Introduction
Our understanding of the supervisor–subordinate relationship, which is central to organizational effectiveness, has been illuminated by recent forward-thinking perspectives on subordinate performance and outcomes, diversity, equity, and inclusion, remote work, and supervisor status (Gong & Sims, Reference Gong and Sims2023; McCluney, Bryant, King, & Ali, Reference McCluney, Bryant, King and Ali2017; Shih, Lin, & Lee, Reference Shih, Lin and Lee2023; Xu, Qin, Dust, & DiRenzo, Reference Xu, Qin, Dust and DiRenzo2019). Transitioning to remote work during the COVID-19 pandemic influenced supervisor–subordinate relationships by increasing feelings of mistrust and psychological contract breach (Gong & Sims, Reference Gong and Sims2023). Shih and colleagues (Reference Shih, Lin and Lee2023) theorized that supervisors would provide mentoring when they perceived a beneficial return on their actions. Finally, Xu et al. (Reference Xu, Loi and Chow2023) found that supervisors’ perspective of subordinate voice may help or harm the supervisor’s status thereby affecting the subordinate’s promotability and visibility. Although these topics represent leading-edge thinking by management scholars, all of these efforts depend on an important foundational perspective: the supervisor’s evaluation of the subordinate’s performance.
Although highlighted as a topic for further research (Warr & Routledge, Reference Warr and Routledge1969), supervisor satisfaction with subordinates has received little attention even though supervisors are responsible for evaluating employee performance, determining employee pay, raises, promotions, growth opportunities, etc., all which impact subordinates’ subsequent attitudes and performance, e.g., in-role performance and extra-role performance on behalf of the organization (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005, Cropanzano, Anthony, Daniels, & Hall, Reference Cropanzano, Anthony, Daniels and Hall2017; Mitchell & Wood, Reference Mitchell and Wood1980). Thus, addressing the lack of attention to supervisors’ satisfaction with subordinates is of critical importance to subordinates’ experiences in organizations and, ultimately, to their work on behalf of organizational goals. Our studies focus on supervisors’ evaluation of subordinate performance based on what subordinates do for them via their work behaviors (Beehr, Weisbrodt, & Zagumny, Reference Beehr, Weisbrodt and Zagumny1994; Weisbrodt, Reference Weisbrodt1991; Locke, Reference Locke and Dunnette1976). As explained by Beehr et al. (Reference Beehr, Weisbrodt and Zagumny1994), supervisors are likely to view subordinates favorably when they help supervisors reach their own goals. Nevison, Comier, Pretti, and Drewery (Reference Nevison, Cormier, Pretti and Drewery2018) studied how the quality and quantity of work affected supervisor satisfaction with subordinates. However, what has yet to be explained is what type of work impact supervisors’ satisfaction with subordinates and if all quantities of work yield positive results.
We assert that supervisors’ perceptions of how subordinate work performance will help or hurt them influences supervisors’ evaluation of subordinate performance and subsequent supervisor satisfaction with the subordinate. We use need theories to explain why and when these relationships between performance and supervisor satisfaction result. Supervisors are responsible for conducting subordinate performance appraisals or providing input for appraisals and their satisfaction will impact subordinate performance, subordinate subsequent career progression, and subordinate contributions to overall organizational performance (Arvey & Murphy, Reference Arvey and Murphy1998; Denisi & Murphy, Reference Denisi and Murphy2017; Mitchell & Wood, Reference Mitchell and Wood1980).
Also, understanding when delivery of promised contributions will help supervisors meet their needs, and when it will not, is important for subordinates. While supervisors are dissatisfied when contributions are deficient of promised amounts, met or exceeded promises do not yield supervisor satisfaction in the same way. Our studies show that, depending on the type of contribution or amount of fulfillment, supervisor satisfaction with subordinates may be lower than subordinates may intuitively expect. Specifically, excess delivery of some contributions yields negative satisfaction from supervisors. This negative effect of excess delivery may indicate that supervisor needs are hindered, which is in line with Warr’s (Reference Warr1994) Vitamin Model. The Vitamin Model stipulates nonlinear relationships between job characteristics like job satisfaction and subordinate well-being (De Jonge & Schaufeli, Reference De Jonge and Schaufeli1998). It is also an example of the meta-theoretical principle presented by Pierce and Aguinis (Reference Pierce and Aguinis2013) called the too-much-of-a-good-thing effect, which also ‘accounts for an apparent paradox in organizational life: ordinarily beneficial antecedents causing harm when taken too far’ (p. 314).
Psychological contracts capture the fundamental agreements between subordinates and employers regarding inducements paid for with subordinate contributions (Rousseau, Reference Rousseau1989). Thus, we study this phenomenon through the lens of psychological contracts for three contribution types: task, organizational citizenship behaviors (OCB), and socioemotional support. Research shows that supervisors’ evaluation of the type of psychological contracts results in varied outcomes (Shih et al., Reference Shih, Lin and Lee2023); we assert that their evaluations of different amounts of contributions (representing deficient, fulfilled, or excess delivery of promised work) will result in varied amounts of supervisor satisfaction. We draw from need theories to develop theoretical arguments for how supervisors evaluate their subordinates’ contributions.
Our paper extends research about supervisor satisfaction with subordinates, psychological contract theory, and performance evaluation. For supervisor satisfaction with subordinates, scholars recognize the lack of research in this area (Nevison et al., Reference Nevison, Cormier, Pretti and Drewery2018; Warr, Reference Warr1994); through our studies, we contribute research to a neglected aspect of supervisor’s job satisfaction, which has downstream impacts on subordinate performance and subsequent organizational success. Also, our research more specifically tests when and why supervisors’ satisfaction with subordinates occurs (i.e., varying types and quantities of contributions) and provides need theories as foundations for hypotheses development that we believe better explain the relationship between work contributions and supervisor satisfaction with subordinates. (Previous papers focused on other theories.) Research has shown that supervisors’ performance evaluations are not mechanistic and often do not represent absolute truth. Many factors may influence supervisor evaluations, such as supervisors’ personal needs (Fried & Tiegs, Reference Fried and Tiegs1995). Additionally, need theories better explain how psychological contracts are evaluated (Lambert, Reference Lambert2011). It follows that supervisor satisfaction with subordinates might also depend on whether the supervisor needs what the subordinate provides.
Our studies extend psychological contract theory beyond subordinates’ experiences of breach and fulfillment by developing theory for how supervisors evaluate their contracts with subordinates. Psychological contracts concern the fundamental building block of relationships between supervisors and subordinates that is essential to doing work in the organization. We use the expanded view of psychological contracts (Lambert, Edwards, & Cable, Reference Lambert, Edwards and Cable2003) to show that supervisors’ view of contributions that exceed promised amounts may result in negative attitudes toward the subordinate. Relatedly, we contribute to an understanding of performance evaluation in two ways; first, by making explicit a standard by which supervisors may evaluate subordinates’ performance, i.e., promised contributions, and second, showing that excess contributions do not necessarily translate into supervisors’ increased satisfaction with subordinates. It depends on what supervisors need. Our results challenge the assumption that supervisors always desire larger amounts of work from their subordinates, adding an important qualification to theories of performance, which implicitly assume that more is always better.
Literature review/hypothesis development
Supervisors’ satisfaction with subordinates’ contributions
We treat supervisors’ satisfaction with subordinates as a concept parallel to the conceptualization of subordinates’ job satisfaction (Locke, Reference Locke1969), capturing both affective and cognitive evaluations of their subordinates in the context of their work role. These evaluations align with the dimensions of job satisfaction introduced by Locke (Reference Locke1969) and expanded upon by Beehr et al. (Reference Beehr, Weisbrodt and Zagumny1994) – and align with who the supervisor perceives the subordinate to be (affective), and what the subordinate can do for the supervisor (cognitive; Beehr et al., Reference Beehr, Glaser, Beehr, Beehr, Wallwey, Erofeev and Canali2006). Supervisor satisfaction is a likely attitudinal outcome of informal and formal performance evaluation and is related to important outcomes such as supervisors’ willingness to distribute discretionary subordinate rewards, e.g., compensation, praise or recognition, mentoring, social or career support, choice assignments or preferential treatment (Bretz, Milkovich & Read, Reference Bretz, Milkovich and Read1992; Campbell & Wiernik, Reference Campbell and Wiernik2015; Ferris, Judge, Rowland, & Fitzgibbons, Reference Ferris, Judge, Rowland and Fitzgibbons1994), all outcomes that affect subsequent subordinate performance and organizational success. For these reasons, we view supervisors’ satisfaction with their subordinates as an important outcome.
The work subordinates perform can be loosely categorized into three types of contributions (Feldman, Reference Feldman1981; Hill, Reference Hill1987; Judge & Ferris, Reference Judge and Ferris1993). Task contributions include activities for which subordinates are formally rewarded by their organizations, e.g., meeting specified performance requirements and complying with rules and regulations (Borman & Motowidlo, Reference Borman and Motowidlo1997; MacKenzie, Podsakoff, & Fetter, Reference MacKenzie, Podsakoff and Fetter1991; Schleicher, Baumann, Sullivan, & Yim, Reference Schleicher, Baumann, Sullivan and Yim2019; Williams & Anderson, Reference Williams and Anderson1991). OCB contributions are subordinates’ contextual behaviors that contribute to the effective social and psychological functioning of the organization including volunteering and assisting others with enthusiasm and perseverance (Borman & Motowidlo, Reference Borman and Motowidlo1997; Organ, Reference Organ1997; Podsakoff, Whiting, Podsakoff, & Blume, Reference Podsakoff, Whiting, Podsakoff and Blume2009; Schleicher et al., Reference Schleicher, Baumann, Sullivan and Yim2019). Socioemotional support is demonstrating appreciation, approval, and concern for the well-being of others (Bowling, Beehr, & Swader, Reference Bowling, Beehr and Swader2005; Eisenberger, Rhoades Shanock, & Wen, Reference Eisenberger, Rhoades Shanock and Wen2020; Larson, Hunt & Osborn, Reference Larson, Hunt and Osborn1976; Semmer et al., Reference Semmer, Elfering, Jacobshagen, Perrot, Beehr and Boos2008). All individuals, including supervisors, have a need for relatedness or to form positive attachments with others (Baumeister & Leary, Reference Baumeister and Leary1995; Ryan & Deci, Reference Ryan and Deci2000); ordinary expressions of caring and support from subordinates may provide an emotional boost to supervisors, satisfying this need and increasing their personal resources (Sheridan & Ambrose, Reference Sheridan and Ambrose2020). Although previous studies focused on subordinate work as one variable (Beehr et al., Reference Beehr, Glaser, Beehr, Beehr, Wallwey, Erofeev and Canali2006) or as various attributes of the subordinate (Nevison et al., Reference Nevison, Cormier, Pretti and Drewery2018), we focus on task, OCB, and socioemotional support contributions because they represent subordinate contribution types that are essential to organization performance and are widely used types of performance in organizational studies (Astakhova & Ho, Reference Astakhova and Ho2018; Fischmann, De Witte, Sulea, & Iliescu, Reference Fischmann, De Witte, Sulea and Iliescu2018; Kamdar & Van Dyne, Reference Kamdar and Van Dyne2007; Schleicher et al., 1976). They represent a more in-depth and nuanced look at predictors that influence supervisor satisfaction with subordinates than were measured in Nevison et al. (Reference Nevison, Cormier, Pretti and Drewery2018) – quality and quantity of performance – providing deeper insight into what type(s) and what quantity of subordinate work result in supervisor satisfaction.
Theoretical foundation for supervisors’ evaluation of contributions
Role theory provides a framework for acknowledging that supervisors simultaneously inhabit two critical roles in the organization – agent and subordinate – and that each role embodies different interests and perspectives (Aldrich & Ruef, Reference Aldrich and Ruef2006; Eisenhardt, Reference Eisenhardt1989; Katz & Kahn, Reference Katz and Kahn1978; Sluss, Van Dick, & Thompson, Reference Sluss, Van Dick, Thompson and Zedeck2011).Footnote 1 First, supervisors have been delegated the authority to adopt an organizational perspective and to guide and direct their subordinates’ contributions toward organizational goals (Ambrose, Schminke & Mayer, Reference Ambrose, Schminke and Mayer2013; Krasikova & LeBreton, Reference Krasikova and LeBreton2012; Rousseau & McLean Parks, Reference Rousseau, McLean Parks, Cummings and Staw1993; Shore, Porter, & Zahra, Reference Shore, Porter, Zahra, Coyle-Shapiro, Shore, Taylor and Tetrick2004). Accordingly, supervisors acting as organizational agents are typically responsible for aspects of leading, offering technical guidance, coaching and, among other tasks, evaluating the performance of their subordinates (Bass & Bass, Reference Bass and Bass2008; Borman & Brush, Reference Borman and Brush1993).
The second role enacted by supervisors is their role as subordinates. Supervisors are accountable to their superiors, just as direct reports are accountable to their supervisors (Ambrose et al., Reference Ambrose, Schminke and Mayer2013; Katz & Kahn, Reference Katz and Kahn1978; Lichtman & Hunt, Reference Lichtman and Hunt1971; Sluss et al., Reference Sluss, Van Dick, Thompson and Zedeck2011). Supervisors have entered an employment relationship where they direct their own contributions of intellectual and physical effort toward organizational goals. In return, supervisors’ employing organizations reward them by delivering promised inducements; in short, supervisors have their own psychological contract of contributions exchanged for inducements (Bordia, Restubog, Bordia, & Tang, Reference Bordia, Restubog, Bordia and Tang2010, Reference Bordia, Restubog, Bordia and Tang2014; Rousseau & McLean Parks, Reference Rousseau, McLean Parks, Cummings and Staw1993).
We use a psychological contracts framework to examine subordinate performance because it enables us to assess quality and quantity of performance for specific work types. Previous studies established that subordinate performance and supervisor satisfaction are, indeed, separate constructs (Beehr et al., Reference Beehr, Weisbrodt and Zagumny1994). Thus, using a psychological contracts lens where performance of specific contributions is the antecedent to supervisor satisfaction with the subordinate is appropriate.
As indicated earlier, our studies focus on what the subordinate can do for the supervisor. Thus, we invoke need theories to explain how subordinate contributions satisfy supervisors’ personal needs (e.g., making their work life easier and reaching work goals; Beehr et al., Reference Beehr, Weisbrodt and Zagumny1994), physiological needs (e.g., inducements which provide for basic survival needs like food and shelter; Maslow, Reference Maslow1943), and psychological needs (e.g., competence, self-esteem, integrity, relatedness; Lambert et al., Reference Lambert, Edwards and Cable2003; Montes & Irving, Reference Montes and Irving2008; Ryan & Deci, Reference Ryan and Deci2000), resulting in supervisor satisfaction. Need theories have been used to explain how subordinates react to psychological contract fulfillment and breach (Irving & Montes, Reference Irving and Montes2009; Lambert, Reference Lambert2011). Specifically, subordinates may appraise their psychological contracts based on how well the terms meet their varied psychological needs (Lambert, Reference Lambert2011; Lambert, Bingham, & Zabinski, Reference Lambert, Bingham and Zabinski2020; Lambert et al., Reference Lambert, Edwards and Cable2003). Need theories have also successfully predicted when the effects of excess inducements lead to positive or negative outcomes (e.g., excess pay and variety lead to increased and decreased satisfaction, respectively; Lambert et al., Reference Lambert, Bingham and Zabinski2020, Reference Lambert, Edwards and Cable2003). Likewise, supervisors may evaluate promised and delivered contributions from subordinates to determine whether these contributions facilitate or hinder their needs fulfillment in two ways. First, delivered contributions from subordinates may directly influence supervisors’ need fulfillment in their role as agents of the organization because evaluations of promised to delivered contributions are directly relevant to attaining organizational goals.
Second, subordinates’ contributions may indirectly influence supervisors’ needs fulfillment because these contributions serve supervisors’ own interests in their roles as subordinates. Subordinates’ contributions may be bundled with, serve as the foundation for, or otherwise facilitate supervisors’ contributions to organizational goals, which indirectly influence supervisors’ inducements delivered by the organization (Coyle-Shapiro & Parzefall, Reference Coyle-Shapiro, Parzefall, Cooper and Barling2008) and thus their ability to meet physiological needs as well as needs for status (Lambert et al., Reference Lambert, Bingham and Zabinski2020). For example, superiors’ evaluation of the supervisor’s effectiveness (and likely the supervisor’s inducements) will undoubtedly be tied to subordinates’ abilities to deliver contributions as promised. As agents of the organization, supervisor competence is tied to being effective in the role (Mintzberg, Reference Mintzberg1989). As a result, supervisors may feel competent and effective as a leader when subordinates deliver contributions. Supervisors may also value subordinates who publicly voice expressions of support for organizational policies, perhaps easing their job of motivating and monitoring subordinates’ performance (Whiting, Podsakoff, & Pierce, Reference Whiting, Podsakoff and Pierce2008; Xu, Qin, Dust, & DiRenzo, Reference Xu, Qin, Dust and DiRenzo2019), which in turn may positively influence supervisors’ wellness (Vansteenkiste, Ryan, & Soenens, Reference Vansteenkiste, Ryan and Soenens2020). Just as need theories are a viable theoretical lens to explain subordinates’ appraisal of their psychological contracts (Lambert, Reference Lambert2011; Lambert et al., Reference Lambert, Bingham and Zabinski2020; Montes & Irving, Reference Montes and Irving2008), need theories suggests that supervisors may assess their subordinates’ delivery of promised contributions by the extent to which their own personal, physiological, and psychological needs are met.
Study 1 hypotheses
The hypotheses development for Study 1, focusing on task and OCB contributions, is presented below. Study 2 replicates a test of the Study 1 hypotheses and adds the contribution of socioemotional support.
Fulfillment
Promises of contributions are fulfilled when what was delivered equals promised amounts, but fulfillment may occur at low amounts or high amounts. When supervisors perceive that promised contributions of task and OCB performance have been delivered, subordinates did what they promised. However, supervisors may not be as satisfied with low fulfillment as they are with high fulfillment. While low fulfillment indicates promises have been kept, the low amount of contributions may be insufficient to meet supervisors’ needs. Subordinates’ work represents a key resource to help supervisors reach organizational and personal goals (Eisenberger et al., Reference Eisenberger, Shoss, Karagonlar, Gonzalez‐Morales, Wickham and Buffardi2014; Larson, Reference Larson1986). Fulfillment at low amounts represents a small amount of contributions that may hinder supervisors’ abilities to deliver their own promised contributions, thereby threatening supervisors’ future inducements from the organization and their ability to meet personal and physiological needs. They may also send a signal to supervisors’ superiors that they are not effective as a leader because they are not managing, guiding, or inspiring subordinates to delivery high amounts of promised contributions, thus threatening supervisors’ feeling of competency, a psychological need. Thus, low subordinate fulfillment of task and OCB contributions may not enable supervisors to meet their needs, resulting in low supervisor satisfaction.
When subordinates fulfill high promises of contributions, supervisors may be pleased because their need for competence is more likely to be met. That is, when subordinates perform well, supervisors may be seen as effective leaders. They have received greater subordinate contributions that they can apply to their own contributions to the organization, likely helping to secure future inducements for personal and physiological needs. The subordinates’ organizational contributions may involve tasks (e.g., meeting productivity quotas) or OCBs (e.g., arranging a successful company event). Thus, for both task and OCB contributions, we hypothesize the following:
Hypothesis 1a, b: Controlling for inducements, when supervisors perceive that promised and delivered (a) task and (b) OCB contributions are equal, their satisfaction with the subordinate will increase as the absolute levels of promised and delivered contributions increase from low to high.
Deficiency
When supervisors perceive that subordinates have delivered task and OCB contributions deficient of promised amounts, supervisors’ satisfaction with subordinates is likely low. Deficiency deprives supervisors of labor and task accomplishment toward organizational objectives, threatening supervisors’ ability to meet organizational goals as an agent of the organization (Morrison & Robinson, Reference Morrison, Robinson, Coyle-Shapiro, Shore, Taylor and Tetrick2004) and subsequently, their feeling of competence as an effective manager. They also may feel that their integrity is in question with superiors because they promised (via their own psychological contract with superiors) to provide contributions to meet goals and were not able to do so because subordinates did not meet promises to them. Supervisor inducements are in jeopardy when goals are not met, which could result in them not being able to meet personal and physiological needs. Our reasoning applies to both task and OCB contributions because, despite the distinctions between them, when contributions were promised, supervisors may have relied on both types of contributions to meet their own and organizational goals (MacKenzie et al., Reference MacKenzie, Podsakoff and Fetter1991; Organ, Reference Organ1988; Priesemuth & Taylor, Reference Priesemuth and Taylor2016; Rousseau & McLean Parks, Reference Rousseau, McLean Parks, Cummings and Staw1993). For these multiple reasons, supervisors’ satisfaction with subordinates will likely be low in response to deficient task and OCB contributions.
Excess
Intuitively, it might seem that supervisors would view excess task and OCB contributions favorably because they may be able to use these additional contributions as resources for organizational goals and, perhaps, personal goals. Instead, we argue that excess contributions may hinder supervisors’ efforts to satisfy their own needs as agents and as subordinates, leading to lower satisfaction with subordinates. More contributions than promised result in, at the least, wasted supervisor and subordinate energy and resources because there is no gain in supervisor satisfaction and, at worst, leads to detrimental effects for the supervisor, which leads to detrimental effects for the subordinate as supervisor satisfaction decreases. For example, excess task contributions may not be timed with demand, or may be at the expense of attention to other tasks or domains, indicating interference with meeting other obligations (Bergeron, Shipp, Rosen, & Furst, Reference Bergeron, Shipp, Rosen and Furst2013; Meier & Semmer, Reference Meier and Semmer2018). They can contradict quoted unit cost assumptions creating a problem for the supervisor. In a negotiated contract based on cost per unit, excess production, especially when consistent over time, suggests supervisors may lack insight into subordinates’ productivity and abilities. In consulting or law practices, excess time spent on client work beyond those hours contractually agreed upon is lost revenue for the supervisor. Supervisors may view exceeding production requirements, e.g., making too much food at a restaurant that leads to waste and unnecessary costs, as problematic. Such excess delivery described in these examples that illustrate the Vitamin Model (Warr, Reference Warr1994) and the too-much-of-a-good-thing effect (Pierce & Aguinis, Reference Pierce and Aguinis2013) may reflect badly on supervisors and result in reduced feelings of competence and effectiveness. Superiors may wonder if supervisors are monitoring subordinates sufficiently or explaining business consequences like lost revenue and increased costs when subordinates provide more than contracted or needed. As these examples illustrate, excess task contributions may add no value.
Excess OCB contributions may be unnecessary, unwelcome, and at the expense of required task contributions (Bergeron et al., Reference Bergeron, Shipp, Rosen and Furst2013; Thompson & Bolino, Reference Thompson and Bolino2018). For example, instead of helping a coworker with a task as promised, a subordinate may do the task for the coworker. This is above and beyond behavior that may benefit the organization but may not serve other legitimate objectives such as the supervisor guiding a coworker toward independent task completion (e.g., helping subordinates meet needs for autonomy and competence). Such OCBs may create additional work for the supervisor in terms of training, monitoring, and mentoring, which could negatively affect their wellness, ability to meet developmental goals of their subordinates. Excess citizenship contributions may resolve an immediate problem but perhaps at the expense of addressing a systematic or recurring need, one that if left unresolved could affect how effective the supervisor is viewed by superiors and the supervisor’s feelings of competence. Thus, we predict that excess task and OCB contributions will be associated with supervisors’ dissatisfaction with subordinates.
Combining our predictions for deficiency and excess creates an inverted U-shaped relationship where satisfaction with the subordinate becomes lower as delivered contributions increasingly deviate from fulfillment of promised contributions, whether toward deficiency or toward excess.
Hypothesis 2a, b: Controlling for inducements, when supervisors perceive that subordinates’ delivered (a) task contributions and (b) OCB contributions are less than promised, supervisor’s satisfaction will be low. Satisfaction will increase as delivered contributions rise to equal promised amounts, but as delivered contributions exceed promised amounts, supervisor’s satisfaction will decline.
Study 1 method
All data, analysis code, and research materials are available at https://osf.io/cvx4n/?view_only=d5b9c798c5b3461db4d4de2e04322eb1.
Sample and procedure
Supervisors managing students employed in a work study program in diverse departments and facilities (e.g., information technology, groundskeeping, physical science labs, student center) across a major southeastern university were surveyed through campus mail at the beginning of their employment relationship with their subordinate and 12 weeks later. Their student subordinates worked a minimum of 10 hours per week. Supervisors reported the amount of contributions that were promised and delivered by their subordinates and their satisfaction with subordinates in Study 1. Ninety-nine supervisors provided data at both time periods. Seventy-two percent of the sample was female; the average age was 41.9 years; 84% were White, 10% Black, and 6% other. All participants granted consent to be surveyed. Data for Study 1 were collected before institutional review board procedures were started; thus, no institutional review board number is available.
Measures
Task and OCB contributions, and satisfaction with subordinates were each represented by three item measures, following the model of Lambert et al. (Reference Lambert, Bingham and Zabinski2020). All respondents were asked two questions for each contribution item, i.e., “How much of this aspect do you believe that your subordinate has promised to give you in this job? and ‘how much of this aspect did you actually receive from your subordinate?’ Compensation rates were set by the university, and were not determined by supervisors in our sample, so we controlled for supervisors’ perspective of inducements by measuring their promised and delivered recognition, meaning praise for their subordinates. To avoid the possibility that contributions items might stimulate thought processes about supervisors’ satisfaction with their subordinates, we positioned items measuring the dependent variable before the independent variables in the survey, thus eliminating a potential deleterious cause of common method variance.
Plan of analysis
Analyses were performed using polynomial regression and response surface modeling to assess the joint relationship promised and delivered contributions with supervisors’ satisfaction with subordinates. (Edwards, Reference Edwards, Drasgow and Schmitt2002; Lambert, Reference Lambert2011). The following equation was used where DI, PI, DC, and PC represent delivered inducements, promised inducements, delivered contributions, and promised contributions, respectively:
Our tests of hypotheses focused on contributions (DC, PC, DC2, DC × PC, PC2) but our equation controlled for the interaction of inducements and contributions and the polynomial effect of promised and delivered inducements (PI × PC, DI × DC, DI, PI, DI2, DI × PI, PI2) to address internal validity concerns. Hypothesis 1 regarding fulfillment corresponded to a test of the DC = PC line; support is indicated by a positive and significant slope combined with null curvature. Hypothesis 2 concerns the breach line and would be consistent with negative curvature and nonsignificant slope along the DC = -PC line of the response surface.
We collected data on promised contributions at Time 1 and Time 2 (12 weeks later) and chose to use Time 2 data. Our decision was based on the reasoning that psychological contracts may be renegotiated and mentally revised over time such that supervisors’ perceptions of promised amounts may evolve (Morrison & Robinson, Reference Morrison and Robinson1997; Rousseau, Hansen, & Tomprou, Reference Rousseau, Hansen and Tomprou2018) just as supervisors recall subordinate behavior to assess performance (Solomonson & Lance, Reference Solomonson and Lance1997). It follows that supervisors, when evaluating promises made and contributions delivered, are likely to use their current, perhaps constructed, memory of this information, rather than retain a perfect memory of what occurred in the past (Schacter, Reference Schacter2012). Our choice to use Time 2 data to measure both promised and delivered contributions, along with satisfaction with subordinates, means that our data is cross-sectional but more closely corresponds to the process of psychological contract appraisal and performance evaluation processes. Accordingly, we paid special attention to the construct validity of our measures in the results.
Study 1 results
Descriptive statistics, correlations, and internal consistency reliabilities (coefficient alphas) are presented in Table 1. All coefficient alphas were above 0.84, indicating reliable measures. Promised task and OCB contributions were highly correlated with each other as were delivered task and OCB contributions (r = .70, .68, respectively). Similarly, promised task and OCB contributions were highly correlated with delivered task (r = .58) and OCB contributions (r = .71). Delivered task and OCB contributions were also highly correlated with supervisor satisfaction (r = .70 and r = .61, respectively). These correlations are consistent with reasoning that supervisors’ performance ratings may be influenced by factors such as subordinate dependability and friendliness as well as supervisor affect toward the subordinate and the supervisor’s general experience in the position (Borman, White, & Dorsey, Reference Borman, White and Dorsey1995; Ferris et al., Reference Ferris, Judge, Rowland and Fitzgibbons1994). Supervisor satisfaction as a proxy for performance ratings could reasonably be expected to correlate highly with delivered tasks and OCBs. Of the 99 cases and 21 items (2079 data points), there were no missing data. Inspection of the scatterplots of promised and delivered terms indicated that there were sufficient data in all quadrants satisfying a requirement for testing deficiency, excess and variations in fulfillment.
a The inducement was recognition (i.e., praise).
b Cronbach’s alphas are shown in parentheses on the diagonal.
c N = 99 for all variables.
* p < .05. **p < .01.
Confirmatory factor analysis
We conducted confirmatory factor analyses using Mplus (8.5) to further assess construct validity paying special attention to construct discrimination. Because the items for promised and delivered terms contain the same content but differed only by the framing of promised or delivered (e.g., how much did the subordinate promise to contribute and how much did they actually contribute), residuals of like promised and delivered items were correlated, e.g., promised task item #1 was correlated with delivered task item #1 (Cole, Ciesla, & Steiger, Reference Cole, Ciesla and Steiger2007). The results show our theoretical model had a significant chi square test indicating poor fit but an acceptable CFI and RMSEA (χ2 = 310.9, df = 160, p < .000; CFI = .93, RMSEA = .10 [.08, .11]) (Hu & Bentler, Reference Hu and Bentler1999). Latent variables were allowed to correlate, the standardized loadings were high, ranging from .72 to .98 and averaging .87, and inspection of the residuals indicated no problems with model specification. However, the number of parameters for the full model exceeded the sample size and we were concerned that model misfit might be camouflaged by the small sample size. To more fully test our measurement model for signs of misfit, we followed a strategy of testing a series of smaller models nested within our larger, target model to probe the construct validity of the measures. Thus, each smaller model contained a subset of the full set of variables but there was more power to examine relationships. Supervisor satisfaction was estimated with promised and delivered recognition, and with promised and delivered task contributions (Model 2), and with OCB contributions (Model 3), Model 4 tested promised terms (recognition, task, and OCB) along with satisfaction, and Model 5 included the delivered terms and satisfaction. Models 2 through 5 exhibited smaller, but still significant chi square values, and RMSEA values that ranged from .06 to .11 and acceptable CFI statistics ranging from .94 to .98. For all models, the loadings were high (ranging from .68 to .99 with an average of .86) and the results showed no signs of model misspecification. We concluded that, despite predictably strong correlations among our variables, there was evidence for construct validity.
Hypothesis testing
The seven terms involving inducements were significant as a set, as shown in Table 2, suggesting that supervisors’ satisfaction was unrelated to promised or delivered recognition. The variance explained by the set of terms involving contributions was significant (R 2 = 0.32 for task contributions and R 2 = 0.24 for OCB contributions). (Results from Time 1 and Time 2 data were similar to cross-sectional data and are provided in Appendix A.) Tests of the response surface features presented in Table 3 show a positive slope along the fulfillment line in combination with null curvature for both task and OCB contributions. These results indicate that satisfaction with the subordinate increased as fulfillment increased from low to high. We conclude support for Hypotheses 1a and 1b.
Estimate represents unstandardized regression coefficient.
LL = lower limit; UL = upper limit.
DC = delivered contributions; DI = delivered inducements (recognition).
PC = promised contributions; PI = delivered inducements (recognition).
R 2I, R 2C, R 2M refer to the variance explained by the inducements terms.
(PI × PC, DI × PI, DI, PI, DI2, DI × PI, DI2), and by the contributions terms (DC, PC, DC2, DC × PC, PC2), and in the total model, respectively.
* p < .05.
* p < .05; **p < .01.
Slope and curvature of fulfillment are associated with Hypotheses 1a and 1b; slope and curvature of breach are related to Hypotheses 2a and 2b.
b1–b5 represent the coefficients in the estimated regression coefficients reported in Table 2.
Excess was tested at the point on the breach (D = −P) line, where D = +1SD and P = −1SD.
Full support for Hypotheses 2a and 2b would be indicated by negative curvature along the breach line combined with a negative, nonsignificant slope. For both task and OCB contributions, the curvature of the breach lines was negative and significant but was accompanied by a positive and significant slope. We further tested the slope of the breach line at a point in the region of excess (+1sd-delivered, −1sd-promised) and found that for both task (slope = 0.07) and OCB (slope = −0.27), the slopes were nonsignificant. Yet the curvature of the breach line remained significant and negative for both contributions (task curvature = −.60**, OCB curvature = −.34*). In combination, these results indicate that when delivered task and OCB contributions fell short of promised amounts, satisfaction with subordinates was low but increased as delivered amounts increased to promised amounts. As excess delivered contributions increased to high amounts supervisors’ satisfaction decreased. The effect was asymmetrical such that deficiency was more strongly related to low satisfaction than excess contributions. Hypotheses 2a and 2b were partially supported. The graphed response surfaces in Figs. 1 and 2 illustrate the asymmetrical inverted U-shape along the breach line and the positive slope along the fulfillment line.
Discussion of Study 1 and prelude to Study 2
Testing our hypotheses in another sample would lend greater credence to our results and enable us to test another form of contribution – socioemotional support. Socioemotional support is usually observed from the subordinate’s perspective, e.g., how much caring and support does the supervisor provide the subordinate (Eisenberger et al., Reference Eisenberger, Rhoades Shanock and Wen2020; Ilies, Lanaj, Pluut, & Goh, Reference Ilies, Lanaj, Pluut and Goh2018; Lin et al., Reference Lin, Wu, Dong, Chen, Wei and Duan2022). Such caring and support relate to the universal beliefs that everyone should be treated with respect, dignity, and human warmth (Baumeister & Leary, Reference Baumeister and Leary1995; Dweck, Reference Dweck2017) and can help people fulfill their need for relatedness. As such, caring and support can be given by subordinates to supervisors and help supervisors fulfill key psychological needs. Subordinates may express socioemotional support to supervisors merely by asking about their day, acknowledging the difficulty posed by corporate policies, or coping with annoying events at work. Just as such offers of support from supervisors lubricate working relationships, we argue that supervisors’ views of subordinates are enhanced when they receive support from their subordinates (Sheridan & Ambrose, Reference Sheridan and Ambrose2020). Supervisors may infer promises of socioemotional support from formal and informal discussion with subordinates in hiring interviews in the same manner that supervisors evaluate promises of other types of contributions. We include socioemotional support because it is a form of contribution that may be important to supervisors’ satisfaction with subordinates.
The general predictions we developed for task and OCB contributions should hold when supervisors evaluate promised and delivered socioemotional support as well. Supervisors should be satisfied with fulfillment of socioemotional support at both high and low amounts, but we anticipate that fulfillment at high amounts is related to higher satisfaction with subordinates than fulfillment at low amounts. Deficient delivery of socioemotional support may rob the supervisor of the ability to meet relatedness needs, resulting in low supervisor satisfaction with subordinates. Similarly, we reason that excess socioemotional support may also result in low amounts of satisfaction. Excess socioemotional support may be perceived as threatening (Judge & Bretz, Reference Judge and Bretz1994) to supervisors’ emotional coping skills perhaps implying to supervisors that their subordinates believe they are unable to manage their work or personal lives, i.e., indicating that they are not competent in some way.
Vansteenkiste et al. (Reference Vansteenkiste, Ryan and Soenens2020) describe relatedness as follows: ‘Relatedness denotes the experience of warmth, bonding, and care, and is satisfied by connecting to and feeling significant to others.’ We propose that supervisors do not need excess socioemotional support because it may make supervisors feel that subordinates have overstepped the boundaries of their work role and generally be seen as awkward, superfluous, disingenuous (Eastman, Reference Eastman1994; Long, Reference Long2021; Vonk, Reference Vonk1998), or inappropriate. Accordingly, we predict that:
Hypothesis 1c: Controlling for inducements, when supervisors perceive that promised and delivered socioemotional contributions are equal, supervisor’s satisfaction with the subordinate will increase as the absolute levels of promised and delivered contributions increase.
Hypothesis 2c: Controlling for inducements, when supervisors perceive that subordinate’s delivered socioemotional contributions are less than promised, supervisor’s satisfaction with the subordinate will be low. Satisfaction will increase as delivered contributions rise to equal promised amounts, but as delivered socioemotional contributions exceed promised amounts, supervisor’s satisfaction will decline.
In Study 2, we test our predictions for socioemotional support along with repeating tests for our predictions regarding task and OCB contributions from Study 1.
Study 2 method
All data, analysis code, and research materials are available at https://osf.io/cvx4n/?view_only=d5b9c798c5b3461db4d4de2e04322eb1.
Sample and procedure
Study 2 employed a variation of a snowball sampling design. Students in a part-time MBA program at a large southeastern US university were invited to take a survey if they supervised another subordinate or were asked to recruit their own supervisor. For participating in the study, the MBA students were offered course extra credit. One-hundred and six supervisors completed surveys.
The mean age of the sample was 36 years and 63% were female. Fifty-seven percent of the sample was White, 21% Black, and 22% other. Supervisors came from diverse occupations including directors and managers in finance, marketing, and operations. Because of the sampling design, supervisors reported on only one subordinate and there were no nesting effects. All participants granted consent to be surveyed in accordance with institutional review board No. H14195 granted by Georgia State University.
Measures
We followed the model of Lambert et al. (Reference Lambert, Bingham and Zabinski2020) when developing items to measure contributions and inducements. For each contribution and inducement item, respondents were asked two questions, ‘One year ago my subordinate promised to: …’ to measure how much was promised and ‘Over the past year, he/she: …’ to measure how much was contributed. As in Study 1, we measured the dependent variable before the independent variables to avoid contaminating supervisors’ reports of their satisfaction with thoughts of subordinates’ promised and delivered contributions.
Hypotheses were tested using Eq. (1), as in Study 1, and estimated for each of the three contributions. We again controlled for the influence of inducements, pay in this study, by including all terms (DI, PI, PI × PC, DI × PI, DI2, DI × PI, PI2) related to inducements to address internal validity concerns.
Study 2 results
Correlations, descriptive statistics, and Cronbach’s alphas are presented in Table 4.
a The inducement was pay.
b Cronbach’s alphas are shown in parentheses on the diagonal.
c N = 108 for all variables.
* p < .05. **p < .01.
All coefficient alphas were above 0.86, indicating reliable measures and good internal validity. Confirmatory factor analyses were performed in Mplus (v8.5). A small amount of missing data was handled by within-person mean substitution, i.e., the missing value of item #1 was replaced with the mean of items #2 and #3. This procedure affected 0.1% of the data in the sample (Roth, Switzer, & Switzer, Reference Roth, Switzer and Switzer1999), affecting 108 values. We again inspected the scatter plots of the delivered by promised terms to ensure there was an adequate amount of data in all quadrants of the graph.
Confirmatory factor analysis
As in Study 1, we allowed the residuals for corresponding contributions items and inducements items to correlate (Cole et al., Reference Cole, Ciesla and Steiger2007). The hypothesized target model of nine latent variables (supervisor satisfaction, four promised and four delivered terms (pay, task, OCB and support) exhibited a significant chi-square test (475.8, df = 278, p < .001), but the CFI (.95) and RMSEA (.08 [.07,.09]) suggested reasonable fit. Again, we were dealing with a sample size that was smaller than the number of estimated parameters; as such, we further inspected nested subsets of the larger hypothesized model to test for signs of model misfit with a larger number of observations per parameter. Models 2–4 tested a single promised and delivered contribution (task, OCB, and socioemotional support, respectively) with supervisor satisfaction, and promised and delivered pay. Model 5 tested supervisor satisfaction with promised pay and three promised contributions, and Model 6 included satisfaction with delivered pay and the three delivered contributions. For each model, the chi-square statistics remained significant, but the CFI statistics exceeded .96 and the highest RMSEA statistics was .09 (.07, .11). The standardized loadings ranged from a low of .70 to a high of .98 (average of .90), and further inspection revealed no evidence of a mis-specified model.
Hypothesis testing
Table 5 and 6 along with Figs. 3–5 provide the polynomial regression results and graphs of the response surfaces. The inducements terms (7) were significant as a set in all three equations explaining only 16% to 18% of the total amount of variance explained in the equation, indicating that the promised and delivered contributions (variance explained ranging from 31% to 68%) were the primary influence related to supervisors’ satisfaction with their subordinates.
Estimate represents unstandardized regression coefficient.
LL = lower limit; UL = upper limit.
DC = delivered contributions; DI = delivered inducements (pay; control).
PC = promised contributions; PI = promised inducements (pay; control).
R 2M, R 2I, R 2C refer to the variance explained by the inducements terms (DI, PI, DI2, DI × PI, DI2, PI × PC, DI × PI), by the contributions terms (DC, PC, DC2, DC × PC, PC2), and in the total model, respectively. Standard errors are in parentheses.
* p < .05.
* p < .05; **p < .01.
Slope and curvature of fulfillment are associated with Hypotheses 1a–1c; slope and curvature of breach are related to Hypotheses 2a–2c.
b1-b5 represent the coefficients in the estimated regression coefficients reported in Table 5.
Excess was tested at the point on the breach (D = −P) line, where D = + 1SD and P = −1SD.
As seen in Table 6, the slope of the surface along the fulfillment line for OCB contributions and socioemotional support was significant and positive, and combined with the nonsignificant curvature, supported Hypotheses 1b and 1c indicating that fulfillment at high amounts was associated with more supervisors’ satisfaction with subordinates than was fulfillment at low amounts. Contrary to our predictions, the slope along the fulfillment line for task contributions was nonsignificant. Pertaining to Hypothesis 2, the negative curvature of the surface along the breach line for task and OCB contributions was significant, indicating that supervisors’ satisfaction declined as contributions were in excess. Our slope tests at a point in the region of excess (+1SD-delivered, −1SD-promised) were nonsignificant for task contributions and OCB contributions. In combination, the surfaces were asymmetrical such that the effects of deficiency were more severe, i.e., lower satisfaction, than were the effects of excess. We conclude partial support for Hypotheses 2a and 2b. Contrary to predictions, the positive slopes along breach for socioemotional support, combined with null curvature, indicated no support for Hypothesis 2c. Instead, for socioemotional support, deficiency was associated with low satisfaction and excess was related to increased satisfaction with subordinates. The response surface graphs in Fig. 3 for promised and delivered task contributions exhibit the asymmetrical negative effects of deficiency and excess. The apparent negative curvature along the breach line for socioemotional support (Fig. 5) was nonsignificant.
Discussion
Psychological contracts have traditionally been studied from the subordinate’s perspective, specifically focusing on the effects of fulfillment and breach of promised and delivered inducements (Bunderson, Reference Bunderson2001; Lambert, Reference Lambert2011; Lambert et al., Reference Lambert, Edwards and Cable2003; Montes & Irving, Reference Montes and Irving2008; Woodrow & Guest, Reference Woodrow and Guest2020). We shifted attention to supervisors’ appraisal of their psychological contracts with subordinates, specifically focusing on their perspective of breach and fulfillment of subordinates’ contributions. We conducted two studies to test how supervisors’ perceptions of promised and delivered contributions from their subordinates are related to their satisfaction with those subordinates. Controlling for relevant inducements (recognition in Study 1 and pay in Study 2), our results revealed that supervisors appeared to focus on what they receive from their psychological contract with their subordinates; that is, supervisors’ view of their subordinates’ promised and delivered contributions. This finding is consistent with findings on subordinates’ psychological contracts, which show that subordinates tend to focus on what they receive in the exchange relationship, namely their inducements, and that their views on their own contributions matter less (Lambert, Reference Lambert2011). Likewise, supervisors’ appraisal appears to be little influenced by the inducements they promise and deliver compared to the effect of their subordinates’ contributions.
Supervisors must simultaneously serve as agents of the organization, by managing their subordinates’ contributions to organizational goals, and as subordinates, seeking to optimize their own receipt of inducements from the organization. We reasoned that supervisors, as they enact both their agent and subordinate role, evaluate subordinates’ contributions against promised amounts according to need theories. That is, when evaluating subordinate performance, supervisors also evaluate the extent to which delivered contributions further or hinder their own interests (e.g., personal needs like making their work life easier and reaching work goals; Beehr et al., Reference Beehr, Weisbrodt and Zagumny1994, physiological needs like inducements which provide for basic survival needs like food and shelter; Maslow, Reference Maslow1943), and psychological needs like competence, self-esteem, integrity, and relatedness (Lambert et al., Reference Lambert, Edwards and Cable2003; Montes & Irving, Reference Montes and Irving2008; Ryan & Deci, Reference Ryan and Deci2000).
Psychological contract theory underscores the benefits of fulfillment and our results found that supervisors were in general satisfied with subordinates when delivered contributions equaled promised contributions. However, contrary to common interpretations of psychological contract theory, we found support (for four of five tests) that supervisors’ satisfaction with subordinates was lower when fulfillment was low compared to when fulfillment was high. It is possible that low amounts of contributions, even though they represent kept promises, may not provide supervisors with the amount of contributions needed to meet their own physiological, psychological, or personal needs. High amounts of such contributions, on the other hand, may help supervisors reach both personal and organizational goals, fulfilling needs for survival (via inducements), competence, integrity, status, and well-being. Thus, all fulfilled promises may not be equal.
Consistent with findings in performance evaluation (Crant & Bateman, Reference Crant and Bateman1993; Ilgen, Mitchell, & Fredrickson, Reference Ilgen, Mitchell and Fredrickson1981; Thompson & Bolino, Reference Thompson and Bolino2018), our results showed that supervisors’ satisfaction was low when contributions of all three types (task, OCB, and socioemotional support) were deficient of promised amounts. The negative effects of deficiency can be explained by psychological need theory, since deficiency may impair supervisors’ well-being because they have not received needed resources (Porter, Reference Porter1963; Vander Elst, Van den Broeck, De Witte, & De Cuyper, Reference Vander Elst, Van den Broeck, De Witte and De Cuyper2012) and could indicate ineffective leadership of subordinates, threatening supervisors’ needs for competence and integrity.
Despite the intuitive notion that excess contributions should be related to supervisors’ increased satisfaction, our results show that for excess task and OCB contributions (both studies) satisfaction with subordinates did not increase but leveled off or declined. Subordinates’ excess contributions may suggest they have inappropriately allocated their time and effort to tasks that impede or obstruct goal achievement, perhaps on other performance dimensions, for both the subordinate and by extension, the supervisor. For example, excess production may mean that supervisors have excess inventory without customers (e.g., in a florist’s shop, preparing more floral arrangements than ordered is wasteful) and may indicate poor management effectiveness to both the supervisor and superiors. As such, the supervisors’ need for competence might be negatively affected. Contrary to our predictions, excess socioemotional support (Study 2) was related to increased supervisors’ satisfaction with subordinates. It may be that excess socioemotional support satisfies needs for relatedness and strengthens relationships between supervisors and subordinates. Perhaps, at extremely high amounts, satisfaction may decrease because such excess feels intrusive or superfluous, but our data did not include such high amounts. We suggest further investigations of the effects of very high socioemotional support on supervisor satisfaction with subordinates.
Theoretical implications
Nearly six decades ago, Warr and Routledge (Reference Warr and Routledge1969) pondered why more research had not been conducted to explore the facet of job satisfaction from the supervisor’s perspective. Since then, few articles have answered that call (for exceptions, see Reference Beehr, Glaser, Beehr, Beehr, Wallwey, Erofeev and Canali2006; Beehr et al., Reference Beehr, Weisbrodt and Zagumny1994; Nevison, Reference Nevison, Cormier, Pretti and Drewery2018). Our studies contribute to this body of work to further demonstrate specific antecedents of supervisor satisfaction with subordinates using psychological contracts as a lens and need theories as its foundation. Using a psychological contracts lens demonstrates that subordinates’ promises of contributions as perceived by the supervisor may be an appropriate standard for subordinate performance evaluations. We introduce need theories to explain how what the subordinate can do for the supervisor (Beehr et al., Reference Beehr, Glaser, Beehr, Beehr, Wallwey, Erofeev and Canali2006) results in whether supervisors are satisfied (or not) with subordinates for the promise and delivery of task, OCBs, and socioemotional support that contribute to meeting supervisor physiological, psychological, and personal needs. Supervisors’ evaluations of their subordinates’ performance is likely to be a powerful influence on subordinates’ work experiences, opportunities, and well-being and is thus deserving of more study.
Our results add to the psychological contract literature by suggesting a framework by which supervisors evaluate their subordinates’ promised and delivered contributions. Our findings indicate that supervisors’ appraisal of their psychological contract with subordinates may mirror aspects of subordinates’ appraisal of their psychological contracts. That is, both parties are primarily focused on what resources they obtain from their exchange relationship. Moreover, our results are consistent with prior work that suggests that subordinates are concerned with more than promise-keeping (Lambert et al., Reference Lambert, Bingham and Zabinski2020; Rousseau & Tijoriwala, Reference Rousseau and Tijoriwala1998; Van der Schaft, Lub, Van der Heijden, & Solinger, Reference Van der Schaft, Lub, Van der Heijden and Solinger2020) – meaning that supervisors’ attitudes toward their subordinates may vary depending on whether subordinates kept promises at low or high amount, and that the effects of excess may depend on the type of contribution involved. By drawing from need theories, we suggest that it is necessary to account for the types of contributions involved, and how fulfillment and breach, whether by deficiency or excess, may facilitate or hinder supervisors’ psychological needs.
We also contribute to psychological contract theory by acknowledging that supervisors inhabit two roles simultaneously, an organizational agent role and a subordinate role. In each role, the theoretical principles associated with psychological need fulfillment may drive evaluations of their relationships with subordinates.
Our work contributes to understanding performance evaluation by showing that the comparison of delivered to promised contributions can serve to denote an important inflection point in supervisors’ satisfaction, indicating that supervisors’ perceptions of promised contributions may be a valuable standard. This finding was consistent with our contention that it may be worthwhile to jointly model not only the performance level (i.e., delivered contributions) but also the standard against which performance is compared (i.e., promised contributions).
Finally, our work will aid in theory development and testing. As scholars continue to study supervisor–subordinate relationships, our findings serve to help them refine various contexts and research questions about supervisors’ evaluation of subordinates’ performance. Supervisors do not always desire more from their subordinates in terms of work contributions. Researchers can now refine research questions to determine under what boundary conditions and through what additional mechanisms this foundational finding may be true.
Practical implications
Managers may find it useful to coach supervisors in how to provide feedback regarding their subordinates’ contributions. Supervisors should not only be able to guide subordinates who deliver deficient contributions but should be able to explain to subordinates when excess contributions are inappropriate and may even impede progress toward organizational goals.
Managers might also be alert to the possibility that they have inadvertently created conflicts between supervisors’ roles as agent and subordinate. For instance, there is an inherent conflict if supervisors’ inducements (subordinate role) increase as a result of subordinates’ excess contributions, but organizational agents are tasked with managing contributions to meet demand (agent role), avoiding excess and stockpiling. Or in another example, when supervisors are rewarded (subordinate role) for developing new business, perhaps at the expense of their attention devoted to supervisory duties, it may limit their efforts to develop and improve the quality of their subordinates’ contributions (agent role).
Subordinates should be aware that contributing work in excess of promised amounts may not be universally desirable from their supervisors’ perspective. Just as supervisors should provide structure and guidance to subordinates, subordinates may benefit from seeking feedback from supervisors regarding the quantity, quality, and utility of their contributions (Ashford, Reference Ashford1986; Tagliabue, Sigurjonsdottir, & Sandaker, Reference Tagliabue, Sigurjonsdottir and Sandaker2020).
Limitations and future research
A limitation of our work is that the data were cross-sectional and single source, albeit with largely consistent findings over two studies. Yet, in line with other person–environment fit studies and psychological contract studies (Edwards, Reference Edwards1996; Raja, Johns, & Ntalianis, Reference Raja, Johns and Ntalianis2004; Weber & Bauman, Reference Weber and Bauman2019), this design corresponds with the likelihood that respondents cannot rely on a photographic memory of promises but instead use a reconstructed memory of promises made in the past. Moreover, common method variance cannot be responsible for the curvilinear effects found in both of our studies (Siemson, Roth, & Oliveira, Reference Siemson, Roth and Oliveira2010). Future work might employ experimental designs to study the causal relationships occurring in psychological contract appraisal (Deng, Coyle-Shapiro, & Yang, Reference Deng, Coyle-Shapiro and Yang2018; Lambert, Reference Lambert2011; Montes & Zweig, Reference Montes and Zweig2009).
We examined only three types of contributions, but task performance and OCB performance are arguably the most examined types of performance (Katz, Reference Katz1964; Van Dyne & LePine, Reference Van Dyne and LePine1998). Including socioemotional support from subordinates to supervisors acknowledges that the important social exchange relationship between supervisors and subordinates is likely to be reciprocal. Yet additional studies may generate findings useful for promoting subordinates’ innovative behavior, problem solving, and decision-making skills.
We inferred the influence of supervisors’ roles as agents and as subordinates but did not empirically tease apart when and how each role was invoked during psychological contract appraisal. It may be that the agent role and subordinate role are generally aligned toward the same objectives but that in some organizational contexts or circumstances there may be conflict such that optimizing the agent role harms the interests of the supervisor from the perspective of the subordinate role, or vice versa. Future research might specifically theorize and examine such possibilities.
Promises of contributions may not be the most commonly used comparison standard for evaluating performance. Supervisors may be more likely to use explicitly and implicitly defined role requirements, or their memories of how other subordinates have performed, as comparison standards. But our results suggest that promised contributions are a viable standard because promised terms explained variance and provided inflection points. Future research may investigate not only a more comprehensive set of contributions but may also jointly model performance and other possible comparison standards (e.g., role requirements, peers’ performance; Posthuma, Campion, & Campion, Reference Posthuma, Campion and Campion2018) that supervisors may be implicitly or explicitly using to evaluate subordinate performance.
Although we collected two distinctly different samples, both were small and likely did not reflect the full heterogeneity of the population of subordinates with supervisory responsibilities. Supervisors in our sample were not high-level managers but tended to be in the lower rungs of their organizations (in both the work study and MBA samples) and their subordinates were simultaneously engaged in higher education. It may be that agent and subordinate roles are differently enacted depending on the characteristics of their organizations. Despite our largely supportive results, research should investigate future questions in larger and more diverse samples.
Our results suggest additional research from the supervisor’s perspective of the psychological contract is warranted. The effects of breach and fulfillment may influence not only supervisors’ satisfaction with subordinates but may also be related to supervisors’ performance, their well-being on the job, and perhaps their own career progression as well as related to important subordinate outcomes, such as subordinate performance, promotions, pay, and career progression.
Competing interest
The authors declare none.
Tanja R. Darden, PhD, is an assistant professor of management at Towson University. Her work focuses on the employment relationship, psychological contracts, and diversity.
Paata Brekashvili, PhD, is an associate professor of Business Administration at Buller School of Business of Providence University College. His research focuses on employment relationships and interpersonal exchange.
Dr Lambert is a professor and a William S. Spears Chair of Business at Oklahoma State University. Her scholarship has focused on the employment relationship, leadership, psychological contracts, person–environment fit theory, and research methods.
Ryan Currie, PhD, is an assistant professor at Abraham Baldwin Agricultural College. His work focuses on trust, development, and pedagogy.
Greg F. Hardt, MS, MA, is an assistant professor at Mount Saint Joseph University. His research focuses on trust and culture.