Abstract
Scholars have long argued clear organizational goals enhance performance whereas ambiguous goals diminish it. Yet, such arguments often overlook the beneficial aspects of ambiguous goals. Therefore, this study proposes and tests a new goal taxonomy that examines the dynamic, mixed effects of goal ambiguity on employees in public organizations. The taxonomy proposed accounts for varying degrees of goal ambiguity and political support. It also considers how different combinations of political support and goal ambiguity affect employees’ role clarity. Findings from a latent class analysis support the proposed taxonomy and indicate role clarity is at its lowest levels when employees confront unsupportive political bodies and ambiguous goals. However, the level of role clarity employees experience in an organization differs when other combinations of political support and goal ambiguity are considered.
INTRODUCTION
One of the most commonly held assumptions in organization theory, research, and practice is that clear goals are a prerequisite to enhanced individual and organizational performance (Chun and Rainey 2005a, 2005b; Jung 2013, 2014; Jung and Lee 2013; Locke and Latham 2002; Pandey and Rainey 2006; Taylor 2013; Wright 2004). This assumption is based on a large body of work, indicating goals are significant for at least two reasons. First, when employee goals are clear as well as difficult but attainable, workers are more motivated to achieve them (Locke and Latham 2002). Second, clear goals help connect employees to organizations by specifying what an organization expects from its workers, how employee action relates to important organizational outcomes, and how workers will be evaluated and rewarded for their job-related performance (Pandey and Wright 2006; Rizzo, House, and Lirtzman 1970; Stazyk, Pandey, and Wright 2011, 610; Wright 2004). Based on these findings, organizations have increasingly pursued and implemented various management systems (e.g., strategic planning, Management by Objectives, and performance measurement schemes) that are intended, at least in part, to clarify goals and employee roles and thereby motivate workers (see e.g., Behrman, Bigoness, and Perreault 1981; Jung and Lee 2013; Pandey and Rainey 2006; Taylor, 2013).
Unfortunately, it appears that the inherently political environment within which public organizations operate serves as a primary source of goal and role ambiguity, the antithesis of goal and role clarity, respectively, (for employees) in the public sector (Lee, Rainey, and Chun 2009; Pandey and Rainey 2006; Pandey and Wright 2006; Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011).1 Not only do public organizations frequently lack market signals that might otherwise help clarify organizational goals (and subsequently, employee roles), but they must also respond to vague and conflicting political mandates that arise from efforts to generate political compromise (Dahl and Lindblom 1953; Lee, Rainey, and Chun 2009; Lowi 1979; Pandey and Wright 2006; Wilson 1989). Owing to these factors, the goals of public organizations are purportedly more numerous, vague, and conflicting than those of private organizations. Efforts to clarify organizational goals and employee roles are subsequently less likely to be successful, thereby undermining efforts to enhance individual and organizational performance (Nutt and Backoff 1993; Rainey 2003; Wright 2001; but see, Boyne 2002; Rainey and Bozeman 2000).
Although the potential costs and pitfalls of ambiguous goals (and, to a lesser extent, roles) have received considerable attention and are well documented in public administration research and practice, much of the existing research overlooks two relatively important points. First, organizations, leaders, and managers confronted with externally derived sources of ambiguity can and often do take steps to clarify goals and roles for employees (Locke and Latham 2002; Milkovich and Wigdor 1991; Rainey and Thompson 2006; Stazyk and Goerdel 2011). Second and more importantly, there is good reason to assume ambiguous goals may, at times, be strategically beneficial to (public) organizations—particularly when “issues are unknown and standards contested” (Noordegraaf and Abma 2003, 869). In this sense, theory provides reason to anticipate higher levels of ambiguity (rather than clarity) may be optimal in certain circumstances. Yet, existing research consistently provides a narrow, one-dimensional view of ambiguity—a view that overlooks, at least empirically, its potential benefits. In practice, we know little about the specific conditions under which ambiguity enhances (rather than diminishes) individual and organizational performance as well as how healthy forms and sources of ambiguity affect other important individual and organizational constructs. To the extent this is true, research and practice may be overly and inappropriately focused on stamping out any source of ambiguity without considering what is lost when doing so.
Consequently, this article offers a first step toward advancing a multidimensional view of ambiguity. Unlike past research, we assume goal ambiguity is likely to be harmful in some circumstances and helpful in others. Drawing on existing goal ambiguity research, we also assume the degree and nature of political support afforded public organizations is a key determinant of whether and when ambiguous goals will be beneficial or harmful to organizations and individuals (see e.g., Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011). Applying this logic, we generate and test a new goal ambiguity taxonomy wherein each category is defined by different combinations of political support and ambiguity. After empirically validating the veracity of our new taxonomy, we then examine how a broader understanding of goal ambiguity relates to employee role ambiguity. Past research indicates goal and role ambiguity are fundamentally related, but distinct constructs; in this sense, linking our taxonomy to role ambiguity is a natural first step in testing the implications of a multidimensional interpretation of (goal) ambiguity. We conclude our article by considering the value of our findings for existing theory and practice as well as how our approach might be adapted for use in other research settings.
ORGANIZATIONS, POLITICAL SUPPORT, AND GOAL AMBIGUITY
All organizations—public and otherwise—depend on externally derived sources of support to survive and thrive. In large part, this need stems from the fact that organizations require resources attainable only through interactions with their external environments (e.g., the market, consumers and other attentive publics, political bodies) to achieve their goals and objectives (see e.g., Carpenter 2001; Lee and Whitford 2013; Pfeffer and Salancik 2003; Rourke 1984). Resources may be tangible, including financial and physical resources, or intangible, such as information, feedback, and reputation (e.g., Carpenter 2001; Granovetter 1985; Lee and Whitford 2013; Pfeffer and Salancik 2003; Rourke 1984; Thompson 1967). Yet, in either case, the character and magnitude of support provided to (or withheld from) an organization serves several important purposes. Besides affording the practical means necessary to accomplish organizational tasks, the nature and extent of support offered helps define and set boundaries around what an organization can and should do (Thompson 1967). When external support and resources are withheld from organizations, it sharply constrains their ability to achieve stated goals and objectives and may signal a need for an organization to adapt to changing environmental exigencies (Carpenter 2001; Gilad, Maor, and Bloom 2013; Granovetter 1985; Lee and Whitford 2013; Mintzberg 1979; Parsons 1956; Pfeffer and Salancik 2003; Thompson 1967).2
Although private and nonprofit organizations generally rely on market signals to assess support for their goals and objectives, public organizations are primarily dependent on sources of political support for evidence of the importance of their missions and overall performance (see e.g., Moynihan and Pandey 2005; Pandey and Wright 2006; Rainey and Steinbauer 1999; Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011; Warwick, Meade, and Reed 1975).3 Consequently, the extent to which politicians perceive value and meaning in an organization’s mission, goals and objectives, and subsequent efforts likely reflects the degree of support politicians are willing to provide an organization. Political support, in this sense, can be understood as “a resource external to the organization paid in the form of financial backing, social support and legitimacy, and policy commitments and direction” (Stazyk and Goerdel 2011, 649). When politicians disagree with an organization’s current direction, they can employ multiple mechanisms—both direct and diffuse—to alter its behavior and mission; such tools commonly include monitoring, (threat of) budget cuts, creation and enactment of administrative rules and procedures, and, in some cases, use of political appointments and appointees (Pandey and Wright 2006; Rainey and Steinbauer 1999; Stazyk and Goerdel 2011; Wolf 1993).
Construed in this manner, the value and importance of political support is difficult to understate. Not only do public organizations operate in an environment characterized by diverse viewpoints and multiple stakeholders, but, when high levels of political conflict are present, public organizations must find some way to mollify disparate interests to attain (or even retain) political support. One method of doing so rests in how organizations construct, shape, and convey their goals (Bozeman 1987; Pandey and Wright 2006; Rainey and Steinbauer 1999; Stazyk, Pandey, and Wright 2011; Wilson 1989; Wolf 1993). Organizational goal ambiguity, which has been defined “as the extent to which an organizational goal or set of goals allows leeway for interpretation,” may be useful here (Chun and Rainey 2005a, 2; Feldman 1989). Ambiguous goals can provide the flexibility and latitude necessary to help organizations respond to and limit political conflict and maintain a supportive resource environment (Chun and Rainey 2005a; Feldman 1989; Pandey and Wright 2006; Stazyk, Pandey, and Wright 2011).
Yet, even the most politically adroit public organization cannot be “all things to all people” (Stazyk, Pandey, and Wright 2011, 610); nor can it fully avoid externally derived sources of ambiguity. In such cases, too much ambiguity can harm organizational performance by making it difficult for employees to understand what an organization hopes to accomplish as well as how it intends to arrive at its goals and objectives (Chun and Rainey 2005a, 2005b; Jung 2014; Jung and Lee 2013; Locke and Latham 2002; Pandey and Rainey 2006; Taylor 2013; Wright 2004). Consequently, although goal ambiguity may be a useful coping mechanism for public organizations operating in unsupportive political environments, the vast majority of research assumes goal clarity is preferable to ambiguity due to its apparent links to individual employee performance.
POLITICAL SUPPORT, GOAL AMBIGUITY, AND EMPLOYEE ROLE CLARITY
One of the primary reasons scholars and practitioners have expressed a preference for goal clarity over ambiguity in organizations is because research consistently demonstrates employees exhibit higher performance when they work toward the completion of goals that are specific and difficult but attainable (Locke and Latham 2002, 2006). Performance gains appear to be a function of two overarching factors. First, employees with clear goals are more likely to be committed to realizing them (Locke and Latham 2002). Second, employees who find value in a goal exert more energy and persistence as they work toward achieving it (Locke and Latham 2002). In this case, clear goals enhance employee motivation by improving individuals’ sense of self-efficacy regarding their ability to actually accomplish tasks that will produce expected outcomes (Bandura 1977; Locke and Latham 1990, 2002; Wood and Bandura 1989).
Interestingly, research also suggests clear goals exist only when managers are involved in shaping and setting goals with and for employees. As Locke and Latham (2002) note, goal clarity and attainment are curtailed when managers rely on generic “do-your-best-goals. . .[that] have no external referent and thus are defined idiosyncratically. This allows for a wide range of acceptable performance levels, which is not the case when a goal level is specified” (706; see also Bandura 1997). In other words, even though employees should play a pivotal role in setting their own goals, goals must be situated concurrently by managers within the broader organizational context and mission to be viewed as legitimate and meaningful by workers. Only when goals are placed in this broader context do they provide a clear sense of direction and purpose for employees. Not surprisingly, research and practice abound with recommendations and strategies managers can employ to clarify and position goals accordingly (see e.g., Locke and Latham 2002; Milkovich and Wigdor 1991; Stazyk and Goerdel 2011).
Unfortunately, because public organizations operate in a complex political environment characterized by widely disparate and often conflicting interests, public administration scholars have raised questions about the efficacy of applying goal-setting strategies and practices originating in the private sector directly to public organizations (e.g., Chun and Rainey 2005a, 2005b; Maynard-Moody and McClintock 1987; Noordegraaf and Abma 2003; Pandey and Wright 2006; Perry and Porter 1982; White 1982; Wright 2001). Concerns reflect a persistent—albeit unproven—belief that public organizations are subject to goals that are more numerous, vague, and conflicting (i.e., ambiguous) than those of private organizations (see e.g., Boyne 2002; Jung 2013, 2014; Rainey and Bozeman 2000; Stazyk and Goerdel, 2011; Taylor 2013; Wright 2001). High levels of goal ambiguity are assumed to make it more difficult for public organizations and managers to know what constitutes an acceptable organizational mission as well as how to craft and delineate organizational expectations; this, in turn, makes it challenging to clearly define job tasks and assess employee performance and can lead to higher levels of job dissatisfaction and turnover intention (Jung 2013, 2014; Jung and Lee 2013; Pandey and Wright 2006; Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011; Taylor 2013; Wright 2001, 2004).
When the effects of goal ambiguity cascade down to the job level, employees may find it arduous to understand what is expected of them in their individual roles—a concept known as role ambiguity (Kahn et al. 1964; Katz and Kahn 1978; Pandey and Wright 2006; Rizzo, House, and Lirtzman 1970; Stazyk, Pandey, and Wright 2011; Wright 2004).4 At the very least, ill-defined goals and roles limit the likelihood an employee will accept and view as legitimate (i.e., doable) any job task or role conferred upon them by the organization (Katz and Kahn 1978). In large part, this is because role ambiguity decreases employees’ sense of self-efficacy (Bandura 1997).
Self-efficacy declines in the face of role ambiguity for two overarching reasons. First, role ambiguity creates an environment in which individuals “lack. . . information concerning the proper definition of the job, its goals, and the permissible means for implementing them” (Kahn et al. 1964, 94). In essence, employees are deprived of key pieces of knowledge that are fundamental in any effort to successfully complete job tasks. Second, ambiguity leaves individuals uncertain of their “standing in the eyes of others and about the consequences of. . .[their] actions” when working within the confines of their roles toward goal attainment (Kahn et al. 1964, 94). Such uncertainty increases the likelihood that employees will experience negative psychological and emotional outcomes on the job, such as anxiety, depression, tension, anger, and fear (Kahn et al. 1964; Rizzo, House, and Lirtzman 1970). Moreover, when considered together, these two factors increase occupational stress and produce deleterious employee-related outcomes, such as lower job performance and increased incidences of turnover, absenteeism, physical ailments and illness, and chronic diseases (Bandura 1997; Rizzo, House, and Lirtzman 1970; Van Sell, Brief, and Schuler 1981).5
Although research examining the interplay between politics, goals, and roles is relatively new, it is worth noting that recent public management research also posits direct links between political support and goal ambiguity as well as goal ambiguity and role ambiguity (see e.g., Erera 1989; Pandey and Wright 2006, 517; Stazyk, Pandey, and Wright 2011; Wright 2001, 2004). In this case, the argument is two-fold. First, the degree of political support provided to an organization shapes the relative clarity or ambiguity of organizational goals. Unsupportive political bodies generate policies and laws that are often viewed as ambiguous by employees (see e.g., Erera 1989; Pandey and Rainey 2006; Pandey and Wright 2006; Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011). Second, the “net result of organizational goal ambiguity is to increase information deficiency about job processes and salient outcomes” (Pandey and Wright 2006, 517). In other words, organizational goal ambiguity increases employee role ambiguity.
RATIONALE FOR THE NECESSITY OF A NEW GOAL TAXONOMY
Given the apparent downsides of ambiguity, it is of little surprise that calls to clarify organizational goals and employee roles abound in public management research and practice. Yet, by focusing only on the disadvantages of goal ambiguity, scholars and practitioners risk generating theories and concrete recommendations that fail to account for its possible benefits. A broader review of the research on organizations and management, however, underscores the seemingly multidimensional nature of ambiguity. In fact, (public) management scholars and practitioners have long noted that ambiguous goals afford organizations the flexibility necessary to adapt to environmental changes and uncertainty (Maynard-Moody and McClintock 1987; Mintzberg 1979; Noordegraaf and Abma 2003; Thompson 1967), take advantage of new opportunities (Gupta, Raj, and Wilemon 1986a, 1986b; Moore 1995), and experiment with policy implementation to improve outcomes (Berman 1978; Maynard-Moody and McClintock 1987; Palumbo, Fawcett, and Wright 1981; Sabatier and Mazmanian 1979). Ambiguity may also allow organizations and organizational members to (re)interpret goals in ways more likely to satisfy the expectations of diverse bodies of politicians and citizens (Matland 1995; Maynard-Moody and McClintock 1987; Moore 1995).
Even within organizations, ambiguity provides certain strategic advantages. For instance, ambiguity may allow organizational leaders to convey a universal message to employees who are otherwise engaged in highly diverse tasks—as is often common in public organizations (Barley, Leonardi, and Bailey 2012; Rainey and Steinbauer 1999). In such cases, additional clarity may be practically impossible or undesirable (Barley, Leonardi, and Bailey 2012). More importantly, ambiguity may afford employees the leeway necessary to “tailor their interpretations of the directive to their differing work functions” (Barley, Leonardi, and Bailey 2012, 284: citing Keleman 2000), thereby providing workers greater job autonomy and discretion (Baier, March, and Saetren 1988; Ginger 1998; Meyer, Becker, and Vandenberghe 2004). Ambiguity can also assist in facilitating collaboration between organizational members working on diverse tasks (Barley, Leonardi, and Bailey 2012; Eisenberg 2007; Gioia and Chittipeddi 1991). Finally, there is some evidence that ambiguity can be used by organizational leaders and managers to limit employee conflict and resistance in the face of large-scale organizational change (see e.g., Barley, Leonardi, and Bailey 2012; Gioia and Chittipeddi 1991; Sitkin and Bies 1993).
That existing research indicating goal ambiguity can be both helpful and harmful to individuals and organizations points to the real need to treat ambiguity as a multidimensional concept. As a multidimensional concept, the pitfalls and benefits of goal ambiguity are likely to be highly contextualized. In the case of public organizations, the nature of (political) support provided will likely have a dramatic impact in shaping whether ambiguity is beneficial or harmful. Consequently, in the next section, we offer a new taxonomy that expands our understanding of the relationship between goal ambiguity and political support. Our taxonomy presumes goal clarity and ambiguity can be both helpful and harmful to individuals and organizations, depending primarily on the degree of political support provided. We also discuss how an expanded goal taxonomy likely relates to employee role ambiguity.
FOUR CONCEPTUAL CATEGORIES OF POLITICAL SUPPORT AND ORGANIZATIONAL GOAL AMBIGUITY AND THEIR INFLUENCE ON INDIVIDUAL ROLE CLARITY
Public management scholarship generally assumes those organizations lacking supportive political bodies struggle in their efforts to cultivate clear goals and roles for employees and suffer performance losses as a consequence (Chun and Rainey 2005a, 2005b, 2006; Jung 2013, 2014; Meier 1997; Pandey and Rainey 2006; Rainey and Steinbauer 1999; Stazyk and Goerdel 2011; Stazyk, Pandey, and Wright 2011). Yet, this assumption overlooks instances where goal ambiguity is, for example, accompanied by supportive political environments. Theoretically, political support and goal ambiguity likely exist in various combinations. As such, figure 1 presents four distinct categories based on several possible combinations of political support and goal ambiguity. Various streams of research in public management—beyond goal ambiguity studies—suggest these four categories exist. However, public management scholarship has yet to develop a coherent framework for studying expected relationships both within and across each category.
Taxonomy Connecting Political Support, Goal, Ambiguity, and Role Clarity
The horizontal axis in figure 1 reflects two overarching goal characteristics: goal clarity and ambiguity. The vertical axis captures whether an organization’s political environment is generally supportive or unsupportive. Together, these factors result in four distinct categories that likely encompass the full spectrum of possible combinations of political support and goal ambiguity, ranging from relatively high degrees of political support and goal clarity to relatively low degrees of political support and goal clarity. Notably, the taxonomy presented in figure 1 is also useful for assessing how various combinations of political support and goal ambiguity will affect employee role clarity. We anticipate each category in the taxonomy will be associated with a predictable level of role clarity (or ambiguity) for public employees. It is, however, important to note that we have adopted a relative conceptualization of role clarity. For example, although Categories III and IV may not have low goal clarity in an absolute sense, they are likely to have relatively lower levels of goal clarity when compared to Categories I and II. We further expect these variations affect employee action and behavior in different and significant ways. Finally, figure 1 also presents illustrative examples of research subsumed under each category. Each category is described below.
Category I: Implementation under Political Solidarity
The first category presented in figure 1 illustrates instances when public employees believe elected officials exhibit a strong, relatively concise policy preference. Category I situations are likely to occur when the solution to a policy problem is straightforward and generally agreed upon despite any broader political differences (e.g., between Republicans and Democrats) or when the political costs of disagreeing with a particular policy solution outweigh the benefits (Moe 1980). In such cases, bureaucrats are likely to perceive strong support for the organization as well as to believe organizational goals are relatively clear (unambiguous).
One example of implementation under instances of political solidarity can be found in cases of “morality politics” or the “politics of sin” (Meier 1999, 1994; Sharp 2005). According to Meier (1999), morality politics occurs when political actors approve one set of values deemed socially acceptable over alternative values deemed perverse or inappropriate. For example, policy issues such as those involving pornography (Smith 1999), alcohol, and drugs (Meier 1994) tend to fall under the heading of the politics of sin. When politicians identify policy issues as involving the politics of sin, they perceive overwhelming public support for more stringent policies and enforcement and are likely to identify only a single viable policy alternative—one likely to be strict on deviant behavior (Meier 1999).
As the morality politics example illustrates, bureaucrats operating in Category I environments receive ample political support and relatively clear organizational goals. If politicians perceive overwhelming public support for a given policy, they are likely to afford the organization the support necessary to strictly enforce or implement policy goals established by political principals. Any public statement from politicians perceived as even remotely opposed to bureaucrats involved in enforcement would likely be politically disastrous. Furthermore, any political directives imposed on bureaucracy under these circumstances are limited to relatively clear instructions for more strict policy enforcement. The roles of bureaucrats within these policy arenas are likely to be relatively clear (Meier 1999). Thus, we offer the following hypothesis:
Hypothesis 1: When bureaucrats believe elected officials are supportive and goals are clear, they will report the highest degree of role clarity.
Category II: Bureaucracy as Technical Expert
Category II reflects situations where public managers operate with the support of elected officials and ambiguous goals. Public management scholars often argue that this combination of characteristics exists when highly professionalized bureaucrats are required to confront extremely technical, complex policy problems (Friedrich 1940; Mosher 1982; Radin 2006; Romzek and Dubnick 1987). In this case, politicians generally lack the time and knowledge necessary to provide clear guidance to an organization and are subsequently more likely to support and trust expert bureaucrats because they believe employees are well trained and highly competent.
Furthermore, ambiguous goals are a precondition of organizational life for employees tasked with addressing complex, technical problems. As Lipsky (1980) notes, “public institutions generally have conflicting or ambiguous goals for good reason. They embrace ambiguity, contradictions, and complexity because. . .society is unable and unwilling to abandon certain fundamental aspirations and expectations in providing public services” (147). Simply, goal ambiguity can be necessary for high-quality service delivery because, at times, bureaucrats rather than politicians better understand the nuances of a policy problem. In this case, ambiguity affords public employees the flexibility, discretion, and autonomy necessary to define and implement public policies on the basis of their unique knowledge, skills, and abilities; this, in turn, is expected to improve outcomes for citizens. Moreover, employees with greater autonomy and discretion are often more highly motivated (see e.g., Bandura 1997; Langfred and Moye 2004).
It is important to note, however, a professional bureaucrat’s individual role within an organization is likely to become somewhat uncertain under prolonged instances of heightened goal ambiguity. Although public employees may initially perceive politicians as granting a high degree of deference to the organization, over the long run, employees must understand what is expected of them as well as the specific criteria against which their job performance is evaluated. As Moynihan and Pandey (2005) argue, clear tasks and goals are essential in efforts “develop a mission-oriented culture, and reduce the potential for rival, confusing, and contradictory management systems and actions” (427). In the presence of protracted task or goal ambiguity role clarity likely suffers. As such, we hypothesize:
Hypothesis 2: When bureaucrats believe elected officials are supportive and goals are ambiguous, they will report relatively lower degrees of role clarity as compared to Category I.
Category III: Legislative Micromanagement of Bureaucracy
The third category captures instances when public employees simultaneously perceive unsupportive political environments and clear organizational goals. As Behn (1995) notes, a lack of trust—one major element of political support—between the legislative and executive branches of government can result in the imposition of numerous rules and regulations on the bureaucracy. Untrusting elected officials, in turn, establish clear, unambiguous goals but also deprive bureaucracies of resources necessary for achieving organizational missions. This trend inevitably results in dismal organizational performance. As organizations continue to underperform, legislatures become more distrusting of bureaucrats and impose a greater number of rules and regulations on bureaucracy. According to Behn (1995), this “micromanagement cycle” encourages the continual decrease of bureaucratic performance.
Although legislative micromanagement may lead to “inevitable bureaucracy” (Lynn 1981) and poor performance, there is little reason to anticipate that any rules and regulations derived from micromanagement also generate ambiguous organizational goals. Rather, organizational goals may grow more specific and clear over time as elected officials impose additional rules specifically intended to further constrain and pattern bureaucratic behavior in ways expected by political principals. However, even in the face of clear organizational goals, public employees confronting an unsupportive body probably experience some degree of role ambiguity. An overabundance of rules and regulations can, in and of itself, create uncertainty for employees, leaving workers unsure which rule or set of rules is most important to follow in any given situation (Frank 1963/4; Hickson 1966; House and Rizzo 1972; Kahn 1964; Morris, Steers, and Koch 1979; Nicholson and Goh 1983; Pandey and Wright 2006; Pearce 1981; Ramaswami, Agarwal, and Bhargava 1993; Rizzo, House, and Lirtzman 1970). Moreover, when legislative micromanagement results in frequent, newly generated rules and regulations, workers may believe job tasks and expectations are volatile and subject to recurrent, unpredictable changes (Frank 1963/4; Hickson 1966; Pandey and Wright 2006; Pearce 1981; Rizzo, House, and Lirtzman 1970). Consequently, we offer the following hypothesis:
Hypothesis 3: When bureaucrats believe elected officials are unsupportive and goals are clear, they will report relatively lower degrees of role clarity as compared to Category I.
Category IV: Deliberative Failure
The final category reflects instances when public employees perceive both an unsupportive political environment and high degrees of goal ambiguity. This dimension most likely occurs under cases of “deliberative failure,” and may be characterized as the worst-case scenario for organizations and individuals. As Meier (1997) notes, politicians often campaign against bureaucracy in their pursuit of office, arguing bureaucracy is too pervasive and unwieldy. Under these circumstances, public employees are inclined to believe elected officials view them as incompetent and untrustworthy (Whorton and Worthley 1981). A general lack of political support need not undermine bureaucratic performance, unless it occurs at times when politicians are otherwise unable to effectively debate and resolve complex policy problems, thereby preventing elected officials from generating and conveying definitive goals to an organization (Finer 1941; Meier 1997; Moe 1982).
As Wildavsky (1988) argues, a lack of consensus in electoral institutions leads to “profound disagreements over equality, democracy, and hence the role of government, disagreements that create conflicting expectations that no conceivable cadre of civil servants can meet” (753). Thus, when public employees work in an environment characterized by unsupportive elected officials and ambiguous organizational goals, organizational performance tends to be drastically undermined. Performance suffers not only because high degrees of subjectivity and uncertainty are injected into the policy arena, but also because bureaucrats themselves are provided little information about what constitutes appropriate organizational action and behavior. In such cases, it is highly unlikely public employees will possess the information required to adequately and accurately comprehend organizational goals and job tasks, as well as how individual action will be evaluated and relates to organizational purposes (see e.g., Moynihan and Pandey 2005; Rainey et al. 1995). Therefore, we expect:
Hypothesis 4: When bureaucrats believe elected officials are unsupportive and goals are ambiguous, they will report the lowest degree of role clarity.
DATA, MEASURES, AND METHODOLOGY
Data for this study come from Phase IV of the National Administrative Studies Project (NASP-IV). NASP-IV included an online survey administered to a nationwide sample of public managers in US municipal jurisdictions with populations of at least 50,000 residents. In addition to city managers and assistant city managers, NASP-IV surveyed department heads in Finance/Budgeting, Public Works, Personnel/Human Resources, Economic Development, Parks and Recreation, Planning, and Community Development. The International City/County Management Association (ICMA) aided in the development of sample design and survey construction. ICMA compiled a contact list for potential respondents based on study criteria. This initial list was augmented by the NASP-IV team to (1) verify information accuracy, (2) update and correct the list as needed, and (3) compile working e-mail addresses for respondents.
These efforts resulted in 3,316 respondents in the study sample. The study protocol was reviewed and approved by the Institutional Review Board at the University of Kansas. Each respondent in the sample received an initial letter through US mail, which introduced the study, provided a secure study participation code for the online survey, and furnished details about participation. Respondents were directed to the study Web site, which provided information about their rights as a study participant. This note highlighted that participation was voluntary, there were no foreseeable risks, and the study team would take all necessary steps to protect the confidentiality of respondents, including conducting analyses and reporting results at the aggregate level only. After the initial letter, multiple methods were used to follow-up with respondents, including e-mail, fax, and phone calls. When the study concluded, 1,538 of the 3,316 had responded, for a response rate of 46.4%. The 1,538 respondents come from 545 different jurisdictions—with 1 respondent from 126 jurisdictions, 2 respondents from 130 jurisdictions, and 3 or more from 289 jurisdictions.
Table 1 illustrates select demographic characteristics of respondents. The mean age of respondents was 51.4 with an interquartile range of 10 (25th percentile being 47 and 75th percentile being 57). As expected in this sample, a sizable majority were male (70.6%), white (85.9%), highly educated (more than 60% with graduate degrees), and well compensated (68.3% with salaries over $100,000). The table also displays the functional specialization of managers: 28.6% were general managers (city managers or deputy/assistant city managers) and the rest managed specific city departments and/or functions. This distribution closely matched the distribution of functional specializations in the sample.
Demographic Characteristics (n = 1,538)
| Percent | |
|---|---|
| Gender | |
| Female | 29.4 |
| Male | 70.6 |
| Education | |
| Some college | 2.9 |
| Bachelors | 32.5 |
| Master’s in public affairs (MPA, MPP) | 37.2 |
| Other graduate degree | 27.4 |
| Race | |
| Hispanic origin | 4.1 |
| Black or African American | 6.7 |
| White | 85.9 |
| Asian | 2.3 |
| Salary | |
| Less than $50,000 | 1.4 |
| $50,000–$75,000 | 6.9 |
| $75,000–$100,000 | 23.5 |
| $100,000–$150,000 | 50.2 |
| $150,000 or more | 18.1 |
| Functional specialization | |
| City manager/chief administrative officer | 14.0 |
| Deputy or assistant city manager | 14.6 |
| Finance/budgeting | 12.3 |
| Public works | 11.6 |
| Personnel/HR | 10.9 |
| Economic development | 7.7 |
| Parks and recreation | 13.7 |
| Planning | 10.9 |
| Community development | 4.3 |
| Percent | |
|---|---|
| Gender | |
| Female | 29.4 |
| Male | 70.6 |
| Education | |
| Some college | 2.9 |
| Bachelors | 32.5 |
| Master’s in public affairs (MPA, MPP) | 37.2 |
| Other graduate degree | 27.4 |
| Race | |
| Hispanic origin | 4.1 |
| Black or African American | 6.7 |
| White | 85.9 |
| Asian | 2.3 |
| Salary | |
| Less than $50,000 | 1.4 |
| $50,000–$75,000 | 6.9 |
| $75,000–$100,000 | 23.5 |
| $100,000–$150,000 | 50.2 |
| $150,000 or more | 18.1 |
| Functional specialization | |
| City manager/chief administrative officer | 14.0 |
| Deputy or assistant city manager | 14.6 |
| Finance/budgeting | 12.3 |
| Public works | 11.6 |
| Personnel/HR | 10.9 |
| Economic development | 7.7 |
| Parks and recreation | 13.7 |
| Planning | 10.9 |
| Community development | 4.3 |
Demographic Characteristics (n = 1,538)
| Percent | |
|---|---|
| Gender | |
| Female | 29.4 |
| Male | 70.6 |
| Education | |
| Some college | 2.9 |
| Bachelors | 32.5 |
| Master’s in public affairs (MPA, MPP) | 37.2 |
| Other graduate degree | 27.4 |
| Race | |
| Hispanic origin | 4.1 |
| Black or African American | 6.7 |
| White | 85.9 |
| Asian | 2.3 |
| Salary | |
| Less than $50,000 | 1.4 |
| $50,000–$75,000 | 6.9 |
| $75,000–$100,000 | 23.5 |
| $100,000–$150,000 | 50.2 |
| $150,000 or more | 18.1 |
| Functional specialization | |
| City manager/chief administrative officer | 14.0 |
| Deputy or assistant city manager | 14.6 |
| Finance/budgeting | 12.3 |
| Public works | 11.6 |
| Personnel/HR | 10.9 |
| Economic development | 7.7 |
| Parks and recreation | 13.7 |
| Planning | 10.9 |
| Community development | 4.3 |
| Percent | |
|---|---|
| Gender | |
| Female | 29.4 |
| Male | 70.6 |
| Education | |
| Some college | 2.9 |
| Bachelors | 32.5 |
| Master’s in public affairs (MPA, MPP) | 37.2 |
| Other graduate degree | 27.4 |
| Race | |
| Hispanic origin | 4.1 |
| Black or African American | 6.7 |
| White | 85.9 |
| Asian | 2.3 |
| Salary | |
| Less than $50,000 | 1.4 |
| $50,000–$75,000 | 6.9 |
| $75,000–$100,000 | 23.5 |
| $100,000–$150,000 | 50.2 |
| $150,000 or more | 18.1 |
| Functional specialization | |
| City manager/chief administrative officer | 14.0 |
| Deputy or assistant city manager | 14.6 |
| Finance/budgeting | 12.3 |
| Public works | 11.6 |
| Personnel/HR | 10.9 |
| Economic development | 7.7 |
| Parks and recreation | 13.7 |
| Planning | 10.9 |
| Community development | 4.3 |
We use all available survey items in NASP-IV to assess the attitudes of respondents regarding political support and goal and role clarity (or, conversely, ambiguity)—each has been validated in previous research. First, two survey items adapted from the work of Gianakis and Wang (2000) are employed to evaluate whether public managers believe they operate in supportive political environments (α = .9378). Second, three items developed by Rainey and colleagues (e.g., Pandey and Rainey 2006; Rainey 1983) are used to examine whether managers perceive their organization as characterized by clear goals (α = .8285). Finally, managers’ attitudes about the relative clarity of work roles is assessed on the basis of three items (α = .8552) from the work of Rizzo, House, and Lirtzman (1970). A detailed review of study measures can be found in the appendix.
The statistical method employed in this article, latent class analysis (LCA), relies on a series of survey items to construct categorical (as opposed to continuous) latent variables. This technique provides at least two advantages in the current study. First, LCA models are particularly useful when researchers intend to examine the extent to which unobserved subgroups in heterogeneous populations influence individual behaviors and attitudes (Goodman 2002; Hagenaars and McCutcheon 2002). Researchers often incorrectly assume that all population heterogeneity has been captured in study variables (Hagenaars and McCutcheon 2002). In the event that respondent differences exist due to unobserved subgroups, an extension of LCA—latent class cluster analysis—affords researchers the ability to sort respondents into unobserved categories based on specific response patterns within the data (Vermunt and Magidson 2002). Second, after establishing clusters, researchers can treat unobserved clusters as model variables in an effort to determine if attitudes or behaviors vary across subgroups. In this sense, latent class analysis is uniquely suited to the goals and objectives of this study insofar as it provides a tool for empirically sorting respondents into, and evaluating the validity of, each class illustrated in figure 1. Furthermore, LCA allows us to test the hypotheses articulated above by examining mean-level differences in role clarity across each subgroup.
FINDINGS
Prior to testing study hypotheses, we evaluated whether there were at least four identifiable latent groups or classes based on the combinations of political support and goal ambiguity illustrated in figure 1. Unfortunately, many of the common fit statistics utilized in traditional latent variable models (e.g., structural equation models) are unavailable in latent class models, complicating efforts to determine absolute model fit. However, measures of relative fit are useful when comparing nested models and selecting a model with an appropriate number of classes.
To determine whether a four class model was empirically justifiable, we assessed model fit using recommendations from existing research. First, we evaluated Bayesian information criterion (BIC) values for models ranging from one to four latent classes (for further discussion, see Vermunt and Magidson 2002). Second, we examined differences between the three and four class model using three significance tests: the Vuong-Lo-Mendell-Rubin likelihood ratio test, the Lo-Mendell-Rubin adjusted likelihood ratio test, and the bootstrapped likelihood ratio test (see Nylund, Asparouhov, and Muthen 2007; Asparouhov and Muthen 2012). Each test provides a significance level useful in comparing fit as additional classes are added to a model. Table 2 provides BIC values as one moves from a one to four class model; table 3 provides results from each significance test.
Model BIC Values
| Number of Classes | ||||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| BIC | 24,957.443 | 22,676.114 | 21,942.181 | 21,297.428 |
| Number of Classes | ||||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| BIC | 24,957.443 | 22,676.114 | 21,942.181 | 21,297.428 |
Model BIC Values
| Number of Classes | ||||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| BIC | 24,957.443 | 22,676.114 | 21,942.181 | 21,297.428 |
| Number of Classes | ||||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| BIC | 24,957.443 | 22,676.114 | 21,942.181 | 21,297.428 |
Significance Results for Likelihood Ratio Tests
| Significance Test | Difference in Parameters | Change in Two Times the Log Likelihood | p |
|---|---|---|---|
| Vuong-Lo-Mendell-Rubin | 6 | 688.777 | <.001 |
| Lo-Mendell-Rubin | 6 | 673.479 | <.001 |
| Bootstrapped Likelihood Ratio Test | 6 | 688.777 | <.001 |
| Significance Test | Difference in Parameters | Change in Two Times the Log Likelihood | p |
|---|---|---|---|
| Vuong-Lo-Mendell-Rubin | 6 | 688.777 | <.001 |
| Lo-Mendell-Rubin | 6 | 673.479 | <.001 |
| Bootstrapped Likelihood Ratio Test | 6 | 688.777 | <.001 |
Significance Results for Likelihood Ratio Tests
| Significance Test | Difference in Parameters | Change in Two Times the Log Likelihood | p |
|---|---|---|---|
| Vuong-Lo-Mendell-Rubin | 6 | 688.777 | <.001 |
| Lo-Mendell-Rubin | 6 | 673.479 | <.001 |
| Bootstrapped Likelihood Ratio Test | 6 | 688.777 | <.001 |
| Significance Test | Difference in Parameters | Change in Two Times the Log Likelihood | p |
|---|---|---|---|
| Vuong-Lo-Mendell-Rubin | 6 | 688.777 | <.001 |
| Lo-Mendell-Rubin | 6 | 673.479 | <.001 |
| Bootstrapped Likelihood Ratio Test | 6 | 688.777 | <.001 |
Findings in table 2 illustrate a four class model outperforms models with fewer classes. Although BIC values report the relative performance of models, they do not provide information about statistical significance. Table 3 indicates that model fit improves significantly (at the .001 level) with the inclusion of a fourth class. Taken together, results demonstrate that a four class model fits the data better and is more appropriate than a three class model. It is important to note that it is possible to justify a greater number of classes based on sample data, but, as Asparouhov and Muthen (2012) note, researchers can find significantly better fit by adding additional classes to a model. As such, we weighed empirical justifications for including additional classes against existing theory and a desire to generate a parsimonious taxonomy before settling on the four class model.
Even though our findings indicate there are four identifiable classes of respondents, it is still necessary to determine whether each class behaves as articulated in the taxonomy. In other words, it is possible that respondents could fall in categories with combinations of characteristics different from those predicted. To determine if respondents fell into the theoretical categories presented in figure 1, we examined the means for the political support and goal ambiguity items in each cluster. Figure 2 provides a graphical representation of each latent class where the x axis represents each questionnaire item and the y axis represents the group mean response for a given item. Generally speaking, response patterns hold relatively close to our expected theoretical categories. One category perceives supportive political environments and clear organizational goals, and another perceives unsupportive political environments and ambiguous organizational goals. The remaining two groups are perhaps more interesting, because one perceives supportive environments and ambiguous goals whereas the other perceives unsupportive political environments and clear goals.
Category Means for Questionnaire Items
Figure 2 also provides information on the proportion of respondents that fall within each category. The largest proportion of respondents falls in Category I (56.8%), which is characterized by supportive political environments and clear organizational goals. The lowest proportion of respondents falls in Category IV (8.4%), which reflects unsupportive political environments and ambiguous goals. Approximately 19% of respondents report ambiguous organizational goals and supportive political environments (Category II), and nearly 16% report clear organizational goals and unsupportive political environments (Category III). These percentages, however, represent model estimations. Assigning respondents to their most likely latent class based on estimates results in 886 respondents in Category I, 290 respondents in Category II, 232 respondents in Category III, and 129 respondents in Category IV.
Given results empirically justify the four theoretical classifications in figure 1, we next examined differences in perceived role clarity across each of the four categories. In this analysis, role clarity is a latent variable measured by three questionnaire items—each of which is discussed in greater detail in the appendix. Table 4 provides the factor loadings associated with the role clarity construct, mean levels of role clarity for each latent class, and the significance levels for mean-level differences between each category.
Factor Loadings and Mean Differences
| Role Clarity Factor Loadings | ||||
|---|---|---|---|---|
| Estimate | SE | Est./SE | p | |
| Role Clarity 1 | 1.000 | 0.000 | 999.000 | 999.000 |
| Role Clarity 2 | 1.127 | 0.050 | 22.752 | 0.000 |
| Role Clarity 3 | 1.302 | 0.049 | 26.778 | 0.000 |
| Category I Mean | ||||
| Role Clarity | 0.000 | 0.000 | 999.000 | 999.000 |
| Category II Mean | ||||
| Role Clarity | −0.581 | 0.067 | −8.612 | 0.000 |
| Category III Mean | ||||
| Role Clarity | −0.397 | 0.072 | −5.513 | 0.000 |
| Category IV Mean | ||||
| Role Clarity | −1.109 | 0.115 | −9.661 | 0.000 |
| Role Clarity Factor Loadings | ||||
|---|---|---|---|---|
| Estimate | SE | Est./SE | p | |
| Role Clarity 1 | 1.000 | 0.000 | 999.000 | 999.000 |
| Role Clarity 2 | 1.127 | 0.050 | 22.752 | 0.000 |
| Role Clarity 3 | 1.302 | 0.049 | 26.778 | 0.000 |
| Category I Mean | ||||
| Role Clarity | 0.000 | 0.000 | 999.000 | 999.000 |
| Category II Mean | ||||
| Role Clarity | −0.581 | 0.067 | −8.612 | 0.000 |
| Category III Mean | ||||
| Role Clarity | −0.397 | 0.072 | −5.513 | 0.000 |
| Category IV Mean | ||||
| Role Clarity | −1.109 | 0.115 | −9.661 | 0.000 |
Factor Loadings and Mean Differences
| Role Clarity Factor Loadings | ||||
|---|---|---|---|---|
| Estimate | SE | Est./SE | p | |
| Role Clarity 1 | 1.000 | 0.000 | 999.000 | 999.000 |
| Role Clarity 2 | 1.127 | 0.050 | 22.752 | 0.000 |
| Role Clarity 3 | 1.302 | 0.049 | 26.778 | 0.000 |
| Category I Mean | ||||
| Role Clarity | 0.000 | 0.000 | 999.000 | 999.000 |
| Category II Mean | ||||
| Role Clarity | −0.581 | 0.067 | −8.612 | 0.000 |
| Category III Mean | ||||
| Role Clarity | −0.397 | 0.072 | −5.513 | 0.000 |
| Category IV Mean | ||||
| Role Clarity | −1.109 | 0.115 | −9.661 | 0.000 |
| Role Clarity Factor Loadings | ||||
|---|---|---|---|---|
| Estimate | SE | Est./SE | p | |
| Role Clarity 1 | 1.000 | 0.000 | 999.000 | 999.000 |
| Role Clarity 2 | 1.127 | 0.050 | 22.752 | 0.000 |
| Role Clarity 3 | 1.302 | 0.049 | 26.778 | 0.000 |
| Category I Mean | ||||
| Role Clarity | 0.000 | 0.000 | 999.000 | 999.000 |
| Category II Mean | ||||
| Role Clarity | −0.581 | 0.067 | −8.612 | 0.000 |
| Category III Mean | ||||
| Role Clarity | −0.397 | 0.072 | −5.513 | 0.000 |
| Category IV Mean | ||||
| Role Clarity | −1.109 | 0.115 | −9.661 | 0.000 |
When utilizing empirically derived latent classes as predictor variables, researchers are limited to testing mean-level differences across categories. Although other methodological techniques (e.g., multiple regression with interaction terms) might provide useful information about the influence of each variable on role clarity, such techniques generally leave open the question of whether definitive categories actually exist. Since we primarily seek to empirically validate each category in our taxonomy, our subsequent analysis is limited to comparing mean-level differences across each category. For the purposes of model estimation, average role clarity was fixed to zero in Category I, indicating all mean-level differences are assessed against a politically supportive environment with clear organizational goals. Theoretically, the characteristics associated with Category I should result in the highest levels of individual and organizational performance whereas all other categories likely result in deviations from optimal performance; this makes Category I an optimal comparison category.
Results in table 4 illustrate that the class of respondents associated with supportive political environments and clear organizational goals report the highest levels of role clarity. Conversely, respondents associated with unsupportive political environments and unclear organizational goals report the lowest levels of role clarity, with a mean difference of −1.109. Additionally, respondents in Categories II and III both report lower levels of role clarity relative to Category I. Those individuals who report supportive political environments and unclear goals (Category II) have a mean-level difference −.581 lower than Category I, and individuals who report unsupportive political environments and clear organizational goals have mean-level differences −.397 lower than Category I. Interestingly, respondents in unsupportive political environments with clear organizational goals report greater role clarity than do those individuals in supportive political environments with unclear organizational goals. This finding marks a new and significant contribution to the literature on political support and goal and role clarity—one that offers a more holistic view of the complex interactions between individuals, organizations, and their external environments. Simply, our findings indicate that clarifying organizational goals may be more important than building political support in any effort to remedy the deleterious effects of role ambiguity, supporting arguments that organizations are not left entirely to the whims of external forces (Stazyk and Goerdel 2011).
DISCUSSION AND CONCLUSIONS
In this article, we have examined (1) the complex relationships between political support and goal ambiguity among a large sample of public organizations and (2) how various combinations of political support and goal ambiguity influence employees’ role clarity. Public management literature often asserts that public organizations require a supportive political environment to reach their goals and objectives and connect employees to the organization. Although existing research has contributed significantly to our understanding of the connection between organizations’ political environments, goal clarity, and role clarity, we argue such relationships are likely to be more complex in practice.
Notably, findings indicate a sizable portion of public managers responding to this survey (34.8%) believe they work in environments characterized by either supportive political environments and ambiguous goals or unsupportive political environments and clear organizational goals. That results demonstrate goal ambiguity (or, conversely, goal clarity) varies across supportive and unsupportive political environments marks an important contribution to public management research in and of itself—a contribution that suggests researchers must cultivate a better understanding of the connection between organizations’ external environments and goals.
Two other findings are also intriguing. First, the majority of respondents (56.8%) perceive politicians as supportive and goals as clear. Second, relatively few respondents (8.4%) report politicians are unsupportive and goals are ambiguous. Taken together, these findings seemingly challenge notions that public administrators generally confront uncertain environments and ambiguous goal directives; instead, public managers in our sample view their environments quite optimistically. It is, however, important to note that the data used to develop this taxonomy reflect cross-sectional responses from senior local government managers, raising three important issues.
First, senior managers in any government organization are likely to have more frequent and substantively different interactions with political principals. The nature of these interactions may provide additional context that affects whether managers view politicians as supportive and goals as clear. Employees at lower levels within an organization are likely to lack this information and may therefore perceive their political environments and goals differently (and with greater variability). Second, it is also possible that local governments, which typically lack professionalized politicians (Nalbandian 1999), provide public managers greater opportunity to help shape goals and objectives. In such cases, politicians may grant greater deference to administrators, and local government employees may experience lower levels of goal ambiguity than employees working at the state and federal levels. Third, political circumstances clearly change over time; cross-sectional data limits our ability to examine how and whether changes in an organization’s political environment affect goal (and role) clarity. As such, future research should examine whether the taxonomy performs differently at various levels within organizations, across other levels of government, and when political environments change.
We also do not examine any factors—individual, organizational, or environmental—that cause managers to perceive goals and political environments as they do. For example, certain types of support may be more significant than others (e.g., financial versus nonfinancial). Similarly, the nature of organizational tasks (e.g., routine versus nonroutine) as well as expertise and professionalization of employees likely affect how employees interpret goals and engage with politicians (see e.g., Meier 1999).
Because we have not directly tested the relationships between organizations’ political environments and goals on individual and organizational performance, we are hesitant to suggest supportive political environments and clear organizational goals are always optimal for public employees. Sometimes public organizations embrace ambiguity for good reasons (see e.g., Lipsky 1980). If future research could determine how, when, and which various factors shape employees’ perceptions of their goals and organizational environments, it might be possible to develop specific management interventions intended to mitigate the potentially negative effects of different combinations of support and ambiguity on individuals and organizations.
Although we have not directly tested the relationships between each category in our taxonomy and organizational performance, we do examine how each category relates to employee role clarity—a construct clearly shown to affect employee performance (Kahn et al. 1964; Rizzo, House, and Lirtzman 1970). Here, results suggest important differences may exist across environments characterized as having supportive politicians and ambiguous goals and environments with unsupportive politicians and clear goals. Managers who perceive politicians as supportive and goals as ambiguous tend to indicate their roles are less certain than managers who perceive politicians as unsupportive and goals as clear. This finding is surprising in light of arguments asserting overly stringent constraints on managers compromise effectiveness by limiting professional discretion (e.g., Radin 2006).
Ultimately, our results suggest that efforts to clarify organizational goals may be more beneficial for managers than attempting to cultivate a supportive political environment—at least with respect to employee role clarity. This finding is significant in that public managers have considerable ability to shape organizational processes, policies, and procedures in ways that can improve goal clarity and thereby mitigate role ambiguity for other mid- and lower-level employees (see e.g., DeHart-Davis 2009; DeHart-Davis and Pandey 2005; Stazyk and Goerdel 2011). For example, research indicates managers can alter organizational structures or (re)design rules and procedures in ways that clarify goals and roles for employees (see e.g., DeHart-Davis 2009; Stazyk and Goerdel 2011). Although high-ranking government executives may not be able to formally change the content of organizational goals established by unsupportive political environments, they may be able to improve role clarity for other organizational members.
Given these conclusions, future research might further examine how different combinations of political support and goal and role clarity influence other important workforce behaviors commonly associated with managing people in the public sector. For example, it would be useful to determine whether high levels of perceived political support attenuate the relationships between goal ambiguity and employee absenteeism, turnover, and motivation. If so, managers might begin to more fully account for these relationships in their efforts to attract and retain employees. It would also be beneficial to consider how the relationships proposed in our taxonomy influence employee job satisfaction and commitment. Employees with stronger job satisfaction and organizational commitment often demonstrate greater interest in, and exert more energy toward, accomplishing collective organizational goals. Given these ends, new data, better survey measures, and alternative methodological approaches are necessary to validate many of the claims and constructs discussed in this article.
Future research should also more fully investigate the complex combinations of political support and goal ambiguity described in Categories II and III of the taxonomy. Although additional research is necessary to validate our results, initial findings suggest that the combination of characteristics in Category I are most likely to enhance performance-related factors such as role clarity, whereas Category IV is likely to inhibit performance. The distinctions between Categories II and III, however, are less certain. In other words, whether one of these categories is distinctly preferable over the other may vary considerably. For example, Category III may be preferable over Category II when attempting to facilitate role clarity, whereas Category II may result in greater organizational commitment among workers than Category III. Future research would do well to more fully examine the conditions under which Categories II and III result in superior outcomes for individuals and organizations.
Ultimately, the taxonomy proposed and tested in this article extends goal ambiguity research considerably by providing new and meaningful insights into the dynamic nature of the public sector goals. On the one hand, our findings clearly support the notion that goal ambiguity is a multidimensional concept. On the other hand, by assuming ambiguity is likely to be both harmful and beneficial to individual and organizational performance, our article also raises considerable questions about how and when goal ambiguity matters. That we are unable to fully resolve this issue in our article points to the real need for public management scholars to more fully consider the benefits and shortcomings of goal ambiguity in future research.
APPENDIX
OPERATIONAL DEFINITIONS*
Political Support
Political support is assessed using two items on a scale from one to seven, ranging from strongly disagree to strongly agree. Political support is based on the following items:
Most elected officials trust the organization.
Most elected officials believe the organization is effective.
Organizational Goal Clarity
Goal clarity (or, conversely, ambiguity) is assessed using three items on a scale of one to five, ranging from strongly disagree to strongly agree. The goal clarity measures include the following items:
This organization’s mission is clear to almost everyone who works here.
This organization has clearly defined goals.
It is easy to explain the goals of this organization to outsiders.
Individual Role Clarity
Role ambiguity is assessed using three items on a scale of one to five, ranging from strongly disagree to strongly agree. The measures include:
My job has clear, planned goals and objectives.
I feel certain about how much authority I have.
I know exactly what is expected of me.
*Items used to assess political support were based on a 7-point scale and items used to assess goal ambiguity were based on a 5-point scale. Although responses could have been adjusted to standardize scales, we avoided this in an attempt to maintain the variability in the original data set. Since each taxonomic category assesses political support and goal clarity independently, these scaling inconsistencies likely had little influence on the results.
REFERENCES
———.
———.
———.
———.
———.
———.
———.
———.
———.
———.
As Rainey (2003) notes, organizational goal ambiguity may originate internally (e.g., as a function of managerial decisions or actions) or externally (e.g., from politicians, citizens, stakeholders, media). For many public management scholars, the extensiveness of externally derived sources of ambiguity helps distinguish public organizations from private ones, but also suggests public organizations are subject to unique and challenging constraints.
Of course, organizations are not left entirely to the whims of external actors and their broader environment. Organizations may take a variety of steps to mitigate environmental uncertainty by, for example, securing new clients or co-opting hostile entities and actors (e.g., Pfeffer and Salancik 2003; Selznick 2011). Although an important point, we limit our current discussion to the function support plays in shaping organizational goal and individual role ambiguity.
For public organizations, political support emanates from political bodies as well as other attentive publics, such as citizens, interest groups, and the media (Carpenter 2001; Rourke 1984). We limit our discussion to the role politicians play in conveying support to public organizations, leaving for future research questions about how other attentive publics shape the behaviors we discuss in this article.
Role ambiguity is only one construct useful in assessing the relationships considered in this article. Another is role conflict. As Tubre and Collins (2000) note, “Role ambiguity occurs when the set of behaviors expected for a role is unclear, and role conflict occurs when there is incompatibility between the expected set of behaviors perceived by the focal person and those perceived by role senders” (156: citing Katz and Kahn 1978). We opt to focus on role ambiguity because it has received more scholarly attention than role conflict. However, it would be useful for future studies to examine how role conflict affects the relationships posited here.
When confronted with high levels of occupational stress, employees often engage in a range of coping strategies intended to mitigate ambiguity. Unfortunately, these strategies are also likely to harm organizational performance (see e.g., Kahn et al. 1964; Rizzo, House, and Lirtzman 1970).
