Conservation of resources (COR) is an integral theory within the stress literature. Its basic tenets are that individuals attempt to gather and protect resources (Hobfoll 1989) in an effort to minimize resource loss. When resources are threatened or lost, individuals will experience stress. Recent work has begun to examine the theoretical implications of COR outside the domain of stress (e.g., Ferris et al. 2007), arguing that the “acquisition and protection of resources parallels the motivational goals of the political processes in organizations” (p. 301). Our research explores the role of COR within the impression management (IM) literature, helping expand COR theory beyond stress-related consequences, such as burnout, job tension, and somatic complaints. Further, researchers are beginning to take interest in one’s ability to achieve higher levels of resources, rather than only considering resource loss and its typically negative consequences (e.g., Fritz and Sonnentag 2006; Wayne et al. 2007). Hence, it is the purpose of this paper to examine the implications of COR theory on non-stress-related outcomes in the workplace, focusing on IM as a mechanism employed in a conscious, goal-directed effort to reach higher levels of resources in organizational settings.

We propose that IM, which is influencing the image one portrays (e.g., Bolino and Turnley 2003), is instrumental in reaching higher levels of social and material resources, and that these higher resource amounts lead to enhanced performance ratings at work. However, we argue that not all individuals who use IM will achieve increased levels of resources. The effort of IM (i.e., self-presentation) can potentially tax resources (Vohs et al. 2005) and subsequently hinder one’s ability to reach desired workplace outcomes. Specifically, “when people must effortfully plan and alter their behavior so as to convey the intended image of self, then success at self-presentation will depend heavily on effective self-regulation” (Vohs et al. 2005). For example, research has shown that politically skilled persons who use high levels of IM receive higher performance ratings from their managers; however, persons low in political skill who use high levels of IM receive lower performance ratings (Harris et al. 2007). In order to further investigate individual differences, we explore one’s ability to regulate and manage resources (e.g., Hochwarter et al. 2008, 2007b). To clarify, IM is considered a potentially resource-draining behavior for many individuals. Energy is considered to be a resource (Hobfoll 1989) and IM takes energy (Vohs et al. 2005), thus potentially depleting one’s resources. However, IM is necessary to acquire other desirable workplace resources (e.g., reputation and positive relationships), ultimately relating to higher performance ratings. We argue that the ability to manage resources (Hochwarter et al. 2008, 2007b) is an individual difference (characteristic or capability) that will allow the effective and efficient use of IM and will lead to desirable resources.

Our research contributes to the extant literature in three ways. First, we are examining an individual difference construct in the ability to manage resources. Researchers in COR theory have explicitly called for future research to examine the impact of individual differences in the COR process (Hobfoll and Shirom 2000). It is our argument that certain individuals are able to manage their resources more successfully and will be more efficient when expending the energy necessary for using IM. This capability or skill set will be related to higher levels of resources compared to those unable to manage their resources. Further, higher resource levels will be associated with positive outcomes for individuals in the form of performance.

Secondly, we examine resources outside of the specific stressor-strain process by examining non-strain related outcomes, such as leader–member exchange (LMX), reputation, and performance. Much of the research on COR has focused on the stressor-strain relationship. We propose that individuals who have higher levels of resources, not only avoid strain responses, but also have higher performance (Wayne et al. 2007). Specifically, as depicted in Fig. 1, we argue that individuals who use IM and are able to manage their resources will have a more positive reputation (Study One) and higher levels of LMX (Study Two). Finally, higher levels of reputation and LMX should be associated with higher performance ratings.

Fig. 1
figure 1

Hypothesized model

Full size image

Thirdly, we use COR theory as an underpinning for the effects of IM in organizations. In their recent review on IM, Bolino and colleagues argued that IM researchers should apply theories in their investigations because one of the major shortcomings of current IM research is that it lacks theoretical grounding (Bolino et al. 2008). We attempt to address this by using the COR framework to guide our arguments about why IM is related to higher levels of resources. Furthermore, as suggested by Bolino et al. (2008), we incorporate Liden and Mitchell’s (1988) model of risk assessment to help explain how one’s ability to manage resources will lead to more desirable outcomes (and thus other desirable resources) when engaging in IM. Additionally, Bolino and colleagues’ (2008) illustrated how certain IM tactics “complement each other in facilitating desired images.” In our studies, we examine the role of positive IM in creating desirable images. That is, we explicitly examine an array of tactics (i.e., ingratiation, exemplification, and self-promotion) that are designed to create a favorable impression on others (Bolino and Turnley 2003).

Theoretical Grounding

Conservation of Resources

According to Hobfoll’s (1989, 2001) COR theory, people attempt to obtain, maintain, and preserve those things that they deem valuable (Hobfoll 2001; Wright and Hobfoll 2004). COR theory states that resources—defined as objects (e.g., food), personal characteristics (e.g., job tenure), conditions (e.g., esteem of others), and energies (e.g., time)—are valuable to individuals (Hobfoll 1989). Therefore, the value of these resources motivates people to invest or expend current resources (e.g., time and energy) in order to gain additional resources (e.g., esteem or reputation), thereby increasing their resource pool and decreasing the chance of future resource depletion (Hobfoll 1989).

The first principle of COR theory states that loss of resources is disproportionately more salient than resource gain (Hobfoll 2001). Individuals must invest resources at their disposal (i.e., resource pool) to limit the loss of current resources. Thus, individuals who lack resources are more vulnerable to continued loss (Hobfoll and Shirom 1993). Within the stress literature, loss of resources has led to such issues as emotional distress (Hobfoll 2001) and job burnout (Wright and Hobfoll 2004). With regard to non-stress-related outcomes, loss of resources (under conditions of perceptions of politics) led to decreased performance (Treadway et al. 2005a).

Although resource gain is not as salient as resource loss, individuals strive to increase resources to minimize future losses. According to the second principle of the COR theory, individuals invest resources to protect and gain other resources (Hobfoll 2001). A continual gain results in an increase in resources at one’s disposal (i.e., resource caravans, Hobfoll 2001). Individuals who have strong resource caravans are better able to gain additional resources, reducing net resource loss (Hobfoll and Shirom 1993). This implies that individuals can build up an arsenal of resources even when no stress is present. In fact, it is believed that “individuals have natural tendencies to grow, develop, and achieve the highest levels of functioning for themselves and the systems in which they participate including families and organizations” (Wayne et al. 2007, p. 66). This natural tendency leads people to attempt to acquire resources that enable their functioning in specific roles (Wayne et al. 2007). The successful use of IM in organizations has been correlated with increased resources in the form of pay raises and extrinsic success (Higgins et al. 2003). Therefore, we argue that individuals engaging in successful IM will have higher levels of resources at their disposal.

Impression Management

Individuals use IM to influence their images and to project different identities to various significant others, and it is considered to be a subset of influence (Rosenfeld et al. 1995; Turnley and Bolino 2001; Wayne and Liden 1995). Although there are many reasons proffered about why individuals engage in IM (e.g., Baumeister and Leary 1995; Festinger 1954; Geen 1991), Geen (1991) argued that IM can be used to manipulate the audience for some immediate social or material gain. Within the workplace, people often use IM strategically over time to “influence salient outcomes such as performance ratings, compensation, and promotions” (Wayne and Liden 1995, p. 235), much in the same way individuals might engage in other common influence tactics, such as assertiveness or exchange (Higgins et al. 2003).

IM can include more automatic, unconscious behaviors; however, it also involves cognitively effortful goal-directed behaviors to influence the impressions of others (Bolino et al. 2008; Schlenker 2003). The primary focus of our research will be grounded within this body of literature—e.g., goal-directed, cognitively effortful behavior (Schlenker 2003) that includes conscious IM with the intention of acquiring resources. Tedeschi and Melburg (1984) argued that individuals strategically use IM to gain more resources without necessarily thinking about how these resources will be used in the future. Further, Liden and Mitchell (1988) suggested that individuals go through a risk assessment process when deciding when and how to use influence. This process involves assessing the target’s susceptibility, the perceived costs and benefits of the attempt, and the situational conduciveness. Thus, IM can be a time and resource consuming process (e.g., requiring time, energy, and effort) given that crafting an impression takes away from time and energy necessary for other roles within the workplace and is viewed as a cognitively effortful and demanding process. Indeed, recent research has shown that routine or more habitual self-presentations are more efficient, whereas effortful self-presentations are argued to consume more resources (Vohs et al. 2005).

This study will rely on three of the positive IM behaviors captured in one of the most widely used conceptualizations of IM by Jones and Pittman (1982). Specifically, and in line with research by Bolino and Turnley (2003), we will focus on the intentional positive IM behaviors of ingratiation, exemplification, and self-promotion. Noting that influence tactics are rarely utilized in isolation (i.e., relying only on one tactic) and individuals typically employ multiple tactics that form an influence style (e.g., Kipnis and Schmidt 1988), Bolino and Turnley (2003) demonstrated support for what they called IM “profiles.” One of these profiles, called positives, utilized high levels of ingratiation, exemplification, and self-promotion. These tactics are used with the intention of creating a positive impression to others (Bolino and Turnley 2003), such as attempts to be seen as likeable or hard-working. By design, the IM literature (and subsequently, the set of constructs used in this research and described in the measures section) discusses deliberate behaviors that acknowledge the intent of the tactic—perception management.

Ingratiation involves favor rendering, praise of others, and paying attention to others by taking an interest in their livelihood (Bolino and Turnley 1999). Exemplification consists of performing extra-role behaviors, such as staying late (and making such acts visible), in an effort to be seen as dedicated, busy, and hard-working (Bolino and Turnley 1999). Lastly, self-promotion involves expressing achievements and abilities in order to appear competent (Bolino and Turnley 2003). These IM tactics are conscious, goal-directed activities that are described as “appearing likeable” and “acting” in ways that might be contrary to one’s reality.

Bolino and Turnley (2003) found that when compared with individuals using aggressive tactics, such as intimidation, those using positive IM were seen as “more desirable work colleagues” (p. 155). Indeed, in a meta-analysis of influence behaviors (Higgins et al. 2003), outcomes of ingratiation were generally positive yet highly variable across studies. Exemplification, when viewed by targets, has been found to be positively related to perceptions of transformational leadership, leader effectiveness, follower satisfaction and the promotion of group cohesion, and positive feelings of group relationships (Gardner and Cleavenger 1998; Rozell and Gundersen, 2003). Self-promotion has been found to be strongly related to positive appraisals in the interview process (Higgins et al. 2003).

Although the goal of these three tactics is to procure a positive image in the eyes of others, this is not always successful. In a review of the social influence process and relevant outcomes (Ferris et al. 2002), ingratiation and friendliness did not always lead to positive outcomes. For example, Thacker and Wayne (1995) found a negative relationship between the use of ingratiation and promotability. Additionally, Bolino et al. (2006) did not find a significant relationship between exemplification-type behaviorsFootnote 1 and supervisor ratings of organizational citizenship behaviors. Furthermore, self-promotion has also been negatively related to performance evaluations by managers (Higgins et al. 2003).

Considering that not every attempt of positive IM is successful, it becomes necessary to examine individual differences and potential moderating variables (Ferris et al. 2002) in order to understand who is more capable of using IM to generate positive outcomes (without depleting one’s energy or existing resources). We propose that the ability to manage resources might be an individual difference construct that will impact the capacity to use positive IM behaviors through a more efficient use of the risk assessment process proposed by Liden and Mitchell (1988).

Ability to Manage Resources

A resource represents “physical, psychological, social, or organizational aspects of the job” (Schaufeli and Bakker 2004, p. 296) that help with job demands not only in order to accomplish workplace objectives, but also are important in their own right (Hobfoll 2002). However, individuals differ in their ability to control and manage physical, social, and organizational resources when under daily workplace demands (Hochwarter et al. 2007b, 2008). The ability to manage resources refers to the explicit ability to regulate oneself at work by using variations of exertion, respite, pace, flexibility, utilization of equipment/personnel, and task assistance (Hochwarter et al. 2008). This ability is viewed as a form of control (Hochwarter et al. 2007b, 2008), given that “those with high sense of control and mastery tend to use their resources judiciously, relying on themselves when this is most appropriate and calling on others when this is necessary” (Hobfoll and Shirom 2000, p. 62). Specifically, when those with a high ability to manage their resources are taxed, they are able to implement other resources and/or strategies to ensure that they do not deplete themselves entirely, can conserve their energy, and can reduce the demands being levied on them.

Research with regard to ability to manage resources has been grounded within the stress literature. Specifically, when interacting at work, this individual difference construct has been viewed as a form of proactive coping believed to translate into well-being due to one’s perceived ability to accumulate resources, regulate behaviors, and more effectively utilize available resources (Hochwarter et al. 2007b). Ability to manage resources has been found to mitigate the traumatic effects of hurricane-induced workplace stress (Hochwarter et al. 2007b), minimize the effects of work-induced guilt (Hochwarter et al. 2008), and buffer against the stressor of accountability (Zellars et al. 2011).

We argue that individuals with the ability to manage resources, especially when they are being taxed in some manner (e.g., by using resources to gain other resources), are able to regulate their behaviors and alter their own responses (Baumeister et al. 1994) in order to maximize the gain of resources while minimizing depletions (Hochwarter et al. 2007b), allowing for a more efficient use of IM in building resources caravans. IM behaviors, particularly those behaviors that are less automatic, are believed to be resource draining (Vohs et al. 2005) as they require time, energy, and effort. Thus, consideration of the risks/rewards, susceptibility of the target, and the conduciveness of the situation (Wayne and Liden 1995) are necessary in order to be effective. Those with a high ability to manage their resources, because of the self-regulatory abilities this affords them, should be more effective at engaging in the risk assessment process of IM, allowing them to better assess target susceptibility, the perceived costs and benefits of the attempt, and the situational conduciveness. Therefore, ability to manage resources should allow for the better selection, altering, and implementation (Thoits 1994) under the effortful, cognitively demanding use of IM, resulting in maximized outcomes (e.g., other resources) while minimizing resource depletion (i.e., energy expended on unsuccessful IM attempts).

Hypotheses Development

Taking a political perspective of resource theories (Ferris et al. 2007), we argue that IM are used in an effort to retain, protect, and build resources (Hobfoll 1989, 2001) that are personally relevant to one’s success within an institution. Two ways to build resources are to create a positive reputation and to have a higher quality relationship with your leader. Ferris and colleagues (2003, p. 215) defined reputation as “a perceptual identity reflective of the complex combination of salient personal characteristics and accomplishments, demonstrated behavior, and intended images presented over some period of time.” Framed within the resource theory literature, positive identity and the esteem of others are valuable resources that individuals aim to acquire (Hobfoll 1989).

LMX theory is based on role-making and social exchange theories and suggests that leaders develop differing relationships with each of their followers as a result of follower behavior, competence, and personal characteristics (e.g., Blau 1964; Graen and Scandura 1987). These relationships result in different yet important outcomes for both members of the dyad. For example, it has been shown that leaders can expect more extra-role behaviors and higher organizational commitment from followers who enjoy higher quality relationships (Illies et al. 2007; Major et al. 1995). Followers, on the other hand, can expect more support, rewards, interaction from their leader, and to be more visible when in higher quality relationships (Cogliser and Schriesheim 2000; Dienesch and Liden 1986; Harris et al. 2005). Thus, developing a positive reputation and higher quality LMX relationships are both dependent upon follower characteristics and behaviors. Further, both should result in higher social capital and additional valuable resources. Interestingly, Liden and Maslyn (1998) conceptualized their multidimensional version of LMX with a professional respect dimension defined as the mutual respect both parties have for each other’s professional capabilities. Sin et al. (2009) cite this dimension as “a function of personal reputation instead of common exchange episodes between both parties” (p. 1049). Thus, LMX is even conceptualized as partly reputational in nature.

Positive IM behaviors, which are used to create positive impressions, would be expected to be associated with workplace resources in the form of enhanced reputation and higher levels of LMX. Specifically, ingratiation is performed to appear likable by praising and giving favors to the target (Bolino and Turnley 1999). Exemplification involves acts of extra-role behaviors such as staying late in attempts to appear dedicated and hard-working. Finally, self-promotion comprises advertising one’s accomplishments to appear competent (Bolino and Turnley 1999). Being seen as likable, hard-working, and competent should be associated with a more favorable reputation. Indeed, as mentioned earlier, when used in conjunction, ingratiation, self-promotion, and exemplification have been found to increase the desirability of the person as a workgroup colleague (Bolino and Turnley 2003), indicating that these behaviors likely enhance one’s reputation.

Moreover, we expect that these positive impressions will lead to higher LMX quality. Using social exchange and upward influence theories, Dienesch and Liden (1986) suggested that the use of influence tactics (e.g., IM) by the subordinate can alter and enhance LMX quality. Successful management of one’s impression can positively change leaders’ attributions toward followers, thus resulting in more favorable assessments and higher quality relationships (Wayne and Liden 1995). Followers who want to maintain higher quality relationships with their leaders need to exhibit positive attitudes in response to task assignments, conform to work rules, and clarify expectations (Waldron 1991). IM is one way to accomplish this. In preliminary support, numerous studies have found a positive relationship between ingratiation and LMX (e.g., Yukl and Michel 2006; Yukl et al. 2008), and a meta-analysis on LMX has revealed a significant positive relationship between self-promotion and ingratiation and LMX (Dulebohn et al. 2012).

However, the mere use of IM in order to enhance reputation and LMX, particularly tactics that are inherently described as conscious or “acting,” can be energy-draining and place one at risk of losing other resources (Vohs et al. 2005). The process of using IM is potentially draining regardless if it is effective or not. That is, whether or not the target perceives the actor as being sincere does not change the fact that engaging in IM uses resources, such as time and energy. The drain in resources by using IM occurs when resources are expended on unsuccessful IM attempts, wherein no resources are gained as a result. Indeed, several studies of influence tactics, IM, and LMX have reported mixed results (i.e., Deluga and Perry 1991; Dockery and Steiner 1990), suggesting that the exertion of energy through the use of IM behaviors may not always result in the desired resource gains.

Thus, the ability to regulate oneself at work in order to use resources more judiciously is a necessary skill for the effective use of IM. As previously argued, an individual’s ability to manage their resources (Hochwarter et al. 2007b, 2008) allows for resource regulation, which should enable a more efficient use of IM without potential resource losses. That is, IM behaviors are demanding for most individuals, particularly if they are conscious. Those who are able to handle a myriad of demands at work properly pace themselves when things are hectic and are able to alter their behavior to ensure they do not “run on an empty tank.” Hence, they should be more successful when expending resources in the IM risk assessment process, allowing a more complete and accurate picture of the target, the environment, and the cost/benefit ratio. This will lead to using IM more effectively to increase positive impressions. Ultimately, they will be more likely to have a more positive reputation and higher ratings of LMX.

Hypothesis 1

The relationship between positive IM and (a) reputation and (b) LMX quality will be moderated by the ability to manage resources, such that the relationship will be negative at lower levels of ability to manage resources and will become positive as the ability to manage resources increases.

Given that reputation is largely perceptual in nature (Ferris et al. 2003), it is unlikely that this is the ultimate goal of individuals who engage in IM. It is more likely that the goal is improved performance ratings (the ultimate form of institutionally relevant financial rewards and resources), with reputation providing a perceptual backdrop against which one’s interpersonal behaviors are interpreted (Hochwarter et al. 2007a; Liu et al. 2007). Those with more favorable reputations are viewed to be effective and competent (e.g., Bromley 1993; Liu et al. 2007; Tsui 1984), and these views should be related to higher performance evaluations. In fact, in three separate studies (including one study using structural equation modeling), and based on signaling theory (Spence 1974), reputation served as an important antecedent to enhanced self- and supervisor-reported performance ratings, “confirming prior notions that job performance can be at least partially a function of social construction” (Liu et al. 2007, p. 162). Similarly, in a two-study test, reputation served as an important moderator in the relationship between political behaviors and job performance ratings (Hochwarter et al. 2007a).

Further, higher quality LMX relationships have often been related to higher follower performance. Research has shown that performance ratings can be the result of factors other than objective performance, encompassing things such as affect or liking (Cardy and Dobbins 1986). According to COR theory, resources can be used to gain other resources (Hobfoll 1989; Thoits 1994). Higher quality relationships provide individuals with support, more frequent or quality interactions, and essentially, more resources, which increase one’s ability to perform on the job. Indeed, many studies have reported significant, positive relationships between LMX quality and performance ratings (e.g., Duarte et al. 1994; Gerstner and Day 1997; Wayne et al. 1997). Therefore, the use of positive IM is not directly related to performance evaluations, but rather operates through reputation and the LMX relationship. Thus,

Hypothesis 2

(a) Reputation and (b) LMX will be positively related to ratings of performance.

Hypothesis 3

The relationship between the interaction of positive IM and ability to manage resources and performance will be mediated by (a) reputation and (b) LMX.

Method

We conducted a two-study design as a constructive replication in order to enhance the validity of the results (e.g., Lykken 1968). To provide further insight into the types of resources that are associated with IM in organizations, Study One focused on reputation, whereas Study Two investigated the role of LMX.

Sample and Procedures

Study One

We acquired the sample with the assistance of undergraduate students attending a university in the Southeast United States. As one option for extra credit, students could email the survey link to family and friends who were working. To ensure students did not feel coerced to assist, an alternate extra credit opportunity was available to students who chose not to share the survey link. Respondents provided their name and phone number, and students were told that the respondents could be contacted to verify authenticity. The collection of phone numbers was utilized as a deterrent to student fabrication of data through their own participation in the survey. This method of data collection has been used successfully in a number of studies (Liu et al. 2004; Treadway et al. 2005a, b) and is conducted in an attempt to increase the generalizability of findings across multiple contexts (Hochwarter et al. 2007b).

Data in Study One were collected at two time points. Antecedents (positive IM and ability to manage resources) were collected at Time 1 and outcome data (reputation and performance evaluations) at Time 2, approximately 1 month later. We received 286 completed surveys in Time 1 and 264 in Time 2. Respondents who completed data collection in Time 1 and Time 2 were matched based on demographics and phone numbers. After matching fully completed surveys, a total of 213 surveys were retained. In order to compare response characteristics of those responding to both waves, we collected a few demographics at both Times 1 and 2. In Time 1, 86 % reported working full time, whereas in Time 2, 88 % reported working full time. In Time 1, 42 % were male and in Time 2, 44 % were male. Finally, in Time 1, the average age was 38 years (SD = 13.66) and in Time 2, the average age was 40 years (SD = 10.94). Race and Tenure were only collected in Time 2, and race was primarily white (78 %), with a mix of Hispanic (8 %), African American (7 %), and Asian (4 %). Average organizational tenure was 7.5 years (SD = 8.9).

Study Two

Data were collected from several organizations from leaders and their followers, thus reducing the concern of common method variance (Podsakoff and Organ 1986; Podsakoff et al. 2003). The organizations were contacted via local Society for Human Resource Management chapters and personal contacts of the authors. All organizations were located in the United States and included such industries as medical, technology, and manufacturing. Surveys were mailed, emailed, or made available online using Survey Monkey, depending on the specific needs and requirements of the organizations. In all but one case, participants were all employees in the organization. In the one case, a subset of departments was used instead of all the employees due to the size of the organization. Leaders and followers were matched based on the identification codes selected by the researchers. All participants were informed that their responses were confidential, and their participation was voluntary and unpaid.

In all, 64 leaders and 474 followers were contacted with 43 leader respondents and 233 follower respondents, resulting in a 67 and 49 % response rate, respectively. From the surveys received, there were 84 useable follower surveys and 37 usable leader surveys, resulting in 84 matched dyads (usability indicating that the surveys could be matched and did not contain excessive missing data). The majority of surveys that were excluded were due to an inability to match, with 13 being excluded for missing data. As suggested by Downey and King (1998), cases missing more that 20 % of their data were dropped from further analyses. Leaders rated on average 3.13 subordinates.

The leader respondents were 92 % Caucasian, 2.7 % African American, and 2.7 % Asian, predominantly female (60.0 %), averaged 51 years in age, and had 12.5 years in organizational tenure. The follower respondents were 81 % Caucasian, 11.3 % African American, 1.7 %, Asian, and 3.5 % Hispanic. The follower respondents were mostly female (78.3 %), averaged 44 years in age and 10 years in organizational tenure.

Control Variables

Age Age was self-reported in years and treated as a continuous variable. Based on theory and previous research (e.g., Bolino and Turnley 2003; Thacker and Wayne 1995), age was included as a demographic control variable in the data analysis to eliminate noise and potentially spurious relationships. For instance, a meta-analysis conducted by Sturman (2003) illustrated that a curvilinear relationship exists between age and job performance.

Measures

Positive IM

IM behaviors that are positive in nature were collected using the Bolino and Turnley (1999) measure, including the three sub-dimensions of ingratiation, exemplification, and self-promotion. This measure captures intentional IM, not automatic IM. Respondents used a scale of 1 “never behave this way” to 5 “often behave this way” [Study One (α = .89); Study Two (α = .91)]. Example items include, “I compliment my colleagues so they will see me as likable,” “I stay at work late so people will know I am hard working,” and “I make others aware of my accomplishments.” Exploratory and Confirmatory Factor Analyses (EFA, CFA) were conducted on both sets of data (i.e., Study One and Study Two) to justify collapsing the sub-dimensions of ingratiation, exemplification, and self-promotion into one variable. In both studies, data demonstrated two clear factors of IM behaviors. Items comprising ingratiation, exemplification, and self-promotion all clearly loaded on to the first factor with no cross-loadings. Items comprising the remaining two sub-dimensions of IM (supplication and intimidation) loaded on to the second factor with no cross-loadings. Thus, the clustering of positive IM was retained for further analysis.

Ability to Manage Resources

Ability to manage resources was collected using the six-item measure developed by Hochwarter and colleagues (2007b), using a scale of 1 “strongly disagree” to 5 “strongly agree” [Study One (α = .76); Study Two (α = .69)]. The measure was designed to capture a variety of resources, primarily individual characteristics and behaviors. The measure includes the following six items: “When work is stressful, I am able to conserve my energy,” “I have enough equipment and personnel at my disposal to fill in for me at work,” “When I feel like my battery is run down at work, I can get others to pick up some of the load,” “When work gets overwhelming, I am able to get away long enough to regain my strength,” “I am able to pace myself at work when things get hectic,” and “I can change my behavior at work to make sure that I don’t run on an empty tank.” This measure has shown satisfactory reliability in five data sets (α = .80, .82, .82, .78, .75) and showed convergent and discriminant validity with anticipated outcomes such as job tension, job satisfaction, and hurricane-related stress (Hochwarter et al. 2008). In another two datasets, ability to manage resources had satisfactory reliability (α = .82, .87) and exhibited convergent and discriminant validity with anticipated outcomes such as job satisfaction and life satisfaction (Hochwarter et al. 2007b).

Reputation (Study One)

Respondents completed a 12-item reputation scale that was developed by Hochwarter et al. (2007a) based on the theoretical underpinnings of Ferris et al. (2003). This scale consists of items such as “I am regarded highly by others,” “My colleagues see me as a person of high integrity,” and “People expect me to consistently demonstrate the highest performance” (α = .95). Responses were recorded using a scale of 1 “strongly disagree” to 7 “strongly agree.”

Leader–Member Exchange (Study Two)

LMX was measured by leaders using the 7-item unidimensional scale proposed by Graen and Uhl-Bien (1995) (α = .81). Sample items include, “I understand my subordinate’s job problems and needs,” and “I would characterize my working relationship with this subordinate as effective” and leaders rated these items using a scale of 1 “strongly disagree” to 5 “strongly agree.”

Performance Evaluations (Study One)

Respondents were asked to share their most recent performance appraisal ratings. We explained “Below are 6 statements that are related to the dimensions of many company performance reviews. Thinking about your last performance appraisal, circle the number that most closely reflects how your boss rated you for each measure of performance.” We used the six-item performance measure by Kipnis and Schmidt (1988) scored on a 1 “very poor” to 7 “outstanding” scale. Sample items include “ability to work independently,” “ability to solve problems,” and “overall performance” (α = .90).

Leader-Rated Subordinate Performance Evaluations (Study Two)

Subordinate evaluations of performance were measured from the leader’s perspective using Wright et al.’s (1995) 10-item performance evaluation on a scale of 1 “strongly disagree” to 5 “strongly agree” (α = .91). Sample items include, “On the job, this subordinate exhibits an underlying concern for doing things or tasks better, for improving situations,” and “This subordinate always gets things done on time.”

Results

Means, standard deviations, and intercorrelations for Study One and Study Two are reported in Table 1. In both studies, there was a significant negative correlation between age and positive IM [Study One (r = −.29, p < .01); Study Two (r = −.31, p < .01)]. In Study One, the associations between age and reputation and age and performance were positive (r = .15 and r = 0.16, respectively, p < .05). However, in Study Two, age was not related to LMX (r = −.09, ns) or performance (r = −.13, ns).

Table 1 Means, standard deviations, and intercorrelations
Full size table

Preliminary Results

Although Study One was collected at two time points, it included self-reported data collected from a single source, which could be subjected to common method bias (Podsakoff et al. 2003). To test if this was an issue in our sample, the full measurement model was tested using a confirmatory factor analysis (CFA) to statistically differentiate between the four study constructs: positive IM, ability to manage resources, reputation, and performance. Although the four-factor model (χ 2 = 1545.42, p < .01, df = 588, χ 2/df = 2.63, CFI = .81, RMSEA = .09, SRMR = .08) displayed acceptable fit by most standards (e.g., Hair et al. 1998; Hu and Bentler 1999; Browne and Cudeck 1992), modification indices were reviewed to understand if better fit could be achieved. Based on the modification indices, 10 errors within scales were correlated with each other. No parameters were estimated across factors. Simulation studies suggest that correlating errors within factors does not lead to seriously biased estimates in multiple indicator models (see Johnson and Creech 1983), and we estimated model fit using the adjusted four-factor model. This adjusted four-factor model resulted in better levels of fit across all the fit indices (χ 2 = 1150.93, p < .01, df = 518, χ 2/df = 2.22, CFI = .89, RMSEA = .07, SRMR = .08) and fit the data significantly better than the original four-factor model (Δχ 2 = 395.50, p < .01). These CFA results suggest that common method bias does not pose a significant problem in Study One.

Because Study Two data were collected from both leaders and followers, and a single leader-rated multiple followers, we tested if non-independence was an issue in our data. To do this, we calculated the interclass correlation (ICC), which “measures the extent to which values of a dependent variable are similar for individuals belonging to the same group” (Diez 2002, p. 590). The ICC score for the DV in Study Two is .26 suggesting that performance ratings are somewhat dependent on the leader and indicating that the data are non-independent.

Hypothesis Tests

We tested our hypotheses using syntax introduced by Preacher et al. (2007) for MPlus statistical software which tests conditional indirect effects utilizing the principles of path analysis (also called moderated path analysis; MPA). MPA is a highly recommended practice because it compensates for the drawbacks of the causal steps approach introduced by Baron and Kenny (1986; see Preacher et al. 2007; Hayes 2009; Rucker et al. 2011; Kline 2011). By using tests more akin to the Sobel (1982) tests, we are able to achieve more accurate estimates of direct and indirect effects (e.g., Edwards and Lambert 2007; Preacher and Hayes 2004). MPA tests conditional indirect effects, or when “the strength of an indirect effect varies across the levels of another variable,” the moderator (Kline 2011, p. 334).

Because of differences in the type of data from both studies (i.e., Study One reports single source data and Study Two reports dyadic data), we implemented different functions while analyzing each study to maximize the robustness of each test. Namely, in Study One, we tested conditional indirect effects while utilizing bootstrapping techniques. Bootstrapping is “one of the more valid and powerful methods for testing intervening variable effects” as demonstrated by the simulation research (Hayes 2009, p. 412) because it accounts for non-normality of data (Williams and MacKinnon 2008). In Study Two, because of the nested nature of dyadic data, we utilized an MPlus multi-level function that “[takes] into account stratification, non-independence of observations, and/or unequal probability of selection” (Muthén and Muthén 2010, p. 594), allowing us to control for non-independence of the nested data. MPlus does not yet have the capability to utilize bootstrapping methods in any multi-level method (Muthén and Muthén 2010); thus, bootstrapping methods were not conducted in Study Two.

Hypothesis 1a proposed that ability to manage resources would moderate the relationship between positive IM and reputation such that those who engaged in positive IM with a high ability to manage resources would have a more favorable reputation than those with a low ability to manage resources. Table 2 shows that, after controlling for age, the interaction was significantly and positively related to reputation (β = .79, p < .05). Supporting hypothesis 1a, Fig. 2 demonstrates that when ability to manage resources is high, the increased use of positive IM is associated with more favorable assessments of reputation. However, when lacking the ability to manage resources, increased use of positive IM is associated with less favorable reputation assessments.

Table 2 Conditional indirect effects of positive impression management on performance through reputation at different levels of ability to manage resources—Study One
Full size table
Fig. 2
figure 2

Ability to manage resources moderating positive impression management and reputation. Study One

Full size image

Hypothesis 1b proposed that ability to manage resources would moderate the relationship between positive IM and LMX such that those who engage in positive IM with a high ability to manage resources will have higher LMX relationships than those with a low ability to manage resources. Table 3 shows that after controlling for age, the interaction was significantly and positively related to LMX (β = 1.64, p < .01), providing support for hypothesis 1b. Graphical results, shown in Fig. 3, demonstrate that when ability to manage resources is high, the increased use of positive IM leads to higher LMX. However, when ability to manage resources is low, the increased use of positive IM leads to lower levels of LMX.

Table 3 Conditional indirect effects of positive impression management on performance through leader–member exchange at different levels of ability to manage resources—Study Two
Full size table
Fig. 3
figure 3

Ability to manage resources moderating positive impression management and leader–member exchange. Study Two

Full size image

Hypothesis 2a suggested that reputation would be positively related to performance. The results, shown in Table 2, demonstrate that reputation is a significant predictor of performance when controlling for positive IM and ability to manage resources (β = .51, p < .01), supporting Hypothesis 2a. Hypothesis 3a argued for the conditional indirect effect of positive IM on performance through reputation as moderated by the ability to manage resources. Bootstrap results demonstrated that the indirect effect of positive IM on performance through reputation is marginally significant at M + 1 SD of ability to manage resources (B = .09, p = .06). Given that the interaction term was significant, and the slopes at the traditionally used M ± 1 SD were not, we investigated a broader SD range away from the mean to more fully understand how the interaction functions (Kline 2011). Results indicated that the indirect effect of positive IM on performance through reputation is significant at high levels (M + 1.5 SD) of ability to manage resources (B = .13, p < .05). Figure 2 provides a graphical representation of these results. Taken together, these results indicate that those who are able to manage their resources leverage positive IM tactics in such a way that increases their reputation and ultimately impacts their performance evaluations (Mathieu and Taylor 2006), providing partial support for Hypothesis 3a.

Hypothesis 2b suggested that LMX would be positively related to performance. Table 3 shows that LMX is a significant predictor of performance when controlling for positive IM, ability to manage resources, and age (β = .56, p < .01), supporting Hypothesis 2b. Hypothesis 3b argued for the conditional indirect effect of positive IM on performance through LMX as moderated by the ability to manage resources. Conditional indirect effects of positive IM on performance through LMX were marginally significant at M −1 SD of ability to manage resources (B = −.11, p = .06), but not significant at M +1 SD of ability to manage resources (B = .08, ns). Again, we investigated a broader range of values away from the mean. The conditional indirect effect was significant at values greater than 1.5 SD away from the mean of ability to manage resources [M −1.5 SD (B = −.16, p < .05); M + 1.5 SD (B = .13, p = .05)], providing partial support for Hypothesis 3b. These results, taken with the plotted interpretation in Fig. 3, indicate that individuals who are able to manage their resources experience higher quality LMX relationships and ultimately higher performance. Further, individuals who are not able to manage their resources experience poorer LMX relationships and ultimately lower performance. These results support that LMX has an indirect effect on the interaction of positive IM and ability to manage resources on performance (Mathieu and Taylor 2006).

Discussion

The main purpose of this two-study replication was to investigate the mediating role of social resources in the positive IM-performance relationship for individuals with the ability to manage resources. IM can include effortful goal-directed behaviors (Bolino et al. 2008; Schlenker 2003) that are associated with enhanced organizational resources (Tedeschi and Melburg 1984). However, some individuals appear to be more capable of managing resources and carefully assessing the risk of using particular IM tactics (Liden and Mitchell 1988), allowing these individuals to gain more resources. Overall, results from both studies indicate that those with the ability to manage resources and who use positive IM have higher levels of social resources (i.e., reputation and LMX) that ultimately lead to higher job performance ratings.

A noteworthy result is that the direct IM—reputation and IM—LMX relationships were negative, indicating that IM attempts alone may actually have negative consequences for individuals. As with prior research, individual differences can alter the effectiveness of IM tactic use and can actually have deleterious effects if not exercised with caution (Harris et al. 2007). It is crucial for individuals to execute IM attempts in a strategic and judicious manner (i.e., managing their resources appropriately), engaging in thorough risk assessment processes. Furthermore, we found the ability to manage resources to be the catalyst for the mediating roles of reputation and LMX on the IM-performance relationship. For individuals who have a high ability to manage resources, reputation and LMX fully mediated the relationship between positive IM and performance evaluations. Together, this suggests that only the individuals who are able to manage their resources are able to gain and leverage social resources in their favor. On the other hand, when individuals lack the ability to manage their resources, IM tends to be associated with lower levels of social resources. The continued investment in IM attempts, which drains resources, is associated with lower levels of LMX and reputation (i.e., loss spirals). Lower levels of resources are subsequently associated with lower levels of performance. Indeed, the IM-performance relationship was fully mediated by LMX for individuals low in ability to manage resources such that use of IM led to lower levels of LMX and, consequently, lower performance.

Taken together, it appears that the ability to manage one’s resources is a crucial individual capability that allows individuals to secure positive work outcomes. In this light, our research expands the extant research on COR theory in a number of ways. First, we investigate an individual difference in the COR process, which is lacking in the current literature (Hobfoll and Shirom 2000). In delineating the impact of individual capabilities (e.g., ability to manage resources) on resource depletion and gain, we can begin to understand the methods successful individuals use that lead to efficient use of resources at work.

Secondly, the current study contributes to COR theory by highlighting the need to examine consequences beyond stress-related outcomes within the context of the workplace. These two studies indicate that social resources (i.e., reputation and LMX) can have positive effects on individual performance. Lastly, our research highlights the value of examining IM in light of COR theory. It seems that successful positive IM is enabled by an ability to manage one’s resources, and it is correlated with higher levels of social resources. This is important because extant IM literature has been argued to be lacking a theoretical basis (Bolino et al. 2008). Combining COR theory with Liden and Mitchell’s (1988) model of risk assessment allowed us to ground our proposed relationships in theory and shed new light on the potentially energy-draining IM process that is dependent upon individual differences.

Strengths and Potential Limitations

This paper has strengths and potential limitations that should be noted when interpreting the results. First, we demonstrated these relationships through a constructive replication provided by the two-study design. This gives us stronger support for our hypotheses and enhances the validity of our results (e.g., Lykken 1968), showing that the interaction between IM and the ability to manage resources impacts not only reputation, but also enhanced relationships in the form of LMX. Finding similar results across two different study designs, both with their own weaknesses and strengths, provides stronger support for the overall findings. In essence, the strengths of one study make up for the weakness of the other. The use of a two-time data collection in Study One hints at a potential for causality among the concepts of interest in the study. Additionally, Study Two used dyadic data, assessing leader reactions to subordinate IM in the form of LMX quality and performance.

In light of these strengths, it is also important to note the potential limitations. In Study One, all of the data were self-reported thus leading to the possibility of common rater bias. To minimize common rater bias, data for Study One were collected at two separate time points. It has been suggested that the collection of self-reported data at separate times minimizes common rater bias (Podsakoff et al. 2003). Additionally, results of the Harmon One-Factor test indicated that common method bias was of minimal concern in the data. Moreover, because of the dyadic nature of the data, Study Two has a relatively small sample size and the alpha reliability of ability to manage resources was .69, below the traditional cutoff of .70. Additionally, we sought a constructive replication with Study Two, which does not require that all the constructs be the same. However, it would be interesting to delineate the impact of LMX versus reputation in the proposed relationships. Future research should collect both measures in the same dataset. Thus, researchers would be more able to examine if it is LMX or reputation that accounts for more variance in the relationships in this study. Furthermore, the direction of causality should be tested in future research. For instance, one could argue that higher performance ratings could lead to enhanced reputation. Lastly, we were unable to capture objective measures of performance. As is the case with many organizations that rely on subjective measures of performance for personnel decisions, we had to rely on managers’ subjective evaluations of performance and employee recall of subjective performance. Future studies could expand these results by measuring objective performance, when possible.

Implications for Practice

Our findings indicate that the ability to manage resources is associated with higher levels of additional resources, which are associated with positive workplace outcomes, such as job performance. Specifically, when employees feel that they have the proper equipment, support, and the ability to pace themselves (these are items from the ability to manage resources scale), they are better equipped to use positive impression management and ultimately better at their jobs.

Because resources tend to caravan, in that positive resources lead to more positive resources (Hobfoll and Lilly 1993), training leaders to build higher quality relationships with their followers will also enhance employee performance. By training managers to identify follower needs (Graen et al. 1982), these higher quality relationships will be another way that employee can manage their resources. This should, in turn, lead to higher performance. Further, it has been suggested that managers encourage employees to seek balance in the workplace (Halbesleben et al. 2009). Thus, as part of their interactions with followers, managers should highlight the resources their followers have access to, as well as encourage them to use such resources to prevent loss and/or to build resource caravans. Therefore, managers in the workplace should be trained to identify resource needs, as well as supporting their employees to manage their resources wisely. For example, if an employee had a particularly draining project with a client that required a great deal of positive impression management, managers should not only allow, but encourage this employee to take a long lunch to recover.

Additionally, organizations may find it useful to provide detailed lists of resource saving services offered to employees, in order to increase their perceptions of their own resources as well as allowing the use of these resources in accomplishing task and work activities. For instance, many universities offer editing services to faculty. These services spare faculty the time and effort needed for editing, and the faculty can use those saved resources to begin new studies or improve their teaching. However, if employees are unaware of such services, they cannot take advantage of them, weakening their ability to manage resources.

Directions for Future Research

Future researchers are encouraged to consider using laboratory settings to investigate ability to manage resources and IM, as well as the tactics that might be seen in a more negative light, such as supplication and intimidation. For instance, giving students a task that cannot be finished in the allotted amount of time without managing one’s resources (e.g., by asking for help or consulting previously completed tasks) and then assessing what the use of intimidation and supplication does to reputation and LMX. Further, personality characteristics such as proactive personality may enable individuals to manage their resources more successfully. Additionally, it would be interesting to examine the impact of non-conscious IM or behavioral IM (as captured by observation), rather than relying solely on self-reports of intentional IM.

In addition to experimental designs, similar field research studies should explore the use of both other-ratings of LMX and negative IM tactics such as intimidation and supplication. Our study was concerned with the impact of follower impression management on leader assessments. However, there has been burgeoning work in LMX agreement, with many researchers finding that leaders and followers do not agree on the quality of their relationship (e.g., Gerstner and Day 1997; Joseph et al. 2011). Future studies should examine the impact of positive IM on LMX agreement and the relating impact on performance. It may be that followers who use positive IM and have the ability to manage their resources are more likely to have an accurate understanding of the LMX relationship with their leader, leading to higher levels of performance than those with an inaccurate understanding.

Moreover, negative IM tactics, such as intimidation, would clearly be resource draining in nature, potentially leading to resource loss spirals rather than resource gains. That is, negative IM tactics suggest workplace conditions that would cause one to behave in a more protective or reactive manner. In fact, prior research has shown that the use of intimidation is actually a result of job strain (Gallagher et al. 2008), but that individual differences can mitigate the negative consequences. The notion of supplication (e.g., “Pretending not to understand something to gain someone’s help”) would suggest an overload situation and the notion of intimidation (e.g., “Dealing strongly or aggressively with someone when they interfere in your business”) would suggest situational politics or other barriers to one’s goals. The situational conditions of politics or role stressors, and one’s subsequent use of intimidation or supplication, would likely lead to decreased social resources and lower performance ratings, in addition to the traditional strain outcomes.

Future researchers also should consider examining the impact of other influencing behaviors, such as coalition building, rationality, and consultation (e.g., Yukl et al. 2008). Influence tactics, such as exchange, might also help one manage resources more effectively by allowing the trading of one resource (e.g., time) for another (e.g., help on a project). Does being able to manage one’s resources help one more efficiently use these other forms of influence, or is it only useful when managing an impression? Further, are those using influence tactics attempting to build a resource caravan? If so, are those able to manage their resources more successful at using influence to get resources, such as co-worker support and commitment? Moreover, researchers should examine the use of specific resources and what conditions and contexts certain resources would be most beneficial. For instance, it may be that directing time and effort toward IM may not work in all contexts. In certain contexts, it might be better to use that same time and effort toward something else, such as task performance. Lastly, it would be interesting to look at the actual gain and depletion of resources over time with a longitudinal investigation.

Conclusion

The acquisition of resources within the workplace mirrors the political process (Ferris et al. 2007). As such, our research was designed to understand resource gains through the use of conscious, positive IM tactics—an important gap in our understanding of resources beyond the domain of stress research. We illustrate that even positive IM tactics can be detrimental to important workplace outcomes, unless the individual has the ability to manage other resources and improve outcomes in a goal-directed fashion. The acquisition of social resources in particular, such as enhanced LMX relationships and reputation, can lead to improved performance ratings—arguably one of the most sought after conditions within the workplace.