out the delivery or production of the service or good (Brown and Potoski, 2003; Brown et al., 2008). Measurability does not only affect the decision to contract out, but also the performance of the service (Alonso and Andrews, 2016; Hart, 2003). In this sense, Alonso and Andrews (2016) found that private providers improved service performance when the performance indicators were clearly stated in the contract. In their analysis of UK prisons, they show how the overall conditions of the prisons and the activities that were offered to the inmates were positively affected by the inclusion of a private provider. However, in cases where it had not been possible to specify the contractual-agreed performance point system they observed a deterioration on aspects such as safety and order (Alonso and Andrews, 2016).

Because complex services present these different, thorny issues of incomplete contracting, contracting governments seek to expand information and expertise and, thus, reduce asymmetries between them and the vendor(s) through unilateral/networking activities (Brown and Potoski, 2004). Government must know as much as possible about the potential contractors, their expertise and the ability to scale up their resources to suppliers. It also means that, ideally, there are plenty of potential suppliers (Levin and Tadelis, 2010). If government actors do not have adequate information about suppliers, their past performance and their potential for valuable future performance, they may negotiate a sub-optimal contract. With respect to contract management, managers must look at the right outputs and understand the constraints, in order to know whether the targets are hit and the goods delivered (Globerman and Vining, 1996). They must also have the willingness to sanction non-performing contractors, rather than avoiding confrontation and hoping for the best possible outcome (Domberger and Jensen, 1997). If the government employee monitoring contracting lacks expertise, he/she will be less likely to raise issues with the performance (DeHoog, 1990). The issue of expertise is exacerbated when government does not take a leading, coordinating role on contract management (Warner and Hebdon, 2001).

Having good information for a given government agency is more easily done when government shares its information across departments and agencies (Prager, 1994). Historically, government departments and agencies did not do this and tended to negotiate contracts on their own. This put each department at an informational disadvantage and more at the mercy of the individual contractor, meaning that ultimately it was easier for the contractor to negotiate deals on its own terms. For instance, it has been stated that “the government has not traditionally behaved as ‘one customer’, and has not made the most of its collective buying power to get maximum value from its suppliers” (NAO, 2013a: 12). In other countries, recent efforts have targeted interagency collaboration across different levels of government. For example, in the US, research has shown that government departments will collaborate with other public organizations if they lack the capacity to work alone (Mullin and Daley, 2010).

Within this scenario, the main idea that we seek to develop in this paper is that, in order to enhance the managerial capabilities that governments need to contract out complex public services, governments need to increase the levels of collaboration between their departments and agencies. The governance form of contracting out should not only address the relation between the public and the private provider, but also consider the relationships between different government actors that can strengthen managerial capacity by reducing information asymmetries and pulling together contract management expertise.

Methods

In order to explain the current managerial challenges posed by the UK government when contracting out complex services to external providers, we have undertaken a case study approach. Our epistemological focus is of an interpretative nature, as we approach evidence from the inside, examining the meaning of the actors and the situations to propose theory, rather than the classic positivistic approach that seeks to test pre-hypothesized models (Evered and Louis, 1981; Ospina and Uhl-Bien, 2012). There are many reasons why contract management fails or succeeds, but we are not engaging in traditional hypothesis testing; rather, we are specifically interested in developing our understanding of collective action within government during contract management. In this study, we focus specifically on the electronic monitoring of criminals. Many criminals are released from prison, but still have curfews and are required to be home by evening and through the night. Private contractors create electronic tags for such offenders to wear, so that their movements can be monitored and it can be established that they are not violating the terms of their release (NAO, 2013c). The multi-national companies G4S and Serco have provided this service to the UK government since 2005.

We have chosen this case for both theoretical and practical reasons. First, we consider the electronic monitoring of released criminals on a large scale to be a complex service – a service whose performance requires its own monitoring by the government. Second, the government has spent over £700m on this contracted service to date, indicating the high stakes in this particular contract (NAO, 2013c). Third, the data we gathered speaks to the role of multiple organizations within government, thus enabling us to better generate inference about the presence of collective action problems in contracting. Fourth, the choice of a case study approach also makes sense given that the concepts of collective action and contract management are not easily condensed into single quantitative indicators (Isett et al., 2011). The case study approach allows us to look more deeply at the relationship between collective action within government and contract management. Fifth, we have chosen this case for practical reasons, as we had good access to key individuals involved in the contracting process, and also access to the documents that could help us understand the complexities of the activity. Finally, this case builds upon existing studies of the MoJ’s contract management, such as the seminal work on private participation in prison management by Boin et al. (2006) and the recent work on the performance of contracting initiatives offered by Alonso and Andrews (2016).

To gather case evidence, we have used two main data gathering methods: semi-structured interviews and document analysis. Semi-structured interviews were developed during the period of 2015 to 2016 with five senior leaders of the National Audit Office (NAO) and the Cabinet Office. Interviewees within the Cabinet Office were responsible for coordinating contract negotiations and helping to manage strategic suppliers, while interviewees from the NAO were more responsible for auditing contract performance and ensuring value for money. The sample was purposive, as we intended to find individuals with a high degree of experience and knowledge on the topic of study. In this sense, we followed the recommendations of Eisenhard and Graebner (2007: 27) to find a sample that is “suitable for illuminating and extending relationships and logic among construct”, rather than trying to find a sample that could help us to generalize the results of the study.

While interviews could not be recorded due to the nature of the posts that the participants hold, the research team thoughtfully completed notes of all relevant aspects discussed during the interviews. This material was complemented by the analysis of internal documents provided by the same interviewees and several documents including: reports by the NAO, the MoJ report on electronic monitoring, Public Accounts Committee transcripts of hearings on contracts, Cabinet Office reports and then media reports about the contracts themselves. A grounded theory approach (Glaser and Strauss, 1967) was followed to make sense of the evidence gathered with the interviews, the internal documents and the media reports. The material was revised by two researchers and, rather than codifying it, we discussed the meaning of each case and its relevance, considering the research question of the study.

The following section unpacks, first, the characteristics of the UK Government when approaching the task of delegating of complex services to private providers, and then moves to explain in detail the MoJ’s contract for the electronic monitoring of criminals.

Electronic monitoring of criminals and the MoJ

The private sector’s role in providing public services to UK citizens and taxpayers on behalf of the government has grown steadily over time. In addition to observing an increase in the total scope of contracting, there has also been an increased willingness on the part of government to contract complex public services to the private sector (NAO, 2013a).

As the NAO has reported time and again, the government in Westminster does not treat contract management as a serious priority (NAO, 2014a: 10). The Parliamentary Committee of Public Accounts agreed in a 2014 report when it said: “The Civil Service has prioritized the work involved in letting contracts and deemed the monitoring of contracts as mechanical and unimportant” (Public Accounts Committee, 2014: 3). Consequently, when a service is contracted out, the government frequently acts as though the work is over when, in truth, it has only just begun. The failure to take contract management seriously results in a number of broader management problems. First, when contracts are signed, individual department leaders are likely to kick management of the contract down the chain to junior civil servants who, as already indicated, do not necessarily possess the commercial expertise for contract monitoring (NAO, 2014a: 36). One representative from the Cabinet Office with whom we spoke echoed this idea by saying that, “the wrong people are doing contract management because the job is pushed down to junior staff”. A representative from the NAO told us contracts are managed by “too few senior experienced commercial people and conversely too many junior staff”. This fact is also reflected in the NAO’s criticism that the “government does not have sufficient understanding of the level of risk it is retaining on contracted out services” (NAO, 2014a: 8). In addition to lacking commercial expertise, these civil servants also have a number of other duties to perform, which means managing the performance of the contractor is not likely to get their full attention. As a result, and as one representative from the NAO stated, “the government reviews found substantial weaknesses in the way contracts are managed”.

The second consequence of not treating contract management as a major priority is that actors across government do not collaborate in ways that would leverage information and expertise. When expertise is lacking at the street-level managerial level, it is all the more crucial that expertise be utilized from around government wherever it is available. By pooling information, expertise and resources, the multiple government actors are better able to deal with the contractor when unforeseen circumstances arise. Such collaboration effectively allows government to bring its full negotiating strength to bear in a particular contractual relationship. When departments are working alone and not sharing information, contract discussions work in favor of the private contractor. Indeed, the NAO has praised recent government initiatives that allow departments to “share information about suppliers, discuss performance issues on particular contracts, and agree overall supplier performance ratings” (NAO, 2013a).

Stories of failure to contract services successfully are now abundant in the UK. In 2002, the Department of Health contracted with four separate information technology (IT) companies to upgrade the electronic record-keeping of the National Health Service, but each provider encountered unexpected problems, and the overall project resulted in the expenditure of significantly more time and resources than had originally been expected (Syal, 2013). In 2012, G4S was contracted to provide security at the London Olympics, but the firm failed to provide adequate security, leaving the government to bring in members of the armed forces to fill the gaps (Neville, 2012). In 2006, Serco was contracted to provide out-of-hours medical care to the region of Cornwall in southwest England, but they too failed to provide adequate resources to meet their performance targets. They were later found to have manipulated data to show that they had met key targets, before having the service provision taken away ( BBC News, 2013). Finally, the MoJ contracted with Serco and G4S in 2005 to provide electronic monitoring to convicted criminals who had been released from prison, but still required home monitoring.

Electronic monitoring of criminals released from prison was pursued as a policy idea because it had been established that the government could save significant amounts of money by releasing some prisoners and monitoring them, rather than keeping them in prison (NAO, 2013c). The system works by establishing a curfew, typically to take place at night for particular individuals. A base unit is placed in the individual’s home to ensure they are there during the curfew and an electronic monitoring tag is placed on the individual in order to detect whether they are in the required location during curfew time. Private contractors provide the electronic equipment and they are also responsible for monitoring of the individuals to ensure that they are in the proper place during curfew. This system of electronic monitoring was first introduced in the UK in 1999, and in 2005, contracts to provide the service were awarded to G4S and to Serco, and the MoJ assumed oversight of these contracts when it was created in 2007 (NAO, 2013c: 5). The contracts for electronic monitoring have been highly controversial, as a review conducted in 2011 revealed that G4S and Serco had both substantially overcharged the MoJ, raising questions of whether fraud had taken place (NAO, 2013c).

When the contracts were first drawn up in 2005, G4S and Serco were assigned with providing both the home base units, as well as the electronic tags to be put on the criminal offenders. They were also tasked with monitoring each offender and this meant that the two companies had to be in contact with any offender who was not in his/her designated place during curfew. If the contractor was not able to initially make contact with the individual, then it was necessary to follow up, perhaps even with a warning letter (NAO, 2013c: 9), but if contact was not established or if repeated violations occurred, then G4S or Serco would have to refer the case to a designated authority, such as the police or another public law enforcement organization. To be paid, the two private contractors had to install equipment and ensure it worked properly, report to the MoJ in a timely fashion and provide an uninterrupted and continuous service (NAO, 2013c).

After the new MoJ was created in 2007, it was given responsibility for managing the electronic monitoring contracts with G4S and Serco. The NAO has reported that the MoJ, in most of its contracts with private providers, did not consider how management should proceed during procurement (NAO, 2014b: 16). This lack of attention created problems for the monitoring of contract performance (MoJ, 2013: 21). The MoJ’s Breedon Review, which examined the performance of 15 different contracts, stated broadly about the Ministry’s contract management:

Contract managers on many of the reviewed contracts were unable to speak knowledgeably on how, why and where governance processes are applied in their particular contracts. This implies that either processes are not defined at a contract level, nor documented or are not applied or understood by contract and operations managers. (MoJ, 2013: 26)

It was not clear to civil servants when they should notify senior departmental managers of performance problems, while their lack of expertise and preoccupation with other tasks also prevented them from ever employing sanctions against contractors for poor performance (MoJ, 2013).

The lack of expertise within the MoJ and the lack of coordination both within the department and across government resulted in a number of specific contract management problems which led to the eventual problem of the two contractors overbilling the government. First, there was confusion over how the performance objectives should be interpreted. Originally, when the first contracts were written in 1999, it was understood by the contractors that they could bill for services upon receiving orders rather than after completing the work – in this case, fitting the monitoring tags ( Civil Service World, 2014). The 2005 contracts corrected this particular misunderstanding between the contractors and MoJ, but there were still gaps in the expectations of performance objectives from each side.

First, G4S and Serco were supposed to bill MoJ according to the number of individuals fitted with tags, not the number of orders received. This is crucial because for some complex offender cases, multiple orders were often issued and contractors would bill according to the number of orders, rather than according to the number of individuals tagged (MoJ, 2013). Second, the MoJ was charged for each offender after the first attempted installation of the monitoring tag, whether it was installed correctly or not. This meant that billing would then proceed, whether or not the individual was actually wearing the monitoring tag. Finally, the point which received the most public attention was that billing continued, even if monitoring was no longer necessary, such as in cases where the individual had returned to prison or had died. For each individual, the contractor had to enter an artificial end-date for when the billing would cease. This date was usually so far off into the future that the potential for overbilling large amounts of money was enormous (MoJ, 2013).

These billing problems suggest a lack of attention to monitoring the performance of the contractors. The first two types of overbilling resulted from misinterpretation over the correct performance objectives, but the fact that overbilling occurred in this fashion over a long period of time indicates that nobody in MoJ was properly monitoring the performance of either G4S or Serco. The monitoring system within MoJ which failed to pick up the overbilling practices was partly revealed by a “routine inspection” in 2008 which revealed some of the overbilling practices, but no action was taken at the time (HC Deb 11 July 2013 vol 567). The MoJ Breedon Review also noted that there was a lack of understanding amongst contract managers over when problems should be escalated to senior management (MoJ, 2013: 27). In light of this, it is not surprising that the overbilling practices were flagged up, yet not dealt with.

The G4S and Serco contracts were due to expire in spring 2013, so in 2012 the MoJ requested data from G4S and Serco and found “anomalies…regarding the average length of orders under which subjects were being monitored” (NAO, 2013c: 10). The anomalies were the orders for which billing continued long after it should have ceased. MoJ hired PwC to conduct an audit of both contractors and after an initial audit of each contractor, MoJ asked to conduct further audits, a request with which Serco complied, but G4S did not, which meant its case was referred to the Serious Fraud Office (NAO, 2013c: 11). At around the same time, each contractor withdrew from consideration for a renewed contract.

Discussion and the future of contracting

The state of contracting between G4S, Serco, the MoJ and other actors in the government between 2005 and 2013 is typical of the broad contracting issues the NAO, the Cabinet Office and others have identified. A lack of attention to the issue means that there is little planning around it and that procedures are not well-defined and perhaps haphazard. Junior civil servants who may or may not have experience with contract management end up mostly responsible without sufficient collaborative assistance from within the department in question or across the government. In the case of electronic monitoring, the disputes over billing practices had to do with measurements of contractor performance. The incorrect interpretations of performance standards relied upon by G4S and Serco reveal the fundamental problem of incomplete contracting. Contracts cannot bring complete certainty, errors are made and disputes can occur, which is why ongoing monitoring of contractor performance is needed. While we know less about the specific levels of expertise of these particular contract managers in MoJ, we do know that there was a lack of communication within government, as billing issues were raised in 2008, but no further action was taken.

The lack of willingness to take action when problems are discovered is illustrative of the problems created when there are collective action problems within government. Individual contract managers within departments may be reluctant to enforce penalties because they fear greater fallout over the contract or because they fear penalties may affect the contractor’s financial viability. When multiple government officials view and discuss performance data, this reluctance should decrease, but lines of communication within departments and between departments has traditionally been weak, information has not been shared and performance indicators have been opaque and unclear (MoJ, 2013). All this impedes collaboration across government and prevents government from bringing its full bargaining leverage to bear with the contractor.

The UK Government now recognizes the importance of collaboration as part of contract management, as it has been emphasized by the NAO and the Cabinet Office. Earlier suggestions of reform focused more on how to manage contracts within departments (NAO, 2008); but as the NAO itself admits, government attention to these sorts of reform has not been consistent (NAO, 2014a). Additionally, the NAO more recently has conceded that the strategy of improving commercial capability within departments will always have severe limitations:

…the procurement profession has had a low status in the civil service, while contract management has been seen as low status within the procurement profession. The profession has lacked the sway over colleagues to implement good practice, and struggled to attract the best talent and skills. Furthermore, without a way to measure the value of this deployment, contract management has been vulnerable to administration cuts and under investment. Yet it is doubtful that the government can improve its capability to be able to have the best contract managers on all its contracts. It will not pay either to bring in or retain commercial experts to match the combined expertise of its contractors. (NAO, 2014a: 10).

With this quotation, the NAO is essentially saying that even if the government decides that contract management is a top priority, the change in priorities will not necessarily translate into greater resources, which could help to bring more commercial expertise to individual departments. In the absence of greater resources, people who can offer greater commercial expertise will always be more tempted by the private sector, which offers better compensation and holds the profession in higher esteem. The people at the Cabinet Office with whom we spoke agreed that this was a particularly intractable problem. The MoJ has been no exception to this trend, as the NAO has reported a high number of vacancies in the Ministry’s “commercial directorate”, as it has struggled to pick up skilled recruits for posts in contract management (NAO, 2014b: 18).

Because of this deficiency of resources and expertise, the NAO and the Cabinet Office have started to focus much more on collaboration and coordination both within and across departments in order to improve contract management. The first component of this plan is to ensure greater communication within departments between street-level civil servants responsible for contract management and the senior managers within the same department. This means having regular communication, especially for complex contracts, and having clear procedures for when junior-level contract managers should report a case up the chain to a senior line manager within the department. Setting up clear procedures for this is designed to minimize the risk aversion that might set in if junior civil servants feel alone in the decision of whether to penalize a contractor for not achieving performance objectives. In July 2014, the MoJ established a Contract Governance Board, which brings together multiple representatives from across the Ministry and strengthens the monitoring of contractor performance (NAO, 2014b: 24). The Board, which contains members representing multiple disciplines, ensures that “sufficient contract management arrangements are defined before deciding to proceed with new contracts” (NAO, 2014b: 38).

Enhancing monitoring and communication within departments is done more feasibly when there is clear performance information to be shared. Prior to the most recent reviews, contract managers did not have reliable performance data, partly due to the lack of clear procedures governing contract management. As a result, managers had to rely on data from the contractors themselves which might have been subject to manipulation (NAO, 2014a: 36). In conjunction with this initiative, the NAO and the Cabinet Office have both repeatedly emphasized that contractors should use open-book accounting, thereby making all relevant financial information transparent for contract managers to see and judge (NAO, 2014a) – a sentiment echoed by each of the people with whom we spoke at the NAO. MoJ contract managers have also adopted a “balanced scorecard” approach so that both junior civil servants and senior departmental managers can easily get a sense of whether contract objectives are being met (NAO, 2014a: 45).

Importantly, there have also been significant efforts to strengthen coordination and information sharing across government departments. First, the Cabinet Office has chosen to gather information about a large number of companies, which it labels “strategic suppliers”, among them G4S and Serco. The Office “collects performance information on each contract with the strategic suppliers every six months”, as well as “ad hoc intelligence from departments each month” (NAO, 2013a: 20). Based on this information, the Cabinet Office gives each strategic supplier a traffic-light rating of red, amber or green, based on their overall performance. Additionally, it also plays a stronger veto player role than it previously had, as its approval is now required for departments to negotiate new contracts or redesign existing ones. Finally, the Cabinet Office helped to create the Crown Commercial Service in 2014 which also helps coordinate contract management across departments. The Crown Representatives are “responsible for leading the government’s relationships with a portfolio of strategic suppliers” (NAO, 2013a: 12). Consistent with the focus on managing relationships with strategic suppliers, the Crown Reps are there to provide assistance to departments in complex negotiations and commercial matters (NAO, 2014a: 17).

Conclusion

Contracting public services to the private sector has become commonplace across government departments in the UK. While many contractors perform well and achieve the stated objectives set before them (Alonso and Andrews, 2016), there have also been a number of high-profile failures over the 15 years that have attracted the public’s attention and caused policymakers to wonder whether contracting out the delivery of public services to private providers actually made financial sense. Previous research has largely treated government as a single, monolithic actor in contract negotiation and management, but in fact, government is comprises numerous actors whose behavior requires coordination in order to maximize bargaining leverage.

In this article, we have presented a case study of the MoJ in the UK, and examined MoJ’s management of the electronic tagging contract with G4S and Serco. In a nutshell, our case shows how the lack of coordination within the department and with other departments was a major reason for the overcharging by the two companies. While our findings represent an important step towards a better understanding of how governments should collaborate with other public organizations to better manage their contracting out activities, our results are not without limitations. A first limitation refers to the number of interviews developed in the case. While the number of documents and reports that we could have access to was very large, our efforts to complement them with face-to-face interviews were limited due to the few individuals that participated in the study. A second important limitation of the present article is that we rely on evidence from a single case study, which obviously diminishes the external validity of the study, and, therefore, the overall generalizability of our findings. Future studies should investigate further if other settings with different contingency factors have the same managerial challenges that we have reported in our case.

In closing, we want to highlight that, despite the vast research on how to contract out the delivery of public services, there is a surprising lack of studies analyzing how governments can collaborate across departments to enhance the management of contracting out activities. As governments seek to develop public services by relying on private contractors, studies which systematically examine how best to manage these organizational forms will certainly be of immense value.