1-2 PAGE HRM FINAL ASS. DUE THURSDAY

http://rop.sagepub.com/ Administration

Review of Public Personnel

http://rop.sagepub.com/content/31/4/349 The online version of this article can be found at:

DOI: 10.1177/0734371X11408698

May 2011 2011 31: 349 originally published online 2Review of Public Personnel Administration

Sally Coleman Selden and Robert Wooters State Governments

Structures in Public Human Resource Management: Shared Services in

Published by:

http://www.sagepublications.com

On behalf of:

Administration

Section on Personnel Administration and Labor Relations of the American Society for Public

can be found at:Review of Public Personnel AdministrationAdditional services and information for

http://rop.sagepub.com/cgi/alertsEmail Alerts:

http://rop.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://rop.sagepub.com/content/31/4/349.refs.htmlCitations:

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

What is This?

- May 2, 2011 OnlineFirst Version of Record

- Dec 7, 2011Version of Record >>

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Review of Public Personnel Administration 31(4) 349 –368

© 2011 SAGE Publications Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

DOI: 10.1177/0734371X11408698 http://roppa.sagepub.com

1Lynchburg College, Lynchburg, VA, USA

Corresponding Author: Sally Selden, Lynchburg College, 1501 Lakeside Drive, Lynchburg, VA 24501, USA Email: [email protected]

Structures in Public Human Resource Management: Shared Services in State Governments

Sally Coleman Selden1 and Robert Wooters1

Abstract

This article explains an emerging trend in state human resource management, the adoption of shared services. According to a 2007 survey of state governments (n = 42), approximately 38% of states adopted a shared services model between 2005 and 2007 and another 26% of other states are considering adopting shared services. Shared services is a blended model which addresses challenges associated with dominantly decentralized and centralized human resource management systems by capitalizing on new technologies and sharing expertise. The study finds that both organizational and political factors drive adoption.

Keywords

shared services, centralization, decentralization, HRM services

The human resources function in public organizations continues to play a significant role in government. The effective recruitment, development, organization, and man- agement of human resources are important drivers of organizational effectiveness (Lawler & Boudreau, 2009). As state leaders become more focused on managing for results and coping with the constraints imposed by increasingly tight budgets, the question facing them is how to guide the state’s operating departments and agencies— and the people comprising them—toward the overall goals of the state, while allowing

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

350 Review of Public Personnel Administration 31(4)

these departments and agencies enough flexibility to carry out their unique missions in providing services to the citizens of the state.

Historically, states have sought to reform their human resource management (HRM) function by centralizing or decentralizing decision-making power between the central HRM agency and line agencies. Recently, states have started adopting a model that balances centralized and decentralized delivery of human resources—shared services—which has diffused through the private sector (Ulrich, Younger, & Brockbank, 2008). For example, on February 10, 2009, Governor Gregoire of Washington issued directive 09-02 requiring state agencies to help develop and implement a shared ser- vices model for state government.1 She stated, “I expect that our new shared services approach and governance structure will capture the benefits of economies of scale in a way that ensures good customer service to the client agencies” (State of Washington, Office of the Governor, 2009). The state’s goal is to identify and revamp those pro- cesses that have the greatest potential for improvement and cost savings.

A shared services model of HRM creates a centralized service function that treats employees and agency-based HRM professionals as internal customers. This approach is designed to enable a government to better leverage existing resources, to reduce duplication of HRM activities across state agencies, and to provide more consistent, higher quality services to internal customers by concentrating existing resources and streamlining processes. Building on the existing knowledge base, this article analyzes the adoption of shared services in HRM by state governments. First, the article explains the shared services model. Second, it provides examples of states who have adopted shared services. Finally, using data collected by the 2007 Government Performance Project, it examines what factors drive a state to adopt shared services.

Centralized and Decentralized Structures in State HRM The centralized approach to HRM in the United States developed after the passage of the Pendleton Act in 1883, which created the federal merit or civil service system (Coggburn, 2005). States began following the federal government’s lead by designing centralized personnel systems to ensure stability and to insulate state employees from political influence and corruption (Coggburn, 2005; Hou, Ingraham, Bretschneider, & Selden, 2000; Sylvia, 1989). A centralized HRM system concentrates power and authority over the state civil service or personnel system in a central personnel agency, which supports the standardization of both HRM policies and procedures and offers increased efficiency through economies of scale (Coggburn, 2005). Centralized personnel systems, however, are not without drawbacks. Centralized human resource systems are often criticized for being complex, slow, rigid, and generally nonrespon- sive to specific agency needs, particularly when it comes to public personnel admin- istration (Golembiewski, 1965). For example, critics note that a centralized recruitment process, although in theory may be more effective, takes too long to screen and identify candidates. As a result, the most qualified applicants may have accepted other

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 351

positions by the time a job offer is made. Similarly, a centrally administered grievance and appeals process can be seen as barrier to an effective disciplinary process, where terminating problem employees for performance problems can be time-consuming and almost impossible (Coggburn, 2005).

Decentralized HRM systems offer state agencies more flexibility by delegating personnel authority to state agencies and relaxing controls imposed by the central personnel agency (Kellough & Selden, 2003). When decisions regarding HRM func- tions such as recruitment and selection, compensation, and discipline are decentral- ized, agency managers have the flexibility to modify their HRM processes to fit the specific goals and needs of the agency (Coggburn, 2005). These processes may be more efficient as well; when HRM decisions are made by line managers and agency HRM professionals, they do not have to go through the bureaucracy of the central personnel agency, making them more timely (Ban, 1995). In decentralized systems, the central human resources agency takes on a more of a consultative role, offering guidance and support to agencies but ultimately leaving the final decisions up to agency managers and HRM staff (Coggburn, 2005).

There are potential problems associated with a decentralized approach to HRM. At the top of the list is the increased risk of political abuse (one of the main reasons for a centralized HRM system) that may result (Coggburn, 2005). There may be less uniformity and fairness in HRM decisions under a decentralized system (Kellough, 1998). Traditional civil service job protections may be eroded as well, with agency personnel being given significantly greater supervisory powers (Hays & Sowa, 2006). Another possible pitfall is that central HRM professionals that are trained in specific functional areas (such as recruitment or compensation) may not be well equipped to become the HRM generalists that they need to be when the central HRM agency moves into the role of consultant and no longer provides those specific services. Moreover, agency personnel may not have the necessary training to effec- tively carry out their new responsibilities (Coggburn, 2005). Finally, in highly decen- tralized HRM systems, a state may lack the information needed to evaluate both the costs and the results of HRM practices, much less plan for their human resource needs within a statewide context.

The equilibrium of forces (such as pressure for standardization, consistency, flex- ibility, agility, etc.) driving states to adopt a more centralized or decentralized approach to HRM is constantly changing as the relative strength of such forces change (Coggburn, 2005; Lawrence, 1983). In recent years, there has been a trend toward the decentralization of the human resources function in states (Hays & Sowa, 2006), which was fueled, in part, by the National Commission on State and Local Public Service’s 1993 report (The Winter Commission) that advocated the decentralization of personnel systems in state government (Kellough & Selden, 2003). The results of a 2005 patterned interview survey of all 50 state offices of HRM found that 16 states (32%) report that they have a significantly (if not totally) decentralized personnel system, and another 24 (48%) are at least partially decentralized (Hays & Sowa, 2006). Research based on the first iteration of the Government Performance Project in

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

352 Review of Public Personnel Administration 31(4)

1998 found that decentralization is more likely to occur in states with a high demand for services from their citizenry as well as in states where central personnel offices report directly to the governor of the state (Hou et al., 2000). Conversely, decentral- ization is less likely in states with “politically charged” governments (i.e., those that are divided politically between the governor and the legislature) and in states with a high level of union activity (Hou et al., 2000).

Balancing Centralization and Decentralization: Shared Services A relatively new form of HRM delivery is shared services, which bridges a centralized and decentralized approach to HRM (Cooke, 2006). Bergeron (2003) defines it as a

collaborative strategy in which a subset of existing business functions [human resources] are concentrated into a new, semiautonomous business unit [HRM shared services agency] that has a management structure to promote efficiency, value generation, cost savings, and improved service for the internal customers [state agencies] of the parent corporation [state government]. (p. 3)

By centralizing or consolidating HRM activities within one agency, a state can eliminate the duplication of services across state agencies, leverage economies of scale, and make HRM processes more efficient (Bergeron, 2003; Corporate Leadership Council, 2006).

Table 1 below outlines differences between centralized, shared services, and decentralized approaches to structuring HRM within state government. With a fully centralized “functional” HRM system, a staff of specialists design standardized HRM policies and practices for all state agencies. This model offers a high degree of state- level control and economies of scale but at the expense of being less responsive to particular agency needs (Bergeron, 2003). Often, given the economies of scale, states are better able to afford the latest technology (Bergeron, 2003).

On the opposite end of the spectrum is the decentralized model which allows each state agency to design and manage its HRM system (Ulrich et al., 2008). As there is no central locus of control, a state agency has the flexibility to change its HRM practices to the extent allowed legally. Although agencies have more control, it can result in HRM redundancies throughout state government and a lack of consistency of HRM practices across state agencies (Ulrich et al., 2008).

The shared services model is a hybrid approach that shares characteristics with both the centralized model (e.g., economies of scale, access to technology, a staff of experts) and the decentralized model (e.g., responsive to customer/agency needs, agil- ity). The shared services agency is responsible for designing a set of HRM policies and delivering HRM services to all or part of state government. Such a model allows HRM leaders to modify the existing way they deliver HRM services and to incor- porate flexibilities within HRM policies, procedures, and practices, where legally appropriate, so agencies can best meet their HRM needs (Ulrich et al., 2008). The

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 353

shared services agency offers a set of HRM services from which state agencies can select. One feature that distinguishes a shared services model from a centralized HRM model is its service culture (Quinn, Cooke, & Kris, 2000). Typically, services offered by the shared services agency are evaluated regularly and subsequently modified based on feedback from the “customer” agencies (Quinn et al., 2000).

Table 1. Centralized HRM Versus HRM Shared Services Versus Decentralized HRM

Dimension Centralized HRM Shared services Decentralized HRM

Design of HRM policies/ processes

Designed by central HRM specialists

Designed by specialists in shared services unit

Designed by local HRM professionals (standards may vary across local units)

Implementation of HRM practices

Governed by central HRM specialists

Governed by agency HRM professionals who select from a menu of services (designed by HRM shared services agency)

Governed by agency HRM professionals

Accountability Central HRM Split between agency HRM/managers and HRM shared services agency

Agency HRM

Services orientation

Standardized HRM policies and services across state government (agencies expected to adhere)

Tailored to agency; offer standardized set of service options that agencies can choose to fit their needs

Agency needs and priorities

Flexibility Mandated use of central resources

Allow flexibility as governed by shared services unit and permitted by law

Up to agencies

Skill requirements for HRM

Technical expertise in functional design and delivery (HRM specialists)

Design expertise but also consulting and support expertise

“General” HRM knowledge (HRM generalists)

Role Provide HRM services that are consistent with the central HRM agency’s mandates; enforce central HRM rules

Help agencies to determine which available HRM services best meet their needs

Provide HRM services specific to needs of agency

Source: Part of table adapted from Ulrich, Younger, and Brockbank (2008). Note: HRM = human resource management.

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

354 Review of Public Personnel Administration 31(4)

The role of the shared services agency is to help its customers (agencies and employees) to determine which services and options best meet their specific needs. Thus, power and influence in the shared services model is dispersed into state agen- cies, similar to a decentralized HRM model, rather than being concentrated at the top of the hierarchy in the central HRM agency (Ulrich et al., 2008). On closer inspection of the literature, HRM shared services can be directed toward a range of functions that are often divided into two categories: administrative services (also referred to as transactional services) and professional services (also referred to transformational services; Quinn & Cooke, 1999; Quinn et al., 2000; Ulrich, 1995). HRM administra- tive services are primarily routine HRM activities, often targeting employee transac- tions (such as payroll), whereas professional shared services encompass a broader set of operational and strategic HRM activities (such as performance management and organizational development).

HRM administrative services. HRM administrative services include all of the prac- tices involved in the day-to-day administrative management of the workforce and are generally routine activities which often involve employee transactions. Typi- cally the services are high volume, are able to be standardized and centralized, and are enabled by adoption of new technologies. These include benefits and payroll administration, training scheduling and registration, employee records activities, and some staffing activities such as employment verification, job posting, and appli- cant flow (Ulrich, 1995).

Typically, this category of shared services offers economies of scale and provides faster, higher quality service to state departments and employees, particularly through the use of information technology (IT) systems (Ulrich et al., 2008). This requires these HRM processes to be standardized, which ensures uniformity of service to all employees, and curtails redundancies and duplications. Benefits and payroll are two processes that are often delivered by HRM shared service function (Corporate Leadership Council, 2006). Another example is a standardized procedure for training registration that replaces several different methods, which can result in significant cost savings and more efficient service to registrants (Ulrich et al., 2008).

Technology also enables employee self-service systems that provide employees with 24-hr access to a wide variety of information and transactional services, ranging from submitting benefits claims to tracking vacation days to managing their retirement plans. Through online self-service portals, employees may be able to resolve as many as 60% of their HRM questions or transactions; of the remaining queries, customer service representatives at the service center can generally sort out the great majority, leaving only a handful for case managers to deal with. Estimates of the cost savings realized by this type of “tiered solution” are as high as 50% (Ulrich et al., 2008).

Professional services. Professional services are neither routine nor administrative, and they more fundamentally reshape or transform the relationship between the state agencies and the central HRM agency. These activities encompass operational and strategic processes and include but are certainly not limited to organizational design and effectiveness; compensation programs such as pay for performance, rewards and recognition, and gain sharing; developmental activities including leadership

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 355

development and personal development planning; and staffing activities such as sourcing candidates, succession planning, and career planning (Quinn & Cooke, 1999; Quinn et al., 2000; Ulrich, 1995).

Clearly, the more operational and strategic nature of professional services demands great expertise in these HRM functional areas. These types of activities are handled by a central shared services agency, sometimes referred to as a center of excellence (or center of expertise) or human resources enterprise, where individuals and teams with a high level of knowledge in specific functional areas come together into a shared service agency that state agencies can use to solve human resources problems. These HRM professionals are specialists in their technical area and are able to apply best practices to specific issues as they arise. This allows them to act as advisors and con- sultants to their “customers,” government agencies, and employees (Quinn & Cooke, 1999; Quinn et al., 2000; Ulrich, 1995).

The professionals that staff such a shared services agency have several roles in the organization. As noted in the previous section, they typically create a menu of services that they offer to client agencies, and they diagnose problems and recommend services that are most appropriate for a given situation (Quinn & Cooke, 1999; Quinn et al., 2000; Ulrich et al., 2008). They collaborate with HRM professionals in individual departments or agencies to select and to implement the most appropriate services. Agency HRM professionals are expected to select from the menu of services, which helps ensure consistency in HRM practices statewide. However, if the current slate of services is insufficient to meet the agency’s needs, the HRM shared service agency experts design additional options (Ulrich et al., 2008).

For example, suppose that the Department of Corrections finds that employees are dissatisfied with the quality of the performance feedback they receive from their supervisor. HRM professionals within the Department of Corrections approach the HRM shared services agency for assistance. The central HRM shared services agency has a menu of possible options to address the problem, including conducting a per- formance appraisal workshop for the agency, an online program on performance feedback that supervisors can complete, and a mentoring program. The central HRM shared service agency staff helps the Department of Corrections HRM staff deter- mine the option that best meets their needs. The agency HRM professionals are responsible for making the selection. If the Department of Corrections and the central shared service agency staff do not believe any of the options is sufficient to address the problem, the design experts within the shared services agency will develop a new offering for the agency, which, subsequently, will be added to the menu of services available to the entire state government.

Examining Shared Services in the States: Data and Analysis This study uses data collected by the Government Performance Project in September 2007. Central state human resource leaders completed an online survey about state human resource practices. Forty-two states completed the survey (an 84% response

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

356 Review of Public Personnel Administration 31(4)

rate).2 States indicated whether they had adopted shared services in the past 2 fiscal years or whether they planned to do so in the next 2 fiscal years. Furthermore, states were asked to provide a description of reforms adopted. This analysis draws on both the qualitative data and quantitative data. The appendix presents the descriptive statistics of the variables included in the multivariate analysis. The study uses ordinary least squares regression to analyze factors associated with state adoption of shared services. Before presenting the multivariate analysis, this article classifies states’ efforts into the two aforementioned approaches to shared services: administra- tive and professional.

Of the 42 responding states, 16 (38%) indicated that they have implemented at least some level of shared services, with 11 more planning to do so. Of the states implementing shared services, the approach used by 7% of states is classified as administrative shared services and 31% implemented changes that are classified as professional shared services. On closer inspection, we find that the scope and struc- ture of shared services within the professional services category vary. As shown in Figure 1 and Table 2, we classified each state effort into one of four approaches, which range along a continuum from least to most complex (in scope and structure). Approximately 10% of states created shared professional services for a single HRM operational process (such as training). The scope and structure of such efforts are similar to that of the aforementioned transactional shared services. Thus, we com- bined transactional shared services and professional shared services for a single HRM process/activity into one category. In total, almost 17% of states adopted shared ser- vices for a specific transaction or HRM process. In this category, we expect the shared service entity to demonstrate a high level of competence in a particular administrative or operational process.

States adopting shared services for a subset of agencies will need a central shared services agency that demonstrates competencies across multiple operational and stra- tegic HRM processes. As illustrated in Figure 1, approximately 10% of states adopted a HRM shared services agency for a subset of state agencies (Approach 2). Seven percent of states implemented a free-standing HRM shared services agency focused solely on HRM processes for all state agencies (Approach 3). Two states housed its

Figure 1. Diffusion pattern of HRM shared services in state government

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 357

HRM shared services within a larger shared service function (Approach 4). Next, the article provides examples of the four approaches presented in Figure 1.

Approach 1—Selected HRM Processes/Activities/Transactions Several states have implemented HRM shared service centers, leveraging technology to realize economies of scale in handling transactional activities. Arizona is one example—the state implemented a HRM Service Center to provide centralized sup- port for their Human Resources Information System (HRIS) and to process selected HRM transactions centrally. The center provides help desk support to all users of the HRIS and has begun to assume data entry responsibilities for selected transactions, resulting in improved staff productivity and HRIS data quality. An example of this is new hire transactions; the state estimates that the agency workload for processing new hire actions has been reduced by 90% (Government Performance Project, 2007).

Michigan operates the Human Resource Service Center, which provides a single point of contact for employees to obtain quick and consistent answers to their most common HRM questions and a single point of contact for enrollment in the state’s group insurance programs. The center provides statewide centralized and standardized

Table 2. Shared Services by State

Selected HRM processes, activities or transactions Arizona Louisiana Michigan New Mexico Oklahoma Oregon Wisconsin Shared HRM services for selected agencies Connecticut Indiana Tennessee Virginia Shared HRM services for all agenciesa

Georgia Maryland Massachusetts Utah Shared HRM services within a larger shared services function Iowa

Kansas

Note: HRM = human resource management. a.Massachusetts is excluded from the analysis because they did not fully complete the survey.

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

358 Review of Public Personnel Administration 31(4)

delivery of routine HRM processes and transactions ensuring consistency across department lines (Government Performance Project, 2007). Employees can contact customer service representatives from 7:00 a.m. to 6:00 p.m., Monday through Friday, or access online services 24 hr a day, 7 days a week. The Human Resource Service Center handles transactions such as enrollment in benefit programs, tax withholding changes, address updates, and wage assignments. It is also responsible for the annual statewide open enrollments for the state’s group insurance programs, the State Employees Combined Campaign (SECC), and the flexible spending accounts. In FY 2006, 21,000 transactions were processed (Government Performance Project, 2007).

Approach 2—Shared HRM Services for Selected Agencies Some states have implemented shared services only for smaller agencies. This is an excellent way to support these agencies, which often have very specialized missions but rarely have the resources for a full dedicated HRM staff. For example, Virginia’s Department of Human Resource Management Service Bureau provides HRM services to agency management and employees in small agencies. Services include compensa- tion, recruitment, employee relationships, Equal Employment Opportunity (EEO) policy and procedure, training, benefits administration, and policy development. The bureau currently provides services to 13 agencies and requires a memorandum of understanding with each agency that outlines the program areas supported and duration of the contract.

Another example is Tennessee’s Shared Services Solutions (SSS) that provides fiscal, procurement, and HRM services for small state agencies (those having fewer than 100 employees). SSS began in July 2007 with 2 client agencies and is now working with 16. The state does not mandate that smaller agencies contract with SSS, and it believes that this was one key to the success of the program: letting small- agency managers participate voluntarily based on the value proposition offered by SSS (Government Performance Project, 2007). Although the program was initially subsidized by the state to eliminate high start-up costs for participating agencies, SSS is now nearly 100% funded by client revenues. SSS has earned high ratings in customer satisfaction, with all participating agencies indicating that they were either “satisfied” or “delighted” with the services provided by SSS in the first quarter of 2009 (C. White, personal communication, May 27, 2009). Other key success factors included a full year of planning prior to implementation, regular consultation with agency direc- tors that helped to define the mission of SSS and build trust, and offering initial short- term services that allowed customer agencies to “test drive” the shared services program. SSS has provided these smaller agencies with higher quality services at a lower cost and is now well positioned to assist them in the implementation of a new enterprise resource planning system in the coming months.

A third example is Connecticut’s Department of Administration Services. In 2005, it created the Small Agency Resource Team (SmART), consolidating the personnel,

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 359

payroll, and affirmative action activities of 22 small agencies. Since its inception, SmART has instituted new procedures for in-service training, tuition reimbursement, employee discipline, new employee orientation, and retirement processing; imple- mented new attendance policies at three agencies; and developed generic training and informational materials pertaining to conducting performance appraisals, progressive discipline, recruitment and selection, and workers’ compensation.

Approach 3—Shared HRM Services for All State Agencies There are states that have implemented a fully integrated HRM shared services model that is used by a broader set of agencies throughout the state. With the goal of improving the effectiveness of HRM while simultaneously reducing costs, the Massachusetts Human Resources Division (HRD) has streamlined its HRM service delivery through a shared services model over the past 3 years. The state uses an HRM advisory council, made up of secretariat-level HRM directors, to help imple- ment and facilitate the new collaborative approach to HRM. Agencies typically implement their own HRM reforms, whereas HRD provides assistance through research, development, best practices for implementation, and ongoing consultation and evaluation. The state notes that two-way communication between HRD and agency stakeholders is an essential part of the process. Since the introduction of the shared services model, major initiatives have included management compensation reform, improvements to the hiring process, and the development of a new online performance appraisal system. According to the state, HRD has realized a reduction in overtime costs, duplication, and administrative work and has also seen a reduc- tion in turnover in the HRD workforce, which the agency believes is a result of its employees having more varied and broader responsibilities that make their work more engaging (Commonwealth of Massachusetts, 2005).

Approach 4—HRM Shared Services Within a Larger Shared Service Function As illustrated by the Washington example highlighted in the beginning of this article, states can chose to use a shared services model for other support functions in addition to HRM activities, such as financial services, security, information technology, and procurement. In 2005, the Kansas Department of Administration began migrating toward a shared services model for the delivery of personnel services, accounting and reporting services, facilities management, printing, surplus, and purchasing. The Division of Personnel Services now works in cooperation and partnership with agency HRM professionals, rather than making unilateral HRM decisions for agen- cies. The state believes that the shared services concepts enable the state to provide a higher quality of services to its customers (Government Performance Project, 2007). The Department of Administration leadership team made certain that its employees

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

360 Review of Public Personnel Administration 31(4)

understood the new vision of the department to ensure that state agencies were treated as customers with a diverse set of challenges to be met (Government Performance Project, 2007).

The state of Iowa created the Department of Administrative Services (DAS) as a way to better manage and coordinate the delivery of human resources, general services (such as maintenance, procurement, and the state vehicle fleet), information technology, and accounting services to state agencies, with the goal of improving services, lowering costs, and creating more flexibly to meet agencies’ needs. The state established four “enterprises” as businesses within state government using the “entre- preneurial management” concept, a customer-focused approach to delivering services in a competitive marketplace. The human resource enterprise offers a wide variety of services to state agencies, including labor relationships, employment services, workforce planning, classification and compensation studies, employee recognition programs, and performance appraisal management. These services are categorized as either “utility” or “marketplace” services. Utility services are those that maximize efficiency for the state through economies of scale; state agencies are required to purchase these services from DAS, though they do have input into the rate setting process. Marketplace services are those where agencies can choose to use outside vendors (such as labor relationships or HRM training) and are paid for by the agencies as they use the services (Government Performance Project, 2007).

Explaining the Adoption of Shared Services in the States Given that states vary in terms of using shared services, naturally, the question arises as to what factors predict the diffusion pattern of shared services. Our depen- dent variable is the HRM shared services continuum presented in Figure 1. Building on the work of Kellough and Selden (2003) and Selden (2006), this article explores the impact of organizational and environmental factors on state adoption of shared services. The rationale for including these variables is discussed below.

Organizational Context Studies of HRM reform often use organizational features as key predictors of reform (e.g., Coggburn, 2005; Hou et al., 2000; Nigro & Kellough, 2006). Like Hou et al. (2000), we focus on the organizational context of the HRM agency.

As HRM professional staff are likely to focus on technically driven changes to a state’s civil service (Nigro & Kellough, 2006), this study considers the professional certifications of a state’s central HRM staff . Specifically, it examines the percentage of staff with certifications from the Society of Human Resource Management (SHRM) and the International Public Management Association for Human Resources (IPMA-HR). Certification by both organizations requires mastery and currency of knowledge, including leading trends, in HRM. Thus, we expect HRM expertise to be higher in states with more professionally certified HRM professionals. Moreover,

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 361

employees seeking out such certification are likely to be more actively engaged in the HRM professional communities sponsoring those certifications (Coggburn, 2005). States with more HRM professionals holding either SHRM or IPMA-HR certification are likely to implement more changes to their personnel system. This study also exam- ines whether the central human resource structure, the number of HRM staff per 100 state employees and the reporting structure of the central HRM department’s director, affect the adoption of shared services. We expect states with more HRM expertise, as measured by having a larger representation of HRM professionals, to be more likely to implement shared services (Coggburn, 2005).

The study also considers to whom the director of the central HRM agency reports. We posit that states whose central HRM director reports to the governor possesses more administrative influence, will cultivate more momentum and support for change, and therefore will be more likely to adopt shared services.

Like Selden (2006), this study includes the 2005 GPP People grade as a proxy mea- sure of the performance of a state’s human resource system. Selden (2006, 2009) found that higher performing states were more likely to implement and to experiment with HRM reforms. By altering human resource operations, reformers expect to improve government performance. Similarly, Brudney, Hebert, and Wright (1999) found that state structural reform efforts appeared to facilitate implementation of reinvention reforms in state agencies. This study expects that states with higher HRM performance, as measured by the 2005 GPP, will be more likely to adopt a HRM shared services model. There is no overlap between the 2005 GPP People grade and the initiatives examined in this study because this study predicts HRM shared services implementa- tion between FY 2006 and FY 2007, which is after the data-collection period covered by the 2005 GPP.

Finally, because the Corporate Leadership Council (2002) found that technology advancements are a primary driver behind the growth of HRM shared services, we include a measure indicating whether a state has adopted a HRM web portal. States with more advanced technologies in place, such as web portals, which promote information sharing and a single access point for employees and agencies, have already begin to change the relationship between the central HRM agency and state agencies and their employees. Organizations adopting HRM web portals have aggregated content from different sources to build a platform for sharing HRM knowledge and information (Ebrahim & Irani, 2005; Nielsen, 2008). As far back as 1958, Leavitt and Whisler forecasted that automated data-processing technologies would result in recen- tralization. Lawrence (1983) highlighted that information has been an important force influencing pendulum swings between centralization and decentralization. Thus, we expect that states that use a HRM web portal are more likely to adopt shared services.

Environment Context Although two objectives of shared services are to provide better quality and more cost-effective HRM services to state agencies and their employees, decisions to adopt such reforms are often political rather than technical (Hou et al., 2000). According to

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

362 Review of Public Personnel Administration 31(4)

Nigro and Kellough (2006), a second driving force for state civil service changes is political. Reforms serve as a means for elected officials to demonstrate to their con- stituents that they are improving the efficiency and effectiveness of their state civil service systems (Nigro & Kellough, 2006). Governors with more institutional power are more likely usher in reforms (Hovey & Hovey, 2005). Furthermore, we expect states with a greater percentage of political appointees serving in the executive branch to be more likely to adopt shared services.

Both practitioners and scholars often mention that reforms are more difficult to implement in states with more unionization or collective bargaining (e.g., Selden, 2006). States that allow collective bargaining need union support to implement many types of HRM reforms. Moreover, states that allow collective bargaining often operate “dual, and sometimes, conflicting personnel systems,” which creates a more complex environment to navigate change (Kearney, 2006, p. 82). Collective bargain- ing personnel systems often face criticisms, such as being rigid, slow, and protective of their employees, similar to complaints made about civil service systems (Kearney, 2006). This study posits that percentage of state government employees covered by collective bargaining agreements will be negatively associated with the adoption of shared services.

Finally, a state’s economic environment may influence the pressures placed on a state from its citizenry. States with less unemployment may be performing better financially, and therefore government officials may have access to resources needed to implement new changes (Kellough & Selden, 2003). However, tight labor markets may drive states to implement reforms to compete more effectively for labor (Selden, 2006). Selden (2006) found that states with higher unemployment were more likely to make changes to personnel authority and hiring policies.

Results As illustrated in Table 3, the independent variables included in the multivariate analysis explain a moderate degree of the observed variations in the shared services continuum (R2 = .46, adjusted R2 = .30). Two of the organizational variables included in this analysis significantly influence state adoption of shared services. As we expected, states that performed better, as measured by the 2005 GPP human resources grade, are more likely to adopt shared services. This finding seems to suggest that states that have been identified as leaders are more likely to accomplish structural reform of their HRM service delivery. One possible explanation is that state human resource leaders may pay more attention to emerging practices and may be driven to implement a shared services model in an effort to continuously improve HRM ser- vices. Furthermore, leaders in these states may have cultivated not only a culture of innovation but also the support of key stakeholders who are quicker to take action (Brudney et al., 1999).

A HRM web portal also has a relationship to the adoption of shared services in states in the predicted direction. As hypothesized, the regression coefficient for the

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 363

HRM portal demonstrates a positive and significant association with adoption of shared services. States without such information sharing technologies may have addi- tional obstacles to overcome if they want to implement shared services (Corporate Leadership Council, 2006; Florkowski & Olivas-Luján, 2006).

In terms of the environmental factors considered in this study, three factors are statistically significant: the institutional power of the governor, the percentage of political appointees in the executive branch, and the percentage of state employees covered by a collective bargaining agreement. States whose governors have more institutional power are significantly more likely to adopt more complex approaches to shared services. In addition, states with a greater proportion of political appointees in the executive branch are significantly more likely to implement structural changes to the way they deliver their HRM systems. It is likely that the implementation of shared services is aligned with the governor’s agendas, which may include improving economies of scale, efficiency, and customer service. In interviews with state

Table 3. Analysis of States Adopting Shared Services

B SE

Organization Percentage of HRM professionals with

IPMA or SHRM certification 0.02 0.02

Number of HRM professionals per 1,000 executive branch employee

−36.19 55.88

State HRM director reports to governor 0.20 0.44 2005 GPP grade 0.93*** 0.38 Utilization of HRM web portal 0.92** 0.52 Environment Institutional power of governor 0.96** 0.54 Percentage of executive branch

employees who are political appointees 0.28*** 0.10

Percentage of employees covered by a labor union

−0.01** 0.00

Unemployment rate 2006 −0.09 0.24 Constant −6.6** 3.30 Fv 2.5** R2 .48 Adjusted R2 .30

N 42

Note: HRM = human resource management; IPMA = International Public Management Association for Human Resources; SHRM = Society of Human Resource Management. *Significant at .10 level. **Significant at .05 level. ***Significant at .01 level.

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

364 Review of Public Personnel Administration 31(4)

political and administrative officials in 2007, the Government Performance Project found that a number of governors were particularly attuned to state government man- agement and were personally involved in improving management of their states agencies. However, unlike Hou et al.’s (2000) work, we did not find a relationship between central HRM directors reporting to the governor and adoption of shared services. However, states with more powerful governors and a larger share of politi- cal appointees in the executive branch are better able to leverage changes in the political and administrative process, which are required to implement more complex approaches to shared services.

As shown in Table 2, unionization is negatively associated with state adoption of shared services. This finding suggests that the likelihood of using shared services is likely to decline as the percentage of unionized employees increases. This finding is consistent with earlier work that demonstrates that unionization is associated with less innovative HRM practices (Donahue, Selden, & Ingraham, 2000; Kellough & Selden, 2003). The result suggests that unions may be a barrier to structural reform and hence the adoption of shared services in states.

As shown above, this study supports Nigro and Kellough’s (2006) observation political motives are key drivers of changes to state’s HRM systems. Furthermore, our study demonstrates that unions continue to challenge states’ abilities to implement changes to their HRM system.

Conclusion This research demonstrates that both the political and management infrastructure of a state’s HRM system influences whether a state implemented shared services within its HRM function. This study found that about 38% of states adopted shared services between FY 2006 and FY 2007 and many others were considering adoption of shared services. A shared services approach seeks to combine the best of a centralized HRM system and a decentralized approach to HRM. A HRM shared services agency func- tions as a single source of information, expertise, and support for agencies, with the goal of improving the quality and consistency of the way HRM issues are addressed within state agencies. Future studies should examine if states that adopt less complex approaches to shared services and that experience success in performing specific administrative or professional shared services are more willing subsequently to employ a more comprehensive shared services approach to HRM and other manage- ment functions in the state.

This approach illustrates how partnerships are being formed within state govern- ments to provide better services to internal customers. Under the shared services model, central HRM professionals work in collaboration with agency and department HRM professionals to determine what services they need. Agencies are granted greater autonomy to make HRM decisions, whereas the central HRM office takes on a more supportive, consultative role. This approach is not as radical as reforms identified by

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 365

Battaglio and Condrey (2006), which threaten the job security of public employees. Instead, the shared services model seeks to modernize state HRM systems by using new technologies and partnerships and providing greater flexibility and better sup- port services to agencies and departments. This movement is consistent with the trend of adopting reforms based on private sector practices to make government more busi- ness-like, which as Bowman (2009) has observed may or may not best serve the needs of government and its citizens. One feature of shared services that has not widely dif- fused at the state level is that of allowing agencies to decide whether to use the shared services unit or contract with an outside vendor for the service. When state shared services units have to compete with outside vendors, it may provide more incentive for them to develop and to deliver higher quality and more cost-effective services. Further studies must examine further the impact and diffusion of shared services, especially when agencies are allowed to select vendors from the marketplace.

Appendix Descriptive Statistics

M SD Minimum Maximum

Adopted shared services 0.76 1.19 0 4 Organization Percentage of HRM

professionals in central office with IPMA or SHRM certification

10.02 12.62 0 57.14

Number of HRM professionals per 1,000 executive branch employee

12.18 4.30 4.72 21.59

State HRM director reports to governor

0.62 0.49 0 1

2005 GPP grade 2.66 0.56 1.3 4.0 Utilization of HRM web portal 0.78 0.42 0 1 Environment Institutional power of governor 3.45 0.39 2.5 4.10 Percentage of executive branch

employees who are political appointees

1.57 2.28 0.03 10.29

Percentage of employees covered by a labor union

44.1 39.3 0 95

Unemployment rate 2006 4.4 0.97 2.9 6.9

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

366 Review of Public Personnel Administration 31(4)

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 would like to thank Pew Center on the States for supporting this project. The views expressed are those of the authors and do not necessarily reflect the views of the Pew Center on the States/ Government Performance Project, or The Pew Charitable Trusts.

Notes

1. The governor’s directive extends beyond just human resource management shared services. 2. The Government Performance Project administered an online survey to states in the sum-

mer 2007. Forty-two states completed the survey as of December 31, 2007. Alaska, Florida, Hawaii, Kentucky, New York, South Dakota, Rhode Island, and Texas did not complete the survey and are excluded from this analysis.

References

Ban, C. (1995). How do public managers manage? San Francisco, CA: Jossey-Bass. Battaglio, R. P., Jr., & Condrey, S. E. (2006). Civil service reform: Examining state and local

government cases. Review of Public Personnel Administration, 26(2), 118-138. Bergeron, B. P. (2003). Essentials of shared services. Hoboken, NJ: John Wiley. Bowman, J. (2009). Turbulence in the civil service: Whither the public service ethos. In

S. W. Hays, R. C. Kearney, & J. D. Coggburn (Eds.), Public human resource management: Problems and prospects (5th ed., 327-338). New York, NY: Longman.

Brudney, J. L., Hebert, F., & Wright, D. S. (1999). Reinventing Government in the American States: Measuring and Explaining Administrative Reform. Public Administration Review, 59(1), 19-30.

Coggburn, J. (2005). The benefits of human resource centralization: Insights from a survey of human resource directors in a decentralized state. Public Administration Review, 65, 424-435.

Commonwealth of Massachusetts. (2005). Human Resource Division annual report. Boston: Author.

Cooke, F. L. (2006). Modeling an HR shared services center: Experience of an MNC in the United Kingdom. Human Resource Management, 45, 211-227.

Corporate Leadership Council. (2002). Optimizing the HR shared services function. Washington, DC: Corporate Executive Board.

Corporate Leadership Council. (2006). HR shared services: Determining the scope, scale, and structure. Washington, DC: Corporate Executive Board.

Donahue, A. K., Selden, S., & Ingraham, P. (2000). Measuring government management capacity: A comparative analysis of city human resources management systems. Journal of Public Administration Research and Theory, 10, 381-411.

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

Selden and Wooters 367

Ebrahim, Z., & Irani, Z. (2005). E-government adoption: Architecture and barriers. Business Process Management Journal, 11, 589-611.

Florkowski, G., & Olivas-Luján, M. (2006). Diffusion of human-resource information-technology innovations in US and non-US firms. Personnel Review, 35, 684-710.

Golembiewski, R. (1965). Men, management, and morality: Toward a new organizational ethic. New York, NY: McGraw-Hill.

Government Performance Project. (2007). [Quality of management in state government]. Unpublished raw data.

Hays, S. W., & Sowa, J. (2006). Changes in state civil service systems: A national survey. Review of Public Personnel Administration, 26(2), 102-117.

Hou, Y., Ingraham, P., Bretschneider, S., & Selden, S. (2000). Decentralization of human resource management: Driving forces and implications. Review of Public Personnel Administration, 20(4), 9-22.

Hovey, K. A., & Hovey, H. A. (2005). CQ’s State Fact Finder 2005. Washington, DC: CQ Press. Kellough, J. E. (1998). The reinventing government movement: A review and critique. Public

Administration Quarterly, 22(1), 6-20. Kellough, J. E., & Selden, S. (2003). The reinvention of public personnel administration: An

analysis of the diffusion of personnel management reform in the states. Public Administration Review, 63, 165-176.

Kearney, R. (2006). The labor perspective on civil service reform in the states. In J. E. Kellough & L. Nigro (Eds.), Civil service reform in the states: Personnel policies and politics at the sub-national level (pp. 77-93). New York: State University of New York Press.

Lawler, E., & Boudreau, J. (2009). What makes HR a strategic partner? People & Strategy,32(1), 14-22.

Lawrence, S. (1983). Centralization and decentralization: The communications connection. Incidental paper of the Center for Information Policy Research, Harvard University, Cam- bridge, MA.

Leavitt, H. J., & Whisler, T. L. (1958). Management in the 1980s. Harvard Business Review, 36(6), 41-48.

Nielsen, J. (2008, July 14). Jakob Nielsen’s Alertbox: Enterprise portals are popping. Retrieved from http://www.useit.com/alertbox/20030331.html

Nigro, L., & Kellough, J. E. (2006). The states and civil service reform: Lessons learned and future prospects. In J. E. Kellough & L. Nigro (Eds.), Civil service reform in the states: Personnel policies and politics at the sub-national level (pp. 315-324). New York: State University of New York Press.

Quinn, B., Cooke, R. S., & Kris, A. (2000). Shared services. Dorchester, UK: Prentice Hall. Quinn, B., & Cooke, R. S. (1999). Adopting and implementing shared services: Manage-

ment accounting guidelines. Hamilton, ON: The Society of Management Accountants of Canada.

Selden, S. (2006). Classifying and exploring reforms in state personnel systems. In J. E. Kellough & L. Nigro (Eds.), Civil service reform in the states: Personnel policies and politics at the sub-national level (pp. 59-76). New York: State University of New York Press.

at WALDEN UNIVERSITY on November 10, 2014rop.sagepub.comDownloaded from

368 Review of Public Personnel Administration 31(4)

Selden, S. (2009). Personnel and human resource management in the states. In S. Hays & J. Coggburn (Eds.), Public personnel administration: Problems and prospectives (5th ed., pp. 46-63). Upper Saddle River, NJ: Prentice Hall.

State of Washington, Office of the Governor. (2009). Directive by the governor 09-02. Olympia: Author.

Sylvia, R. D. (1989). Critical issues in public personnel policy. Pacific Grove, CA: Brooks/ Cole.

Ulrich, D. (1995). Shared services: From vogue to value. Human Resource Planning, 18(3), 12-23. Ulrich, D., Younger, J., & Brockbank, W. (2008). The twenty-first century HR organization.

Human Resource Management, 47, 829-850.

Bios

Sally Coleman Selden is a professor of management at Lynchburg College. She holds the Senior Professional in Human Resources (SPHR) certification and her primary teaching inter- ests are in the areas of public and nonprofit management, human resource management, and leadership. She was selected as the 2009-2010 recipient of the Sydnor Professorship Award for Teaching Excellence in the School of Business and Economics. Dr. Selden is a fellow in the National Academy of Public Administration and author of Human Capital: Tools and Strategies for the Public Sector, and her work has appeared in journals such as American Journal of Political Science, Administration and Society, Review of Public Personnel Administration, Public Administration Review, and Journal of Public Administration Research.

Robert Wooters holds a B.S. in Mathematics from the University of California, Irvine and a Master of Business Administration from Lynchburg College in Lynchburg, Virginia. He was the lead research associate for the People section of the Pew Center on the States’ report Grading the States 2008, a project that evaluated the management processes of state governments. Previous publications include a chapter on Human Resource Management Information Technology in Public Human Resource Management: Problems and Prospects (5th Edition), and contributions to the People Forward series of reports for the Pew Center on the States. He is currently employed as a management consultant in Lynchburg, Virginia.