Printer Friendly

COST ACCOUNTING AT THE SERVICE LEVEL: AN ANALYIS OF TRANSACTION COST INFLUNCES ON INDIRECT COST MEASURMENT IN THE COST ACCOUNTING PLANS OF LARGE US CITIES.

INTRODUCTION

Cost accounting has been used and studied by organizations and researchers for over 100 years. In spite of its long history and potential, cost accounting remains understudied in the context of government organizations (Rivenbark, 2005). Many books give descriptions of how to do cost accounting (Coe, 2007; Finkler, Purtell, Calabrese, and Smith, 2012) and from time to time a small amount of research is published on cost accounting in government (Ammons, Coe, & Lombardo, 2001; Coe & O'Sullivan, 1993; Geiger, 2010; Mullins & Zorn, 1999; Rivenbark, 2005). Generally, the assessment about cost accounting is much the same as it was more than 70 years ago. While cost accounting has received renewed interest by those that want to make government more efficient in recent years, there is little current research on the topic (Lorig, 1936; Rivenbark, 2005).

An exception to this general dearth is a recent study that indicates that management cost accounting continues to have many relevant purposes for U.S. cities (Mohr, 2015). This study found significant relationships between cost accounting and two common uses: rate setting and cost management. Additionally, cost accounting and service costs have a strong link with benchmarking and performance (Ammons & Rivenbark, 2008), grant overhead cost recovery (Coe, 2007), and decisions to contract out services (Stein, 1990). As the recent study on the uses of cost accounting notes in its discussion, the big unanswered question is what is impeding the general use and further development of cost accounting in government? This present research presents a partial explanation to this question by analyzing the effects of transaction costs on cost accounting plans.

This paper uses transaction cost theory to explain the use of indirect cost measurement for specific services in public cost accounting plans. The public organizations are large U.S. cities that are utilizing centralized cost accounting plans to allocate indirect costs down to individual products and to measure costs of individual services. The statistical models of Table 3 show that transaction costs are related to fewer indirect cost measures for high asset specific services, but services with outcomes that are more difficult to measure are actually related with a greater likelihood of indirect cost measurement. This may be explained by the observation that services that have outcomes that are difficult to measure may be more appropriate to measure at the input or cost level. This conclusion is further supported by cost measurement's relationship to performance measurement. This research provides a plausible explanation for what may be limiting cost accounting in U.S. local governments, but it also highlights the need for understanding the different dimensions of transaction costs for cost accounting and performance measurement.

COST ACCOUNTING AND TRANSACTION COSTS LITERATURE AND THEORY

Cost Accounting: An Introduction

Cost accounting has been discussed as an important innovation in government financial management for over 100 years (March & Olsen, 1989; Pinkett, 1965). The reformers of Theodore Roosevelt's Committee on Departmental Methods thought about how cost accounting could be implemented in the government so that managers could understand their unit costs. This preoccupation with unit costs was a hallmark of scientific management and cost accounting was an important precursor to modern budgeting and performance (Williams, 2003; 2004). Even today, cost accounting and cost measurement are important, if seldom discussed topics, of public administration and public financial management (Rivenbark, 2005). This may be partially attributable to the observation that the use of advanced cost accounting techniques such as Activity Based Costing (ABC) to measure a service's direct and indirect costs can lead to organizational resistance and rejection of the cost accounting system. However, cost accounting is often utilized during periods of fiscal stress to better allocate resources and manage costs (Brimson, Antos, & Collins, 1999; Kehoe, 1995; Levine, 1985). So, given the current fiscal environment, the topic has received more attention recently as governments are being asked to do more with less (Mohr, 2015).

Cost accounting fills an important gap in the development of costs beyond just basic budgets or financial records. Departmental budgets, which are often regarded as the amount spent to provide a service, are insufficient to determine the expense of a service or activity for three basic reasons (Coe, 2007; Finkler et al., 2012). First, overhead services that are necessary for service production such as administration, payroll, accounting, information technology and other overhead costs are not typically included in the departmental budget. Second, capital costs such as building expenses and equipment expenses may be included in a separate capital improvement budget. Finally, departments generally have many services that they perform for all services, or indirect costs, which need to be distributed down to individual services within the department to get an accurate product or service cost (Kaplan & Cooper, 1998). Cost accounting allows an organization a method or system for developing better estimates of the cost of services than using only the budget or basic accounting records.

Unlike businesses that need to have good cost accounting to correctly price their services, governments use cost accounting generally for five reasons: to collect grant overhead costs, to measure performance, to manage their costs, to allocate resources efficiently and to develop the cost for services that are paid with user charges (Coe, 2007; Geiger, 2010; Rivenbark & Carter, 2000). These purposes are especially relevant for handling fiscal stress. Getting more resources from other units of government through grants and user charges are traditional ways that cities raise resources during periods of fiscal stress. Likewise, performance measurement and management may be useful for increasing the efficiency of the organization over the long term. When governments are being asked to do more with less and when local governments are given opportunities to shift some of the overhead costs to the upper levels of government through grants, cost accounting may receive more attention from management.

One of the most recent innovations in cost accounting is the use of ABC to improve the cost of products and services (Kaplan & Cooper, 1998). The insight of ABC is that the traditional cost accounting systems are not accurate enough to measure costs sufficiently for a competitive global environment. To address these challenges Cooper and Kaplan developed ABC to measure costs more accurately and give managers the ability to control their costs. In resource scarce environments, organizations need to be as cost effective and efficient as is reasonably possible.

The key understanding and innovation of ABC is the recognition that indirect costs, besides just traditional overhead, importantly influence the cost of products, such as the cost of set up, the cost of departmental managers' time, and other departmental or service related indirect expenses (Kaplan & Cooper, 1998). The recognition of these expenses allowed for more accurate estimates of individual product and service costs.

These costs were also to be measured in ways that would be more useful to the managers to control the total cost of their products and services. This was accomplished by using unambiguous cost drivers to link the consumption of indirect resources to the production costs of activities. ABC was heralded as a key innovation for saving American industry and government (Turney, 2010) because accurate product and service costs were seen as an essential element of allocating an organization's scarce resources. Individual products and services could be compared in a more meaningful way to determine profitability or other values to the company (Kaplan & Cooper, 1998; Kaplan & David, 1992).

Important contributions toward the use of ABC in government were noted in cities like Indianapolis and Charlotte where governments used the information from ABC to evaluate whether services could be contracted out to the private sector (Brimson et al., 1999; Weiss, 1997). Experience showed that when governments knew their costs they could work to manage their costs and make their services competitive with outside contractors (Weiss, 1997). ABC was also criticized for being inappropriate for government contracting decisions because it fails to take into account contract monitoring and maintenance costs (Mullins & Zorn, 1999). While it would be insufficient to rely solely upon ABC to inform the decision of whether to contract a public service, the information that ABC provides can highlight product and service costs, and managers can take actions to reduce them (Geiger, 2010). Along with the use of performance information, ABC has been argued to be an important consideration for government financial management in areas such as budgeting (Melkers, 2003; Premchand, 2006) and performance management (Rivenbark & Carter, 2000).

In spite of the arguments for the efficacy of ABC and the extensive measurement of indirect costs, implementation and utilization remains a problem for government cost accounting (Mohr, 2015). First of all, extensive cost accounting, like ABC, intrudes upon the managers and employees of the services and processes that it intends to measure. If organizational leaders want to measure indirect costs, then they have to know the processes of the departments sufficiently to understand and link the indirect costs with their consumption during the production of the good or service. This leads to a typical three-step procedure of activity analysis, development of cost drivers, and then linking them together to develop a cost plan or cost system. This is accomplished by interviewing departmental managers and key employees (Kaplan & Atkinson, 1998 Chapter 4). The cooperation of managers and employees is assumed, as is their willingness to divulge important process information. Research in businesses around the world has shown that this is actually a tenuous assumption for organizational leaders to make about such an extensive evaluation of departmental product processes and individual service costs (Anderson & Young, 1999)

Traditional cost accounting was much less intrusive because it was generally spreading overhead costs over the production departments using generic cost drivers (Kehoe, 1995). The generic cost drivers could be taken from traditional information systems such as personnel systems or capital inventories and did not require employee cooperation. The activity and cost driver analysis relies upon interviews with production managers and key personnel to understand the process, which creates a system that has much more information about the production processes. More production knowledge is beneficial from the perspective of organizational leaders, especially if it is very detailed and actionable as in the case of ABC.

More production knowledge may be regarded as a benefit to upper level managers, but departmental managers do not want to give up their valuable organizational capital. It has long been known that efforts by management to gain information on the production process are often met with resistance (Crozier, 1964; Simon, Guetzkow, Kozmetsky, & Tyndall, 1954), but this appears to be especially the case with ABC. ABC, as a general concept, was introduced in the United States (U.S.) automotive industry in the 1990s and found support at General Motors in the division where it was originally developed. In spite of its corporation-wide adoption, ABC was abandoned within a decade because departmental managers had a hostile response to the ABC system and failed to use the information (Anderson & Young, 2001). The experience of ABC at Chrysler was even more dramatic as the organization used outside consultants to impose ABC on the organization, which generated severe resistance at the departmental level. The study of the implementation and utilization of ABC has been a concerted academic effort because the use of ABC has generally underperformed its expectations (Turney, 2010). The list of factors that influence the implementation and utilization of ABC is quite long and informs the current study of transactions costs in government cost accounting practice (19). The thrust of all of these studies is that ABC generates significant resistance by the managers that do not want to give up their valued knowledge of processes to their superiors and that there are certain contextual or institutional factors that can facilitate the long term utilization of cost accounting.

Previous studies of ABC in government have tended only to focus on ABC's deficiencies relative to the ideal cost and resource allocating systems. Mullins and Zorn (1999) note that ABC does not account for additional costs that are necessary to consider in the contracting process such as contract monitoring and maintenance. Brown, Myrring and Guard (1999) note that the budget works upon departmental appropriations and not upon total resource consumption. It may seem unreasonable for some managers to want to invest their time in a system for which they get little reward and that may understate costs and benefits. A managerial perspective suggests that having cost systems that account for all of the relevant organizational costs is beneficial. It may not account for all of the relevant costs of public service provision such as externalities, but managerial cost accounting can give a much better estimate of what individual product and service costs are to the organization. Departmental managers may want to resist the use of cost accounting because cost accounting may not be integrated with the budget (20), but organizational leaders can create separate systems for holding departmental managers accountable for the full organizational expense of individual products and services (Geiger, 2010). Advanced forms of cost accounting that measure indirect costs of products and services may be important, as the proponents of ABC argue, but previous studies of cost accounting in government have failed to consider the question of what determines whether a product or service's indirect costs actually get measured?

Transaction Cost Theory

While the literature of ABC implementation is suggestive of many factors influencing cost accounting within an organization, transaction costs have not been explored as potential factors that affect the use of cost accounting within organizations. Transaction cost theory generally suggests that many organizations' activities are influenced profoundly by the cost of developing and maintaining transactions. The original articulation of the concept of transaction costs comes from Coase's theory of the firm and exchange within the market (1937, 1960).
"In order to carry out a market transaction it is necessary to
discover who it is that one wishes to deal with, to inform people that
one wishes to deal and on what terms, to conduct negotiations leading
up to a bargain, to draw up the contract, to undertake the inspection
needed to make sure that the terms of the contract are being observed,
and so on." (Coase, 1960 pg 15)


This foundation has led to two different concepts of transaction costs which find support within the literature (Langlois, 2006). The first is the view most often found in the literature on public organizations that tends to define transaction costs as the costs of organizing, searching and carrying out transactions that leads to a type of friction or loss to the system. Examples of this type of transaction cost trace back through Dahlman (1979) and can be found in the public contracting literature that tends to view transaction costs as simply unrecognized costs to contracting (Johnston & Girth, 2012). Others have used transaction costs to explain the loss of efficiency in exchanges such as the public bond market (Marlowe, 2009).

Dahlman (1979) suggests that information search, transfer, and setup costs were all representative of the transaction costs that affect organizational or bureaucratic transactions (Baldwin, 2008; Langlois, 2006). These transaction costs would generally be well known to managers and can be factored into the cost of developing a cost accounting system or plan. Organizations and their leaders can decide whether these costs are worth incurring and design a system that maximizes the benefits from cost accounting relative to its cost. While it might be a bit of a strong assumption that organizational leaders know all of the production costs of cost accounting, it is likely that they have a fairly good idea about these costs relative to the more difficult concept of economic transactions costs.

The second type of transaction cost is the economic transaction costs that comes from Williamson (1975; 1985). This version of the theory states that the key variables of asset specificity and uncertainty arising from opportunism increase costs of transactional relationships beyond the frictional transactions costs previously described. In the economic transactions cost version of the theory, the economic characteristics of a transaction create a risk premium that must either be paid or the parties must develop rules of governance to regulate the relationship. In public administration and management, the use of transactions cost theory has been to show that contracting services in local units of government is profoundly affected by these economic transactions costs (Brown & Potoski, 2003, 2005; Clingermayer & Feiock, 1997; Levin & Tadelis, 2010).

In economic transactions cost theory the key variable that creates the need for a risk premium and leads to an economic transaction cost is asset specificity (Williamson 1985, p.30). Asset specificity is defined as technology, equipment, knowledge, or characteristics that must be acquired to perform an activity, provide a service, or produce a product and that has little or no value in the market outside the relationship between the supplier and consumer (21). Asset specificity makes the supplier and consumer mutually dependent upon one another, which creates a "lock-in effect" that makes the parties reliant on each other's continued cooperation and good faith (Levin & Tadelis, 2010). The tendency for asset specificity to create a lock-in effect leads those that need to develop asset specific resources to demand either a risk premium before getting into the relationship or extensive governance arrangements to govern this mutual dependency afterwards. In market based organizations, the key asset specific resource is often the technology to create a product or to provide a service, but in the public organization the technological assets needed for service provision rarely create lock in effects. Public organizations are unlike private organizations in that they must rely upon people much more for their productive purposes (Baumol, 1993) and asset specificity is much more likely to be found within the human dimension of public service. For the public organization, essential assets might be extensive knowledge about norms, customs, and habits of behavior of specific publics or the willingness to risk one's life for a public service such as police or fire. An example of asset specificity in a public organization is put forward by Williamson when he discusses the importance of probity or loyalty in the Department of State (1999). He points out that characteristics such as probity are a part of individual asset specificity for the state department where loyalty is necessary to insure confidentiality and continued cooperation in a difficult work environment.

The second dimension of transaction costs in the property rights version of transaction cost theory is uncertainty arising from the possibility of exploitation of specific assets. This exploitation is thought to arise from one or both parties being able to act opportunistically in the relationship. Uncertainty stemming from opportunism may not necessarily arise if there is perfect trust between the contracting parties. If both parties are reasonably sure that their investment in specific assets will not be exploited by their partner, then transaction costs need not arise. Both parties can invest in specific assets and both parties can be made better off by a mutual cooperation in the production of the good or service. The opposite occurs when the parties suspect that a potential partner will not act in a completely trustworthy manner. The transaction cost arises in the latter situation when parties must extensively monitor each other's performance or do not engage in mutually beneficial cooperation. Uncertainty creates an opportunity cost for the organization to either engage in an activity or service, or it must create governance arrangements to overcome the opportunity cost (Williamson, 1985).

In the measurement of the cost of services, two competing uncertainties arise. As the organization seeks to measure the service more accurately, the organizational leader's uncertainty about the processes and costs of their departments declines. This means that organizational leaders have less transactions costs as the organization measures the costs of more services. Cost measurement of services gives organizational leaders additional information about processes and procedures that they can use to understand their organizations. To organizational leaders, transactions costs from uncertainty decrease with greater cost measurement (Figure 1).

In contrast, the departmental or service managers may experience significant losses of production knowledge and control (Crozier, 1964; Stein, 1990). The production managers may not want to give up their valuable information to upper level managers who can then change the processes and eliminate the significant leverage that departmental managers can accrue with time. Production managers in public organizations may fear that they will face greater uncertainty in the budget process and that they will not have sufficient resources or control of the programs for which they will be held accountable. These managers face increasing uncertainty about their processes and control of the perception of cost effectiveness of their services.

Especially in the presence of asset specificity, uncertainty at the departmental level will lead to transaction costs and resistance by employees. Without asset specific resources, the employees have little leverage in a bargaining relationship. They can be easily replaced by other workers that can be found in the employment market. If the employee or the department has a specific asset that cannot be easily transferred to other employees or other departments, the production department has significant leverage in the cost measurement relationship. With high asset specific resources, the transaction costs experienced by departmental managers and employees will tend to take precedence over the lower transaction costs and benefits of cost accounting experienced by organizational leaders. Transactions costs throughout the hierarchy will be an important determinant of the use of cost accounting to measure specific departments, activities, and services.

The resistance to cost measurement may reveal itself through departments or services that do not have to measure the indirect costs for service provision. Cost measurement can be resisted by getting the government employer not to measure the indirect costs of products or services. This would be the best possible scenario from the perspective of the production managers and employees because the production employees would retain their production knowledge and leverage. The production employees and managers would also be able to minimize the perception of the cost of their services. Conversely, upper level managers would have much less knowledge about the processes and product costs. This situation will arise most often in the presence of high asset specificity and uncertainty about how to evaluate the output. As the transaction cost theory notes, departments and employees without asset specificity or uncertainty are unlikely to have the leverage to remove the measurement of indirect costs in organizational cost measurement.

OTHER GOVERNANCE FACTORS AND COST ACCOUNTING USE

Transactions costs theory suggests that governance structure evolves to mitigate and minimize the effects of transactions costs over time (Williamson, 1975; 1985). Several governance variables may assist in overcoming transaction costs such as leadership, slack resources and other governance factors that lead to more service level cost accounting (Figure 2). In the cost accounting literature on ABC the contextual variables that lead to successful implementation are assumed to be exogenous to transactions costs in the short run; although this need not be the case over the long run, as organizations may adapt their governance arrangements to reduce transactions costs. In the short run the organizational arrangements may be thought to be fixed and not determined directly by the choice of cost accounting; however, this may be a strong assumption when one considers the longer term of cost accounting development and use.

One extensively researched factor that influences organizational innovation and cost accounting specifically is the characteristics of organizational leaders (Fernandez & Rainey, 2006; Julnes & Holzer, 2001; Moynihan & Pandey, 2005; Rogers, 2003). Leadership may be able to mitigate some of the effects of transactions costs by being able to define the initiative in a way that it can be supported by the organization and reduce uncertainty. Leaders who are predisposed to organizational change and committed to implementing the innovation are much more likely to have organizations that adopt and implement their preferred innovations. Studies of ABC note that the actions and the example set by leaders are critical to determining the implementation of ABC in a business (Anderson, 1995; Anderson & Young, 1999, 2001; Foster & Swenson, 1997; Innes & Mitchell, 1995; Malmi, 1997; Shields, 1995).

Leaders that demonstrate commitment to the initiative can have a strong and positive effect upon the extent to which it is used. Along with the example that they set, leaders who have formal goals for cost management may be especially likely to implement and further develop cost accounting systems. If the organization resists the cost accounting of services, leaders who have committed themselves to extensive service level cost accounting are more likely to reduce the uncertainty surrounding the socially constructed value of the difficult innovation (Krackhardt, 2001). As employees look to the leader to overcome their resistance, leaders that support the cost accounting will be much more likely to present a vision to the employee that minimizes transactions costs and increases the perception of spillover benefits from the system. In particular, professional managers are known to prefer cost accounting techniques such as activity based costing (McCue, 2001; Purtell & Fossett, 2010). So, professional managers in local government are likely to support more cost accounting by providing solid and stable goals and expectations for cost system development, and showing employees how the system will be used by the organization to achieve other valuable objectives such as performance or strategic management improvement.

One leadership related trait that is related to cost accounting is a leader's interest in using cost accounting for performance measurement. Cost accounting has been noted as being important to performance measurement because outcomes between similar programs can only be compared when there is a reasonable basis for comparison of the level of resources that have been devoted to the program or service (Ammons & Carter, 2000; Rivenbark & Carter, 2000). It does not make sense to benchmark the outcomes of a program with high levels of resources to one that has low levels of resources. Even if the programs are the same in all other respects, the low resource program is likely to have a lower level of outputs and a lower impact than the high resource program. Programs like the North Carolina Benchmarking program focus on developing uniform cost accounting to be able to compare programs. Performance measurement is valued by organizational leaders (Ho, 2006) and is being implemented by nearly all levels of government (Berman & Wang, 2000; Melkers & Willoughby, 1998, 2005).

Besides the leadership incentive to measure cost accounting along with performance measurement, performance measurement can also tell us about how difficult to measure the service might be. Services that do not have output or outcome measures might not have measures of performance because they are inherently difficult to measure. This measurement problem is similar to the problem of service measurability uncertainty. For this reason, performance measurement should be measured at both the organizational level where performance measurement can give some indication of organizational leader's incentive to measure services and performance should be measured at the service level where service measurability can also be measured. Both are expected to lead to greater indirect cost measurement of services.

Another governance factor that influences transactions costs is organizational structure, including centralization, formalization, and hierarchical design. The more centralized organization is one that has power and decision making authority concentrated in a relatively few individuals. Formalization is the degree to which formal rules or procedures are created and followed. Hierarchy is the number of levels of an organization between decision makers and the people that produce goods or services. These factors are generally believed to inhibit the search function of the organization looking for innovations and reduce the amount of innovations considered for adoption (Rogers 2003). Organizational centrality, formalization and hierarchy tend to decrease the search and overall level of organizational innovativeness; however, these structural characteristics reverse their effect during implementation. Centralization has been found to be positively correlated with ABC implementation in private organizations (Anderson, 1995; Gosselin, 1997).

Likewise, formalization and hierarchy have been shown to be a determinant of ABC implementation success (Gosselin, 1997), particularly when a chain of command is strictly followed (Geiger, 2010). These may be thought of as brute force methods of overcoming transactions costs. If a hierarchical, formal, and highly centralized organization wants to implement extensive cost accounting, it may be able to overcome resistance by simply forcing compliance through formal rules or authoritative decisions that come down through the hierarchy. Williamson (1985) notes that the governance arrangements such as the decision to bring contracts in house or under the governance arrangements of a hierarchy may be economically rational because they reduce transactions costs. In this case, the level of organizational structure is not an either-or decision but one of relative degree. Those organizations with relatively more hierarchy, formalization, and centralization will have a relatively higher level of governance control and a lower level of transactions costs. Ceteris paribus, this may result in more cost accounting and more indirect cost measurement.

Slack resources are another governance factor that can be related to the ability to overcome transaction costs and is often an important variable in the implementation of organizational innovation such as cost accounting. The ability to use more intensive forms of cost accounting and measure more indirect costs is facilitated by slack organizational resources. Slack organizational resources can overcome the transactions costs for cost accounting by providing additional incentives through higher wages and benefits for cooperation, providing opportunities for learning and training, reducing informational uncertainty, and providing information technology to aid the process. Cities that are experiencing fiscal stress have been found to need slack resources to carry out processes of retrenchment (Levine, Rubin, & Wolohojian, 1981). So, while fiscal stress may motivate searches for cost reduction strategies, some amount of slack resources may necessarily mediate a more extensive indirect cost measurement for services.

In their review of the amount of time that it took to develop an ABC system, Anderson and Young (2001) note that ABC training is associated positively with task significance and team cohesion, which is negatively related to development time. It seems that additional resources, especially when used for training and development purposes would be related with more indirect cost measurement of services. Furthermore, many other studies of ABC implementation in the private sector note that slack resources are positively associated with successful implementations (Anderson & Young, 1999, 2001; Foster & Swenson, 1997; Innes & Mitchell, 1995; Shields, 1995). Slack resources are therefore also to be expected to have a positive impact upon indirect cost measurement of services

Another governance factor that may affect which services get measured is the size of the organization. Size is often found to relate positively with an organization's ability to implement innovative practices (Rogers, 2003). Size is a relevant variable for cost accounting because larger organizations often have more processes and have greater need for understanding where they are using indirect resources. The use of ABC was found to be positively related to organization size in the accounting literature (Baird, Harrison, & Reeve, 2007; 2004). These studies also note that smaller organizations likely will not use more advanced forms of cost accounting such as ABC because it is difficult and impractical for smaller organizations. More service level cost accounting is also likely to be found in larger organizations that may want to have more information on their larger number of services.

A final governance factor that has been found to be significant in the use of extensive cost accounting is the relationship of an organization to its unions. Anderson and Young (Anderson & Young, 1999, 2001) note that unions can effectively block the implementation of ABC. Having good relationships with the union is seen as critical to having a good implementation process for intensive forms of cost accounting. As unions can influence their members and the members can influence the government, the presence of unions and management's relationship to them may be a factor that limits the use of extensive cost accounting. Moe (2009) finds that the strength of collective bargaining contracts affects organizational outputs in the area of education. If the members of the union are threatened by cost accounting, the union will work to defeat cost accounting of their services, especially if the union is strong and if it does not have good relationships with management. Management's positive relationships with unions or strong laws that limit the power of unions should be associated with more service's indirect costs being measured by cost accounting.

A Transaction Costs Model of Indirect Cost Measurement of Services

Cost accounting has been proposed as an important reform to our financial management practices to address public agencies' fiscal stress. These views often only consider cost accounting as a rational response for the organization to deal with fiscal stress and related cost management activities. A view of cost accounting that considers governance factors and transactions costs suggests that cost accounting may not develop uniformly throughout an organization to deal with issues of fiscal stress and may even undermine efforts to deal with fiscal stress. It has been argued that transactions costs limit the utilization of indirect cost measurement for services that have high asset specificity. Governance structure characteristics such as leadership, performance measurement, organizational structure, size, slack resources, and an organization's relationship with its unions will positively affect the use of indirect cost measurement of services. Based on the above theoretical understanding, the proposed study examines the impact of these factors on the use of indirect cost measurement of individual services (Figure 2).

With the exception of transactions costs, all of the preceding variables receive support from previous studies of cost accounting in for-profit businesses. The unexplored theoretical contribution of this theory is that transaction costs are likely to be related to the measurement of indirect costs for a service. Where asset specificity is high and the uncertainty surrounding the measurement of services is high, there should be an inverse relationship to service level indirect cost measurement and where measurement uncertainty is theoretically opposing expectations. The theory suggests that the indirect cost measurement of a service is a complex organizational choice that is influenced by more than just the standard technical criteria and subject to differences in transaction costs throughout the hierarchy of the public organization. This observation is similar to that made by Julnes and Holzer in their study of performance measurement. They note that performance utilization is often influenced by factors beyond the typical rational-technocratic factors (Julnes & Holzer, 2001 697). If this is the case, more attention needs to be paid to non-technical factors, such as transaction costs and governance factors, which can have long term impacts on which services measure cost.

RESEARCH EXPECTATIONS FOR THE DETERMINANTS OF COST ACCOUNTING OF SERVICES

The research expectations to be tested come from a simplification of the theoretical model (Figure 2). While the full model shows that slack resources and governance factors mediate the effect of fiscal stress on the use of cost accounting, the available cross sectional data on cost accounting practices will not allow for a test of interdependence between the concepts. The effect of slack resource and governance factors influencing transaction costs would be difficult to test with only 30 cities. Furthermore, the long term relationship between indirect cost measurement at the service level reducing fiscal stress for the organization is not well suited to test by cross-sectional means. Ideally, this relationship would be tested using longitudinal data; however, there is almost no longitudinal data on cost accounting practices. For these reasons, I have chosen to test only the relationships between concepts marked by the solid lines in Figure 2.

First, high levels of fiscal stress are thought to positively influence the use of cost accounting and more services having their indirect cost measured. Where organizations have high levels of fiscal stress, governments will tend to have more indirect costs measured. Furthermore, the presence of slack resources will also facilitate the use of more service level indirect cost measurement because the organizations with slack resources should be able to buy off dissenters or provide funding for other valued objectives to the department and service level managers. While there are practical reasons for a negative relationship between fiscal stress and slack resources, this interdependent relationship is simplified to one of individual effects.

At the service level it is expected that economic transactions costs will be negatively related to indirect cost measurement of services. High asset specific services are expected to have lower likelihood of indirect cost measurement. Furthermore, the uncertainty dimension of service measurability is also expected to lead to a lower likelihood of indirect cost measurement because services that are difficult to measure should also be more difficult to develop accurate indirect cost drivers to measure the indirect costs. For both dimensions, theory predicts that indirect cost measurement is more likely to be negatively related to transactions costs.

The governance factors discussed will also have an effect upon the use of indirect cost measurement of services. Professional leadership, hierarchical organizational structure, larger organizations, and those with positive union relationships are all more likely to use more extensive indirect cost measurement. Performance measurement by both the organizational leaders and department managers is also likely to lead to more indirect cost measurement because they have indicated their willingness to measure a service.

EMPIRICAL MODEL OF SERIVCE LEVEL COST ACCOUNTING

The forgoing analysis uses the cost accounting plans of 30 large US cities to determine which services measured their indirect costs at the service level. The 64 services that I analyze and the 30 cities create the problem of a multi-level or hierarchical data structure. The probability of a service being selected in some cities for cost measurement is greater than in other cities, which can cause the error terms to be correlated and violate the traditional assumptions of regression. Estimating hierarchical or multilevel models is the preferred estimation strategy for statistical models with such nested data structures (Gelman & Hill, 2007; Raudenbush & Bryk, 2002) (22). The following equations model the service level and group level estimators on the probability of a service having an indirect cost measure in the cost accounting plans for each service.

Equation 1:

The Hierarchical Logistic Regression Model of Indirect Cost Measurement

Pr([Y.sub.ij] = 1) = [logit.sup.-1]([[beta].sup.0] + [[beta].sup.AssetSpec] * [AssetSpec.sub.i] + [[beta].sup.uncertainty] * [Uncertainty.sub.i] + [[beta].sup.Expend] * Expend i + [[beta].sup.Deptperf] * [DeptPerf.sub.i] + [mathematical expression not reproducible] [mathematical expression not reproducible] ~ N ([[beta].sup.OrgHier] * [OrgHier.sub.j] + [[beta].sup.Manager] * [Manager.sub.j] + [[beta].sup.personnel] * [Personnel.sub.j] + [[beta].sup.Orgperf] * [OrgPert.sub.j] + [[beta].sup.TaxLimit] * [TaxLimit.sub.j] + [[beta].sup.RTW] * [RTW.sub.j] + [[beta].sup.GovSlack] * [GovSlack.sub.j] [mathematical expression not reproducible])

The statistical model is a hierarchical logistic regression to fit the probability that a service(i) has a cost driver in its cost accounting plan with the predictor variables indicated by theory. The service level predictor variables, which are described more fully in the data section, are asset specificity (AssetSpec), outcome measurement uncertainty (Uncertainty), departmental expenditure (Expend), departmental performance (DeptPerf), and a random effects term to account for the variation the city level. At the group or city level, organizational hierarchy (OrgHier), city manager forms of government (Manager), personnel (Pers), organizational performance measure (OrgPerf), cities operating under tax limits (TaxLimit), cities in right-to-work states (RTW), and government slack as measured by fund balance (GovSlack). The statistical estimates of the model were conducted with the lme4 package of the R statistics package version 2.15.0 (Bates & Sarkar, 2007).

DESCRIPTION OF THE DATA

This analysis uses data from two requests made to the finance directors of large U.S. cities over 100,000 citizens to provide their organizational cost accounting plans. The first request was sent in November 2011 and the follow up request was sent two weeks later. The response to these two requests was 30 cities out of the 272 total cities. The response rate of just over 11% is fairly typical for cost accounting research (23). The method of multi-level modeling has been shown to be efficient and useful for generalizing information to the population with groups as small as 5 and units per group of more than 10 (Austin, 2010).

This sample of cities comes from all regions of the U.S. and the size or citizen range of the cities varies from cities with just over 100,000 citizens to the largest cities in the country with over a million citizens. Table 1 shows that the sample cities tend to be slightly larger than the overall population of large U.S. cities, but the sample is not significantly different. Also, there are more cities from the south than other regions of the U.S., but this difference is also not significantly different.

The outcome variable is the presence or absence of an indirect cost driver in the cost accounting plans for 64 services that have been studied in relationship to transaction costs (Brown and Potoski 2003; 2005). The presence of a cost driver to measure the indirect costs of the service was coded as a 1 and the absence of the cost driver in the cost plan was coded as a 0. Because cities provide a variety of service and some cities do not provide all of the services, the analysis is run on only services that that can be found in the city's budget. This removal of services that the cities do not provide results in 1122 remaining services to be analyzed.

The explanatory variables of interest are the levels of asset specificity and uncertainty for 64 services taken from Brown and Potoski's work on service contracting (Brown & Potoski, 2003). Their work indicates that services that have high levels of specialized equipment and training have high levels of asset specificity. For example, the highest levels of asset specificity are for services such as hazardous materials disposal and for the operation of airports and hospitals. These services have the highest level of asset specific technologies and skills. The lowest asset specific services are those for which little training or technology is required such as secretarial service and parking lot operation and maintenance. The measure of asset specificity is taken from managers' perceptions of asset specificity and is measured on a 1 to 5 scale. Table 2 shows that the mean asset specificity for the remaining 1122 services is 3.05, which is slightly less than the mean for all services of 3.12.

The measure for uncertainty are also taken from Brown and Potoski's work. Their service measurability scale asked managers to measure how difficult it is to measure a service on a 1 to 5 scale with 1 being easy and 5 being difficult. Services such as public health and mental health services are some of the highest uncertainty services because it is so difficult to measure the outcome of these programs. It would require separate and extensive evaluation to determine the outcome of these types of programs. Conversely, building and grounds maintenance are some of the lowest uncertainty programs because these programs' inputs, processes, outputs, and outcomes can be easily measured. Table 2 shows that the mean measurement uncertainty for the sample is 2.61, which is slightly less than the mean for Brown and Potoski's scale of 2.70.

The other variables were taken from a variety of sources that included city budgets, comprehensive annual financial plans, and other sources. The measure of slack is unreserved fund balance per total expenditure taken from the governmental statements of the city's Comprehensive Annual Financial Report (CAFR). The fiscal constraint measure was measured by the presence or absence of a tax and expenditure limitation on property tax (Yusuf, Fowles, Grizzle, & Liu, 2012). Expenditure is the budgeted departmental expenditure of the service and generally proxies the size of the service. The other variables for the theoretical model were taken from several sources. The variables for manager and hierarchy were taken from the city budget.

The manager variable is a city manager that has executive responsibility and that is not elected. The variable that measures the hierarchy of the organization is the number of levels in the organization chart taken from the budget between the public and the law enforcement department. For example, an elected sheriff would be "1" because there is only one level between the law enforcement department and the public. A mayor with a sheriff that reports directly to her would have two levels. A mayor that has a chief of staff or a city manager that reports to her would be a "3" and so forth. This measure does not measure the hierarchy of individual departments but generally proxies the hierarchy of the organization because almost all local units of government have law enforcement departments and generally give them the most direct access to top elected officials and the public of nearly all departments.

The organizational personnel variable measures the size of the organization and is the number of full-time-equivalents taken from either the budget or the CAFR. The departmental performance measure is measured by the presence or absence of performance measures in the departmental budget for a service. Organizational performance is whether the organization includes performance measurements in the introductory sections of their budget document (Ho & Ni, 2005). The union variable comes from the National Labor Relations Board list of states that have right-to-work laws (NLRB, 2012).

ANALYSIS OF HIERARCHICAL LOGISTIC REGRESSION MODELS

The analysis of transaction costs influences on service level indirect cost measurement reveals that asset specificity is negatively related to the measurement of indirect costs at the product or service level in all models (Table 3). Services that have greater uncertainty surrounding outcomes were thought also to be negatively related to cost measurement because of transactions costs; however, the results show that services that are more difficult to measure actually have a positive and significant relationship with indirect cost measurement of services in all three models. An explanation for these results and those concerning performance measurement is presented in the discussion and conclusion sections.

In terms of the specific estimates of the models, table 3 presents three models that describe the determinants of indirect cost measurement of individual services in the general cost accounting plans. Model 1 is the complete model with all of the variables of interest in the theoretical discussion. Models 2 and 3 are presented because there was a high correlation between cities that have city manager forms of government and the hierarchy of the organization. These two variables are important institutional features that may influence the measurement of indirect costs of services. Therefore, the full model was run with the variable manager and hierarchy omitted. Model 2 omits the variable manager and shows that hierarchy is approaching standard levels of significance. Model 3 omits the variable for hierarchy but the variable for manager stays insignificant at standard levels.

The log-likelihood of the three models shows that Model 3 has worse fit than Models 1 or 2. Models 1 and 2 fit the data equally well, but because Model 2 has fewer variables than Model 1 it has a slightly lower AIC. This suggests that the best fitting model for the data is Model 2. The variance of the city level random effects is minimized in both Models 1 and 2. For these reasons, the preferred model is Model 2 and the results discussed will refer to this model.

In all of the models, asset specificity is significantly and negatively related to the measurement of indirect costs of services and uncertainty is positively and significantly related to indirect cost measurement. In Model 2, the coefficient for asset specificity is -.86 which corresponds to a change in the odds of having an indirect cost measure for the service decreases by a factor of .43. To visualize the effect of asset specificity over the range of measurements (Figure 3), the predicted probability over the range is .323 to just .054. In other words, the probability of cost measurement for the lowest asset specific services is just under one in three. The highest asset specific services measure indirect costs in a little more than 1 in 20 services.

The measure for outcome uncertainty is also statistically significant but in the wrong direction predicted by standard transactions cost theory. When outcome uncertainty increases, the measurement of indirect costs for services increases by 1.03 on the logistic scale. Log transformed into an odds ratio, an increase of 1 unit on the Brown and Potoski measurement uncertainty scale makes indirect cost measurement 2.8 times more likely. In Figure 3 the predicted probability of indirect cost measurement ranges from just .048 to .393. This increasing usage of indirect cost measurement with increasing measurement uncertainty suggests that uncertainty may be a different consideration than asset specificity, a finding which is discussed in the next section.

The use of performance measurement at both the departmental level and at the organization level was significant. However, performance measurement at the departmental level is the opposite sign of performance measurement at the organizational level. The departmental performance measurement variable, which indicated whether the city reported performance measures in the departmental budget for the service, was negative and significant by standard levels of significance. The odds ratio for departmental performance is .51, which means that departments that report output or outcome performance measures decrease the odds of observing an indirect cost measurement by a factor of .51. In contrast, organizations that report performance measures in the introduction or general sections of their budgets are 2.4 times more likely to measure their service level indirect costs. Both variables are significant and will be addressed in the subsequent discussion.

The other governance variables were only modestly significant or were not significant at standard levels. The hierarchy variable of Model 2 approaches the standard .05 level of significance. Substantively the variable is not large but the odds of measuring indirect costs of a service increase by 1.36 times if the hierarchy variable increases by one unit. The other variables in the model are insignificant. Larger departmental expenditure, organizations with more personnel, and cities that are in right-to-work states tend toward less indirect cost measurements. These results are opposite of that predicted by theory but are not significantly different from zero; so, there is little reason to believe that these variables substantively influence the use of indirect cost measurement of services. Professional managers, tax limits, and governmental slack are of the correct sign but are also insignificant.

DISCUSSION OF MODEL RESULTS

These results partially confirm the transactions cost theory and compliment the literature of cost accounting in businesses. First, the results confirm that asset specificity is likely to significantly reduce the development of cost accounting in public organizations. Practitioners and researchers should pay attention to the elements of asset specificity to determine how it will influence transaction costs when using cost accounting. Asset specificity predictably leads to less indirect cost measurement as managers that have asset specific resources can bargain with those resources to reduce cost measurement of their services. These results confirm the theoretical effect of transaction costs for asset specificity.

Service measurability uncertainty was found to be positively related to indirect cost measurement of public services. One interpretation of this finding may be that services with output or outcomes that are more difficult to measure tend to be ones that organizations are interested in measuring their costs more accurately. Programs such as mental health are difficult to measure because there are many factors that influence the final outcomes. These programs may not be ideal for outcome measurement. Instead, it may make more sense to accurately measure the actual cost of providing services and comparing these costs to other available alternatives or a remediableness criterion (Williamson, 1999). So instead of focusing on effectiveness, cost efficiency may be the primary concern of the management. If the service is cost competitive, then the service can be justified by the remediableness criterion and the values determined by the political process (Lindblom, 1965; Wildavsky, 1964). The empirical finding that cost measurement may be substituted for outcome performance measurement is one that should be taken seriously by scholars that advocate a theoretical reason for always measuring the final outcomes of programs (Hatry, 2006; Poister, 2003).

The other interpretation of the finding surrounding uncertainty of service measurability is that the concept of service measurability uncertainty is not fully developed. Likely, there are many different dimensions of uncertainty within service measurability. It is not clear from Brown and Potoski's description and definition of service measurability what part of the programmatic logic model the concept is measuring. Service measurability may be related to the output or the outcome of the service but that is not delineated in the concept. Even the level of process may be captured by this measurement of service measurability. The result is that this concept needs to be explored further and future work should incorporate better measures of uncertainty.

The findings regarding performance measurement also support the first interpretation of the finding for service measurability. It seems logical that managers would want to measure outputs and outcomes. In fact performance management scholars note that measures of efficiency are particularly underdeveloped (Ho & Ni, 2005; Julnes & Holzer, 2001). This is likely because managers have a strong aversion to being held accountable for service costs, which they may not fully control (Simon et al., 1954). In contrast, organizational leaders have a strong incentive to measure costs as rigorously as possible. It reduces their uncertainty in budget and financial management of organizational subunits and it reveals important process related information that is useful to the organization. So, while department and service level managers resist cost measurement, organizational leaders that are interested in data driven management will push to have a more thorough understanding of the direct and indirect costs that lead to the total cost of programs and services. More qualitative and case study work should be done on the competing incentives among managers in the preferred tradeoff between cost and performance measurement.

CONCLUSION

The results of this study extend the understanding of transactions costs theory into new domains for public organizations. Asset specificity reduces cost measurement at the service level as predicted by theory, but the finding that service measurability uncertainty led to more service level indirect cost measurement is also interesting. Departmental managers and organizational leaders have different preferences about what should be measured. Services that have outcomes that are difficult to measure might be more suitable for cost measurement of the resources that are used to produce the service. More work and theory should be developed around the concept of remediableness for public resource allocation, and accurate cost measurement is a necessary first step towards a remediableness criterion. To fully understand remediableness though, we will need to fully understand the transactions costs in the cost measurement process. This research has taken a first step in this direction by showing that indirect cost measurement is related to transactions costs at the service level.

This research is a first step in understanding cost measurement in public organizations and one that is important given the continuing fiscal stress that cities and public organizations continue to experience. The most logical extension for this work is to continue to look at transactions cost dimensions and future work should focus on the different dimensions of uncertainty. Future work should also look at how training and other governance factors might importantly mediate the effect of transaction costs on cost accounting. Additionally, transaction cost theory might be extended to other financial management tools beyond cost accounting such as performance measurement and the general budget process.

REFERENCES

Ammons, D., Coe, C, & Lombardo, M. (2001). Performance-comparison projects in local government: Participants' perspectives. Public Administration Review, 61(1), 100-110.

Ammons, D. N., & Carter, L. (2000). Benchmarking as a performance management tool: Experiences among municipalities in North Carolina. Journal of Public Budgeting, Accounting, and Financial Management, 72(1), 106-124.

Ammons, D. N., & Rivenbark, W. C. (2008). Factors influencing the use of performance data to improve municipal services: Evidence from the North Carolina benchmarking project. Public Administration Review, 68(2), 304-318.

Anderson, S. W. (1995). A framework for assessing cost management system changes: The case of activity based costing implementation at General Motors. Journal of Management Accounting Research, 7(1), 1-51.

Anderson, S. W., & Young, S. M. (1999). The impact of contextual and process factors on the evaluation of activity-based costing systems. Accounting, Organizations and Society, 24(1), 525-559.

Anderson, S. W., & Young, S. M. (2001). Implementing management innovations: Lessons learned from activity based costing in the US automobile industry: Springer Netherlands.

Austin, P. C. (2010). Estimating multilevel logistic regression models when the number of clusters is low: a comparison of different statistical software procedures. The International Journal of Biostatistics, 6(1).

Baird, K., Harrison, G., & Reeve, R. (2007). Success of activity management practices: the influence of organizational and cultural factors. Accounting & Finance, 47(1), 47-67.

Baird, K., Harrison, G. L., & Reeve, R. C. (2004). Adoption of activity management practices: a note on the extent of adoption and the influence of organizational and cultural factors. Management Accounting Research, 15(4), 383-399.

Baldwin, C. Y. (2008). Where do transactions come from? Modularity, transactions, and the boundaries of firms. Industrial and Corporate Change, 17(1), 155-195.

Bates, D., & Sarkar, D. (2007). Ime4: Linear mixed-effects models using S4 classes.

Baumol, W. J. (1993). Health care, education and the cost disease: a looming crisis for public choice. Public Choice, 77(1), 17-28.

Berman, E., & Wang, X. (2000). Performance measurement in US counties: Capacity for reform. Public Administration Review, 60(5), 409-420.

Brimson, J. A., Antos, J., & Collins, J. (1999). Driving value using activity-based budgeting: Wiley.

Brown, R. E., Myring, M. J., & Gard, C. G. (1999). Activity based costing in government: possibilities and pitfalls. Public Budgeting & Finance, 19(2), 3-21.

Brown, T. L., & Potoski, M. (2003). Managing contract performance: A transaction costs approach. Journal of Policy Analysis and Management, 22(2), 275-297.

Brown, T. L., & Potoski, M. (2005). Transaction costs and contracting: The practitioner perspective. Public Performance & Management Review, 28(3), 326-351.

Clingermayer, J. C., & Feiock, R. C. (1997). Leadership turnover, transaction costs, and external city service delivery. Public Administration Review, 57(3), 231-239.

Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386-405.

Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1-44.

Coe, C. K. (2007). Governmental and nonprofit financial management. Viena VA: Management Concepts Inc.

Coe, C. K., & O'Sullivan, E. (1993). Accounting for the hidden costs: A national study of internal service funds and other indirect costing methods in municipal governments. Public Administration Review, 53(1)9-64.

Crozier, M. (1964). The bureaucratic phenomenon: University of Chicago Press.

Dahlman, C. J. (1979). The problem of externality. Journal of Law and Economics, 22(1), 141-162.

Ernst and Young. (2003). 2003 Survey of Management Accounting. IMA. Retrieved from http : //www.imanet.org/docs/default-source/general/2003surveyofmgtaccting-ey.pdf?sfvrsn=2

Fernandez, S., & Rainey, H. G. (2006). Managing successful organizational change in the public sector. Public Administration Review, 66(2), 168-176.

Finkler, S. A. Purtell, R, Calabrese, T., & Smith, D. (2012). Financial management for public, health, and not-for-profit organizations (4 ed.): Prentice Hall.

Foster, G., & Swenson, D. W. (1997). Measuring the success of activity-based cost management and its determinants. Journal of Management Accounting Research, 9, 109-142.

Geiger, D. R. (2010). Cost management and control in government, fighting the cost war through leadership driven management. New York: Business Expert Press.

Gelman, A., & Hill, J. (2007). Data analysis using regression and multilevel/hierarchical models: Cambridge University Press New York.

Gosselin, M. (1997). The effect of strategy and organizational structure on the adoption and implementation of activity-based costing. Accounting, Organizations and Society, 22(2), 105-122.

Hatry, H. (2006). Performance measurement: getting results: Urban Institute Press.

Ho, A. (2006). Accounting for the value of performance measurement from the perspective of Midwestern mayors. Journal of Public Administration Research and Theory, 16(2), 217.

Ho, A., & Ni, A. (2005). Have cities shifted to outcome oriented performance reporting?--A content analysis of city budgets. Public Budgeting & Finance, 25(2), 61-83.

Innes, J., & Mitchell, F. (1995). A survey of activity-based costing in the UK's largest companies. Management Accounting Research, 6(2), 137-153.

Johnston, J. M., & Girth, A. (2012). Government contracts and "managing the market": Exploring the costs of strategic management responses to weak vendor competition. Administration and Society, 44(1).

Julnes, P. L., & Holzer, M. (2001). Promoting the utilization of performance measures in public organizations: An empirical study of factors affecting adoption and implementation. Public Administration Review, 61(6), 693-708.

Kaplan, R. S., & Atkinson, A. A. (1998). Advanced management accounting (Vol. 3): Prentice Hall Englewood Cliffs, NJ.

Kaplan, R. S., & Cooper, R. (1998). Cost & effect: using integrated cost systems to drive profitability and performance: Harvard Business Press.

Kaplan, R. S., & Norton, D.P. (1992). The Balanced Scorecard-Measures That Drive Performance, Harvard Business Review, January-February, 71-79.

Kehoe, J. (1995). Activity-based management in government: Coopers & Lybrand.

Krackhardt, D. (2001). Viscosity models and the diffusion of controversial innovation. In A. Lomi & E. Larsen (Eds.) Dynamics of organizations: Computational modeling and organization theories: AAAI Press.

Langlois, R. N. (2006). The secret life of mundane transaction costs. Organization Studies, 27(9), 1389-1410.

Levin, J., & Tadelis, S. (2010). Contracting for government services: theory and evidence from US cities The Journal of Industrial Economics, 58(3), 507-541.

Levine, C. H. (1985). Police management in the 1980s: From decrementalism to strategic thinking. Public Administration Review, 45, 691-700.

Levine, C. H., Rubin, I., & Wolohojian, G. G. (1981). The politics of retrenchment: How local governments manage fiscal stress: Sage Publications Beverly Hills, CA.

Lindblom, C. (1965). The intelligence of democracy: Free Press New York.

Lorig, A. N. (1936). Cost accounting and the classification of municipal expenditures. The Accounting Review, 11(3), 291-295.

Malmi, T. (1997). Towards explaining activity-based costing failure: accounting and control in a decentralized organization. Management Accounting Research, 8, 459-480.

March, J. G., & Olsen, J. P. (1989). Rediscovering institutions: The organizational basis of politics: Free Press.

Marlowe, J. (2009). Method of Sale, Price Volatility, and Borrowing Costs on New Issue Municipal Bonds. Retrieved from: http://ssrn.com/abstract=1333396 or http://dx.doi.org/10.2139/ssrn.1333396

McCue, C. P. (2001). Local government accountants as public managers: An evolving role. State & Local Government Review, 33(2), 144-157.

Melkers, J. (2003). Incorporating performance measures into the budget process. In T. H. Poister (Ed.), Measuring Performance in Public and Non-Profit Organizations. San Francisco: Josey-Bass.

Melkers, J., & Willoughby, K. (1998). The State of the States: Performance-Based Budgeting Requirements in 47 out of 50. Public Administration Review, 58(1).

Melkers, J., & Willoughby, K. (2005). Models of performance measurement use in local governments: Understanding budgeting, communication, and lasting effects. Public Administration Review, 65(2), 180-190.

Moe, T. M. (2009). Collective bargaining and the performance of the public schools. American Journal of Political Science, 53(1), 156-174.

Mohr, Z. (2015). An analysis of the purposes of cost accounting in large US cities. Public Budgeting and Finance, 35(1), 95-115.

Mohr, Z. (2016) Performance Measurement and Cost Accounting: Are They Complementary or Competing Systems of Control? Public Administration Review, 76(4), 616-625.

Moynihan, D. P., & Pandey, S. K. (2005). Testing how management matters in an era of government by performance management. Journal of Public Administration Research and Theory, 15(3), 421.

Mullins, D. R., & Zorn, C. K. (1999). Is activity based costing up to the challenge when it comes to privatization of local government services? Public Budgeting & Finance, 19(2), 37-58.

NLRB. (2012). What is a right to work state? Retrieved 7/11/2012, from http://www.nlrb.gov/faq/questions/what-right-work-state

Pinkett, H. T. (1965). The Keep Commission, 1905-1909: A Rooseveltian effort for administrative reform. The Journal of American History, 52(2), 297-312.

Poister, T. (2003). Measuring performance in public and nonprofit organizations: Jossey-Bass Inc Pub.

Premchand, A. (2006). Public expenditure management: Selected themes and issues. In H. Frank (Ed.), Public Financial Management. Boca Raton, Fl: CRC Press.

Purtell, R. M., & Fossett, J. W. (2010). Beyond budgeting: public-service financial education in the 21st century. Journal of Public Affairs Education, 95-110.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (Vol. 1): Sage Publications, Inc.

Rivenbark, W. C. (2005). A historical overview of cost accounting in local government. State & Local Government Review, 37(3), 217-227.

Rivenbark, W. C., & Carter, K. L. (2000). Benchmarking and cost accounting: the North Carolina approach. Journal of Public Budgeting, Accounting, and Financial Management, 12, 125-137.

Rogers, E. M. (2003). The diffusion of innovations (5th ed.). New York: The Free Press.

Shields, M. D. (1995). An empirical analysis of firms' implementation experiences with activity-based costing. Journal of Management Accounting Research, 7(1), 148-165.

Simon, H. A., Guetzkow, H., Kozmetsky, G., & Tyndall, G. (1954). Centralization vs. decentralization in organizing the controller's department: A research study and report. New York: Controllership Foundation.

Stein, R. M. (1990). The budgetary effects of municipal service contracting: A principal-agent explanation. American Journal of Political Science, 34(2), 471-502.

Turney, P. B. (2010). Activity Based Costing: An Emerging Foundation for Performance Management. Cost Management, 24(4).

Weiss, B. (1997). Activity-based costing and management: Issues and practices in local government: Government Finance Officers Association.

Wildavsky, A. (1964). The politics of the budgetary process. Boston/Toronto.

Williams, D. (2003). Measuring government in the early twentieth century. Public Administration Review, 63(6), 643-659.

Williams, D. W. (2004). Evolution of performance measurement until 1930. Administration & Society, 36(2), 131-165.

Williamson, O.E. (1975). Markets and hierarchies: analysis and antitrust implications: a study in the economics of internal organization. New York: Free Press.

Williamson, O. E. (1985). The economic intstitutions of capitalism. New York: Free Press.

Williamson, O. E. (1999). Public and private bureaucracies: A transaction cost economics perspective. Journal of Law, Economics, and Organization, 15(1), 306-342.

Yusuf, J., Fowles, J., Grizzle, C., & Liu, G. (2012). State fiscal constraints on local government borrowing: effects on scale and cost. In H. Levine, J. Justice, & E. Scorsone (Eds.), Handbook of Local Government Fiscal Health. Burlington, MA: Jones and Bartlett.

ZACHARY MOHR

University of North Carolina at Charlotte

(19) For a good overview of this literature see Anderson and Young (1999).

(20) Cost accounting can be integrated with the budget through the use of internal service funds. See Coe and O'Sullivan 1993 for a description.

(21) In the transactions cost literature, it is commonly assumed that the supplier and consumer are exchanging goods and services in a market situation, which is prior to the creation of governance mechanisms that create the multidivisional firm. In the case of extant public organizations, the transactions between supplier and consumer are often between the employee, who supplies human capital, and the public organization that must acquire or consume this capital.

(22) A common measure of the level at which the error terms are correlated is the interclass correlation coefficient or ICC. The ICC for the 30 city cost accounting data exceeds 0.05 or greater. Therefore, a multilevel model is appropriate for this analysis.

(23) Cost accounting research typically has a low response rate. One of the largest cost accounting surveys of businesses and government had a response rate of just over 9% (Ernst and Young, 2003).
Table 1:
City Characteristics

2010 Citizen Range (*)  Sample Count  Region (**)     Sample Count
 (Population %)          (Sample %)   (Population %)   (Sample %)
100,000-200,000              13       Midwest              4
 (60.52%)                 (43.33%)    (15.50%)          (13.33%)
200,000-500,000              10       Northeast            2
 (27.31%)                  33.33%     (10.33%)          (6.67%)
500,000-1,000,000            5        South                13
 (8.86%)                  (16.67%)    (35.06%)          (43.33%)
1,000,000 +                  2        West                 11
 (3.32%)                  (6.67%)     (39.11%)          (36.67%)
Total                        30       Total                30
 (100%)                    (100%)     (100%)             (100%)

(*) Citizen difference is not significant using the t-test of the means
(**) Region difference is not significant using the chi-squared tests

Table 2:
Variable Descriptive Statistics

Variable                  n     mean      sd        min     max

Asset Specificity         1122      3.05      0.64    1.75       4.22
Service Measurability     1122      2.61      0.51    1.53       4.29
Departmental Expenditure  1122  27092     79853       3     816123
Departmental
Performance (*)           1122      0.81      0.39    0          1
Manager (*)               1122      0.69      0.46    0          1
Organizational Hierarchy  1122      3.11      0.83    2          5
Personnel                 1122   3688      3400     731      15038
Organizational
Performance
Measurement (*)           1122      0.68      0.47    0          1
Tax Limit (*)             1122      0.61      0.49    0          1
Right-to-work (*)         1122      0.64      0.48    0          1
Government Slack          1122      0.08      0.09    0.13       0.26

Table 3:
Hierarchical Logistic Regression Model of Indirect Cost Measurement in
CAP

                         Model 1             Model 2
                         -Complete Model     -Without Manager

Determinants             b      t            b      t
Asset
Specificity              -0.86  (-5.5) (**)  -0.86  (-5.5) (**)
Outcome
Measurement
Uncertainty               1.03   (6.0) (**)   1.03   (6.0) (**)
Expenditure              -0.01  (-0.0)       -0.01  (-0.0)
Department
Performance              -0.68  (-2.3) (*)   -0.68  (-2.3) (*)
Organizational
Hierarchy                 0.29   (1.1)        0.31   (1.7) +
Manager                   0.05   (0.1)        -       -
Personnel                -0.04  (-0.8)       -0.04  (-0.8)
Org.Perf.
Measures                  0.80   (2.1) (*)    0.81   (2.1) (*)
Tax Limit                 0.16   (0.5)        0.15   (0.5)
Right to work            -0.24  (-0.7)       -0.25  (-0.7)
Government
Slack                     1.27   (0.7)        1.24   (0.7)
Int.                     -2.78  (-3.4) (**)  -2.80  (-3.5) (**)
Random effects:
Groups                     Variance             Variance
City                          0.29                0.29
Number of observations:    1122
Number of groups:            30
AIC                         933.50              933.50
log-likelihood             -454.8               -454.8

                         Model 3
                         -Without Org. Hier.

Determinants             b      t
Asset
Specificity              -0.85  (-5.5) (**)
Outcome
Measurement
Uncertainty               1.03   (6.0) (**)
Expenditure              -0.09  (-0.1)
Department
Performance              -0.63  (-2.1) (*)
Organizational
Hierarchy                 -       -
Manager                   0.42   (1.3)
Personnel                -0.03  (-0.7)
Org.Perf.
Measures                  0.79   (2.1) (*)
Tax Limit                 0.25   (0.9)
Right to work            -0.24  (-0.8)
Government
Slack                     1.75   (1.1)
Int.                     -2.29  (-3.4) (**)
Random effects:
Groups                       Variance
City                          0.3
Number of observations:
Number of groups:
AIC                          934.70
log-likelihood               -455.3

(**) (Pr|z|>0)<.01, (*) (Pr|z|>0)<.05, +(Pr|z|>0)<J,
COPYRIGHT 2017 Southern Public Administration Education Foundation, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Mohr, Zachary
Publication:Public Administration Quarterly
Article Type:Report
Geographic Code:1USA
Date:Mar 22, 2017
Words:11644
Previous Article:BENCHMARKING PROPERTY TAXES IN A METROPOLITAN AREA.
Next Article:TWENTY YEARS OF BENCHMARKING IN NORTH CAROLINA: LESSONS LEARNED FROM COMPARISON OF PERFORMANCE STATISTICS AS BENCHMARKS.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters