The Costs of Cooperation

Abstract

Service collaborations often must confront risks arising from problems of coordination, division, and defection. U.S. scholars have focused on understanding the efficacy of three general strategies to reducing these risks. First, the use of adaptive and restrictive contracts to reduce the risks from service characteristics has received a lot of attention. Second, scholars have studied how the use of different institutional arrangements reduces the risks of collaborative service provision. Third, attention has been devoted to understanding how the social networks of administrators and elected officials mitigate risk in sharing services. This article concludes with suggestions for future research on this topic.

Keywords

shared services contracting collaboration municipality

Introduction

The fiscal realities many local governments face represent a “new normal” in which accountability is a public agenda priority (Martin, Levey, and Cawley 2012; Elling, Krawczyk, and Carr 2013). For many communities, delivering high-quality services within the financial constraints they now face is a serious concern. These pressures have increased interest in establishing more shared service delivery arrangements and focused attention on the difficulties involved in managing the complex relationships among participants in cooperative efforts (Zeemering 2007, 2008; Hawkins 2009; LeRoux, Brandenburger, and Pandey 2010; LeRoux and Carr 2010; Scholz and Feiock 2010). The polycentric nature of many U.S. regions creates more opportunities for service sharing but also increases potential risks because of the diversity of potential partners, range of services provided, and the complexity of service arrangements. How risks from service collaborations can be managed has been the focus of a growing base of academic research on local government service arrangements (Feiock 2007, 2013; Hawkins and Carr in press). The difficulties of defining outcomes that indicate achievement of the objectives and the costs of monitoring and enforcing participant contributions to the agreement are important obstacles to the creation of shared service arrangements. Dealing with problems of monitoring and enforcing contracts can be particularly challenging in polycentric settings where information asymmetries and uncertainty abound (Steinacker 2004).

The purpose of this article is to review research examining shared service delivery in the United States to identify major themes about the obstacles to establishing and managing shared service arrangements. Popular discussions about the efficacy of local government services often focus on the expected benefits of sharing services and state officials, business interests, and the local media often express surprise that collaborative service arrangements are not more widespread.¹ They emphasize the potential of cost reductions gained through increasing the scale of operations in each city or by eliminating what they see as redundant activities across neighboring municipalities. They often see the absence of collaborative service arrangements as indicating a lack of accountability to residents and a desire by local officials to protect their service “fiefdoms.” Yet these accounts often ignore the risks to local governments from these activities and the costs involved in reducing these risks to levels where they are comfortable sharing the responsibility of providing services to their residents with one or more other governments.

Academic researchers have devoted considerable effort to understanding the sources of risk in shared service arrangements and have begun to identify the strategies used by participants to mitigate this risk. There is an emerging literature on this topic rooted in the experiences of local governments in Canada, Australia, and the European Union, but we focus exclusively on the U.S. context.² Understanding the collaboration risks faced by local governments in other countries is important, but the dynamic institutional context and intergovernmental relations within the United States present a unique set of challenges to local government officials in pursuing shared service arrangements. We begin with a brief review of the research that has identified three sources of transaction risks (problems of coordination, division, and defection) and the factors that affect their levels (Heckathorn and Maser 1987; Maser 1998; Hawkins 2009).³ Next, we review the research that has focused on the strategies that local government officials have used to manage these risks. We conclude with suggestions for future research on this topic.

Transaction Costs in Establishing Shared Service Delivery Agreements

Shared service arrangements take many forms, ranging from largely informal agreements of mutual aid to the creation of regional service authorities (Nunn and Rosentraub 1997; McCabe 2000; Feiock 2009; Scholz and Feiock 2010). The adoption of interlocal agreements is commonly examined as an exercise in contracting that frames the decision as a choice among several different production options, including in-house production and a variety of contracting arrangements (Brown and Potoski 2003a, 2003b; LeRoux and Carr 2007; Andrew 2009a; Carr, LeRoux, and Shrestha 2009; Hefetz, Warner, and Vigoda-Gadot, 2012). Much of this research tends to specify economic, political, and other local contextual variables to explain the motivations of local public officials in selecting different production options (Feiock 2009). A few analysts break this topic into the decision to contract for services and the subsequent decision about how extensive the collaboration with external producers will be (Kwon and Feiock 2010; Shrestha and Feiock 2011).

A relatively large and continuously growing body of literature utilizes transaction cost economics to explain local production decisions in general and interlocal cooperation and shared service delivery more specifically (Hefetz and Warner 2012). According to Williamson (1975, 1985), uncertainty, limited information, and the characteristics of the investment are considered the critical components for characterizing a transaction or the “friction” in market exchanges. Williamson links these issues to a conceptual explanation for organizational arrangements of a firm, where transaction costs are defined as the “…comparative costs of planning, adapting, and monitoring task completion under alternative governing structures” (1981, 552–53). This concept has been used in public administration and urban policy to explain administrative decisions and organizational arrangements in a variety of policy domains. Problems of coordination, division, and defection will often arise during efforts to create arrangements to produce services with other organizations (Heckathorn and Maser 1987; Maser 1998; Hawkins 2009). The remainder of the section discusses these three sources of risk in detail. Our review reveals that the costs of coordination and defection have received a lot of attention from U.S. scholars, but division problems have received relatively little attention in the research literature.

Risks due to Coordination Costs

The process of bringing together two or more local governments to create a shared service delivery arrangement is described in the literature as a coordination problem (Hawkins 2009). The costs of negotiating can be relatively high and create a disincentive for local public officials to enter into service agreements with other governments, especially if the extent of any joint gains is unclear (Williamson 1981; Feiock 2013; Hatley 2010). One issue is that local officials must develop agreements that support the service delivery goals of the group, yet also address the needs of their own diverse populations (Kwon and Feiock 2010). This tension can increase the costs of coordinating the agreement. Agreements typically require approval from each city council, and depending on the nature of the transactions, several rounds of negotiation and bargaining may be required to determine the terms and conditions of the agreements. In the presence of serious differences among the parties, serious obstacles may arise during negotiations on these agreements (Hatley 2010).

The severity of coordination costs is affected by several factors. The complexity of organizing tasks among the potential partners and coordinating the partnership is more difficult when the preference of potential collaborators diverge. Preference divergence is expected to increase when partners differ substantially in terms of demographic characteristics, resource levels, or municipal institutions (LeRoux 2008; Oakerson 2004; Steinacker 2004). Oakerson (2004) emphasizes the importance of the homogeneity of residential preferences for services within cities in reducing coordination costs because public officials are more able to “speak with one voice” for the residents, when making governing decisions on their behalf. Several studies have focused on this issue. For example, Bae (2009) found that differences in demographic composition across jurisdictions have a negative effect on the formation of intergovernmental agreements among governments in Georgia’s metropolitan areas. Similarly, respondents to Hawkins’ (2009) examination of joint venture formations reported that a lack of agreement among communities on development goals was a significant barrier for local governments seeking to establish an agreement. Finally, Frederickson’s (1999) concept of administration conjunction suggests that regions with relatively large numbers of professional administrators will have lower coordination costs because these administrators often share a regional perspective and seek to cooperate on public services where possible.

The existence of a singular overriding goal may also help reduce coordination costs. Cost savings and improved service quality are two commonly cited objectives for collaborations among local governments (Feiock and Carr 2001; LeRoux and Pandey 2011), and adding new services or infrastructure may be the easiest way to unify partners around a single goal (Nunn and Rosentraub 1997). However, motivations to collaborate stemming from fiscal stress may also involve significant coordination and division costs. For example, Hatley’s (2010) case analysis of an effort by five cities to replace their fire departments with a single multijurisdictional fire authority illustrates that individual goals of cost savings may increase coordination costs to the point that the effort to share services is abandoned. Among the obstacles that he cited was that administrators generally expected to reap long-run savings from the fire authority, but most of the elected officials expected short-run cost savings.

Any differences in the policy preferences of each jurisdiction must be reconciled through a process of negotiation, and the costs of these activities depend upon the services involved in the agreement. Metrics quantifying what each partner receives through an agreement are more costly to develop for some services, and successful services cooperation depends in large part on the ability of administrators to effectively monitor and manage provider performance (Brown and Potoski 2003a; Feiock 2009; Hawkins 2009). Depending on the services that are to be jointly produced, extensive communication may be required to specify the tasks to be completed jointly or individually by the participants. Some services are better candidates for production through collaborative arrangements because of their political salience with the public and, consequently, elected officials. Williams’ (1971) typology of services is perhaps the most often cited example of coordination costs differing among groups of services. In his analysis of metropolitan politics, he divided public services into two groups: system maintenance and lifestyle functions. System maintenance functions generally relate to the “infrastructure” that supports the creation of multiple residential location options in the region, such as roads, water distribution, and solid waste disposal. Lifestyle services, in contrast, provide access to social and life opportunities for residents and which can vary dramatically in quality from city to city, such as education, public safety, parks and recreation, and economic development. Williams’ argued that local official will have greater difficulties overcoming the coordination costs in the case of lifestyle services such as parks and recreation, housing, education, and economic development that support employment opportunities (Rawlings 2003; Wood 2006; LeRoux 2006; LeRoux and Carr 2010). In contrast, communities will be more willing to provide system maintenance services through interlocal agreements and other shared service delivery mechanisms because these services are neutral to the values of citizens and are perceived to be essential to the functioning of government.

Subsequent research has largely confirmed Williams’ propositions. For example, Rawlings’ (2003) examination of intergovernmental fiscal transfers between localities in metropolitan areas concluded that the transfers were higher for system maintenance services than the lifestyles services she studied. LeRoux and Carr (2010) examined the service delivery networks for eight services among the forty-four local governments in Wayne County, Michigan. They found that the networks for the four system maintenance services were more centralized than the networks linking the governments in the case of the four lifestyle services, but this group also has significant scale economies, and this fact may explain the tendency toward centralized service production. Finally, one study found interlocal contracting patterns that were not consistent with Williams’ propositions. Wood (2006) found that intergovernmental service delivery arrangements among governments in the Kansas city region were just as common in system maintenance services as in the lifestyle services.

Risks due to Division Costs

Division problems occur when local governments agree on the general goals for the collaboration, but encounter difficulty in dividing and distributing the expected benefits among the group (Hawkins 2009). The literature examining shared services often focuses on the issue of coordination—explaining the situations when local governments will contact or otherwise share services with other organizations—but these studies rarely focus on how division costs affect these decisions.

The few studies focusing on division issues have provided important insights into the risks division problems create for the parties to shared service arrangements. Steinacker (2004, 2010) has devoted some attention to these costs in game theoretic discussions of collaboration. She proposed that division costs can be significant obstacles for cooperation on public services that are nonrivalrous and have limited excludability; “cooperation in these situations is dependent on the relationship between the value of the cooperative outcome and each player’s ideal outcome” (p. 49). Her work suggests that participants will accept returns that deviate from their ideal levels, but that the risk of defection increases as the gap grows. Hawkins (2009) has also focused on the obstacles created by division costs in collaborations on local public services. His research surveyed local government officials to assess the extent to which issues of dividing tasks, costs, and benefits were obstacles in their efforts to form joint ventures with other governments. He concluded that division issues in terms of different perspectives on appropriate contributions to an agreement and the expected level of return from these contributions hindered the formation of joint ventures in the twelve metropolitan areas he studied.

Many of the same factors that increase coordination costs are likely to also increase division problems. Steinacker’s (2004) work points to difference in political power, policy preferences, resource levels, and demographic composition as important factors. She observes that the relative power of participants may depend on the specific services to be shared. For instance, she suggests that cities with the largest population may have greater power in agreements involving capital-intensive services such as water or sewage treatment systems. She goes on to suggest that power differentials will usually be less important in the case of labor-intensive services because cooperating cities are on more equal footing when it comes to their own production costs. Hawkins’ (2009) analysis of the formation of joint ventures among cities indicated that differences in socioeconomic characteristics and resource levels increased division problems. Hatley’s (2010) examination of the efforts by five suburban communities in the Detroit metropolitan area to create a multi-community fire and emergency medical services authority also highlighted the importance of division costs. Demographic, political, and fiscal similarities, combined with a past history of fruitful collaboration, suggested these communities were good candidates for interlocal collaboration. Nevertheless, despite several years of work the effort failed. His work suggests that differences over how future savings will be shared among the participants can be a significant obstacle to the creation of shared service arrangements.

Risks due to Defection Costs

Defection problems emerge when one party does not comply with the agreement (Hawkins 2009). Without a credible commitment, one or more of the negotiating parties has an incentive to defect and free ride on the efforts of others (Feiock 2009). Bargaining requires a careful outlining of the terms and conditions of the exchange in order to ensure credibility of commitment, to adapt to environmental uncertainties, and to resolve conflicts that may arise in the future. As Brown and Potoski (2005, 328) note, policy decisions are particularly risky when local governments are faced with “limited information, uncertainty about the future, and the prospect that people or organizations behave opportunistically.” In essence, each participant must be confident that the others have consistent policy preferences, will maintain the underlying goals and objectives of the agreement, and have a commitment to fulfill their obligations (Hatley 2010; Hawkins 2010).

This risk of defection is greatly affected by the risks created by two characteristics of public services: asset specificity and measurement difficulty (Carr, LeRoux, and Shrestha 2009). The concept of asset specificity refers to whether specialized investments are needed to deliver the goods or service. The second dimension, measurement difficulty, indicates the ease at which performance measures can be identified and the extent to which vendors can be expected to meet all their obligations in delivering the service.

Asset Specificity

Highly asset-specific services are difficult to adapt to other uses and therefore only a few vendors will likely be willing to provide the service in the local market. Sellers of these services make specialized investments and cannot easily use them to provide other services. Brown and Potoski (2005, 329) suggest the following examples of specialized investments:

the use of a specific location that is only moveable at great cost;

the use of highly specialized human skills that cannot be put at work for other purposes;

the use of specialized tools or a complex system designed for a single purpose; and

the requirement that the service reach the user within relatively limited period of time or the quality of the service greatly diminishes.

Why do high levels of asset specificity increase the costs of managing service arrangements with external providers? Theoretically, asset specificity creates significant risks for both buyers and sellers and works against the creation of a competitive market for the service. Sellers are vulnerable to decreases in demand for the service because they must make investments in assets that are not easily deployable to other uses. The risk to the supplier of losing customers is great and, as a practical matter, only a few suppliers can survive in the same market. This reality makes buyers vulnerable too, because few sellers are willing to provide the service in the face of this risk. Thus, the specialized investments necessary to provide the service help to create monopolistic conditions that limit viable competitors and put buyers at risk of opportunistic behavior by sellers. Empirical work, however, has provided mixed results of the effect asset specificity has on production decisions. Hefetz and Warner (2012) analyzed asset specificity for sixty-seven services on the production decisions of 118 municipalities. The results of their multivariate model indicated asset specificity is not an important factor for contracting with other governments, for-profit or nonprofit organizations.

There are two direct implications for the cost savings that can be expected from intergovernmental cooperation on highly asset-specific services. First, weakly competitive markets may not produce the production costs savings expected. Vendors can be expected to behave opportunistically by raising prices or reducing service quality because the risk of penalty is minimal. Second, ensuring the contractor does not exploit the opportunity to behave opportunistically requires significant investments in contract management. These activities may include frequent communication with the vendor, regular and extensive monitoring of vendor performance, and the periodic enforcement of contractual penalties (Brown, Potoski, and Van Slyke 2006; Carr, LeRoux, and Shrestha 2009). These added costs reduce—and may completely offset—any production cost savings gained through intergovernmental cooperation. One solution to this problem is to create long-term arrangements that provide for common or joint ownership of highly asset-specific facilities and equipment (Andrew 2009b).

Measurement Difficulty

The concept of measurement difficulty refers to the ability of the contracting organization to evaluate vendor performance or to effectively monitor how the vendor delivers the service. A service is difficult to measure when neither the outcomes to be achieved nor the activities to be performed in delivering the service are easily identifiable. Easily measured services have readily identifiable performance metrics that accurately represent the outputs and/or outcomes of service quantity and quality (Brown and Potoski 2005). It is far less costly, for example, to measure the quality of trash collection than it is to assess the delivery of mental health care services. Developing a contract for the latter services is more costly because expected levels of output often cannot be established in advance and performance of these types of functions requires substantial discretion, making it difficult to clearly specify performance expectations in contract language (Carr, LeRoux, and Shrestha 2009).

How does measurement difficulty increase the costs of providing public services with another organization? Services that are difficult to measure generate serious information asymmetries between contracting organizations and vendors. Thus, effective monitoring may require the development of expensive evaluation systems and the use of complex—and costly—contracts. In these instances, it is more costly to deliver these services with external providers. These added costs reduce and may even completely offset any savings in production costs gained using an external provider. Some analysts suggest the risks from measurement difficulty can be reduced if other governments or nonprofit providers are used instead of for-profit contractors (Jang 2006; Feiock and Jang 2009). These organizations are expected to have cultures more closely in line with the government and be less likely to behave opportunistically than for-profit organizations.

When the measurement difficulty of a service is low, the savings in production costs from external production is likely greater than the cost of monitoring these suppliers. Government officials are thus likely to rely on private and nonprofit contractors when measurement difficulty is relatively low (Brown and Potoski 2003a, 2005; Carr, LeRoux, and Shrestha 2009). This strategy is attributed to the fact that nongovernmental producers usually operate in more competitive markets than governmental suppliers, increasing the likelihood of greater cost savings than when governments are used. However, as measurement difficulty increases, the cost of monitoring service providers increases relative to the gains achieved from a competitive market. When measuring service quality is more difficult, cities will turn to their governmental counterparts because monitoring and enforcement costs will be lower due to the expectation of less opportunistic behavior by another government in comparison to a private provider (Brown and Potoski 2003; Lamothe, Lamothe, and Feiock 2008). When measurement difficulty becomes very high, city officials are expected to opt for direct production rather than spend scarce resources for costly monitoring and enforcement of a contracting agreement. Internal management can significantly reduce negotiation, monitoring, or enforcement costs as these become intraorganizational, rather than interorganizational issues.

Reducing Risk in Service Sharing Agreements

In general, U.S. scholars have focused on three different approaches to reducing the risks in providing shared services described in the previous section. First, the use of adaptive and restrictive contracts to reduce the risks from service characteristics has received a lot of attention. A second area of focus is on the use of different institutional arrangements to reduce the risks of collaborative service provision. Finally, researchers have also devoted attention to understanding how the social networks of administrators and elected officials mitigate risk in sharing services. This section briefly explains the research to date on these three topics and how these different strategies reduce the risk of collaboration on public services.

Reducing Risk through Contract Design

An important strategy for mitigating the risks in service collaborations is through the strategic design of contracts. As the discussion in the previous section explained, risk created by the potential for defection by participants can introduce considerable uncertainty into a collaborative arrangement. Moreover, the transaction costs created by asset specificity and measurement difficulty are important sources of this risk. This section discusses how these risks can be reduced through the use of adaptive and restrictive contracts.

Managing Risk with Restrictive and Adaptive Contracts

The risk created by service characteristics can be managed to some extent through the type of collaborative service agreement (Andrew 2010; Andrew and Hawkins 2013). Adaptive agreements permit local governments to craft flexible arrangements that allow participants to provide joint services but utilize language broad enough to leave room for renegotiation. In this sense, these agreements are created “incomplete” by leaving certain parts of the agreement open for future negotiation and adjustment. Examples of adaptive contracts include memoranda of agreement, memoranda of understanding, and mutual aid agreements (Andrew 2009b).

Mutual aid provides a good example of the benefits of adaptive contracts. These agreements involve commitments of support during emergencies from other local governments and can provide a clear process for them to request reimbursement following the provision of the promised aid (Andrew 2010). The adaptive agreements used for mutual aid permit some uncertainty to exist but allow flexibility to the parties to improve the collaboration over time. Because conditions and events may change quite rapidly, services such as emergency preparedness and response may prohibit local governments from specifying in advance the exact processes and outcomes of joint activities. Adaptive agreements can greatly reduce the transaction costs of writing and implementing agreements supporting service collaborations by providing flexibility in these instances so that resources or tasks may be reallocated among participants in response to changing circumstances. A disadvantage of adaptive agreements, however, is the higher costs associated with verifying outputs related to ambiguous and complex transactions.

In contrast, restrictive contracts specify outputs prior to the point where key investments must be made by the parties to the agreement (Andrew 2010). These agreements are more binding than adaptive agreements and often include detailed procedural safeguards to reduce uncertainty. An agreement is restrictive if the rules governing the transactions are clearly specified and the expected outcomes of the agreement are included in the contract (Andrew 2010). This may be particularly important, for example, when the collaboration involves the development of rules to govern a transfer of total responsibility for the provision of a service to another governmental unit or when exchanges of payments, revenue sharing, or impact fees will be required (Andrew 2009b).

The threat of shirking on contractual obligations imposes serious costs to the parties of a shared service agreement and instability created administrative turnover makes shared service agreements more expensive and risky in comparison to alternative service production options. Extensive turnover in key administrative or policy positions may erode or obscure the original basis of the agreement and lead cities to avoid collaborating with cities experiencing these changes (Feiock, Jeong, and Kim 2003). Indeed, a study of executive turnover by Feiock and his colleagues (2007) showed that cities experiencing high turnover of city leadership engaged in less external contracting irrespective of the type of policy or service. An advantage of a restrictive agreement is that clearly specified rules can reduce uncertainty and risk among parties created by these circumstances. At the organizational level, a restrictive arrangement, for example, between public safety agencies, can produce a relatively strong functionally organized bureaucracy to ensure stability and decisiveness of coordinating efforts (Kettl 2007). Additionally, a stable set of rules governing processes and administrative procedures can reduce future conflicts when coordinating policy across multiple jurisdictions (Andrew and Hawkins 2013).

Reducing Risk through Institutional Design

A second general strategy for reducing the risks of service collaborations is through the creation of institutions that can mitigate the risks to participants identified in the preceding sections. The potential of specific forms of self-organizing institutions to be used to confront problems stemming from the economic and environmental interconnectedness of metropolitan regions is a topic of extensive research (see especially Scholz and Feiock 2010; Feiock 2013).

Feiock (2009) lists six institutional mechanisms that may be used to support shared service arrangements among local governments. The institutions he identifies differ in terms of the autonomy retained by participants and if decisions affecting the partnership are made bilaterally or multilaterally. Self-organizing institutions that permit substantial autonomy to participants reduce coordination costs but may also encounter significant costs from problems of division and defection. At the other end of the scale, institutional mechanisms that reduce autonomy can mitigate defection and division costs but may significantly increase coordination costs. The multilateral institutions identified by Feiock include regional authorities, regional organizations, and collaborative groups/councils. In this section, we focus on the three bilateral institutions discussed by Feiock because they are the most relevant to service collaborations among local governments. These institutions include managed networks, contract networks, and policy networks.

Managed Networks

Managed networks are cooperative agreements among cities that are designed or coordinated by third parties such as state or federal governments to reduce coordination problems and the potential for defection. In these mechanisms, a higher level government provides funding and mandates the formation of collaborative relations among local governmental actors. Existing institutions or actors, such as state or county governments, a council of government, or a policy entrepreneur serve as a broker and assume responsibility for helping to craft an agreement (Wood 2006). Shrestha (2010) concluded that managed networks can be effective mechanisms for collaborations on services for which transaction costs are relatively low.

Contract Networks

Contract networks link individual units through joint ventures, interlocal agreements, and service contracts that require the consent of those involved. This institutional mechanism links local governments in legally binding agreements but preserves the autonomy of local actors while also providing a formalized shared service delivery mechanism (Feiock 2009). Contract networks provide a self-organizing mechanism for local governments to work collaboratively to resolve service provision problems. Contracts formalizing service delivery and that clearly outline the responsibilities of partners also entail transaction costs related to monitoring the contract and evaluating the performance of the service provider, but as Wood (2004) contends, such interlocal arrangements require less frequent interaction by the participants and thus involve lower transaction costs. These may prove to be more stable and, ultimately, more permanent arrangements. Andrew (2009b) contends that interlocal agreements can be an effective way to integrate service production in urban regions. Norms of reciprocity among local government actors is important for the maintenance or establishment of new contract networks because these networks often span functional areas (Shrestha and Feiock 2009; Andrew 2010). This multiplexity is important to reducing transaction risk, particularly opportunism, because information about defections is shared relatively easily through the network and noncompliance in one service contract can affect relationships in other service agreements (Andrew 2010).

Policy Networks

The third bilateral institution is policy networks. This mechanism uses informal, self-organized exchange relationships between local governments to encourage trust and reciprocity among the members of the network (Scholz, Berardo, and Kile 2008; Andrew 2009b, 2010; Shrestha 2010; Hawkins 2011). Andrew (2010) argues that an important feature of a policy network is that participants pay attention not just to a potential partner’s characteristics but also consider the relations of the potential partner within the broader network (Andrew and Carr 2013). The repeated interactions among local government actors help to mitigate the transaction costs of forming a shared service agreement.

Policy networks provide the greatest degree of autonomy in decisions to enter and exit agreements, and this autonomy can affect the risk of sharing services. This flexibility can also increase costs of decisions and may limit the situations where policy networks can be effective. These less formal arrangements may still have relatively high transaction costs due to the attention to bargaining, negotiation, consensus building, and conflict resolution required to initiate and maintain the cooperative effort. Feiock (2009) points out that policy networks often “complement and reinforce other mechanisms” (p. 365) by providing important support to the more formal authority structures that facilitate service collaboration. The support that the “administrative conjunctions” described by Frederickson (1999) is one example of how policy networks may help to mitigate the risks of collaborating through contract networks, regional organizations, and other institutions identified by Feiock. The advantages of policy networks are discussed at length in the next section.

Using Local Networks to Reduce Risk

The role that networks play in reducing the costs of establishing and maintaining collaborative service arrangements has been examined from two general perspectives. The first perspective focuses on how the selection of partners can affect the likelihood those local governments will enter into collaborative service arrangements. This research emphasizes the importance of developing shared values, trust, and reciprocity among the partners linked through service production agreements. The role that affiliation networks of municipal administrators and local elected officials play in stimulating collaborations among local governments has also received attention in this literature. The second approach to this topic focuses on understanding how participants are linked to each other within service production networks affects the risks faced by the participants. This line of research has focused on understanding how the configurations of shared service agreements created by local governments affect information flows and trust among the collaborating organizations in the region. Findings from these two literatures are discussed in this section.

Reducing Risks by Selecting Similar Partners

One way local government officials may reduce the risks involved in sharing services is by establishing collaborative arrangements with similar local governing units, a process known as “homophily” (Andrew 2009b; Andrew and Hawkins in press; Hawkins and Andrew 2011). This expectation is based on the belief that shared values among partners will reduce the risk of opportunism in collaborative arrangements, requiring less time for negotiating the agreements lower monitoring and enforcement costs. Risks stemming from the selection of partners have long been a focus of the literature on local government contracting and many studies suggest risk is reduced when governments choose to collaborate with other governments that have comparable municipal institutions, institutional roles, levels of education and professional training, and the professional affiliations of administrators and elected officials (see especially, Feiock 2013; Frederickson 1999; Hawkins and Carr in press; LeRoux and Carr 2010).

Municipal governance institutions, particularly form of government, play an important role in service production choices (Carr 2013). For example, it is argued that the longer tenure of city managers and other professional administrators in council-manager governments permits them to look further into the future when considering standardizing services and alternatives to managing jurisdictional problems and to consider the consequences of their actions on the overall vitality of the region (Frederickson 1999; Wood 2006). The literature suggests that local governments managed by professional managers are more likely to adopt advanced technology or managerial innovation in order to achieve managerial efficiency than their counterparts without professional managers (Carr 2013).

Research examining the impact of social networks on service collaboration highlights the role of interpersonal relations in building trust among local governing officials (Carr, LeRoux, and Shrestha 2009; LeRoux 2008; LeRoux, Brandenburger, and Pandey 2010) and local policy networks of administrators and elected officials are thought to be crucial in this process (Frederickson 1999). This approach is exemplified by studies that examine the impact of shared membership in professional associations and common educational backgrounds on the likelihood that local governments will enter into service arrangements together (Carr, LeRoux, and Shrestha 2009; LeRoux, Brandenburger, and Pandey 2010). Frederickson (1999) argued that disciplines and professions, for example, membership in the International City-County Management Association or having a public administration degree, function as institutions and impart a shared system of norms and values among members. Carr, LeRoux, and Shrestha (2009) found support for this proposition in their analysis of the service production arrangements used by municipal governments in Michigan. Cities with managers scoring higher on an index designed to capture participation in the public service institutions identified by Frederickson also were more likely to contract for services with other local governments. LeRoux, Brandenburger, and Pandey (2010) reported similar findings for an analysis of interlocal services cooperation in cities across the United States. They suggest regional institutional actors, such as a council of government, provide venues for greater face-to-face interactions that engender trust that is essential for cooperative endeavors. Also, since the service production networks often involve multiple services and governing units, these collaborations may often also be reinforced by norms of reciprocity encouraged by these multiplex relationships (Shrestha and Feiock 2009; Bae and Feiock 2012).

The professional affiliation networks of local government officials provide a second way to think about the power of social networks in reducing transaction costs. Frederickson (1999) described the existence of networks of local government officials that provide the basis for strong levels of interlocal cooperation in the United States. These “administrative conjunctions” are based on an ethic of collaboration based in part on common educational and training in public management and local government administration. His initial conception of administrative conjunctions presumed that administrators networks were more extensive than those created by the elected officials in the same organizations and that administrators were more strongly disposed to favor service collaboration than were elected officials. However, Matkin and Frederickson (2009) did not find evidence that institutional roles (elected legislators, elected executives, appointed administrators, and department heads) affected support for metropolitan cooperation on a hypothetical information system. Other studies have also reported evidence that elected officials are more active in interlocal networks and more supportive of service collaborations than Frederickson’s initial work predicted (Zeemering 2008; LeRoux and Carr 2010).

Reducing Risks through Strategic Use of Network Structures

Local governments involved in collaborative service arrangements are part of a network of organizations involved in service production activities, and they may also be able to reduce the risks from these arrangements if they are strategic in the networks they join. The web of relations created by agreements among organizations can be relatively simple, as in the case of a single jurisdiction in which many local governments have established a tie, or highly complex where local governments form multiple relations with one another across a region (Shrestha 2010). Our understanding of how the creation of interlocal service agreements leads to different forms of network structure is limited, but a literature examining different network configurations and how they affect services cooperation among local governments is emerging. One thing that is clear is that the optimal selection of partners also depends upon the goods and services involved in the collaboration.

There are many potential network structures that may have implications for the level of transaction costs in service collaborations, but “bonding” and “bridging” structures have been studied most often. Andrew (2009b) depicts these two structures as strategic alliances that governments can join when needed. He argues that bridging structures help local governments to reduce risks in collaborative arrangements for services characterized by high asset specificity. He predicts the risks associated with asset-specific problems will lead local governments to enter into service agreements with only a few “high-status” actors; “Establishing contracts with central actors is important for local governments to reduce the costs of crafting and monitoring multiple agreements with other localities independently” (Andrew 2009b, 385). Counties are well positioned to be the central actor in such bridging networks. These sparse networks permit localities with different interests and resources to negotiate in ways that maximize their control over transactions (Scholz, Berardo, and Kile 2008). This type of “bridging” network may allow cities not only to discover a broader set of possible gains from establishing shared service agreements within others that are inside and outside the region but also the opportunity to reap the advantage of innovation or visions that are not available within a highly clustered network (Andrew and Carr 2013). Thus, a sparse network structure may provide information advantages and reduce coordination problems across jurisdictions.

A second form of strategic alliance occurs when local governments enter into agreements with partners of their current partners to mitigate credible commitment problems. Andrew (2010) argued that dense network relationships reduce risks of uncertainty in collaboration on service production, thereby creating social capital. These bonding structures reduce uncertainty by improving the amount and quality of information about the actors in the network and increase the confidence the participants have in the others (Andrew and Carr 2013). This close-knit structure can be particularly useful for collaborations on services with high measurement costs and for confronting policy problems characterized by complex tasks and significant uncertainty over expected outcomes. A highly clustered network reduces the cost of enforcing an agreed set of working rules because any actions taken or not taken by a recipient are easily shared with others in the network. Any threat of collective sanction among the participants in a shared service agreement will enhance the credibility of punishments being imposed. Given this, participation in this kind of structure may signal to the other partners that the local government is willing to take their interests into account (Andrew 2010).

Finally, another strategy local governments can use to reduce risks in service collaborations is to take part in “multiplex” service production arrangements (Shrestha and Feiock 2009; Bae and Feiock 2012). Multiplex relations occur when local governments engage with other units in multiple shared service arrangements simultaneously. In some cases, shared service delivery reflects cooperative arrangements where there are multiple agreements between various city departments between two cities. Multiple service contracts that link more than one service can reduce credible commitment problems and minimize the potential for defection. For example, a city may have a police service contract with another city in the region, as well as shared service contracts for fire or emergency medical services with the same city. Shrestha and Feiock (2009, 806) suggest “[t]he risks involving contractual arrangements in one service area can be mitigated if these contracts are embedded in broader multiplex service relationships.” Multiplexity is important for making decisions in shared service delivery arrangements because they may signify more trust and, therefore, influences future exchanges and the maintenance or expansion of existing shared service delivery arrangements.

Discussion

This review has examined the transaction costs that inhibit shared services delivery and how local governments can craft institutional structures to mitigate these barriers. Polycentric regions provide opportunities for service collaborators, but their large numbers of governments increase coordination problems and the economic and racial segregation in these regions increase division and defection risks. The literature we reviewed provides some insight into the challenges of negotiating and bargaining shared service agreements in light of the transaction costs of cooperation; in particular, the risks associated with asset specificity and measurement difficulty and enforcing and monitoring agreements under less than ideal conditions. These challenges are compounded depending on whether services perform lifestyle or system maintenance functions.

The bargaining and negotiation costs of coming to an agreement on service delivery are not insurmountable. As the literature suggests, local governments can overcome transaction risk by crafting agreements that fit the context of the agreement. Restrictive agreements or flexible and more adaptive agreements, for example, provide alternative arrangements for dealing with the risks associated with changing environmental conditions. Through some types of institutional mechanisms, such as a managed network, third parties play an important role in negotiating agreements and reducing environmental uncertainty that can undermine cooperation. Alternatively, contract networks and policy networks rely more extensively on social capital as a precursor to cooperation. The configuration of shared service delivery networks across a region provides a mechanism to mitigate transaction risk, as do the networks of professional affiliations among local government officials. Our review of the literature on shared service delivery is of course not exhaustive of the large number of studies on this topic. It does provide, however, a discussion of public service production arrangements that complement the research on cost savings and management efficiencies from interlocal agreements.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Andrew

S. A.

2009a. Recent developments in the study of interjurisdictional agreements: An overview and assessment. State and Local Government Review 41:133–42.

Andrew

S. A.

2009b. Regional integration through contracting networks: An empirical analysis of institutional collective action framework. Urban Affairs Review 44:378–402.

Andrew

S. A.

2010. Adaptive versus restrictive contracts: Can they resolve different risk problems? In Self-organizing federalism: Collaborative mechanisms to mitigate institutional collective action dilemmas, ed. Feiock

Richard

Scholz

John

, 91–113. New York: Cambridge University Press.

Andrew

Simon A.

Carr

Jered B.

. 2013. Mitigating uncertainty and risk in planning for regional preparedness: The role of bonding and bridging relationships. Urban Studies 50:709–24.

Andrew

Simon A.

Hawkins

Christopher V.

. 2013. Regional cooperation and multilateral agreements in the provision of public safety. American Review of Public Administration 43:460–75.

Bae

Jungah

. 2009. Institutional choices for local service contracting and collaboration. International Review of Public Administration 14:27–42.

Bae

Jungah

Feiock

Richard C.

. 2012. Managing multiplexity: Coordinating multiple services at a regional level. State and Local Government Review 44:162–68.

Bel

Fageda

Mur

. 2013. Why do municipalities cooperate to provide local public services? An empirical analysis. Local Government Studies 39:435–54.

Brown

Trevor L.

Potoski

Matthew

. 2003a. Transaction costs and institutional explanations for government service production decisions. Journal of Public Administration Research and Theory 13:441–68.

10.

Brown

Trevor L.

Potoski

Matthew

. 2003b. Managing contract performance: A transaction costs approach. Journal of Policy Analysis and Management 22:275–97.

11.

Brown

Trevor L.

Potoski

Matthew

. 2005. Transaction costs and contracting: The practitioner perspective. Public Performance & Management Review 28:326–51.

12.

Brown

T. L.

Potoski

Van Slyke

D. M.

. 2006. Managing public service contracts: Aligning values, institutions, and markets. Public Administration Review 66:323–31.

13.

Carr

J. B.

2013. Form of government and policy choices: Lessons from the empirical literature. Working Paper, L. P. Cookingham Institute of Urban Affairs, University of Missouri at Kansas City, MO.

14.

Carr

J. B.

LeRoux

Shrestha

. 2009. Institutional ties, transaction costs, and external service production. Urban Affairs Review 44:403–27.

15.

Elling

R. C.

Krawczyk

Carr

. 2013. What should we do? Public attitudes about how local government officials should confront fiscal stress. Local Government Studies. doi: 10.1080/03003930.2013.823408

16.

Feiock

2009. Metropolitan governance and institutional collective action. Urban Affairs Review 44:356–77.

17.

Feiock

Richard

Jang

Hee Soun

. 2009. The role of nonprofits in local service delivery. Public Administration Review 70:668–80.

18.

Feiock

R. C.

2007. Rational choice and regional governance. Journal of Urban Affairs 29:47–63.

19.

Feiock

R. C.

2013. The institutional collective action framework. Policy Studies Journal 41:397–425.

20.

Feiock

R. C.

Carr

Jered B.

. 2001. Incentives, entrepreneurs, and boundary change: A collective action framework. Urban Affairs Review 36:382–405.

21.

Feiock

R. C.

Clinger [sic]

J. C.

Shrestha

Dasse

. 2007. Contracting and sector choice across municipal services. State and Local Government Review 39:72–83.

22.

Feiock

R. C.

Jeong

Kim

. 2003. Credible commitment and council-manager government: Implications for policy instrument choices. Public Administration Review 63:616–23.

23.

Frederickson

H. G.

1999. The Repositioning of American Public Administration. PS: Political Science & Politics 32:701–11.

24.

Hatley

2010. The art of collaboration: Interlocal collaboration in the provision of fire services in the Detroit area. Unpublished Dissertation. Wayne State University, Detroit, MI.

25.

Hawkins

C. V.

2009. Prospects for and barriers to local government joint ventures. State and Local Government Review 41:108–19.

26.

Hawkins

C. V.

2010. Competition and cooperation: Local government joint ventures for economic development. Journal of Urban Affairs 32:253–75.

27.

Hawkins

C. V.

2011. Local economic development joint ventures and metropolitan networks. Public Administration Quarterly Spring:60–93.

28.

Hawkins

C. V.

Andrew

Simon

. 2011. Understanding horizontal and vertical relations in the context of economic development joint venture agreements. Urban Affairs Review 47:385–412.

29.

Hawkins

C. V.

Carr

Jered B.

. In Press. The costs of services cooperation: A review of the literature. In Municipal shared services: A public solutions handbook, ed. Sharpe

Alex Henderson. M. E.

30.

Heckathorn

D. D.

Maser

Steven M.

. 1987. Bargaining and the sources of transaction costs: The case of government regulation. Journal of Law, Economics, and Organization 3:69–98.

31.

Hefetz

Warner

Mildred E.

. 2012. Contracting or public delivery? The importance of service, market, and management characteristics. Journal of Public Administration, Research and Theory 22:289–317.

32.

Hefetz

Warner

Vigoda-Gadot

. 2012. Privatization and intermunicipal contracting: The U.S. local government experience, 1992–2007. Environment and Planning C: Government and Policy 30:675–92.

33.

Hulst

J. R.

van Monfront

A. J. G. M.

. 2011. Institutional features of inter-municipal cooperation: Cooperative arrangements and their national contexts. Public Policy and Administration 27:121–44.

34.

Jang

2006. Contracting out parks and recreation services: Correcting for selection bias using a Heckman selection model. International Journal of Public Administration 29:799–818.

35.

Kettl

D. F.

2007. System under stress: Homeland security and American politics. 2nd ed. Washington, DC: CQ Press.

36.

Kortt

Dollery

Grant

. 2012. Regional and local tensions: The role of shared services. Public Policy 7:47–62.

37.

Kwon

Sung-Wook

Feiock

Richard C.

. 2010. Overcoming the barriers to cooperation: intergovernmental service agreements. Public Administration Review 70:876–84.

38.

Lamothe

Feiock

R. C.

. 2008. Examining local government service delivery over time. Urban Affairs Review 44:27–56.

39.

LeRoux

2006. The role of structure, function, and networks in explaining interlocal service delivery: A study of institutional cooperation in Michigan. Unpublished Dissertation. Wayne State Univeristy, Detroit, MI.

40.

LeRoux

2008. Nonprofit community conferences: The role of alternative regional Institutions in interlocal service delivery. State and Local Government Review 40:160–72.

41.

LeRoux

Brandenburger

P. W.

Pandey

S. K.

. 2010. Interlocal service cooperation in U.S. cities: A social network explanation. Public Administration Review 70:268–78.

42.

LeRoux

Kelly

Carr

Jered B.

. 2007. Explaining local government cooperation on public works: Evidence from Michigan. Public Works Management and Policy 12:344–58.

43.

LeRoux

Kelly

Carr

Jered B.

. 2010. Prospects for centralizing services in an urban county: Evidence from self-organized networks of eight local public services. Journal of Urban Affairs 32:449–70.

44.

LeRoux

Kelly

Pandey

Sanjay

. 2011. City managers, career incentives and municipal service decisions: The effects of managerial ambition on interlocal service delivery. Public Administration Review July/August: 627–36.

45.

Martin

L. C.

Levey

Cawley

. 2012. The “new normal” for local government. State and Local Government Review 44:17S–28S.

46.

Maser

S. M.

1998. Constitutions as relational contracts: Explaining procedural safeguards in municipal charters. Journal of Public Administration Research and Theory 8:527–64.

47.

Matkin

David

Frederickson

H. George

. 2009. Metropolitan governance: Institutional roles and interjurisdictional cooperation. Journal of Urban Affairs 31:45–66.

48.

McCabe

B. C.

2000. Special-district formation among the states. State and Local Government Review 32:121–31.

49.

Nunn

Samuel

Rosentraub

Mark S.

. 1997. Dimensions of interjurisdictional cooperation. Journal of the American Planning Association 63:205–19.

50.

Oakerson

R. J.

2004. The study of metropolitan governance. In Metropolitan governance: Conflict, competition, and cooperation, ed. Feiock

Richard C.

, 17–45. Washington, DC: Georgetown University Press.

51.

Rawlings

2003. The determinants of cooperation among local governments in metropolitan areas. Unpublished Dissertation. George Washington University, Washington, DC.

52.

Scholz

John

Feiock

Richard C.

, eds. 2010. Self-organizing federalism: Collaborative mechanisms to mitigate institutional collective Action dilemmas. New York: Cambridge University Press.

53.

Scholz

J. T.

Berardo

Kile

. 2008. Do networks solve collective action problems: Credibility, search, and collaboration. Journal of Politics 70:1–14.

54.

Shrestha

2010. Do risk profiles of services alter contractual patterns? A comparison across multiple metropolitan services. In Self-organizing federalism: Collaborative mechanisms to mitigate institutional collective action, ed. Scholz

John T

Feiock

Richard

, 114–41. New York: Cambridge University Press.

55.

Shrestha

Manoj K.

Feiock

Richard C.

. 2009. Governing U.S. metropolitan areas: Self-organizing and multiplex service networks. American Politics Research 37:801–23.

56.

Shrestha

Manoj K.

Feiock

Richard C.

. 2011. Transaction cost, exchange embeddedness, and interlocal cooperation in local public good supply. Political Research Quarterly 64: 573–87.

57.

Steinacker

2004. Game theoretic models of metropolitan cooperation. In Metropolitan governance: Conflict, competition, and cooperation, ed. Feiock

Richard

, 46–66. Washington, DC: Georgetown University Press.

58.

Steinacker

2010. The institutional collective action perspective on self-organizing mechanisms: Market failures and transaction cost problems. In Self-organizing federalism: Collaborative mechanisms to mitigate institutional collective action, ed. Feiock

Richard C.

Scholz

John T.

, 51–72. New York: Cambridge University Press.

59.

Tavares

António F

Camões

Pedro J.

. 2007. Local service delivery choices in Portugal: A political transaction-costs framework. Local Government Studies 33:535–53.

60.

Williams

1971. Metropolitan political analysis. New York: Free Press.

61.

Williamson

1975. Market and hierarchies. New York: Free Press.

62.

Williamson

1981. The economics of organization: The transaction cost approach. American Journal of Sociology 87:548–77.

63.

Williamson

1985. The economic institutions of capitalism. New York: Free Press.

64.

Wood

2004. Metropolitan governance in urban America: A study of the Kansas City region. Unpublished Dissertation. University of Kansas.

65.

Wood

2006. Scope and patterns of metropolitan governance in urban America: Probing the complexities in the Kansas City region. American Review of Public Administration 36:337–53.

66.

Zeemering

E. S.

2007. Who collaborates? Local decisions about intergovernmental relations. Unpublished Dissertation. Indiana University, IN.

67.

Zeemering

E. S.

2008. Governing interlocal cooperation: City council interests and the implications for public management. Public Administration Review 68:731–41.